SEAFCO - Special Foundations Practice in Bangkok

Current Practices in Deep Foundations and Diaphragm Wall Construction
in Thailand is published to mark the auspicious occasion of SEAFCO

40th Anniversary. It is a compilation of 48 selected published papers in the
international conferences. The case studies and researched works presented 40th Anniversary Celebration of Seafco Public Company Limited
in these papers are mainly from the projects carried out by SEAFCO in
Bangkok and other parts of Thailand. Published papers authored by SEAFCO’s 19 December 2014
geotechnical engineers in other major projects and research studies overseas
are also included.
Current Practices in Deep Foundations and Diaphragm Wall Construction in Thailand

40th Anniversary Celebration of Seafco Public Company Limited
In Thailand, according to the authors’ experience as a contractor, development
in both construction and design aspects of wet-processed deep-seated
bored piles, barrette foundations and diaphragm walls in past four decades
Current Practices in Deep Foundations
are significant. With recognition of technical and economic advantages of
using these high capacity cast-in-place foundations by local practitioners, and Diaphragm Wall Construction in Thailand
it is expected that they will be more popular in the future construction industry
of Thailand and the region. However, in the authors’ opinion, there is much
work to be done with particular focus on constructability issues, concrete Editors:
technology for wet-processed bored piles and barrette, reliable but cost-effective
quality control testing and application of value-engineering in design and Zaw Zaw Aye
construction. Starting from the planning stage, site investigation, design,
construction and inspection should be integrated to enable designers, Aung Win Maung
contractors and construction inspectors to participate as a team with Thayanan Boonyarak
a common goal. Appropriate and practical specifications should be established
jointly by these parties for local soil conditions and construction methods. Chalermpol Taechakumthorn
Continuing education should be promoted for designers, inspection engineers,
and contractors. Nutthachai Prongmanee
SEAFCO is committed to continue its research-minded initiation to link Nutthapon Thasnanipan
theory and practice in deep foundations. We hope that researched works
and findings presented in this book will be useful and serve as a source
of reference to all in the field of geotechnical and foundation engineering,
particularly those who are involved in the construction industry of Thailand.
Seafco Public Company Limited

144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510
TEL : (662) 919-0090 - 7, FAX : (662) 919-0098
www.seafco.co.th 2014
WE BUILD MODERN FOUNDATIONS
40th Anniversary Celebration of Seafco Public Company Limited
19 December 2014
Current Practices in Deep Foundations

and Diaphragm Wall Construction in Thailand
Editors:
Zaw Zaw Aye

Aung Win Maung
Thayanan Boonyarak
Chalermpol Taechakumthorn
Nutthachai Prongmanee
Nutthapon Thasnanipan
Member of
PREFACE
Seafco Public Company Limited is a deep foundation contractor established in 1974. The company
has been involved in numerous foundation and deep excavation projects in Thailand, including
large-scale foundation works, infrastructure projects, and more recently the underground structures
of Bangkok’s second subway project (MRT Blue Line Extension Project). SEAFCO expanded its
operation overseas and successfully took part in the construction of diaphragm walls for Marina
Bay Sands in Singapore in 2007. SEAFCO has also established Seafco (Myanmar) Company
Limited in late 2013 and completed two projects to date.
Consolidating state-of-the-art technology, sophisticated equipment and engineering professionals,
SEAFCO initiated the innovative techniques and solutions in geotechnical and foundation
engineering in Thailand and Myanmar. To improve the currently used methods, material, system,
design concepts and construction techniques, SEAFCO has spent considerable time and effort
on Research and Development activities.
SEAFCO’s R&D team continuously conducts research on bored piles, barrettes, diaphragm
walls, associated testing methods, deep excavation and other geotechnical engineering works.
New findings and research results have been presented in a number of conferences and
seminars organized by various institutions both domestically and internationally. In forty years,
R & D Division of SEAFCO has produced over 100 technical papers (both in English and Thai)
which were presented and published in international conferences held in various countries. The
purpose of this publication is to mark the auspicious occasion of SEAFCO 40th Anniversary. It is a
compilation of selected published papers in the international conferences. The case studies and
research works presented in these papers are mainly from the projects carried out by SEAFCO
in Bangkok and other parts of Thailand. Published papers authored by SEAFCO’s employees in
other major projects and research studies overseas are also included.
I would like express my appreciation to the clients, consultants, designers, universities, institutions,
the Engineering Institute of Thailand under H.M. the King’s Patronage (EIT) and partners who have
been continuously supporting SEAFCO for 40 years. I also would like to acknowledge the effort of
SEAFCO’s R&D team for their commitment and hard work in the research works.
I hope that researched works and findings contained in these publications will be useful and serve
as a source of reference to all in the field of geotechnical and foundation engineering, particularly
those who are involved in the construction industry of Thailand.
Narong Thasnanipan
President & CEO
Seafco Public Company Limited
FOREWORD
First of all, I would like to extend my congratulations to SEAFCO for 40 years of successful
establishment, not only as a business entity but also as a great contributor for construction industry
of Thailand.
SEAFCO is well known for research minded initiatives in foundation engineering. SEAFCO shares
outcomes of its R&D works to the engineering society by publishing technical papers, organizing
conferences, and delivering lectures in various institutions in both Thailand and overseas. I
proudly acknowledge the contributions SEAFCO has been making to strengthen the foundation of
civil engineering society in Thailand and beyond.
Since the time I was working on my doctoral research, I have found SEAFCO’s publications
in various international conferences very useful to understand practical aspects of deep
foundations. SEAFCO puts the innovative theory into practice by equipping itself with knowledge,
experience and professionalism. SEAFCO R&D team has put in commendable effort to compile
all of their published papers and produced the book “Current Practices in Deep Foundations
and Diaphragm Wall Construction in Thailand”. Geotechnical specialists, foundation engineers,
structural engineers and practitioners in construction industry will find this book beneficial.
I, as the President of the Engineering Institute of Thailand under H.M. The King’s Patronage (EIT),
would like to sincerely thank Seafco Public Company Limited for endless support they have been
providing to the various activities of EIT under the leadership of CEO Khun Narong Thasnanipan. In
particular, SEAFCO’s great support in a leading role to Thailand Underground & Tunneling Group
(TUTG) in winning the bid for Thailand to host the World Tunnel Congress 2012 and successfully
organizing the event at Queen Sirikit National Convention Center in Bangkok from 18 to 23 May
2012.
I believe SEAFCO will continue to grow in business with trust and confidence of its stakeholders.
I am also confident that SEAFCO will maintain its position as the leading innovators of foundation
and geotechnical engineering.
I wish SEAFCO all the success.
Prof. Dr. Suchatvee Suwansawat

President
The Engineering Institute of Thailand
Table of Contents

Bored Pile and Barrette: Construction Methods, Design,
Analysis and Performance
Barrettes Founded in Bangkok Subsoils, Construction and Performance
Thasnanipan N., Maung A.W. & Tanseng P. .................................................................. 1
Design, Construction and Behavior of Bored Cast In-Situ Concrete Piles
in Bangkok Subsoil
Thasnanipan N., Teparaksa W., Maung A.W. & Ganeshan B. ........................................ 9
Large Diameter Bored Piles in Multi-Layered Soils of Bangkok
Thasnanipan N., Tanseng P. & Anwar M.A. .................................................................. 19
Effect of Construction Time and Bentonite Viscosity on Shaft Capacity
of Bored Piles
Thasnanipan N., Ganeshan B. & Anwar M.A. .............................................................. 29
Performance Comparison of Bored and Excavated Piles in the Layered Soils
of Bangkok
Thasnanipan N., Anwar M. A., Maung A.W. & Tanseng P. .......................................... 39
Base grouting of wet process bored piles in Bangkok subsoils
Teparaksa W., Thasnanipan N. & Anwar M. A. ............................................................. 51
Review of the Shaft Capacity Degradation of Bored Piles Constructed
with Bentonite Slurry
Thasnanipan N., Anwar, M. A. & Maung A. W. ............................................................. 57
Failure Mechanism of Long Bored Piles in Layered Soils of Bangkok
Thasnanipan N., Anwar, M. A. & Maung A. W. ............................................................ 69
Barrettes : A Versatile Foundation for Transmission Line Towers
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. ............................................ 77
P ractical Installation of Stanchions for Top-Down Construction
in Bangkok Subsoil
Thasnanipan N., Maung A. W. & Z. Z. Aye ................................................................... 85
Concrete for Wet Process Bored Piles
Thasnanipan N., Maung A. W., Tanseng P. & Navaneethan T. .................................... 97
Record Load Test on a Large Barrette and its Performance in the layered
soils of Bangkok
Thasnanipan N., Maung A. W. & Z. Z. Aye ................................................................. 107
Performance of Wet-Process Bored Piles Constructed with
Polymer-Based Slurry in Bangkok Subsoil
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Teparaksa W. ............................... 117
Barrette of Over 50,000 kN Ultimate Capacity Constructed in
the Multi-Layered Soil of Bangkok
Thasnanipan N., Z. Z. Aye & Teparaksa W. ................................................................ 135
Performance of Toe-grouted Large Bored Piles in Sand-Gravel
Formation of Chiang Mai City, Northern Thailand
Thasnanipan N., Z. Z. Aye & Teparaksa W. ................................................................ 145
Application of Polymer-based Slurry for Wet-process Bored Piles
Construction in Mylti-layered Soil of Bangkok
Z. Z. Aye, Singtogaw K. & Submaneewong C. ........................................................... 155
Development in Construction of Bored pile and Barrette for
Infrastructure Projects in Thailand, A Country Report
Thasnanipan N., Singtogaw K., Z. Z. Aye & Pravesvararat S. .................................... 163
Behaviour of Polymer-based Slurry for Deep-seated Bored Piles
in Multi-layered Soil of Bangkok
Thasnanipan N., Z. Z. Aye & Boonyarak T. ................................................................. 175
Effectiveness of Toe-Grouting for Deep-Seated Bored Piles
in Bangkok Subsoil
Thasnanipan N., Z. Z. Aye & Submaneewong C. ....................................................... 185
C urrent Practice and Future Trends of Cast-in-place Deep Foundation
in Thailand
Thasnanipan N., Singtogaw K., Z. Z. Aye & Pravesvararat S. ................................... 199
Development and Achievement of Deep-seated Bored Piles and
Barrettes Construction in Thailand in the Past Forty Years,
A Country Report
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Boonyarak T. ................................. 211
Barrette for High-rise Buildings and Heavy Structures in Bangkok
Sub-soil Thasnanipan N., Z. Z. Aye, Boonyarak T., Kitpayuck P.
& Thasnanipan N. ...................................................................................................... 223
Diaphragm Wall and Deep Excavation
Performance of a Braced Excavation in Bangkok Clay, Diaphragm Wall
Subject to Unbalanced Loading Conditions
Thasnanipan N., Maung A.W., Tanseng P. & Wei S.H. ............................................... 245
Prediction and Performances of Short Embedded Cast In-Situ Diaphragm
Wall for Deep Excavation in Bangkok
Thasnanipan N., Teparaksa W., Maung A.W. & Wei S.H. ............................................ 253
Lessons from the collapse during construction of an inlet pumping station:
Geotechnical instrumentation aspect
Teparaksa W., Thasnanipan N., Maung A. W. & Tanseng P. ....................................... 263
Behavior and performance of diaphragm walls under unbalanced
lateral loading along the Chao Phraya River,
Thasnanipan N., Maung A. W. and Tanseng P. & Teparaksa W. ............................... 271
Analysis of Lateral Wall Movement for Deep Braced Excavation of
Bangkok Subsoils
Teparaksa W., Thasananipan N., & Tanseng P. ....................................................... 279
Diaphragm Wall and Barrette Construction for Thiam Ruam Mit Station Box,
MRT Chaloem Ratchamongkhon Line, Bangkok
Thasnanipan N., Maung A.W. & Ganeshan B. ............................................................. 291
Monitoring of Diaphragm Wall Displacement and Associated Ground
Movement, Braced Excavation Adjacent to Historical Building at the Bank
of Chao Phraya River
Tanseng P., Z. Z. Aye & Submaneewong C. ............................................................... 299
Performance of Buttress-Support Thin Diaphragm Wall for Underground
Car Park in Bangkok
Thasnanipan N., Z. Z. Aye & Submaneewong C. ........................................................ 309
Stability of a Trial Trench Excavated under Polymer Slurry in Bangkok
Soft Clay
Thasnanipan N., Maung A. W., Boonyarak T. & Aye, Z. Z. .......................................... 315
Application of Top-down Construction Method for Deep Excavations
in Urban Area of Bangkok
Thasnanipan N., Z. Z. Aye, Submaneewong C. & Boonyarak T. ................................. 321
Construction of diaphragm wall for basement excavation adjacent to
tunnels in Bangkok subsoil
Thansnanipan N., Maung A. W., Z. Z. Aye, Submaneewong C.
& Boonyarak T. ........................................................................................................... 337
Application of Observational Method in Diaphragm Wall Support
Basement Excavation
Z. Z. Aye, Submaneewong C., Boonyarak T. & Chanchad C. ...................................... 351
Diaphragm Wall Support Deep Excavation in Bangkok
Thasnanipan N., Z. Z. Aye, Maung A. W. Boonyarak T., Kitpayuck P.&
Thasnanipan N. ........................................................................................................... 369
Underpass Construction in Congestive Area of Bangkok
Z. Z. Aye, Boonyarak T., Kitpayuck P., Mahasantipiya S. & Thasnanipan N. ...................... 387
Quality Control Tests on Deep Foundations
Sonic Integrity Test on Piles Founded in Bangkok Subsoil-Signal

Characteristics and Their Interpretations
Thasnanipan N., Maung A.W. & Ganeshan B. ........................................................... 405
Sonic Integrity Test of Piles-integrity Effected by Basement Excavation
in Bangkok Soft Clay
Thasnanipan N., Maung A. W., Tanseng P. & Z. Z. Aye ............................................. 415
Non-Destructive Integrity Testing on Piles Founded in Bangkok Subsoil
Thasnanipan N., Maung A. W. & Navaneethan T. & Z. Z. Aye ..................................... 425
Damages to Piles associated with Excavation Works in Bangkok Soft Clay
Thasnanipan T., Maung A. W. & Tanseng P. ............................................................... 435
Quantifying Pile Head Condition Before Basement Excavation by
Cross-hole Sonic Logging Tests
Thasnanipan N., Z. Z. Aye, Boonyarak T. & Kampananon N. ...................................... 445
Advantages and Limitations of Integrity Tests for Large Diameter Bored
Piles in Bangkok
Thasnanipan N., Z. Z. Aye, Boonyarak T. & Kampananon N. ...................................... 453
Application of Low-strain Integrity Tests for Quality Assurance of
Deep Foundations in Thailand
Maung A. W., Z. Z. Aye & Boonyarak T. ...................................................................... 463
Geotechnical Engineering : Research and Practice
Ground Movement Prediction and Building Damage Risk-Assessment

for the Deep Excavations and Tunneling Works in Bangkok Subsoil
Z. Z. Aye, Karki D. & Schulz C. ................................................................................... 473
Key Factors in Application of Observational Method – Bangkok Experience
Phienwej N. & Z. Z. Aye ........................................... ................................................... 495
Ground Movements Associated with the Underground Excavations
of the First Bangkok MRT Line
Phienwej N., Sirivachiraporn A., Suwansawat S. & Z. Z. Aye ...................................... 525
Geotechnical Risk Assessment and Management for Tunnelling Project
in Congestive Urban Area – Bangkok Experience
Z. Z. Aye, Kitpayuck P. & Phienwej N. ......................................................................... 535
Tunneling Effect on Pile Group Response in Bangkok
Boonyarak T. & Ng C. W. W. ....................................................................................... 547
SEAFCO PUBLIC CO LTD
SEAFCO
Barrettes
Strip Piles
When single pile safe
working load in excess of
2000 tons is required or
conventional bored piling
is impractical, barrettes or
D EEP FOU N D ATI ON S
strip piles are the best al-
ternative. Various dimen-
sions, cross-sections and
orientation can be formed
and arranged to suit de-
sign requirements.
SEAFCO barrettes are
used in foundations for ex-
pressways, MRTA stations
and high-rise buildings.
MEMBER OF
A - 13
A - 14
A - 15
A - 16
A - 17
PILE BASE GROUTING IN GROUTING ASSEMBLY Injection Rates: It is not a
BANGKOK SUBSOILS direct controlling parameter but
Current base grouting practice instead used to achieve maxi-
Cast-in-situ bored piles with base uses two U-circuits consisting of mum pressure in a gradual
grouting are in common use in 12mm diameter PE tubes at- manner. Injection rates of 5-
Bangkok for foundations of heavy tached at the bottom with 12mm 10litres/min are proved to be
structures. Base grouting is usually GI pipes with perforations satisfactory.
employed in bored piles to improve wrapped with rubber sleeves
pile capacity and performance. The (manchettes). They are attached Residual Pressures: Normall y
unavoidable problems arising from with the reinforcement cage to residual pressures of the order
loosening of soil and sedimentation place the manchettes at the base of 20-30% of target maximum
at the base caused by boring can of the pile. pressure are desirable.
be minimized by base grouting.
The first base grouted bored pile in
Bangkok was used in 1985 at the GROUTING CRITERIA PERFORMANCE OF BASE
Rama IX cable-stayed bridge GROUTED PILES
across the Chao Phraya River. Grout Pressures: Maximum
pressure developed is a single Base grouting with tube-a-
Teparaksa, W. (1994) reported that most important parameter. It manchettes technique has been
base grouted piles in the second should be gradually achieved by used by SEAFCO in many pro-
sand layer of Bangkok subsoils low injection rates to avoid hy- jects in Bangkok. Improve-
have an increment in their perform- draulic fracturing. ments achieved have been
ance as below; proved by load tests on both
Grout Volumes: Usually non-grouted and base-grouted
total grout take is not a limiting bored piles on the same project
criteria and is used to decide sites.
Conditions Increment % about the staged grouting. Gen-
erally, the higher the required
In Failure Load 12-24 maximum pressure to be
In Skin Friction 9-27 achieved, the higher will be the
In End Bearing 11-21 grout volume under similar condi-
N on-G route d v s. B ase -G route d P ile s (S ilom)

The above increments were deter- L o ad (to n ) N B B
-10 G G G Soft Clay
mined at fully mobilized skin friction 0 500 1000 1500 2000 2500 3000 3500 4000
A B
0
1.59-1.86% and 1.04% of pile di- -20
Stiff Clay
ameter for base grouted and non- 10 Silty Clay
base grouted piles respectively.

Settlement (mm)
20 -30
Dense Sand
Mechanism of improvement by 30
-40
base grouting involves;

Hard Clay
40
• Replacement of sediments with 50

-50
grout under the pile base. Dense Sand
• Squeezing water out of the 60

N o n -G ro u te d P ile - D ia . 1 .5 x6 0 .0 0 m
-60
sediments. B a s e -G ro u te d P ile A - D ia .1 .5 x5 5 .1 6 m (3 S ta g e s - to ta l 2 0 0 0 ltr)

B a s e -G ro u te d P ile B - D ia .1 .5 x6 0 .0 0 m (3 S ta g e s - to ta l 2 0 0 0 ltr)
• Improving the sand stratum

around the pile base.
• Prestressing of the pile.
• Increasing skin friction in sand
layers.
SEAFCO
26/10 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510. Tel. (662) 919 0090-4, Fax. (662) 919 0098
A - 18
Generally, total volume of grout at pile base Soft Clay
Base-Grouted Piles - CP
Load (ton)
ranged from 500 liters to 4000 liters at single or -10 Clayey Sand
0 500 1000 1500 2000
multiple stage grouting. Grouting pressure ranges Stiff Clay 0
from 25 bars to 60 bars depending on the depth of

-20
Settlement (mm)
Silty Clay
10
pile base level and relative density of sand layer in -30
the project area. Sand

20
-40 Silty Clay
Silty Clay
30
-50
Dense Sand
40
-60
Base-Grouted Pile - Dia. 1.2x55.87m (Grouted Soil)
All rights reserved. No part of this publication may be reproduced in any Silty Clay
Base-Grouted Pile - Dia. 1.2x55.90m (Virgin Soil)
form or by any means, without the prior written permission of SEAFCO CO.,
LTD.
load settlement curves of base-grouted piles, before
and after all adjacent piles have been base-grouted.
Non-Grouted vs. Base-Grouted (V-Rangsit) -CNW

N on-G route d v s. B ase -G route d (Prathu-N am)
Soft Clay
Lo ad (ton)
Soft Clay Load (ton)
-10 -10 0 500 1000 1500 2000 2500 3000 3500
0 500 1000 1500 2000 2500 3000
Stiff Clay Stiff Clay 0
0
-20 10 -20
Age: 62 days
Settlement (mm)
20
Settlement (mm)
-30 Age: 61 days -30 10

30
Age: 82 days
Hard Clay 40 Hard Clay
-40
-40
50
20
Age: 57 days
-50 60
-50
70
Sand
-60 80 Sand
30
-60
No n -Gro u ted Pile - Dia. 1.5x60.2m Non-Grouted Pile - Dia. 1.5x60.2m
Sandy Clay
Bas e-Gro u ted Pile - Dia. 1.5x59.72m (2 Stag es - to tal 1500 ltr) Sandy Clay Base-Grouted Pile - Dia. 1.5x60.3m (2 Stages - total 1500 ltr)
Non-Groute d vs. Ba se -Groute d Pile s (Ra m a III) Non-Grouted vs. Base-Grouted (Silom) - JTC
Soft Clay
Soft Clay Load (ton)
Load (ton) -10
-10 N B 0 500 1000 1500 2000 2500 3000 3500
G G 0 500 1000 1500
0
0 Stiff Clay
Stiff Clay
-20 Age: 115 days
-20
10
Settlement (mm)
Dense Sand
10
Settlement (mm)
-30 20
-30
30
Hard Clay 20 -40
-40
Desne Sand 40
Age: 60 days
30 -50
-50 50
Dense Sand
Silty Clay
60
-60 Silty Clay 40 -60
Dense Sand
Non-Grouted Pile - Dia. 1.0x53.0m Non-Grouted Pile - Dia. 1.5x63.72m
Bas e-Grouted Pile - Dia. 1.0x55.0m (500 ltr 1 Stage)
Base-Grouted Pile - Dia. 1.5x63.72m (2 Stages - total 1000 ltr)
References:
• Teparaksa, W., Thasnanipan, N. and Anwar M. A. (1999), “Base Grouting of Wet Process Bored Piles in Bangkok
Subsoils”, Eleventh Asian Regional Conference, International Society for Soil Mechanics and Geotechnical Engineer-
ing, Seoul, Korea, August 16-20, 1999. Pp. 269-272.
• Lui, J. and Gao W. S. (1999), “The Behavior of Post-grouting Pile Groups under Vertical Load”, Eleventh Asian Re-
gional Conference, International Society for Soil Mechanics and Geotechnical Engineering, Seoul, Korea, August 16-
20, 1999. Pp. 209-212.
• Thasnanipan, N. (1997), “Current Development of Bored Piling Techniques in Bangkok (in Thai)”, Technical Seminar
on Foundation 1997 organized by the Engineering Institute of Thailand, Bangkok, Thailand, February 4, 1997. Pp. 1-
26.
• Teparaksa, W. (1992), “Behavior of Base-Grouted Bored Piles in Bangkok Subsoils”, Piling: European Practice and
Worldwide Trends. Conference organized by the Institution of Civil Engineers, London, April 7-9, 1992, Thomas Tel-
SEAFCO
ford. Pp. 296-301.
26/10 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510. Tel. (662) 919 0090-4, Fax. (662) 919 0098
A - 19


Object
Contiguous bored pile wall and bored piles for con-
struction of basement and foundation of the Rama
Thibodi Hospital.
Project Description
Bored pile wall with temporary braced excavation
method for construction of 3 level basements for car
parking.
Type of Work
Foundation Piling, Contiguous bored pile wall.
Owner
Rama VI Hospital. Layout of contiguous and foundation bored piles.
Designer
Arun Chaiseri Consulting Engineers Co., Ltd.
Main Contractor
Ch. Karnchang Public Company Limited.
Project Schedule
2005-2006
Construction Method
 Excavation with contiguous pile wall using bot-
tom up construction method to –11.5m.
Construction Details Excavation with braced pile wall for construction of footings and basements.
Foundation bored pile: 7 (0.8mx55m)
121 (1.0mx55m)
163 (1.2mx55m)
32 (1.5mx55m)
Contiguous Pile Wall: 364 (1.0mx20.0m)
Subsoil Conditions
Soft clay: 0.0-11.0m
Medium clay: 11.0-14.0m
Stiff clay: 14.0-28.0m
Dense sand: 28.0-30.0m
Hard clay: 30.0-47.0m
Very dense sand: 47.0-55.0m
Exposed contiguous bored pile wall.
No. 144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: seafco@seafco.co.th,
Homepage: www.seafco.co.th.
A - 20
MahaNakhon
Deep Foundation Works
Project Description
Barrettes, bored piles and diaphragm walls for a 75
storey high-end Hotel and Residence Tower with 2
level basements, and a retail complex with 4 level
basements, located on Narathiwas Road.
Type of Work
Foundation piling and diaphragm walling.
Owner
Pace Project One Co., Ltd., Pace Project Two Co.,
Ltd. and Pace Project Three Co., Ltd.
Project Schedule
2011-2012
Layout of the project.

Construction Details
Foundation bored pile: φ0.80mx57-65m = 59 +11
φ1.00mx57-65m = 13+14
φ1.20mx65.0m = 4
0.80mx3.0mx65.0m = 1
1.20mx3.0mx65.0m = 114
Diaphragm Walls: 9,264sqm (0.6m to 0.8m
thick x 16.0m to 22.0m)
Subsoil Conditions
Soft clay: 2.0-8.0m
Very dense sand: 43.0-49.1m
Hard Clay: 49.1-55.0m
General View of the Site.
Very Dense Sand 55.0-67.0m
Rebar Cage Installation for Diaphragm Wall.

Diaphragm Wall Construction in Progress.
A - 21


Object
Construction of basement structures of a 42 storey
office tower on 9 storey retail podium.
Project Description
Construction of bored pile foundations, diaphragm
walls and deep excavation works for 2-3 level base-
ments construction.
Type of Work
Foundation piling, diaphragm walling and substruc-
ture works.
Owner
Land and House Property Company Limited.
Designer
Palmer &Turner (Thailand) Limited.
Project Schedule
2008-2009
Construction Method
 Bored piling and diaphragm walling
 Deep excavation to –13.0m with diaphragm
walls using bottom up construction method.
Construction Details Excavation with braced diaphragm wall for construction of basements.
Foundation bored pile: 0.80mx57m = 22
1.20mx57m = 65
1.50mx57m = 183
1.80mx57m = 25
Diaphragm Walls; 8,172 sq.m. (0.80m thick
x 18m deep)
Voume of Excavation: 95,000 cu.m.
Temporary Bracing: 1,987 tons
Temporary Platform: 3,807 sq.m.
Concrete Volume: 24,295 cu.m. (substructure)
Steel Reinforcement: 4,350 tons (substructure)
Basement Floor: 30,909 sq.m.
Subsoil Conditions
Soft clay: 2.5-9.0m
Stiff to very stiff clay: 15.3-40.0m
Dense Sand: 42.8-60.0m
Completed substructure.
A - 22
ASOKE COMPLEX - INTERCHANGE TOWER
Substructure Works
Object
Construction of foundation mats, footings, deep base-
ments for car parking and wastewater treatment
tanks.
Project Description
Diaphragm wall with 2-3 levels temporary bracing for
deep excavation up to –15.90m to construct 3.5 level
basements for car parking.
Type of Work
Deep excavation with temporary bracing, Civil Con-
struction. Layout of temporary bracing.
Owner
B&G Park Company Limited.
Designer
Palmer and Turner (Thailand) Ltd.
Project Schedule
March 2005– November 2006
Construction Method
• Braced excavation to –15.90m with diaphragm
wall using bottom up construction method.
Temporary bracing: 1,478 tons
Earth excavation: 54,000cu.m Rebar fixing for foundation mat in progress.
Basement floor area: 13,000sq.m.
Subsoil Conditions
Very dense sand: 40.0m+
Concrete casting of 3.0m thick foundation mat in progress.
144, Prayasuren Road, Bangchan, Khlong Sam Wah, Bagkok 10510, Thailand. Tel. (662) 919-0090 to 97, Fax. (662) 919-0098, 518-3088. Email: seafco@seafco.co.th,
A - 23
UBC III AND EM2
Deep Foundation Works
Project Description
Bored piles and diaphragm walls for a 32 storey
office tower and shopping complex with 2 to 4 level
basements, located on Sukhumvit Road.
Type of Work
Foundation piling and diaphragm walling.
Owner
Bhiraj Buri Co., Ltd (UBC III) &
Sukhumvit City Mall Co., Ltd. (EM2)
Project Schedule
2011
Foundation bored pile: φ0.60mx18.5m = 126
Bored Pile Wall φ0.60mx22-23.0m = 433
Bored Pile Wall φ0.80mx18.5m = 393
Foundation bored pile: φ0.80mx60.0m = 158
φ1.00mx60.0m = 30
φ1.20mx60.0m = 27
φ1.50mx60.0m = 308
φ1.80mx60.0m = 140
0.80mx3.00mx60.0m = 4
1.00mx3.00mx60.0m = 22
Diaphragm Walls: 7,057.5 sqm (0.8-1.0m thick
x18.0-22.0m)
Subsoil Conditions
Dense sand: 32.5-34.0m General View of the Site.
Dense Sand: 53.9-58.0m
Diaphragm Wall Construction in Progress. Drilling for Large Diameter Bored Piles.
A - 24


Object
Construction of a 2 level underground car park under
a football field.
Project Description
Construction of bored pile foundations, diaphragm
walls and deep excavation works for 2 level base-
ments and football field on top.
Owner
Srinakrin Wiroj University
Architects & Engineers

Krunthep Tanakom Co., Ltd. (Designer) Layout of the project (Courtesy of the Designer)
Plan Consultants Co., Ltd.
Kumjornkit Construction Co., Ltd. (MC)
Project Schedule
2008-2010
Construction Method
 Bored piling and diaphragm walling
 Deep excavation to –7.3m with diaphragm
walls using top-down construction method
Foundation bored pile: 0.60mx18.5m = 508
Diaphragm walls; 7,664 sq.m. (0.60m thick
x 16m deep)
Excavation volume: 102,000 cu.m
Basement floor: 36,000 sqm
Excavation with diaphragm wall using top-down construction method for base-
Concrete volume: 39,624 cu.m ment construction.
Reinforcement: 4,218 tons
Subsoil Conditions
Sand: 41.5-46.0m
Substructure works in progress.
A - 25


Object
Deep foundation work for wastewater treatment plant.
Project Description
Bored piles and diaphragm walls for construction of
deep basements for a wastewater treatment plant in
Bangsu, Jatujak area.
Type of Work
Foundation piling, diaphragm walling and soil im-
provement.
Owner
Layout of pile foundation and diaphragm walls.
The Bangkok Metropolitan Administration.
Designer & Engineers

The Bangkok Metropolitan Administration.
CH. Karnchang Public Company Limited (MC)
Project Schedule
2008-2009
Foundation bored pile: 0.80mx57.0m= 971
1.00mx57.0m = 7
1.00mx57.0m = 380
1.50mx57.0m = 46
Diaphragm Walls: 2,460 sq.m (1.5m thick x

24.7m deep ) Bird’s eye view of the project site.
10,225 sq.m (1.2m thick x
24.7m deep )
Soil mixing: 1,734 cu.m
Subsoil Conditions
Stiff to very stiff clay: 13.5-24.0m

Diaphragm wall and bored pile construction in progress.
A - 26



Object
Underground car park with disable access, natural
light and ventilation for the Central Hospital of Bang-
kok Metropolitan Authority. Large open space and
garden on the roof top.
Project Description
Diaphragm wall with top down excavation method for
construction of 4 level basements for car parking and
roof top garden.
Type of Work
Foundation Piling, Diaphragm Walling, Deep Excava-
tion, Civil Construction.
Diaphragm wall and bored pile construction.
Owner & Designer

The Bangkok Metropolitan Administration
Project Schedule
2005-2007
Construction Method
 Excavation with diaphragm Wall using top-
down construction method to –13.0m.
Foundation bored pile: 69 (1.8mx22m)
Diaphragm Wall: 4,900sq.m. (0.8m thick,
20.0m deep)
Basement floor area: 11,200sq.m.
Landscaping: 2,800sq.m. Excavation with permanent RC beams lateral support in progress.
Subsoil Conditions
Very dense sand: 42.0-50.0m
Bird’s eye view of the completed underground car park.

Cross section of underground car park.
A - 27



Object
Deep foundation works for a 47 storey luxury condo-
minium.
Project Description
Bored piles, contiguous pile wall, diaphragm walls
and soil improvement for construction of deep base-
ment.
Owner
Sathorn Park Co., Ltd. & Grace Ivory Co., Ltd.
Designer Layout of pile foundation and diaphragm walls.

Meinhardt (Thailand) Ltd. (Structural Engineers)
Project Schedule
2007-2008
Foundation bored piles: 1.00mx55.0m = 192
1.20mx55.0m = 238
Contiguous bored pile: 1.00mx22.0m = 83
Diaphragm Walls: 6,352 sq.m. (1.0m thick x
20-22.0m deep)
1,377 sq.m. (1.2m thick x
20-22.0m deep)
Soil mixing: 275 cu.m.
Subsoil Conditions
Diaphragm wall construction in progress
Exposed diaphragm wall and contiguous piles. Bored piling in progress.
A - 28
SRINAKRIN UNDERPASS
Civil Works
Project Description
Construction of an underpass for the traffics along
Srinakrin Road.
Type of Work
Foundation piling, diaphragm walling and civil works.
Owner
The Bangkok Metropolitan Administration (BMA).
Project Schedule
2009-2011
Foundation bored pile: φ1.00mx29.0m = 73 Construction of the Underpass in Progress.
0.80mx3.0x25.0m = 3
Diaphragm Walls: 13,672sqm

(0.8m thick x 25.0m)
Structural Concrete: 42,230cu.m

Reinforcement Steel: 5,266tons
Subsoil Conditions
Construction of Diaphragm Wall under the Existing Highway.
Layout of the Project.
A - 29
GENERAL LEGEND :
MRTA ISP (Underground Structures-North)
MRTA ISP (Underground Structures-South)
MRTA Interchange Station BANG SU STATION
MRTA Station
Thiam Ruam Mit (Thailand Cul- SRT System (Hopewell)
BMA System (BTS)
tural Center) Station (S12) is one

of 18 underground stations of the
R
VE
Metropolitan Rapid Transit Au-
RI
A
AY
R
thority (MRTA) Initial System Pro-
PH
AO
THIAM RUAM MIT STATION
CH
ject. The station is located on
Ratchadapisek Road. The initial
system project or Blue Line is the
first link in the network and 20km.
HUA LAMPHONG STATION
Long. The perimeter walls of the

station box are made of cast in-
situ concrete diaphragm walls
1.0m and 1.2m thick. The south
wall of the station is a 1.2m thick
diaphragm wall and has openings Location of Thiam Ruam Mit Station.
for three tunnels linking the south WORK UNDERTAKEN
and north underground stations BORED PILES: Dia. 0.6mx19.5-23.5m 12 nos.
to the Depot. A Tunnel Boring Dia. 1.0mx36.0m 95 nos.
Machine (TBM) driving shaft for Dia. 1.2mx36.0m 9 nos.
twin tunnels is also included in Dia. 1.5mx36.0m 5 nos.
the station box on the northern BARRETTES: 1.2mX3.0mX44.5-55.0m 63 nos.
end. STEEL STANCHION: 859 tons
DIAPHRAGM WALL: 24,338.9 sq.m. (1.0m Thick)
Diaphragm walls are embedded 1,612.8 sq.m. (1.2m Thick)
32-35m deep in sand layer for
24.5m deep excavation work. 63 20.25 6.03 6.03 20.25
barrettes (1.2mx3.0m) embedded

Foot Path
+101.000
44.5-55.0m deep in conjunction TOP SLAB +98.900
with pre-placed steel stanchions

at the top are used to facilitate Soft Clay
CONCOURSE SLAB
top down construction and to

support the station box. ORANGE LINE PLATFORM
ORANGE LINE SLAB
Glass Fibre Polymer Reinforce- Stiff Clay DIAPHRAGM WALL
ment was used in diaphragm wall BASE SLAB +80.450

panels at the tunnel opening lo- +78.450
cations to allow direct break Silty Sand
BARRETTE
through for TBM.
C
L TRACK C
L TRACK
Cross section of station box (Schematic).
TYPE OF WORK: Foundation Piles, Diaphragm Wall and Barrettes

OWNER: The Metropolitan Rapid Transit Authority
MAIN CONTRACTOR: Nishimatsu Construction Co., Ltd. (ION Joint Venture)
D04/99
DESIGNER: Ove Arup and Partners International Limited

PERIOD: 1998-1999
SEAFCO
144 Prayasuren Road, Bangchan, Khlong Sam Wah, Bangkok 10510. Tel. (662) 919 0090-7, Fax. (662) 919 0098
A - 30
2-SC
THAI
R.S.TOWER
N
INSURANCE
2-SC 2-SC
OPEN
2-SC
2-SC
4-CF
Mr. P FRESH MILK
ENTRANCE
AREA 4-CF ENTRANCE
12.1 12.2
CL
TO RAMA IX
CL CL
TBM
DRIVE 27.000 TO LATPHRAO
ROAD CL
SHAFT ROAD
LINE
CL
RATCHADAPHISEK RD. CL
ORANGE
ENTRANCE
12.4 ENTRANCE
VEN 12.3
T. VEN
SHAFT
T.
SHAFT
1
2
359.500
Layout of station, diaphragm wall and barrettes.

U n d ra in e d S h e a r S tre n g th SPT
S u (t/m 2 ) N (B lo w s /ft)
0 5 10 15 0 20 40 60 80
0 0
F IL L
B H -3 B H -2
B H -3
B H -3 7
SO FT C LAY -1 0 -1 0
S T IF F C L A Y -2 0 -2 0
M E D IU M -3 0 -3 0
Depth (m)
Depth (m)
DENSE SAND
-4 0 -4 0
HARD CLAY
-5 0 -5 0
-6 0 -6 0
-7 0 -7 0
Subsoil conditions at the site.

General view of the site (diaphragm wall
and barrette work).
Panel excavation with grab.
A large volume of bentonite slurry sup- Lowering a reinforcement cage with glass
ply, storage and treatment facilities for Placing a strip of water-stop fibre polymer reinforcement soft eye for
an excavation up to 300 cu.m. in the stop-end plate. tunnel opening.
Concrete pouring using double tremie A wall panel being cast with a wa- Exposed diaphragm wall in-
sets for T-Shaped panel. ter-stop between panel joints. side (TBM) driving shaft.
A - 31
A scene of barrette construction activity. Barrette construction in progress – (SEAFCO Guide
Frame with extended section to control verticality of
stanchion in the foreground)
Left – Lowering rebar cage at-

tached with sonic logging access
tubes.
Centre – Setting up SEAFCO

guide frame for steel stanchion
to be installed with great accu-
racy.
Right – Placing a steel stanchion

in position.
Concreting a barrette with pre-placed steel stan-

chion, using double tremie sets .
Steel stanchions on site, Verticality of stanchion can also be
ready to be installed. checked against tall buildings nearby.
A - 32
Checking trench verticality and dimen-
sion with Koden drilling monitoring
Attaching a vibrating wire strain gauge
equipment..
Drilling monitoring record in the reinforcement cage of diaphragm
showing good verticality of wall.
trench and dimensions.
All rights reserved. No part of this publica-

tion may be reproduced in any form or by
any means, without the prior written per-
mission of SEAFCO PUBLIC CO., LTD.
Monitoring and Checking Bentonite A jack-out pressure cell to be installed

Slurry Quality. in the reinforcement cage of diaphragm
wall.
Checking concrete quality of barrette

with sonic logging method. Testing on barrette capacity with fully
instrumented load test.
Sonic logging test record show-

ing good quality of concrete .
References
• Teparaksa, W., Thasnanipan, N., Maung, A. W. and Wei, S. H. (1998), “Prediction and Performances of Short Em-
bedded Cast In-situ Diaphragm Wall for Deep Excavation in Bangkok Subsoil”, Fourth International Conference on
Case Histories in Geotechnical Engineering, St. Louis, Missouri, March 9-12, 1998. Pp. 686-692.
• Thasnanipan, N., Anwar, M., A., Maung, A. W. and Tanseng, P. (1999), “Performance Comparison of Bored and
SEAFCO Excavated Piles in the Layered Soils of Bangkok”, Symposium on Innovative Solutions in Structural and Geotechnical
Engineering in Honor of Professor Seng-Lip Lee, AIT, Bangkok, May 14-15, 1999. Pp. 345-353.
A - 33
GENERAL
Rangsan Silom Precious (present name is Royal Charoen Krung)

building is located at the corner of Silom and Charoen Krung Roads,
Bang Rak, Bangkok. It is a 63-storey tower with six basement levels
constructed by using top-down method. Foundation bored piles of
1.5m in diameter founded in sand layer at depths up to 60m, support
the building. The piles were base-grouted. A 1.02m thick cast in-situ
diaphragm wall (toe depth 36.0m below ground level) was used for
basement excavation and construction. The maximum excavation
depth was about 20.0m. 95 built-up steel stanchions pre-founded in
bored piles were used for supporting the basement slabs and parts of
superstructure during simultaneous construction of basements and
the super structure. Three basement slabs, B1, B3 and B5 with
openings were used as lateral supports to the diaphragm wall to allow
excavation reach to the required depths.
WORK UNDERTAKEN
BORED PILES: Dia. 1.0mx36.0m 5 nos.

Dia. 1.5mx60.0m 529 nos.
BASE GROUTING: 534 piles
STEEL STANCHIONS: Built-up Sections 95 nos.
Section of building. DIAPHRAGM WALLS: 13,748 sq.m. (1.02m Thick)
INSTRUMENTATION: 5 Inclinometer Tubes in the wall and 1
Tube in the soil behind the wall.
I-1 I-2
I-6
I-3
LEGEND
BORED PILE
BORED PILE WITH STANCHION
I-5 DIAPHRAGM WALL PANEL
INCLINOMETER TUBE
0 15 25 50m
I-4
View of basement construction with top-down method. Diaphragm wall and pile layout.
TYPE OF WORK: Foundation Piles and Diaphragm Walls

OWNER: Rangsan Silom Precious Co., Ltd.
SUPERSTRUCTURE
CONTRACTOR: Samsung (Thailand) Development Co., Ltd.
DESIGNER: Siam Engineering Consultants Co., Ltd.
PERIOD: 1991-1992 SEAFCO
A - 34
Undrained Shear Strength SPT
Su (t/m2) N (Blows/ft)
0 10 20 30 40 0 20 40 60 80 100
Top Fill 0 0
Soft to Medium Clay
-10 -10
Stiff to Hard Clay

-20 -20
Depth (m)
Depth (m)
-30 -30
Dense Sand
-40 -40
Hard Silty Clay
-50 -50
BH-1
BH-1
Very Dense Sand BH-2
BH-2
-60 -60
Subsoil conditions.
All rights reserved. No part of this publication may

be reproduced in any form or by any means, without
the prior written permission of SEAFCO PUBLIC
CO., LTD.
Diaphragm wall panel excavation.

Lowering a rebar cage section
N on-G route d v s. B ase -G route d P ile s with box-outs for basement slab
L o ad (to n ) connections. 4th FL. +15.30
0 500 1000 1500 2000 2500 3000 3500 4000 3rd FL. +10.80
0
B1 -2.70
10
B3 -7.90
Settlement (mm)
B5 -13.10
20 B6 -15.70
MAT FOUNDATION
30
D-WALL D-WALL
-36.0m. -36.0m.
40
50
60
N o n -Gro u te d P ile - D ia . 1 .5 x6 0 .0 0 m
B a s e -Gro u te d P ile A - D ia .1 .5 x5 5 .1 6 m (3 S ta g e s - to ta l 2 0 0 0 ltr)
B a s e -Gro u te d P ile B - D ia .1 .5 x6 0 .0 0 m (3 S ta g e s - to ta l 2 0 0 0 ltr)
Floors and basements constructed

Pile load test results.
Installing a 24.0m-long built-up Preparation for connections be-

Excavation below the basement slab.
steel stanchion in the borehole. tween basement floor and dia-
phragm wall.
References:
• Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), “Analysis of Lateral Wall Movement for Deep Braced Exca-
vation in Bangkok Subsoils”, Civil and Environmental Engineering Conference – New Frontiers & Challenges, AIT,
Bangkok, November 8-12, 1999.
• Lee, S. L. and Yong, K. Y. (1992), “Base Grouting of Bored Piles”, 2nd International Conference on Deep Foundation
Practice incorporating PILETALK 1992, Singapore, 4-6 November 1992. Pp. 147-153.
SEAFCO
A - 35
GENERAL DON MUANG
The Din Daeng Underpass is the

first of its kind in Bangkok Me- N
tropolis to ease the traffic conges-
M EN T
tion. It is an alternative to flyovers THE V
IC TORY
M ONU
FROM
and overpasses. The underpass
route extends for 415m with two
lanes with one way traffic from the
Victory monument. It runs along TO
AS
OK
GE
the middle of the Asoke-Din 5 15 35 E-
DIN
AY TA
DA
Daeng Road, and passes under
SW T S
EN
G
ES FIRS
the elevated First Stage Express-
EX THE
way as a cut and cover tunnel.
PR
ED
AT
82m long approach sections at
EV
EL
both ends of the underpass were
constructed with 0.45m thick rein- Layout of Din Daeng Underpass.
forced concrete walls. The 252m
long middle section of the under-
pass, which includes portals and WORK UNDERTAKEN
cut and cover tunnel, is con-
structed with 1.0m thick diaphragm DIAPHRAGM WALL: Thickness 1.0m
walls having toe level 21.0m Toe Depth –21.0m
deep. Cantilever walls and rein- 10,545sq.m.
forced concrete strut-supported INSTRUMENTATION: 8 Inclinometer Tubes in the wall
walls are used in the portal sec- 2 Inclinometer Tubes in the ground
tions. Excavation depths for base behind the wall, 8 Sets of VWSG. and
slab construction varied from 4.9m 2 Vibrating Wire Piezometers
to 6.5m below the pavement level.
0.00m
Back Fill
Soft Clay
-6.50m
Drain Pipe
-21.00m
Stiff Clay Diaphragm Wall
Schematic section of cut and cover tunnel. Diaphragm wall panel excavation in progress.
TYPE OF WORK: Diaphragm Wall

OWNER: Bangkok Metropolitan Administration.
D02/99
MAIN CONTRACTOR: Unique Engineering and Construction Co., Ltd.

DESIGNER: Epsilon Co., Ltd.
PERIOD: 1996
SEAFCO
A - 36
U n d ra in e d S h e a r S tre n g th SPT
2
S u (t/m ) N (B lo w s /ft)
0 5 10 15 20 25 30 0 10 20 30 40 50 60
0 0
F IL L
B H -1
B H -1
B H -2
B H -2
B H -3
-5 -5 B H -3
B H -4
B H -4
B H -5
B H -5
SOFT CLAY
-1 0 -1 0
Depth (m)
Depth (m)
-1 5 -1 5
-2 0 -2 0
S T IF F C L A Y
-2 5 -2 5
M E D IU M
DENSE SAND -3 0 -3 0
Subsoil conditions at the Site.

Diaphragm wall panel excavation in the middle of
traffics.
All rights reserved. No part of this publica-

tion may be reproduced in any form or by
any means, without the prior written per-
mission of SEAFCO PUBLIC CO., LTD.
Diaphragm wall panel excavation with

modified grab crane under the existing
expressway. Lowering a rebar cage section.
General view of Din Daeng Underpass. Portal section. Tunnel section.
References:
• Thasnanipan, N., Maung, A. W. and Singtogaw, K. (1996). “ Experience in Construction of an Underpass Tunnel using
Diaphragm Wall – Bangkok (in Thai)”, Technical Seminar on Construction of Basements 1996 organized by Engineering
Institute of Thailand, Bangkok, 18 September 1996. Pp. 55-70.
D02/99
• Teparaksa, W., Thasnanipan, N. and Tanseng, P. (1999), “Analysis of Lateral Wall Movement for Deep Braced Excava-
tion in Bangkok Subsoils”, Civil and Environmental Engineering Conference – New Frontiers & Challenges, AIT, Bang-
kok, November 8-12, 1999.
SEAFCO
A - 37
GENERAL +14.80
+11.30
0 6 10 20M
The Thammasat Administration +7.80
Vane Shear
Building project site is located on +4.30 SPT-N (Blows/ft)
Strength (kPa)
+0.80 +0.70 0 20 40 60 80 0 50 100
Tha Prachan Campus, Bangkok -3.20
0
Phase 1
nearby the Chao Phraya River.

Phase 2
-5.80 Soft to
Medium 5
The site is surrounded by a his-

-8.40 CLAY
10
torical building and other existing 15

structures. The construction of 0.80m Thick D-Wall Bored Pile
Stiff to Very
this building commenced about

Stiff CLAY 20
one year after completion of li- -28.00 (D-Wall Tip)

Medium to
25
brary building with three base- Very Dense

SAND
30
ments being separated by an ex- -48.00 (Pile Tip) 35
isting building. Bored piles with Basement section and soil profile.
pile tip in dense sand layer are
used for supporting the building. WORK UNDERTAKEN
A diaphragm wall of 0.8m in thick-
ness was designed for excavation BORED PILES: Dia. 0.8mx48.0m 17 nos.
9.7m deep with two levels of tem- Dia. 1.0mx48.0m 33 nos.
porary bracing. The diaphragm Dia. 1.2mx48.0m 5 nos.
wall toe was embedded down to Dia. 1.5mx48.0m 23 nos.
28.0m to achieve the overall sta- DIAPHRAGM WALL: 7,742sq.m. (0.8m Thick)
bility of the excavation on the river EARTH WORK: 59,592cu.m.
bank. Various types of instrumen- TEMPORARY BRACING: 586.50 tons
tation were installed in the wall INSTRUMENTATION: 6 Inclinometer Tubes, 10 Tiltmeters,
and existing buildings to observe 5 Vertical Beam Sensors, VWSGs in
the ground movements and re- one Panel, 4 Earth Pressure Gauges,
sponse of the buildings during 20 Surface and Deep Settlement
excavation. Plates.
Excavation and basement construction. Diaphragm wall panel excavation.
TYPE OF WORK: Foundation Piles, Diaphragm Wall and Excavation

OWNER: The Thammasat University
MAIN CONTRACTOR: Ch. Karnchang Public Co., Ltd.
DESIGNER: SJD-3D Co., Ltd.
PERIOD: 1999
SEAFCO
A - 38
Chao Phraya River
S4 D10 S10
D4 S3
D9 S9 I-2
I-1 D3
D1 T10
T1
I-6 I-3 EL5
Building
5 Storey
S1 T9
5 Sto
EL1 S8
S2 D8
rey B
T2
D2
uildin
T8
g
D7 EL4
T3 EP-1
I-5 S6 D5 I-4 D6
S5 S7
T7 EL3
T4
T5 EL2 T6 5 Storey Building
5 Storey Building
Legend 0 6 10 20m
Inclinometer in D-Wall Deep settlement plate (5.0m)
Tilt plate
Vertical beam sensor Earth pressure gauge
Surface settlement plate (1.0m) Pre-load location
Temporary bracing and working platform.

Layout of instrumentation and the project site.
All rights reserved. Bored Pile φ0.8x5.98m

No part of this
Load (Tons)
publication may be
reproduced in any 0 200 400 600 800
form or by any 0.0
means, without the
prior written per-
5.0
Settlement (mm)
mission of
SEAFCO PUBLIC
CO., LTD.
10.0
15.0
20.0
Excavation completed to the final depth. Pile load test results.
Earth pressure gauges for checking

Pre-loading the struts. Preparation for base slab casting.
the strut force.
References:
• Thasnanipan, N., Maung A. W. and Tanseng, P. (1999), “Behavior and Performance of Diaphragm Walls under Unbal-
anced Lateral Loading along the Chao Phraya River”, 5th International Symposium on Field Measurements in Geome-
chanics– FMGM99, 1-3 December, 1999, Singapore. Pp 267-272.
SEAFCO • Thasnanipan, N., Maung A. W., Tanseng, P. and Wei, S. H. (1998), “Performance of a Braced Excavation in Bangkok
Clay, Diaphragm Wall Subject to Unbalanced Loading Conditions”, 13th Southeast Asian Geotechnical Conference,
16-20 November, 1998, Taipei, Taiwan, ROC. Pp 655-660.
A - 39
GENERAL I-3
Excavation -15.5m
The Sathorn Complex project is
planned for a multipurpose office Bored Pile
Barrette
and shopping complex. It is lo- Inclinometer Tube
I-4
cated at the Sathorn and Rama
IV Road intersection. The foun- I-2
dation bored piles and barrettes 89.35
are embedded 60m deep in sand
layer to support the building. A
0.8m thick cast in-situ diaphragm
wall 18.0m in depth, with three-
level temporary bracing was ini-
tially planned for construction of a
four-level basement. After com- Excavation -14.0m
I-1
120.25
pletion of diaphragm construction,
basement excavation was later
modified with four level-bracing to Foundation and diaphragm wall layout.
allow excavation down to 15.5m
below ground level for construc-
tion of five basement levels. For
the foundation in the area of the
main tower, barrettes were em- WORK UNDERTAKEN
ployed together with bored piles
to increase load bearing capacity. BORED PILES: Dia. 1.2mx60.0m 443 nos.
Bored piles and barrettes were BARRETTES: 0.8mX2.7mX60.0m 50 nos.
also incorporated as legs in the DIAPHRAGM WALL: 7,002sq.m. (0.8m Thick)
diaphragm wall to carry the load INSTRUMENTATION: 4 Inclinometer Tubes in the wall
of building.
16
+1.70
BORED PILE Φ 1.2m. TOE DEPTH 60.17m.
-1.50
-2.40
15 0 175 350 525 700 875 1050 1225 1400 1575 1750 1925
2 14
0
1 1 - 4.20
13
-5.10 3 12 -5
-6.90
11
-10
-9.00 4 10 -9.60
SETTLEMENT (mm.)
9 -15
-11.80 5 8 -12.30
-20
6 6 -13.30
-14.00
6 7 -25
-15.50
-30
-18.00 -18.00
-35
LEGEND
2 CONSTRUCTION STAGE
-40
TEMPORARY BRACING
LAOD (Tons.)
Basement excavation and construction sequence (Schematic). Pile load test result.
TYPE OF WORK: Foundation Piles, Diaphragm Wall and Barrettes

OWNER: Quality Houses Public Company Limited.
MAIN CONTRACTOR: Kay-Thai Co., Ltd.
DESIGNER: K.C.S. Associates Co., Ltd.
PERIOD: 1995-1996
SEAFCO
A - 40
Su (t/m2) SPT N
(Blows/ft)
0 5 10 15 20 25 0 10 20 30 40
0.0 0.0
Top Fill
-3.0
BH-1
-3.0
BH-2
-6.0 -6.0
Soft Clay -9.0 -9.0
Depth (m)
Depth (m)
-12.0 -12.0
-15.0 -15.0
Medium to Very
-18.0 -18.0
Stiff Clay
-21.0 -21.0
-24.0 -24.0
Dense Sand
-27.0 -27.0
Subsoil conditions at the site.
Lowering a reinforcement cage attached with incli-

nometer tube void former.
All rights reserved. No part of this publication may be reproduced in

any form or by any means, without the prior written permission of
SEAFCO PUBLIC CO., LTD. Work in progress.
Exposed diaphragm wall. Exposed barrette top sections.

References:
• Thasnanipan, N. and Singtogaw, K. (1996), “Construction of Diaphragm Wall and Bored Piling for Sathorn Complex
Project (in Thai)”, Technical Seminar and Site Visit on Construction of Deep Basements using Short Embedded Dia-
phragm Wall in Bangkok, organized by Engineering Institute of Thailand. Bangkok, 19 December 1996. Pp. 129-143.
• Teparaksa, W., Thasnanipan, N., Maung, A. W. and Wei, S. H. (1998), “Prediction and Performances of Short Embed-
ded Cast In-situ Diaphragm Wall for Deep Excavation in Bangkok Subsoil”, Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louis, Missouri, March 9-12, 1998. Pp. 686-692.
SEAFCO vation in Bangkok Subsoils”, Civil and Environmental Engineering Conference – New Frontiers & Challenges, AIT,
A - 41
GENERAL
I-6
The Rajavej Hospital is located I-7

on Phayathai Road, Rajatheewi,
Bangkok. It has five basement
levels constructed with top-down I-5
method for car parking and

rooms for utilities. Foundation
bored piles of 0.8m, 1.2m and
1.5m in diameter founded in sand I-4
layer at depths of 58m, support INDEX

BORED PILE Φ 0.80m.
the building. A 1.0m thick cast I-3 BORED PILE Φ 1.20m.
I-1
in-situ diaphragm wall (toe depth BORED PILE Φ 1.50m.
STEEL STANCHION
21.0m below ground level) was INCLINOMETER TUBE
used for basement excavation 0 6 10 20m
and construction. The maximum

I-2
excavation depth was about
14.5m. 80 steel stanchions pre-
founded in bored piles were used Diaphragm wall and foundation layout.
for supporting the basement
slabs and parts of superstructure WORK UNDERTAKEN
during simultaneous construction
of basements and the super BORED PILES: Dia. 1.0mx58.0m 5 nos.
structure. Two basement slabs, Dia. 1.2mx58.0m 114 nos.
B1 and B3 with openings were Dia. 1.5mx58.0m 56 nos.
used as lateral supports for the STEEL STANCHIONS: 414x405-498x432mm 80 nos.
diaphragm wall to allow excava- DIAPHRAGM WALL: 6,058 sq.m. (1.0m Thick)
tion to the required depths. INSTRUMENTATION: 7 Inclinometer Tubes in the wall.
0.0m
B1
-6.7m B3
D-WALL
-11.8m
-14.5m
BORED
PILE
-21.0m
Lowering a steel stanchion into

Basement section.
borehole before casting the pile.
TYPE OF WORK: Foundation Piles and Diaphragm Wall

OWNER: Rajavej Hospital Co., Ltd.
SUPERSTRUCTURE
CONTRACTOR: SAE Thailand Co., Ltd.
DESIGNER: P & CYGNA Consultants Co., Ltd.
PERIOD: 1994-1995 SEAFCO
A - 42
Undrained Shear Strength SPT
Su (t/m2) N (Blows/ft)
0 10 20 30 40 0 20 40 60 80 100
Top Fill 0 0
Soft to Medium Clay
-10 -10
Stiff to Hard Clay

-20 -20
Depth (m)
Depth (m)
-30 -30
Dense Sand
-40 -40
Hard Silty Clay
-50 -50
BH-1
BH-1
Very Dense Sand BH-2
BH-2
-60 -60
Subsoil conditions.
All rights reserved. No part of this publication may be

reproduced in any form or by any means, without the
prior written permission of SEAFCO PUBLIC CO.,
LTD.
21m-long reinforcement cage

Diaphragm wall panel excavation. with box-outs for basement slab
connections.
BORED PILE DIA. 1.2mX58.16m

Applied Load (Tons)
0 500 1000 1500 2000
0
OPENING 10
Pile Head Movement (mm)
B3 SLAB
OPENING
20
30
0 6 10 20m 40
50
Basement slab with temporary open- Pile load test result.

ings for excavation work.
Installing a 18.0m-long steel

stanchion in the borehole.
Excavation at final depth of 14.5m
Excavation below the basement slab. below ground level.
References:
• Thasnanipan, N. and Singtogaw, K. (1996), “Bored Piling and Diaphragm Wall Work for Excavation of Five-level Base-
ment of Rajavej Hospital (in Thai)”, Technical Seminar and Site Visit on Experience in Construction of Deep Base-
ments in Bangkok Soft Clay, organized by Engineering Institute of Thailand. Bangkok, 26 March 1996. Pp. 104-140.
SEAFCO
A - 43
KABAR AYE EXECUTIVE RESIDENCE
Bored Piling & Diaphragm Wall Works
–
-
A - 44
PROCEEDINGS OF THE THIRTEENTH
SOUTHEAST ASIAN GEOTECHNICAL
CONFERENCE
November 16-20 1998, Taipei, ROC
Barrettes Founded in Bangkok Subsoils, Construction and Performance

N. Thasnanipan, A. W. Maung and P. Tanseng
Seafco Co., Ltd., Bangkok, Thailand
Sponsored by;
Chinese Institute of Civil and Hydraulic Engineering

Southeast Asian Geotechnical Society
1
Thirteenth Southeast Asian Geotechnical Conference, 16-20 November, 1998, Taipei, Taiwan, ROC
Barrettes Founded in Bangkok Subsoils, Construction and

Performance
N. THASNANIPAN SEAFCO Co., Ltd., Bangkok, Thailand

A. W. MAUNG SEAFCO Co., Ltd., Bangkok, Thailand
P. TANSENG SEAFCO Co., Ltd., Bangkok, Thailand
SYNOPSIS Barrettes become not an uncommon foundation element used to support multi-storey buildings, elevated
expressways and underground station boxes in Bangkok, Thailand. The demand for barrettes is necessitated mainly by
constraints of site conditions, construction method, equipment and extensive bearing capacity requirement, etc. This paper
presents construction practice and the performance of barrette constructed in Bangkok metropolis. Trial trenching near a canal
for assessment of trench stability and soil deformation are presented. Choice of barrette and common defects found in barrette
are also discussed.
INTRODUCTION slurry used for trenching in Bangkok subsoil usually comply

with DFCP-4 of the Oil Companies Materials Association
Recently published papers indicate that barrettes become a and the Institution of Civil Engineers’ (ICE) Specification
common foundation type used in various soil conditions to for Piling and Embedded Retaining Walls. The usual
carry very high load for sky scrapers. Development of bentonite slurry properties specified are 1.10g/ml for density,
properties and infrastructures such as construction of multi- 30-50sec for viscosity (Marsh’ cone), 8-11.5 for pH and not
storied buildings, elevated expressways, subway railway more than 4% for sand content. Usually 3m to 6m long
stations, etc. demand deep foundations with large bearing sections of steel tremie pipe of 25cm in diameter are joined
capacity in Bangkok metropolis, Thailand. Bangkok is well together to form a continuous pipe with required length for
known for its subsoils constraints especially, by the presence concrete pouring under the slurry.
of thick soft clay on the top, followed by a series of
alternating stiff clay or hard clay and dense or medium dense Cast-in-situ reinforced concrete, pre-cast concrete and
sand layers. Due to not only the subsoil conditions, but also temporary steel box guide walls appropriate for site
non-availability of adequate construction area and conditions are used for trenching. In normal site conditions,
overlapping developments above and underground, barrettes guide walls with a minimum depth of 1.2m are commonly
become common to replace the usual large diameter bored used. Base grouting is also used to improve the capacity and
cast in-situ piles in circular shape. The earliest barrettes of performance of barrettes if necessary.
0.82mx2.98m-3.0m with toe depth of 50m were constructed
in Bangkok around 1985 for International Trade Centre Generally an average rate of excavation about 10-15m2/hr is
building in conjunction with diaphragm wall for basement achieved with mechanical grab in Bangkok subsoil. Usually,
construction. The first static load test on a barrette excavation rate decreases with increment in excavation
(0.82x2.7x61.8m) in Bangkok was carried out in 1992 for a depth, having non-linear relationship between excavation
55 storey building on Silom Road, Bangkok. time and depth. Figure 1 illustrates excavation time and rate
vs. excavation depth for barrettes of 1.50x3.00x55.00m. with
no site constraint and idling time.
CONSTRUCTION METHOD
Normally a cable hung grab mounted on crawler crane is

used for trenching under supporting fluid which is
commonly bentonite slurry. Concrete pouring is carried out
under bentonite slurry using tremie pipes. Properties of
573
3
24.00 TRIAL TRENCH PERFORMANCE
B15
20.00 B20
B3
B10
A trial trench (0.8mx2.7mx14m), with above mentioned soil
16.00 B19 reinforcement close to the canal and high slurry head, was
Time (Hours)
Best Fit
12.00 carried out to study the feasibility in construction of barrettes
to support the elevated expressway across the Canal. The
8.00
major concern was stability of 11.0m thick soft clay layer
4.00 (Fig. 2) which has a very low shear strength less than 15 kPa
0.00
and natural water content of up to 80%. The planned trench
0.00 10.00 20.00 30.00
Depth (m)
40.00 50.00 60.00 was about 1.5m away from the canal water and four
reinforced sheet piles were installed between the trench and
30.00
the canal prior to excavation. In addition to the reinforced
25.00
B15
B20
sheet piles, a temporary sheet pile wall was also constructed
B3 for protection of slurry spillage into the canal as the local
Excavation Rate (m2/hr)
20.00 B10
B19 authorities strictly required.
Best Fit
15.00
Slurry
Level
Guide Wall
Creep (mm) Creep (mm)
10.00 0 100 -100 0
0.0 0.0
TOP SOIL Canal

5.00
0.00
0.00 10.00 20.00 30.00 40.00 50.00 60.00 2.0 2.0
Depth (m)
Steel
Sheet Pile
1 Soft clay Stiff clay Med. sand V. stiff clay Dense sand 4.0 4.0
Depth (m)
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
SOFT
6.0 6.0
Figure 1. Excavation time, rate vs. depth for barrettes
Depth (m)
CLAY
1.5mx3.0m.
Usually 1.0m or more over casting of concrete above

8.0 8.0
designed cutoff level is practiced. Observation of 3

completed projects indicates that concrete over break of 10.0
barrettes was found to be in a range of 6% to 14% for 10.0
barrettes with about 2.0m over casting.

12.0
12.0
MEDIUM
TRENCH STABILITY CLAY After 24 Hours
After 48 Hours
In general site conditions, Bangkok soft clay layer with

14.0
14.0
undrained shear strength value of 15- 20 kPa does not pose a

particular problem for trench stability during barrette Figure 2. Soil movements on opposite faces of trial trench
excavation under bentonite slurry. However trenching in the
close vicinity of the river or canal/khlong required careful The trial trench filled with bentonite slurry was open for 48
assessment of soil properties, especially sensitivity and hour after excavation down to 14m. Lateral movements of
natural water content of the soft clay and presence of trench face along the depth were studied with graphs of the
permeable soil layers connecting the river bed with high pore trench profiles plotted by a Koden drilling monitoring
pressure. Normally, higher guide wall with adequate slurry equipment stationed at a fix position on the guide wall at
pressure and reinforcement of the surrounding soft clay by every 6 hour interval. Trench profiles plotted were
sheet piles placed perpendicular to the alignment of the compared and changes were manually measured using the
planned barrette and driven to a stiffer soil layer, overcome reference scales shown on the graph. The recording
the instability of the trench. accuracy of the equipment is in a range of +/- 0.2% (in this
case the accuracy would be within 2-3mm). 20mm to 30mm
lateral soil movement inward the trench was observed on the
monitoring records on non-reinforced soil side after 24 hours
574
4
and 48 hours respectively in the soft clay layer. For trench capacity with safety factor of 2.0 to 2.5. The toe depth of
face close to the canal , observed movements of the soil constructed barrettes ranges from 45.0m to 60.0m. Concrete
being reinforced with sheet piles are generally less than those cube strength of 30MPa to 40MPa are commonly used for
of non reinforced soil and some movements are considered barrettes. Minimum reinforcement of 0.5% of the cross
induced by the sheet pile wall installed for canal water sectional area of barrette is usually provided for either full
protection. A profile of trench showing soil movement length or partially as per load requirements.
plotted from the monitoring records is presented in Figure 2.
Barrettes can be constructed with flexible layout plan for
There is no particular problem associated with trench both vertical and lateral loads. The layout pattern of
stability in the sand layers 20 to 45m of typical Bangkok barrettes can be arranged in a continuous row or column,
subsoil below the ground level. The drilling monitoring radial, alternating long and short axis of barrette and a
records show no collapse of trench surface along the depth combination of two or more of such patterns. One type of
section (Fig. 3). equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the
mobilization cost. In addition to the large bearing capacity
requirements, on site difficulties such as limited head room,
also demands the barrettes. Piling rigs cannot be utilized
under such situations like presence of high voltage power
cables overhead, existing overpasses or structures for
elevated expressways and planned subway stations.
A summary of observation on barrettes of 19 projects
completed, with respect to selection criteria is presented in
Table 1.
Table 1. Summary of Barrette Selection
Criterion No. of Remarks

Project
High Load Capacity 4 incorporated with bored
piles
Minimise Construction 6 Alternative for bored
Equipment piles
Limited Head Room 2 Elevated Highways
for Excavation under or above existing
structures/power lines
Figure 3. Trench profile of barrette (1.5mx3.0mx55m)
Combination with 10 As diaphragm wall legs
plotted by Koden drilling monitoring equipment
Diaphragm Wall
Foundation as well as 3 provision for the future
LOAD BEARING CAPACITY OF BARRETTES portion of column requirement
Common type of barrettes constructed in Bangkok area is Since basic shape of the cross section of barrette is
rectangular shape with dimensions ranging from 0.6mx2.7m rectangular, certain dimensions of barrette can provide a
to 1.0mx2.7m. These barrettes have been mostly constructed greater perimeter for skin friction than that of the
incorporated into diaphragm walls for deep basements of conventional bored pile (in circular shape) with equivalent
high rise buildings. However, individual larger barrettes cross sectional area. Hence, for friction piles, more load
(1.2mx2.7m and 1.5mx3.0m) are also currently in use to carrying capacity per unit volume of concrete can be
support large column loads as the grabs of large dimensions achieved by a barrette than that of a circular bored pile.
are currently available in the market. A number of load test
on such large barrettes will be carried out in the near future. Figure 4 illustrates that typical barrette of 2.7m with various
Maximum bearing capacity of barrettes is usually designed thickness can give 12 to 30% higher friction area than that of
up to allowable concrete stress of 5MPa. Toe depth of circular piles with corresponding equivalent sectional area.
barrettes are limited by subsoil conditions and required load
575
5
PERFORMANCE OF BARRETTES
11
10
Available load test results, 4 nos of static load test and one
9
dynamic load test, are summarized below in Table 3.
8
Perimeter (m)
7
Table 3. Summary of Barrette Load Test
6
Perimeter of
5 Square
4 Perimeter of Project Barrette Test Total Permanent
3
Circle
Barrette 2.7x X Dimension Load Settlement Settlement
2
(m) (kN) (mm) (mm)
1 A 0.82x2.70x 14000 4.83 0
0 61.80* 28000 12.56 0
0 1 2 3 4 5 6 7 8 35000 17.63 3.47
Equivalent Gross Cross-sectional Area (m )
2 B 0.82x2.70x 12000 5.65 0.15
44.00 24000 34.15 24.72
C 0.80x2.70x 12900 4.46 0.24
Figure 4. Comparison of shape effect between barrette and 55.00 25800 11.01 2.10
circular pile 30140 12.35 2.17
D 0.80x2.70x 11000 3.07 0.21
50.00 27500 7.97 0.15
QUALITY CONTROL E 1.00x2.70x 20555 14.00 3.0
48.94**
Construction tolerance allowed for barrettes are generally Note: * with base grouting
identical to that for bored piles. Normally verticality of ** load test by dynamic load
1:100 is allowed in barrette construction in Thailand.
Drilling monitoring is employed to check the verticality of All the barrettes tested, except the barrette of project B were
the trench on random barrettes or on all barrettes. founded in dense sand layer with embeddment more than 5m
After construction, integrity testing such as sonic integrity into it. The barrette of project B was embedded into about
(low strain dynamic test) and sonic logging tests and 0.5m in dense sand layer and failed under the maximum test
concrete coring are sometimes applied. Common defects load of 24MN. Base grouting was employed for the barrette
found in barrettes constructed in Bangkok subsoils are of project A. A soil profile and load vs. settlement graph of
presented in Table 2. test barrette of project A are presented in Figures 5 and 6
respectively.
Table 2. Common Defects Found in Barrettes
SPT-N (BLOWS/30cm)
0 50 100
Defect found Possible Causes -4

SOFT
Concrete cover improper cage position, provision of

CLAY
-1 0
inadequate spacers -1 6
STIFF
Bentonite slurry improper tremie concrete pouring, -2 2
CLAY
inclusion or poor improper desanding and recycling of

DENSE
quality concrete slurry, inadequate concrete slump
-2 8
SAND
Reinforcement inadequate supervision -3 4
SILTY
cage deviation -4 0 CLAY
Uplift of improper tremie concrete pouring -4 6 DENSE

SAND
reinforcement cage (with very long tremie pipe -5 2
embedment in concrete poured) in SILTY

CLAY
partially reinforced barrettes

-5 8
-6 4 DENSE
SAND
-7 0
Figure 5. Soil profile at project A site
576
6
Load (MN)
0 10 20 30 40
0
10
Settlement (mm)
15
20
25
Cycle - 1
30
Cycle - 2
35 Cycle - 3
40
Figure 6. Load vs. settlement graph of barrette load test - Figure 8. Project E -Simulated load settlement curves (PDA)
Project A.
SPT-N (BLOWS/30cm)
DEPTH
(m) 0 50 100
Barrettes of project E were constructed on both side of water

0 FILL
-3
supply canal and one barrette was tested with high strain -6 SOFT
CLAY
-9
dynamic load test as no anchor piles were available. A -1 2
200kN steel hammer was dropped at a height of 2.6m above -1 5
-1 8
the pile cap. A graph of dynamic load test signal analysis -2 1
and simulated load settlement curve are shown in Figures 7 -2 4

-2 7
and 8 respectively. -3 0
-3 3 SILTY
CLAY
-3 6
-3 9
-4 2
-4 5
-4 8
-5 1
-5 4
SILTY SAND
-5 7
-6 0
-6 3
-6 6
Figure 9. A soil profile of Project E site
Load test results listed in Table 3 indicate that all barrettes

tested are able to carry the design load with Factor of Safety
of 2 or higher. From the available load test results, the end
bearing capacity of test barrettes cannot be estimated.
Figure 7. High strain dynamic load test signal (force and However, all barrettes are considered as friction piles, except
velocity) and PDA analysis - Project E for project A in which base grouting was used to improve the
end bearing of barrette. For test barrette of project A, no
permanent settlement was observed under test load of 28MN
which is 200% of the design load and base grouting is
considered contributed in good performance of barrette.
Barrette load tests with full instrumentation devices such as

strain gauges and magnetic extensometers are planned to be
carried out for subway stations in the near futures. The
577
7
performance of barrettes in Bangkok subsoil can then be
evaluated with a great accuracy.
CONCLUSION
Construction practice adopted in the foundation industry for

barrette installation in Bangkok subsoil has been discussed in
this paper. There is no indication of trench instability under
bentonite slurry in Bangkok subsoil for barrette construction
with proper site preparation. This is evidenced by trial
trenching and profiles plotted by drilling monitoring
equipment. To achieve a good end bearing capacity for
barrette founded in sand, base grouting is recommended.
Performance of barrettes has been discussed with available
load test results, and found to be satisfactory with load
bearing performance.
ACKNOWLEDGEMENT
The authors express their appreciation to the colleagues,

especially to Mr. Ganeshan Baskaran and Mr. Muhammad
Ashfaq Anwar for their invaluable suggestion and assistance
in preparation of this paper.
REFERENCES
Schlosser F., Simon B. and Morey J., (1989) “High

Capacity Barrette in a Region of Old Underground
Quarries”, Proceedings of the International Conference on
Piling and Deep Foundations. London, 15-18 May 1989.
Pp 119-130.
Woo S. M., Lee K.H. and Hsieh K.J. (1993) “The Causes of
Cast-In-Situ Diaphragm Wall Defects”, Eleventh Southeast
Asian Geotechnical Conference, 4-8 May, 1993,
Singapore, Pp 793-798.
Ho C. E. (1993) “Deep Barrette Foundation in Sigapore

Weathered Granite” Eleventh Southeast Asian
Geotechnical Conference, 4-8 May, Singapore, Pp 529-
534.
Ho C. E. and Lim C. H. (1998) “Barrettes Designed as

Friction Foundations: A Case History”, Fourth
International Conference on Case Histories in Geotechnical
Engineering, St. Louis, Missouri, USA. March 9-12, 236-
241.
Dibiagio E and Myrvoll F. (1983) “Full Scale Field Test of a
Slurry Trench Excavation in Soft Clay”, Proceedings of the
Conference on Instrumentation in Civil and Geotechnical
Engineering, 4-5 July 1983 Singapore. Pp 49-58.
578
8
Fourth International Conference on
Case Histories in Geotechnical

Engineering
March 8 – 12, 1998, St. Louis, Missouri, USA
Editor : Shamsher Prakash
Design, construction and behavior of bored

cast-in-situ concrete piles in Bangkok subsoils
Thasnanipan N. , Maung A. W. and Ganeshan B.
SEAFCO Co., Ltd., Bangkok, Thailand
Teparaksa W.
Chulalongkhon University, Bangkok, Thailand
University of Missouri-Rolla
Rolla, Missouri
9
Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering,
St. Louis, Missouri, March 9-12, 1998
DESIGN, CONSTRUCTION AND BEHAVIOUR OF BORED CAST IN-SITU CONCRETE PILES IN BANGKOK SUB SOIL
Narong Thasnanipan Wanchai Teparaksa Aung Win Maung Ganeshan Baskaran Paper No.1.20
Seafco Co., Ltd. Civil Engineering Dept. Seafco Co., Ltd. Seafco Co., Ltd.
Bangkok 10320 Chulalongkorn University Bangkok 10320 Bangkok 10320
Thailand Bangkok 10330 Thailand Thailand
Thailand
ABSTRACT
This paper briefs design considerations, construction methods and materials, static load test results and pile integrity test results of the
cast in-situ bored piles constructed as foundation elements for a wastewater treatment plant. The treatment plant is located in
Bangkok, capital of the Kingdom of Thailand and is to serve about 195,000 inhabitants of the city. A total of 402 piles, 296 nos. of
1500 mm diameter and 96 nos. of 1000 mm diameter, were designed to carry the whole structural load of the treatment plant.
Maximum working load on individual piles was designed to be 1000 tons and 500 tons for 1500 mm and 1000 mm diameter pile
respectively with a safety factor of 2.5. Initially the pile toe depth was designed to be 60m below the ground level, and embedded into
hard clay for 1500 mm diameter piles while the smaller diameter pile embedded into second sand layer at a depth of 53m and base
was not decided to be grouted. Pile load test results on fully instrumental pilot piles showed the failure well before reaching the
designed test load. Piles were re-designed and pile toe depth was set to be 55m, embedded into dense sand layer for both sizes and
base grouting was decided to be done for all piles. Proof load test results performed on four working piles(contract piles), two for
each size, produced well acceptable results conforming to the design and structural specifications and requirements. However,
settlements observed were higher than that observed for other projects in the vicinity. Base Grouting was done using tube - a
manchette method. A maximum volume of 500 liter of cement grout (w/c ratio ~ 0.5) or maximum grouting pressure of 40 bars was
used as limiting criteria. However, maximum pressure of 40 bars was hardly achieved. From the proof load test results ultimate load
capacity was estimated and compared using different methods suggested by previous researchers. Integrity test was decided to be
performed on each pile and results available currently are summarized.
KEYWORDS
Bored Piles, Base Grouting, Static Pile Load Test, Integrity Test, Pile Capacity
281
11
INTRODUCTION sensitive and has an-isotropic and time dependent stress strain
properties. Below is stiff to very stiff clay layer underlain by
The city of Bangkok has been emerged as one of heavily hard clay layers. These clay layers extend down to about 42m
populated urban cities in the world. Due to its rapid economic in depth. However, the subsoil condition below 36m from the
growth, the construction industry has been challenged to ground level in the project area is variable - a sand layer with
develop the infrastructure to cope with the booming economy. a thickness of about 13m is present sporadically. In some
As a result, while multi storied private owned buildings locations within the project site, a stiff to hard dark gray clay
become most attractive sky scrapers in this city of Angels, is also found at the depth between 40m and 49m. Generally a
huge portion of the Government budget is swallowed by new thick dense sand layer is located below 50m from the existing
projects involving construction of highways and railways, ground level. However, at this area alternated layers of dense
both elevated and at-grade, sub ways, large span bridges and sand and hard clay of 5m thick also was encountered. Fig. 2
water and wastewater treatment facilities. Construction shows the general bore log data observed at the site.
industry is further challenged by the city’s subsoil condition
as it is plagued with a lot of constraints such as, presence of
thick soft clay as top strata, land subsidence and high risk of
flooding. These constraints compel most structures be relied
on pile foundation usually, on large diameter bored cast in-situ
piles, embedded into first or second sand layers. Though, the
first large bored pile was installed nearly two decades ago in
this city and hundreds of thousands of piles have already been
constructed, design, construction methods, performances and
effectiveness of these piles are seldom studied and rarely
reported.
The project site is located at the Rama III road, with east
boundary along the Chao Phraya river which is the largest of
its kind in the kingdom. The piles are usually of 1500mm and
1000 mm diameters. The maximum design working load on
piles are 500 tons and 1000 tons for 1000 mm and 1500 mm Data from 3 bore holes within the project site
piles respectively. Pile toe was designed to be embedded into
second sand layer at depth 55m from the ground level. Fig. 1
shows the general layout of the piles. All the piles are cast in
Fig. 2 Soil profile at the project site
situ bored piles under wet process and base grouted.
PILE CAPACITY AND DESIGN CONSIDERATIONS
The load capacity of the piles were estimated for the soil
failure using the available soil data. Safety Factor of 2.5 was
used for both end bearing and shaft bearing capacities
considering the negative skin friction as well. The estimated
safe working load in tons, using individual three bore hole
data are 480, 508 and 583 for 1000 mm piles and 839, 880
and 993 for 1500 mm piles.
The estimated load capacity, with required safety factor, from

the soil bore log results is, in many cases, below the designed
safe working load. Hence, means of improving pile bearing
capacity was considered. The pile bearing capacity can be
improved either by extending the pile length further or by
Fig. 1 General lay-out of bored piles grouting the pile or by enlarging the pile base. Usually,
enlarging the base for long piles is hardly practiced in
SUBSOIL CONDITIONS Bangkok. Piling contractors rather prefer to extend the length
of the shaft and additional required capacity is achieved by the
The geotechinical investigation at the site indicates that a shaft friction. For the bored piles designed at this project site,
strata of 14 -15m thick soft clay layer occurs in the project if an extra bearing capacity of 250 t/m2 could be achieved by
area as common in Bangkok sub soil. This soft clay is
282
12
base grouting, the grouted piles would be capable of carrying Slurry properties
the designed safe working loads with adequate safety factor.
The limiting values for the bentonite slurry according to the
Based on the results from a project just near by the site and contract specification were 1.04 - 1.20 for density; 30-40s for
past experiences in the vicinity it was proposed that the above viscosity (Marsh’s cone); not less than 7 for pH and not more
mentioned required capacity could be achieved by base than 3% for sand content.
grouting the piles with tip founded at about 55 m from the
ground. On average 3.5 % of bentonite concentration was used.
Bentonite properties were monitored before concreting for all
The option of extending the pile length was given up as the piles. Arbitrary checking was done during drilling. From the
clay layer is found to be at about depth of 60 m and thus bentonite results, no significant changes were observed. On
possibly higher initial settlement was expected if the piles are average, the viscosity observed was 32.8s (Marsh cone); with
embedded into the clay layer. Moreover, the end bearing maximum of 40s and a minimum of 29s. The average values
capacity would be reduced and this reduced end bearing of 1.11g/cc, 8.9, 1.62% were noted for density, pH and sand
capacity might influence the shaft bearing capacity too. This content respectively. The minimum values are 1.02g/cc, 7,
conclusion was reinforced by the previous pile load test 0.1% and maximum values are 1.11g/cc, 9 and 1.55
results on a pilot pile, founded at 60m and non base grouted. respectively. Eventhough, the sand content was set to be
This non grouted and longer pile of diameter 1500 mm, within 3%, at times it was difficult to achieve this. In this
constructed by some other piling contractor, failed at about cases complete recycling of the slurry was done.
1750 tons, while the designed maximum test load was 2500
tons. The 1000mm diameter pilot pile was, though not Concreting and Reinforcement
plunged to failure, settlement observed was too excessive for
the structural requirements. Load settlement characteristics of Concrete cube strength of 350 ksc (35 Mpa) was used with
these pilot test piles have been shown in Fig. 4(a) and Fig. minimum cement content of 400 kg/m3. Concreting was done
4(b). Finally the previous piling contractor was disqualified by usual tremie method of 250mm tremie pipe. Concreting
and the new contractor was called upon. was requested to be done with in twenty four hours after final
excavation. Since a concrete batching plant was established
within the project area, concrete supply was usually regular
CONSTRUCTION METHODS, MATERIALS AND and timely supplied otherwise it would have been a major
PROBLEMS concern in Bangkok because of its ill fated traffic congestion.
Average concreting rate was about 25 - 30 cu.m per hour.
Wet process After pouring the concrete to the desired level, the temporary
casing was withdrawn.
The usual wet process using the bentonite slurry was adopted.
Temporary casing of length 15m was used to maintain the Reinforcement was provided to full length of the pile
hole stability in top soft layer. The holes were drilled using considering the possible tensile forces during construction
augers and then continued using the bucket method. The period of the structure. Maximum of 0.5% of cross sectional
slurry was introduced upon reaching the first sand layer and area was used for both sizes of the piles.
maintained all the time higher enough to support the bore. The
slurry was monitored continuously for the properties of All the main bars used for reinforcement are of high strength
density, viscosity, pH and sand content. deformed bars of 4000 ksc (400 Mpa) and round bars of 2400
ksc (240 MPa) were used for spirals. Reinforcement was
Construction of bored piles started during the rainy season and provided by fabricating in the form of cages as usual and the
therefore the site needed additional requirements to improve maximum length of a cage was set to be 12 m. A minimum lap
the working condition. As the water level was high in the length of forty times the diameter of the bars was used at the
Chaophraya river flood protection was needed. Because of the joints of the cages.
nature of the Bangkok sub soil and the project site was
previously used as urban waste dumping area, the site was Inclusions in pile top and concrete level below trimming level
very swampy during the rainy season gave a lot of
inconveniences in solving problems for drainage of water Few cases were observed where the concrete level was below
accumulated. The situation was further aggravated as the the designed cut off level and/or inclusions in pile top portion.
excavated soil was not able to be removed on time. All these Mainly, this has been contributed by the high un- predicted
reduced the productivity rate drastically. However, all the slumping down of concrete upon extracting the temporary
piles were cast within the time frame set according to the casing. As Fleming et al. (1985) pointed out concrete would
contract. normally be retained by cohesive soils with shear strength in
excess of 15 kN/m2 and this limit is rarely achieved by the top
soft clay of Bangkok sub soil.
283
13
Final sound concrete level has been a talking point between the grout volume was the limiting factor. Usually the grouting
the main contractor and piling contractor for both cases where pressure was within the range of 20 - 30 bars.
the final concrete level is above and below the designed trim
level. Usually for the first case the main contractor suffered to Injection rate of grout was 10 - 15 liters/ min. Experiments on
trim the long pile heads but in the latter cases the piling grout samples collected from three of the bored piles
contractor was called upon for remedial works. However, it (1500mm dia.) in this site was performed by Anwar (1997).
should not be neglected that the possibility of failure to over Comparison of base grouted and soft base (non base grouted)
flush enough concrete at the pile head to ensure that all piles were studied and summarized in Table 1. It was
concrete present below the trim level is of full strength and recommended that low injection rate about 1-2 liters/min
thus having the inclusions at the pile top. would improve the effectiveness of base grouting. However,
practicality and sustainability of grout pump, especially for
Verticality and Position Errors large volume of grout, can hardly justify this requirement.
Though the direct comparison of effect of base grouting is not
Verticality tolerance was set to be 1 in 100 and no piles have possible as the compared piles are founded in different soil
been found to be out of this tolerance. However, four piles strata other parameters in Table 1 are worth enough to be
(about 1% of the total) have been found to be out of position noted.
deviating the specified tolerance of 7.5cm plus the allowance
for verticality. However, many of them were found to be
under the mat on which the columns were to be erected. Table 1 Load distribution and settlement behavior of base
Hence, dealing with these erroneous locations was not grouted and soft base piles.
dramatic and modification of reinforcement for the mat have
been done according to the redistributed load. Base Soft
Grouted Base
GROUTING Diameter(mm) 1500 1500
Base Depth (m) 55 60
The effectiveness of base grouted piles in Bangkok sub soil Design Load Qd (ton) 1000 1000
was studied by many previous researches. Teparaksa (1994) Failure Load (ton) 2725 * 1750
suggests base grouting mainly aims to increase soil stiffness Load at 20mm total settlement (QL) 2290 1520
beneath the pile base which was affected by the boring
% increase in QL 50.66
process. Also he noted that for these piles the displacement at
Portion of QL carried by shaft (QS) 1720 1450
fully mobilized skin friction is in the order of 1.59 to 1.86
percent of the diameter. Test results in this project too agrees % increase in QS 18.6
the range suggested. Portion of QL carried by base (Qb) 570 70
% increase in Qb 714
The PE pipes (0.5”) and tube-a manchette were in-corporated Total settlement at (Qd) (mm) 5 8.13
into the reinforcement cages and lowered into the bore to Total settlement at 1.5 Qd (mm) 7.5 15.08
facilitate the base grouting. The grouting was done after the * - Estimated
concrete is set, usually the base was cracked using a high
water pressure not later than twenty four hours, and grout was PILE LOAD TEST AND RESULTS
injected.
Four piles, two from each size, were selected to perform the
static pile load test. All the test piles and anchor piles are
contract piles. The maintained load test was performed in two
cycles for each pile with the first cycle having maximum test
load of safe design working load. The maximum test load for
second cycle was 150 % of the design safe working load. The
maximum test load for these both cycles were maintained for
24 hours.
The third cycle was performed as a quick test and each

incremental load was maintained for 10 minutes. The
maximum test load on this third cycle was 250 % of the safe
Fig.3 Tube a manchette used for grouting
design working load. The results were satisfactory and
complying the design requirements (Max. allowed settlement
The grouting process was controlled by either maximum grout
of 10mm under working load)
volume of 500 litters or a maximum pressure of 40 bars. But,
the pressure of 40 bars was rarely achieved and thus usually
284
14
Ultimate load interpretation
Table 3 (b) CYCLE - 2
Predicted ultimate loads were based on graphical methods
proposed by Mazurkiewicz (1972), Fuller and Hoy (1970) and Cycle 2 (Maintained Load Test)
Butler and Hoy (1977). It may be appropriate to mention here Base Max. Head Elastic
that ultimate capacity determined from the above mentioned Pile Dia. depth Load Movement Recover
methods generally not yield a load corresponding to a No. (m) (m) (ton) (mm) y
“plunging” failure. The ultimate capacity is defined by the (mm)
load at which the pile moves rapidly (plunging) without any TP - 1 1000 55 750 8.55 7.66
further load increment. Most of the existing methods for TP - 2 1000 55 1250 17.40 12.49
determining the ultimate capacity from pile load test, however, TP - 3 1500 55.3 1500 7.72 6.59
are mostly based on some arbitrary displacement criteria and TP - 4 1500 55.5 2500 27.19 16.00
there is no consensus on the best method of interpreting the
failure load. Fellenius (1980) pointed out that preferred
method of interpretaion often depend on an individual’s
Table 3 (c) CYCLE - 3
experience and he further recommended that several methods
be applied and that a preferred method be chosen on the basis
Cycle 3 (Quick Test)
of the user’s needs and special conditions of the job.
Base Max. Head Elastic
For base grouted piles in Bangkok sub soil, the predicted Pile Dia. depth Load Movement Recover
failure load based on the above methods well agrees each No. (m) (m) (ton) (mm) y
other (Teparaksa, 1994). For this particular project the (mm)
variation of predicted loads by these different methods are in TP - 2 1000 55 1250 17.05 12.43
the range of 2.5% to 19%. It should be noted that the tests TP - 4 1500 55.5 2500 29.26 17.52
were not performed to failure of piles. As the piles were not
tested for failure it was difficult to compare the results with
actual failure load. Load test results on soft base pilot piles
Table 2 Ultimate load predicted by different methods

Max. Ultimate Load (tons)
Test. Mazurk Fuller Butler and
Pile Dia. Load -iewicz and Hoy Hoy
No. (mm) (ton) (1972) (1970) (1977)
TP - 1 1000 750 1625 1585 1575
TP - 2 1000 1250 1700 1480 1460
TP - 3 1500 1500 3600 3200 3185
TP - 4 1500 2500 3250 2775 2725 Fig. 4(a) Load test on pilot pile with non base grouted pile
(Note: Piles were not tested to failure) (Diam. 1500mm, Toe embeddment at depth 60m
below ground)
Summary of pile load test results
Table 3 (a), (b) and (c ) summarize the pile load test results
and Fig. 5 and Fig. 6 show the corresponding load settlement
characteristics.
Table 3 (a) CYCLE - 1

Cycle 1 (Maintained Load Test)
Base Max. Head Elastic
Pile Dia. depth Load Movement Recover
No. (mm) (m) (ton) (mm) y Fig 4 (b) Load test on pilot pile with non base grouted pile
(mm) (Diam. 1000mm, Toe embeddment at depth 53m
TP - 1 1000 55 500 4.69 4.52 below ground)
TP - 2 1000 55 500 4.66 3.98
TP - 3 1500 55.3 1000 4.55 4.47
P-4 1500 55.5 1000 5.47 5.43
285
15
Load test results on base grouted contract piles Almost all the cracks were indicated on piles adjacent to the
sheet pile cofferdam which was used for deep excavation.
These piles in many cases were repaired by coring and cement
grouted. Physical damages to piles located adjacent to the
excavation boundary were a major source of trouble as the
sheet piles used for temporary retaining purpose cannot
completely retain the soil movement and local soil strains
induced could be sufficient enough to cause cracking.
Possibility of providing more main reinforcement bars for
piles adjacent to the excavation zone to prevent these worse
effects would have been worth to have considered.
Fig. 5 (a) (proof) Load test on contract pile (TP - 1)
Dia. 1000mm, Toe depth 55m below Ground Level Table 4 Interpretations by sonic integrity test
Diam. Diam.
1500 mm 1000mm
nos. % nos. %
Available test results 300 100 59 100
Sectional increase (small) 25 8.3 10 16.9
Sectional increase 04 1.3 02 3.4
(prominent)
Sectional Decrease (small) 04 1.3 - -
Intermittent sectional 10 3.3 04 6.8
Fig. 5 (b) (proof) Load test on contract pile (TP - 2) variations (1)
Dia. 1000mm, Toe depth 55m below Ground Level
Cracks (small) (1) 10 3.3 08 13.6
Cracks (prominent) (2) 03 1.0 - -
Notes:
1. These variations were attributed to the sub soil conditions
2. All the cracks have been found on piles adjacent to the
excavation supported by steel sheet piles
CONCLUSION
Fig. 6 (a) (proof) Load test on contract pile (TP - 3) A case history of cast in-situ concrete bored piles in Bangkok
Dia. 1500mm, Toe depth 55.3 m below Ground Level sub soil has been summarized. A total of 402 piles of 1000mm
(96 nos.) and 1500mm (306 nos.) have been constructed as
foundation elements of a wastewater treatment plant.
Base grouting has been considered to improve the bearing

capacity and settlement behavior of bored piles. Maximum
pressure of 40 bars or 500 liter of grout volume has been used
as limiting criteria. Base grouting with ‘tube -a manchette’
method has been adopted and found to be effective.
Static pile load test results on contract piles and pilot test piles
Fig. 6 (b) (proof) Load test on contract pile (TP - 3) have been summarized. Pilot piles with base non grouted,
Dia. 1500mm, Toe depth 55.5 m below Ground Level have shown either high degree of settlement or failed well
INTEGRITY TEST AND SIGNAL INTERPRETATIONS before the designed maximum test load. Ultimate load
capacity has been estimated by different methods and
It was requested to perform the integrity test on all the piles variation of the ultimate capacity found to be within 2.5 to 19
cast. Commonly interpreted defects by the integrity testing percent.
equipment developed by IBBC-TNO are shown in Table 4.
Pile integrity test was decided to be performed on all piles and
The integrity testing results showed clearly the physical commonly indicated signal characteristics have been tabulated
damages in many piles where the ill considered means were for available data.
practiced to cut the piles to trimming level.
286
16
ACKNOWLEDGEMENT
The authors are much thankful for all the staff and employees
of SEAFCO CO., LTD. especially for Mr. Kamol
Yooyuenpattana and Mr. Chetsadar Plongkrathok.
REFERENCES
Anwar M. A [1997], “Base grouting to improve the

performance of wet process bored piles in Bangkok sub soils”,
AIT Master Thesis, Bangkok, Thailand.
Fleming, W. G. K., F. Reiding and P. Middendorp. [1985]

“Faults in cast in place piles and their detection”, Proc.
Second Intern. Conf. On Structural Faults and Repair., ICE,
Westminister, London, pp. 301 - 310.
Joshi, R. C., and H. D. Sharma [1987] “Prediction of ultimate

pile capacity from load tests on bored and belled, expanded
base compacted and driven piles” Proc. Intern.. Symp. On
Prediction and Performance in Geotechnical Engineering,
Galgary, pp. 135 - 144.
Ng Kim Cheng [1983], “The construction problems and

performance of large bored piles in second sand layer”, AIT
Master Thesis, Bangkok, Thailand
Teparaksa, W. [1994], “Newly developed toe-grouted bored

pile in soft Bangkok clay; Performance and Behaviour”,
Intern. Conf. On Design and Construction of Deep
Foundations, FHWA, Florida, pp. 1337 - 1351..
Woo Siu Mun and C. H. Juang [1994], “Analysis of pile load

test results”, Conf. On Deep Foundations and Ground
Improvement Schemes, Bangkok, Thailand.
287
17
PROCEEDINGS OF THE 3RD INTERNATIONAL GEOTECHNICAL SEMINAR ON DEEP
FOUNDATIONS ON BORED AND AUGER PILES/ GHENT/ BELGIUM/19-21 OCTOBER 1998
Deep Foundations on Bored

and Auger Piles
BAP III
Edited by
W.F.VAN IMPE & W.HAEGEMAN
Laboratory of Soil Mechanics, Ghent University, Belgium
Large diameter bored piles in multi-layered soils of Bangkok

N. Thasnanipan, P. Tangseng & M. A. Anwar
Seafco Company Limited Bangkok, Thailand
A.A. BALKEMA / ROTTERDAM / BROOKFIELD / 1998
19
Deep Foundations on Bored and Auger Piles, Vam Impe & Haegeman (eds) © 1998 Balkema, Rotterdam, ISBN 90 5809 022 1
Large diameter bored piles in multi-layered soils of Bangkok
N. Thasnanipan, P. Tanseng, M. A. Anwar

ABSTRACT: Static pile load test results of ten large diameter bored piles founded in the multi-layered soils
of Bangkok are discussed in this paper. All piles were instrumented with vibrating wire strain gauges
(VWSG) and telltale extensometer rods. Range of shaft friction and end bearing values mobilized in different
soil horizons are reported. Progressive type of failure mechanism which is peculiar with the long piles in
multi-layered soils, and found to be initiated from the soil layers present at the middle reach of pile shaft and
then extending towards the top and bottom layers, is also presented. Brittle type of failure observed in some
sand and very stiff to hard clay layers is also discussed. Settlements at design loads calculated from global
factor of safety concept and limit load concept are also compared. Additionally, back calculated
α and β values are also compared with the previously recommended values for similar soils.
1 INTRODUCTION
Bored cast in situ piles are extensively used as a

foundation element in Bangkok due to the
requirements of huge loads to be transferred and
limitations of using driven piles like limited
capacities, associated soil movements and some
unwanted environmental effects. Diameters of these
piles normally fall in the range of 0.80 to 1.50 m and
toe depths down to 60m from ground level are quite
common. Soil profiles of plain of Bangkok are
always evident of the presence of marine Bangkok
Soft Clay (BSC) at the top changing to medium at
about 15 to 18m. First sand layer is usually 5 to 10m
thick and found at 25m to 30m depth, below is a
series of stiff to hard clay and medium to very dense
silty sand layers (Ref. to the Figs. 5 to 14 for typical
soil profiles). Actual pore pressure conditions in the
upper BSC are hydrostatic from circa 1m below
ground level. Then the hydrostatic conditions
changes to piezometric draw down near the bottom Figure 1. Variation of effective overburden pressure
level of BSC. Piezometric draw down resulted in with depth, (after KERDSUWAN, 1984)
increased effective overburden pressure of about 20
ton/m2 in the first sand layer and below as shown in
Figure 1. The under drainage in the Bangkok soils is 2 METHOD OF CONSTRUCTION
attributed to the deep well pumping in the area.
According to the latest reports, piezometric draw Bentonite slurry in conjunction with rotary bucket is
down conditions have started recovering since last the normal drilling procedure of pile construction
few years due to the control over deep well pumping when piles are to be founded in the first sand or
and shifting of industrial areas away from Bangkok below. Top 15 to 18m depth of soft to very soft clay
metropolitan. is almost always temporarily cased to assure the
stability of borehole. Firstly, auger is used to drill
511
21
within the temporary casing followed by rotary with unconfined compression test while for deeper
bucket with bentonite slurry down to final depth of stiff to hard clay layers Cu is indirectly estimated
excavation. Special cleaning bucket or air lift from Standard Penetration Test (SPT-N) data as Cu
technique is normally applied to clean the borehole = C1.SPT-N (ton/m2), with C1 equal to 0.674 and
base of any congregated sediments before lowering 0.507 for high and low plasticity clays respectively.
the reinforcement cage which is followed by tremie C1 values mentioned here are based on the statistical
concreting. Bentonite slurry viscosity is normally analysis carried out by Pitupakorn, 1985. Skin
maintained with in the range of 30 to 50 seconds friction capacities for the sand layers are calculated
(Marsh cone viscosity) and construction time, as fs = σv’.Ks.tanδ, with coeficient of horizontal
starting from the casing driving till the completion earth pressure Ks equal to 0.7 and δ equal to 0.75φ.
of concreting, usually fall in the range of 10 to 20 Angle of internal friction φ is also estimated from
hours excluding some accidental delays due to SPT-N values by first correcting the N values for
equipment break down, unavailability of ready overburden correction (Bowels, 1988). Some
mixed concrete or similar reasons. designers also use an equivalent term of β instead of
Ks.tanδ. It must be noted that the empirical
parameters discussed here are based on the effective
3 DESIGN PARAMETERS overburden pressure σv’ which is calculated with the
assumption of hydrostatic conditions from the
A combination of total stress method for clay layers ground level, ignoring the piezometric draw down
and effective stress method for sand layers is conditions below BSC shown in Figure 1.
normally adopted due to its simplicity and
limitations regarding unavailability of reliable 3.2 End bearing
effective stress parameters. There is no standard
criteria of acceptable settlement for these piles, but a Ultimate end bearing capacity for the sand layers is
limit of 5mm at working loads is normally accepted calculated as Nq.σv’ with bearing capacity factor Nq
for high rise buildings. in the range of 5 to 20 depending upon the relative
density of the sand layers. A limit of 500 ton/m2 for
3.1 Skin friction maximum end bearing is normally used for piles
seated in the second sand layer. For clay layers
Skin friction capacity of clay layers is estimated as fs ultimate end bearing is estimated as Nc.Cu with
= α.Cu, with adhesion factor α varying from 0.9 to bearing capacity factor Nc equal to 9. Global factor
0.3 depending upon the undrained shear strength of of safety of 2.5 on the accumulated ultimate skin
clay layers. Undrained shear strength Cu of shallow friction and end bearing is normally used to
clay layers is usually determined in the laboratory calculate the safe working loads.
Table 1. Principal data of test piles with interpreted failure loads.

Total Estimated QL Gross
SETTLEMENT(mm) AT
Qd1 Mobilized Skin Failure Load (ton) Limit load Qd3 settlement Difference of
at a Design at Qd3 Qd3 as
Dimensions Design Skin Mobil/Estim
Brinch-Hansen's
Average Failure
Mazurkiewicz's
Chin-Konder's
Pile Friction settlement of load (mm) compared to

Qu ave/2.5
Qd2/Qd1
Davisson
1.5xQd1
2.0xQd1
De'Beer
Qu ave
(Dia. x Load Friction Skin

Qd2=
Load
Qd1
No. (ton) 1% of Dia. '=QL/1.5 original Qd1

Depth)m (ton) (ton) Friction (ton) (ton) (%)
1 2 3 4 5 6=(4/5) 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
TP-1 φ1.2x57.10 650 1582 1230 1.29 1435 1715 1575 630 0.97 6.8 13.2 35.5 975 650 6.8 0.0
TP-2 φ1.2x46.25 650 1929 930 2.07 2815 2017 2416 966 1.49 5.5 9.3 15.0 1140 760 5.5 16.9
TP-3 φ1.0x46.51 500 1118 800 1.40 1477 1651 1544 1557 623 1.25 4.5 8.2 12.3 920 610 5.5 22.0
TP-4 φ1.0x49.47 450 970 930 1.04 1172 1270 1138 1193 477 1.06 6.3 12.8 40.5 650 430 5.5 -4.4
TP-5 φ1.0x43.00 450 1106 690 1.60 1399 1642 1347 1463 585 1.30 4.5 11.1 34.5 675 450 4.5 0.0
TP-6 φ1.0x41.00 360 641 620 1.03 759 997 878 351 0.98 3.2 5.0 7.2 830 550 5.0 52.8
TP-7 φ1.2x43.6 450 1499 850 1.76 891 1390 1141 456 1.01 2.5 4.2 5.4 1150 770 4.5 71.1
TP-8 φ1.2x43.5 450 519 640 0.81 1012 1180 978 1057 423 0.94 3.3 5.9 48.5 850 570 4.7 26.7
TP-9 φ1.0x43.5 400 701 530 1.32 867 946 975 929.3 372 0.93 4.0 5.2 36.5 750 500 4.9 25.0
TP-10 φ1.2x54.00 500 799 1100 0.73 3.5 6.4 16.3 1000 670 5.0 34.0
512
22
4 TEST PILES than the estimated skin friction values with the
exception of piles TP8 and TP10 who have 19% and
Test piles discussed in this paper were all 27% less capacity respectively. It must be noted that
constructed following the procedure discussed in TP10 was not tested to failure so ultimate shaft
section 2. All piles were instrumented with VWSG capacity was not yet mobilized.
at five to seven different levels to estimate the shaft Unit skin friction values mobilized in different
friction transferred to different soil horizons. One set soil layers are given in Table 2. It must be noted that
of telltale extensometer rods was also used near the the unit skin friction values reported here are the
tip of each test pile to measure the elastic shortening actual maximum mobilized capacities only. Since
of the pile shaft and finally pile base movement at piles were not instrumented to measure the
different stages of load testing. Static maintained corresponding displacements at the interface of each
load testing method was used for all piles. Test loads soil horizon, minimum values are not reflecting the
were applied using a system of hydraulic jacks ultimate unit friction values.
against the reaction frame of steel girders fixed
against anchored reaction piles. Normally three
cycles of loading are applied with first cycle up to Table 2. Skin friction values mobilized in different
the design load Qd and maintained for 12 hours soil horizons and corresponding α values.
followed by second cycle of loading up to 2 times of Soil Type SPT-N Ave. Skin Friction α
Qd and maintained for 24 hours. Third cycle of Cu Mobilized (ton/m2) (Ave.)
loading is applied up to 2.5 times Qd or maximum Average Maximum
pre-decided test load. Some times a fourth quick Soft to medium Clay 2-12 4.7 6.0 13.9 1.3
loading cycle up to the maximum test load is also Stiff to very stiff 20-35 14.0 7.5 18.8 0.54
applied and maintained for two hours. All test piles Clay
except TP10 were tested to well above the expected Hard Clay 40-46 21.5 7.3 24.6 0.34
failure loads with the concept of sufficiently First silty Sand 18-68 - 11.2 19.9 -
mobilizing the end bearing. Second silty Sand 20->100 - 15.6 25.1 -
5 TEST LOAD RESULTS

Average skin friction values mobilized in the soft to
medium clay layers seems to be even higher than
Test load results with pile dimensions are
average Cu of these layers, most likely reasons for
summarized in Table 1 and Figures 5 to 14. It is
this are: 1) actual diameter in the soft to medium
quite evident that the safe design loads (Table 1,
clay layer is slightly higher than the nominal
col.13) calculated from the average of failure loads,
diameter of the pile due to the use of temporary
interpreted by different available methods (cols.7-
casing; 2) actual Cu of the top few meters of
11), are in reasonable agreement with the estimated
weathered crust is higher than the Cu used in the
design loads (col.3). But the settlements at the
analysis. Values of α reported in Table 2 are quite
design loads (col.15) have a wide scatter between
conforming to the previous recommendations like
2.5mm and 6.8mm. So instead of computing failure
Ng, (1983) and Tomlinson (1995). Skin friction
loads from methods mentioned in Cols. 7-11, if we
values mobilized in the very stiff to hard clay layers
use the limit load concept at settlement of 1% of the
are relatively low, since at failure loads considerable
diameter (col. 18) and then use a partial FOS of 1.5
portion of the pile head settlement is absorbed by the
to calculate design loads (col.19), resulted
long pile shaft itself which reduces the relative
settlements at this design load (col. 20) will be quite
pile/soil interface movement near the deeper soil
uniform and within the limits of allowable
layers which results in partial mobilization of skin
settlement at design loads for high rise buildings.
friction capacity of these soil horizon. In other
Global FOS against overall ultimate failure will still
words, in case of long piles the geotechnical
be in excess of 2.5 against the ultimate failure of the
capacity of deeper soil layers may not be fully
pile which still have not been reached even at large
exploited before they are declared to be failed due to
pile head movements (Buttling, 1992). This can also
the excessive pile head movement, same is the case
be confirmed from the maximum end bearing values
for end bearing of these piles. Average values of unit
mobilized under maximum test loads i.e. only 25%
skin friction mobilized in the first sand and second
of the ultimate end bearing is mobilized (ref. Section
sand layers are 11.2 ton/m2 and 15.6 ton/m2
5.2)
respectively which corresponds to a β value of 0.45
to 0.55, without considering piezometric draw down
5.1 Skin friction discussed in section 1. But if we consider
Actually mobilized shaft friction capacities (ref. piezometric draw down, corresponding β values will
Table 1, col.4) are on the average by 50% higher be 0.23 to 0.28. This confirms the suggestions made
513
23
by Meyerhof (1976) who concluded that β value is considerable portion of the total pile head settlement
also dependent on the length of the pile and can be is absorbed by the elastic shortening of the pile
as low as 0.15 for very long piles. Maximum unit shaft. So even at 3 to 3.5 times working loads
skin friction mobilized in sand layers is also maximum pile tip movement is less than 7% of the
comparable to test results by Reese and O’Neill diameter of the pile. Ultimate end bearing values
(1988) who measured a maximum value of 19.15 which need a pile base settlement of 20% to 30% of
ton/m2 for sand layers. the pile base diameter (Tomlinson, 1995) are still
Dense sand and very stiff to hard clay layers only 25% mobilized.
found to exhibit a brittle type of failure mechanism Maximum mobilized end bearing for piles with
in majority of the cases, a typical family of curves tips in the sand layers are plotted in Figure 4. It is
representing such type of behavior are shown in clear that the estimated end bearing value of
Figure 2. Residual shaft friction in dense sand layers
with increasing butt loads found to be dropped to 14
50% of the maximum mobilized values in some
cases. For very stiff to hard clay layers, such 12 Soft Clay
V. Stiff Clay
Mobilized Skin Friction (ton/m )

reduction is found to be as high as 60%.
2
Dense Sand
10 Dense Sand
Hard Clay
Hard Clay
10 8 Hard Clay
Hard Clay
Dense Sand
9 Dense Sand 6
Hard Clay
Skin Friction Mobilized (ton/m
8 Hard Clay
)
4
2
V.Stiff Clay
7
Hard Clay
6 V.Stiff Clay 2
5
0
4 0 200 400 600 800 1000
Butt Load (ton)
3
2
1 Figure 3: Skin friction mobilization in different soil
0
horizons for TP10
0 200 400 600 800 1000 1200
Butt Load
(ton) Table 3. Maximum end bearing mobilized and
Figure 2: Family of curves showing typical brittle corresponding pile base settlement from telltale data.
failure in dense sand and very stiff to hard clay Pile
Max. Load Elastic Total Net
Transfered to Compressi Settlement Settlement
% of
Max. End Estimate Average
Bearing d End SPT-N Soil Type near
Pile
layers. No. the Base
VWSG (ton)
on of Pile of Pile Head of Pile Base
(mm) (mm) (mm)
Dia.
Mobilized Bearing Near
2 2
the Toe
(ton/m ) (ton/m ) the Toe
TP-1 418 25.7 94.1 68.4 5.7 370 500 100 Clayey Sand (SC)
TP-2 71 16.2 34.7 18.4 1.5 63 500 100 SAND (SP-SM)
Failure mechanism of different soil horizons with TP-3

TP-4
382
230
18.3
16.3
81.6
90.7
63.3
74.4
6.3
7.4
487
293
500
206
70
45
Silty Sand (SM)
increasing butt loads exhibit a progressive type of

Silty CLAY (CH)
TP-5 194 13.0 60.5 47.5 4.7 247 430 35 Silty Sand (SM)
behavior for some test piles. A typical, unit skin TP-6

TP-7
199
101
9.1
15.8
13.8
58.4
4.7
42.7
0.5
3.6
254
89
500
183
50
40
Silty Sand (SM)
Silty CLAY (CL)
friction development with increasing butt load TP-8 481 9.1 90.9 81.9 6.8 426 500 60 Silty Sand (SM)
(TP10) is shown in Figure 3. In number of cases TP-9

TP-10
199
201
14.1
8.0
69.0
16.3
55.0
8.3
5.5
0.7
253
178
500
228
60
50
Silty Sand (SM)
Silty CLAY (CL)
dense sand and very stiff to hard clay layers found to

reach their ultimate unit friction values well before
the maximum test loads which shows that pile head 500 ton/m2 in most of the cases correspond to a base
displacements required to mobilize ultimate shaft settlement of approximately 7% of the pile diameter
friction in these layers are quite small. An other and this level of settlement have rarely been
point to be noted here is that the progressive failure achieved even at such a high pile top displacements.
mechanism found to be started from the layers End bearing mobilized in the sand layers for
present in the middle reach of the pile shaft and then different values of base settlement (ref. Figure 4)
it extended towards top and bottom. Once the sand match well with the recommendations of Reese and
and stiff to hard clay layers started yielding most of O’Neill (1988) who recommended a value of critical
the loads were shifted to the medium and soft clay end bearing equal to 5.76*SPT-N (ton/m2)
layers at the top. corresponding to a base settlement of 5% of the pile
base diameter for sand layers of similar relative
5.2 End bearing density. Since the average value of SPT-N for test
piles founded in sand layers is approximately 60,
End bearing values mobilized under maximum test critical end bearing comes out to be 5.76*60 = 350
loads are given in Table 3. It is quite evident that ton/m2 and from the test piles plotted data Figure 4,
514
24
500 TP-
SPT-N (Blows/ft) Load (Tons)
3 0 20 40 60 80 100 0 500 1000 1500 2000 2500
450 0 Fill Material
TP-
400 8
TP-
)
Soft to Medium
10 CLAY (CH)
1
2
350
Mobilized End Bearing (t/m
.
300
TP2 (1.2m x 46.25m)

20 Stiff to Very Stiff
CLAY
250 TP- TP-
Depth (m)
5 9 Dense SAND
200 30 Very Stiff CLAY
Dense SAND
Very Stiff CLAY
150
40
100 TP- Dense to Very Dense
SAND
2
50
2
R = 0.889 TP-2
50
0
0 1 2 3 4 5 6 7
% (Base Sett./Dia. of Applied Load (Tons)
Pile) 0 500 1000 1500 2000
0
Figure 4: Development of end bearing with base
settlement as % of pile diameter. 20

40

0 20 40 60 80 100 0 500 1000 1500 2000
0
60
Fill Material
80
10
Soft to Medium
CLAY (CH)
100
20
TP-2
.
120
TP1 (1.2m x 57.1m)
Medium to Dense
Depth (m)
SAND (SM or SC)

30
40
Stiff CLAY
Figure 6: Test pile TP-2
(CH)
50
Very Dense SAND 0 20 40 60 80 100 0 400 800 1200 1600
TP-1 0
(SC)
Top Soil
60
Soft Clay
Applied Load (Tons) 10
0 500 1000 1500 2000 Medium to Stiff Clay
0 Medium to Dense
TP3 (1.0m x 46.51m)
Depth (m)
Sand
20
Stiff Clay
20
Medium Dense Sand
30
40
Very Stiff to Hard
Clay
60
40
Dense to Very Dense
Sand
80
TP-3
50
100
TP-1
120 Applied Load (Tons)
0 500 1000 1500 2000
0
Figure 5: Test pile TP-1 20

40
60
critical end bearing corresponding to 5% base
settlement is 270 ton/m2, a difference between the 80
predicted value of end bearing is due to the reason

that end bearing values predicted by Reese and 100
O’Neill (1988) are based on the SPT data with TP-3

120
energy ratio (ER) of 0.55, while SPT-N used for the
test piles are not corrected for ER.
Back calculated average value of Nc for clay Figure 7: Test pile TP-3
515
25
SPT-N (Blows/ft) Load (Tons) SPT-N (Blows/ft) Load (Tons)
0 20 40 60 80 0 400 800 1200 1600 0 20 40 60 80 0 200 400 600 800 1000
0
Medium Stiff Clay
0 Fill Material
Very Soft to Soft

Clay
Soft to Medium
10 CLAY (CH)
Loose Sand 10
Very Stiff Clay
.
TP6 (1.0m x 41.0m)
TP4 (1.0m x 49.47m)
Medium Dense Sand

20
20 Very Stiff CLAY
(CL)
Depth (m)
Depth (m)
Very Stiff Clay
30
30
Dense to Very Dense
Sand
40
Dense to Very Dense
(SM)
40
Hard Clay
50
Dense to Very Dense

TP-6
50
Sand
TP- 4 Applied Load (Tons)
60
0 500 1000
Applied Load (Tons) 0

0 500 1000 1500 2000
0
20
40
20
60
80
100 TP- 6
40
TP-4
120
0 20 40 60 80 0 500 1000 1500 0 20 40 60 80 0 400 800 1200

0 0 Fill Material
Fill Material
Soft to Medium
Soft to Medium CLAY (CH)
10 CLAY 10
.
(1. Very Stiff CLAY

TP7 (1.2m x 43.6m)
0m (CH)
x
20
De 20 43.
Stiff CLAY Very Stiff CLAY
Depth (m)
0m
pth ) . (CL)
(m)
TP
5
30 30
Dense to Very Dense
SAND (SM)
Dense to Very Dense

SAND
40 40
Hard CLAY (CL)
TP-5
Dense to Very Dense
SAND (SM) TP-7
50 50
Applied Load (Tons)

0 500 1000 1500 2000 Applied Load (Tons)
0 0 500 1000 1500 2000
0
20 20
40 40
60 60
80
80
100
100
120 TP- 7
TP- 5
120
Figure 9: Test pile TP-5 Figure 11: Test pile TP-7
26
0 20 40 60 80 100 0 400 800 1200 0 20 40 60 80 100 0 500 1000
0 Fill Material 0 Fill Material
Very Soft to Medium

CLAY (CH)
10
10 Very Soft to Soft
CLAY
.
TP8 (1.2m x 43.5m)
Stiff CLAY (CH-CL)

20
20
Depth (m)
.
Stiff to Very Stiff
TP10 (1.2m x 48.9m)

Medium Dense CLAY
SAND (SM)
Depth (m)
30 Dense to Very Dense
30 SAND
Hard CLAY (CH, CL)
40
40
Dense to Very Dense
SAND (SM)
Very Stiff CLAY (CH) TP-8 Very Stiff to Hard

Clay
50
50
Applied Load (Tons)
0
0 500 1000 1500 2000
TP-10
60
Applie d Loa d (Tons)

20 0 500 1000
0
40
60

80
20
100
120
TP- 8
TP- 10
Figure 12: Test pile TP-8 40

0 20 40 60 80 100 0 500 1000 Figure 14: Test pile TP-10
0 Fill Material
Very Soft to Medium

layers is 8 with 4.5, 6.0 and 12.8 for the three piles
10
CLAY (CH)
seated in clay layers. From Table 3 it can be
confirmed that the end bearing values in clay layer
.
also depend upon the base settlement and values of

TP9 (1.0m x 43.5m)
Stiff CLAY (CH-CL)

20
Nc more than 9 can also be mobilized at high base
settlement ratios. An other possible reason for the
Medium Dense
SAND (SM)
30 value of Nc more than 9 may be the underestimation

Hard CLAY (CH, CL) of undrained shear strength of clay from correlation
40
with SPT-N discussed in section 3.1.
Dense to Very Dense
SAND (SM)
Very Stiff CLAY (CH) TP-9

50 6 CONCLUSIONS
Applied Load (Tons)
0 500 1000 1500 2000 1. Skin friction and end bearing values of wet
0
process bored piles mobilized in the subsoils of
20
Bangkok are reported and compared with the
previous researches made for similar cases.
40 2. Dense sand and very stiff to hard clay layers

often exhibit a brittle type of failure with
60
residual friction capacities as low as 50 to 60%
80 of the peak mobilized values.
3. Failure of long piles in multi-layered soil is
100
dominantly progressive in nature and some soil
TP- 9 horizons reach their ultimate friction capacities
120
well before the over all ultimate failure loads.

4. Calculation of ultimate failure loads from the
Figure 13: Test pile TP-9 normally available methods and then design
516
27
loads by applying a global FOS give widely
scattered settlements at design loads and if we
calculate design loads from limit load concept,
quite uniform settlement at design loads can be
achieved.
AKNOWLEDGEMENTS
Writers gratefully acknowledge the continuous

support and invaluable suggestions made by their
colleagues, especially Mr. Aung Win Maung and
Mr. Ganeshan Baskaran from analysis to final draft
of this paper.
REFRENCES
Bengt B.Broms and M.F.Chang (1988), Bored Piles

in Residual Soil and Weathered Rocks in
Singapore, Deep Foundations on Bored and
Auger Piles, Van Impe (ed.), Balkema,
Rotterdam.
Bowles, J.E.(1988), Foundation Analysis and
Design, McGraw-Hill.
Buttling, S. (1992), Prediction of the Performance
of Some Instrumented Pile Tests in Hong
Kong, Proc. of Symposium on Prediction Vs
Performance in Geotechnical Engineering,
SEAGS, Bangkok.
Meyerhof, G.G. (1976), Bearing Capacity and
Settlement of Pile Foundation, Jour. Of
Geotch. Engg. Div., ASCE Vol102, No GT3,
pp195-228.
Ng, K.C. (1983), The Construction Problems and
Performance of Large Diameter Bored Piles in
Second Sand Layer, M. Engg. Thesis no. GT-
82-26, AIT, Bangkok.
Pitupakorn, W. (1985), Predictions of prestressed
concrete pile capacity in Bangkok stiff clay and
clayey sand, 8th Southeast Asian Geotechnical
conf, Kuala Lumpur.
Reese and O’ Neil (1988), Drilled Shafts: Design
Procedures and Design Methods, US Dept. of
Transportation F. H.A. Office of
implementation, McLean, Va.
Tomlinson, M.J. (1995), Pile Design and
Construction Practice, 4th Ed. Chapman & Hall.
517
28
PROCEEDINGS OF THE 3RD INTERNATIONAL GEOTECHNICAL SEMINAR ON DEEP
FOUNDATIONS ON BORED AND AUGER PILES/ GHENT/ BELGIUM/19-21 OCTOBER 1998

and Auger Piles
BAP III
Edited by
W.F.VAN IMPE & W.HAEGEMAN
Laboratory of Soil Mechanics, Ghent University, Belgium
Effect of construction time and bentonite viscosity on shaft capacity of

bored piles
N. Thasnanipan, G. BAskaran & M. A. Anwar
29
Deep Foundation on Bored and Auger Piles, Avan Impe & Haegeman (eds) © 1998 Balkema, Rotterdam, ISBN 90 5809 022 1
Effect of construction time and bentonite viscosity on shaft capacity of bored

piles
N. Thasnanipan, G. Baskaran & M. A. Anwar

ABSTRACT: Effect of construction time and slurry viscosity on the shaft friction capacity of cast in situ
bored piles have been studied by various investigators in the past but the extent to which these parameters
effect the shaft capacity is still not clear. Major obstacle in this regard is that it is hard to normalize the shaft
capacity degradation against other numerous parameters which are influencing simultaneously. Results from
eleven pile load tests, constructed with different slurry viscosities and construction times, in the layered
alluvial strata of Bangkok are presented in this paper.
It has been concluded that slurry viscosity, do not have significant effect on the shaft load transfer of these
piles but considerably reduced with increase in construction time. Trend of reduction in capacity seems to
follow an exponential decrease with increase in construction time with major part of degradation within first
24 hours of construction time.
1 INTRODUCTION 60 m from ground level are quite usual. Soil profile

of plain of Bangkok generally consists of Quaternary
Shaft friction capacity of bored piles, especially, in alluvial deposits of alternating sand and clay layers
granular soils is greatly affected by the installation as shown in Figure 1, down to the rock face which is
procedures, but the sole effect of bentonite slurry reported to be at least 550 m deep in the area
(slurry) properties and its exposure time is still (Balasubramanium, 1991).
unclear. Results from the previous researches on 2
deleterious effects of slurry properties and its contact

Average Engineering P o re P re s sure (t/m )
Parameters
time on the shaft load transfer of wet process bored 0 20 40 60

0
piles are a bit anecdotal and to some extent
0
Weathered Crust
contradictory. Major problem in this regard is that, Su = 10 - 20 KPa
the effect of either parameter can not be totally γt = 14-16 KN/m3

Bangkok Soft Clay (BSC)
(CH)
10 -1 0
isolated from other influencing variables. Due to

this reason researches have been found to report
almost converse conclusions for the similar type of 20 Medium to Very Stiff Clay Su = 40 - 140 KPa
-2 0
Depth (m)
(CH) γt = 17-21.0 KN/m3

soils.
In order to assess the effect of slurry viscosity and
exposure time on the shaft load transfer, data from 30
Actu
-3 0
SPT-N = 20 - 50
Hy
Dense to Very Dense

γt = 20.0 KN/m3
ros
al P
Sand (SM)
eleven instrumented bored pile load tests is

tati
iez
c
om
Lin
presented in this paper. All test piles were

etric
40 Su = 250 KPa -4 0
Dra
constructed using same type of equipment and Hard Clay (CL)

γt = 21.2 KN/m3
w
Do
procedures, in a similar type of soil conditions. So it

wn
Lin
is assumed that the influence of other variables

e
50 -5 0
would be minimized and the effect of slurry Dense to Very Dense

Sand (SM) SPT-N = 50 - 100
γt = 20.0 KN/m3
viscosity and exposure time, to some extent, could
be estimated. 60 -6 0
2 BORED PILES AND SOIL CONDITIONS

Figure 1. Typical soil profile of Bangkok plain with
Use of bored cast in situ piles in Bangkok, is very piezometric draw down conditions.
common. Diameters of these piles normally fall in
the range of 0.80 to 1.50 m and toe depths down to
171
31
Below the top few meters of weathered crust, a thick noted that major part of the reduction i. e. 43 percent
layer of well known “Bangkok Soft Clay” (BSC) took place in the first 8 hours of contact time and
changing to medium at about 15 to 18 m, is present. only 13 percent (56-43) reduction took place in the
First sand layer is usually 5 to 10 m thick and found rest of 89 hours (97-8), for the second barrette.
at 25 m to 30 m depth, below is a series of stiff to These findings support the concept that major part of
hard clay and medium to very dense silty sand the shaft friction capacity reduction with increase in
layers. Actual pore pressure conditions in the upper construction time, took place in first few hours of
BSC are hydrostatic from approximately 1 m below construction time and further increase in
ground level. Then the hydrostatic conditions construction time have minor contribution towards
changes to piezometric draw down near the bottom the reduction in shaft friction capacity.
level of BSC, Figure 1. Piezometric draw down
resulted in increased effective overburden pressure Corbette (1975), presented the results of load testing
of about 20 ton/m2 in the first sand layer and below. of a diaphragm wall panel excavated in conjunction
The depressurization of sand layers is attributed to with bentonite suspension, and construction time of
the deep well pumping in the area, Nutalaya (1981). approximately 48 hours. He inferred that the
presence of bentonite suspension even for such a
long time has not adversely affected the
3 REVIEW OF PREVIOUS WORK development of skin friction.
Degradation of perimeter load transfer in permeable Littlechild and Plumbridge (1998), concluded from
soil layers is attributed to the formation of filter cake the pile load tests of BERTS project Bangkok, that
which is left in place and get sandwiched at soil-pile shaft friction capacity of bored piles constructed
interface and is not scrapped of by fluid concrete. under slurry tends to decrease as the construction
Scouring capability of fluid concrete depends upon time and slurry viscosity increases.
its shear strength and that of the filter cake and it is
argued that if shear strength of filter cake is more Majano and O’Neil (1993) attempted to model the
than the fluid concrete, it can not be scoured and left formation of filter cake in the laboratory with
in place resulting in the degradation of load transfer different slurry dosages, differential pressures and
capacity of the shaft. In case of impermeable soils exposure times. They stated that perimeter load
reduction in the shaft load transfer is mainly due to transfer is a complex function of the physical and
the softening of the soil at the perimeter of the chemical characteristics of the slurry and the
borehole, because the formation of thick filter cake geomaterials, the roughness of the borehole, the
is not possible as slurry can not permeate in to the fluid pressures exerted by concrete, the shearing
soil, but some researchers e. g. Veder, C. (1963) properties of the soil, and possibly the chemistry of
observed a thin cake of few millimeters even in clay the fluid concrete. They attempted to correlate the
formations and argued that it may be the result of potential degradation of the soil-pile interface to the
electrical forces or chemical reaction of bentonite thickness and shear strength of the cake formed
suspension on the wall of the borehole. against the walls, and argued that the shear strength
and thickness can not be measured with only one or
O’Neil et al. (1992), reported the effect of slurry two parameters like slurry dosage and differential
properties on two drilled shafts constructed in pressure. With a bentonite concentration of 72
similar ground conditions (alternating layers of very kg/m3 and differential pressure of 0.5 psi they
stiff clay and dense sand). Shafts 1 and 2 were achieved a filter cake thickness of 3.14 and 4.5 mm
constructed with Marsh cone viscosity of 37 and 49 after a contact time of 4 and 24 hours. Authors also
sec. respectively, time between opening the compared the laboratory tests with the actual field
borehole and completion of concreting for both piles conditions and proved that due to the presence of
was 5 to 7 hours. Minor difference in load transfer enormously high differential pressures in the field in
of both shafts was found and authors concluded that some cases, thicknesses of filter cakes which would
the difference in slurry properties used for the be considered detrimental for shaft load transfer,
construction for both shafts had a little effect on require few hours to develop (They proved that a
shaft load transfer. filter cake of 6 mm thickness would require only 2.3
hours to develop in the field under a differential
Cernak (1976), performed full-scale load tests on pressure of 10 psi).
three barrettes in sandy gravels, two barrettes were
constructed using slurry with different exposure
times, 8 and 97 hours. Load test results indicated a 4 METHOD OF CONSTRUCTION
decrease of skin friction capacity of 43 percent and
56 percent, respectively, as compared to the barrette All the eleven piles reported in this paper were
excavated dry and concreted immediately. It must be constructed, following the internationally accepted
172
32
guidelines and specifications. Slurry in conjunction must also be noted that the actual contact time for
with rotary bucket is used for all piles. slurry starts, once it is fed to the borehole till the up
Temporary casing, having length depending upon flowing concrete reaches the bottom level of
the thickness of the BSC at the site is driven using a temporary casing. Since the accumulated time
high frequency vibro-hammer. Drilling operation is required for concreting above the bottom level of
commenced with an auger down to the base of the temporary casing and auger drilling inside the
temporary casing. Further drilling below the bottom temporary casing roughly required one hour,
of temporary casing is continued using a drilling construction times given in Table 2 are supposed to
bucket by first filling the borehole with slurry. be appropriate for comparison.
Bentonite, classified as Activated Sodium Bentonite
is usually used and slurry properties are controlled
within the limits given in Table 1. 5 SHAFT CAPACITY ESTIMATION
Table 1. Slurry specifications followed. A commonly accepted design approach which

Slurry Property Control Limits consists of a combination of total stress method for
Viscosity (Marsh cone) 30 - 60 clay layers and effective stress method for sand
layers is adopted to estimate the shaft friction
Density (g/cc) 1.04 - 1.15 capacity reported in Table 2, col. F. Skin friction
pH 7 - 11 capacity of clay layers is estimated as fs = α.Cu,
Sand Content <4%
with adhesion factor α. taken from the curve
proposed by Suchada (1989), Figure 2, for Bangkok
subsoils. It can be noted that value of α used is quite
comparable to the previous recommendations by
Outer diameter of the drilling bucket is slightly less Tomlinson (1957) and Kulhawy (1984).
than the nominal size of the pile to provide an
annular by-pass for the slurry during lowering and
hoisting operations. Required nominal size of the 1.20
Tomlinson,1957
Bored Cast Insitu
pile is achieved using a cutting edge with Piles
sidecutters. Slurry level in the borehole is 1.00
Stas and Kulhawy, 1984
maintained within 1 to 1.5 m from top of temporary

casing. After drilling down to the required tip level 0.80
α
of pile, special cleaning bucket and/or air lift

Adhesion Factor ,
Suchada,1989
technique is normally applied to clean the borehole 0.60 for bored piles in Bangkok soils
base of any congregated sediments before lowering

the reinforcement cage. Concreting is done using a 0.40
tremie pipe at the center of the borehole. Total time

to carry out this operation, starting from the casing 0.20
driving till the completion of concreting, usually fall

in the range of 10 to 20 hours excluding some 0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
accidental delays due to equipment break down, Undrained Shear Strength (ton/m )
2
delay in ready mixed concrete due to traffic

congestion or similar reasons. In most of the cases
whole operation of pile construction is accomplished
within 24 hours, but in some cases, due to the local Figure 2. Comparison of Adhesion factor α used for bored piles
authority regulations which do not permit to operate in Bangkok Subsoils.
during night time hours, borehole has to be left
exposed for 10 to 12 hours unagitated. In such
cases, drilling is intentionally curtailed above the Undrained shear strength Cu of shallow clay layers
final excavation level and resumes next day. We is normally determined in the laboratory with
assume that in such cases, during drilling after a unconfined compression test while for deeper stiff to
delayed period of 10 to 12 hours, any possible filter hard clay layers, Cu is indirectly estimated from
cake on the borehole walls is automatically scraped Standard Penetration Test (SPT-N), as Cu =
of. Construction times more than 24 hours reported C1.SPT-N (ton/m2), with C1 equal to 0.674 and
in Table 2 are examples of such cases. Construction 0.507 for high and low plasticity clays respectively.
time reported in Table 2, col. D is the time elapsed C1 values mentioned here are based on the statistical
between the start of auger drilling to the finishing of correlations suggested by Pitupakorn, 1985. Skin
concreting. Longer construction times for the test friction capacities for the sand layers are calculated
pile reported are usually due to the time consumed in as fs = σv’. Ks.tan.δ with coefficient of horizontal
instrumentation installation with the rebar cage. It earth pressure Ks equal to 0.7 and δ equal to 0.75 φ.
173
33
Table 2. Construction parameters and shaft capacities of test piles.
Bentonite Pile Age at

Pile Slurry Construction the Time of Estimated Actual Shaft Actual/Estimated
Pile No. Dimensions Viscosity Time Load Testing Shaft Capacity Shaft Capacity
(m) (sec.) (hrs) (days) Capacity (ton)
(ton)
A B C D E F G H=(G/F)
TP-1 φ1.2x57.1 38 43.0 32 1230 1200 0.98
TP-2 φ1.2x46.3 38 13.1 29 930 1300 1.40
TP-3 φ1.0x46.5 37 9.8 25 800 1250 1.56
TP-4 φ1.0x49.5 37 38.7 15 930 750 0.81
TP-5 φ1.0x43.0 38 26.0 19 690 700 1.01
TP-6 φ1.0x41.0 45 11.8 32 620 800 1.29
TP-7 φ1.2x43.6 55 16.8 32 850 1100 1.29
TP-8 φ1.2x43.5 38 12.3 24 640 750 1.17
TP-9 φ1.0x43.5 37 11.3 18 530 700 1.32
TP-10 φ1.2x62.0 41 32.4 27 1534 2000 1.30
TP-11 φ1.2x54.2 38 30.0 39 1050 1300 1.24
for 12 hours followed by second cycle of loading up

Value of Ks used is in good agreement with the to 2 times of Qd and maintained for 24 hours. Third
recommended values by different researchers like cycle of loading is applied up to 2.5 times Qd or
Fleming (1977), who suggested a value not more maximum pre-decided test load. Occasionally, a
than 0.75 to account for the possible degradation due fourth quick loading cycle up to the maximum test
to the formation of filter cake. Angle of internal load is also applied and maintained for two hours.
friction φ is also estimated from SPT-N values by All test piles, except TP-2 and TP-10, were tested to
first correcting the N values for overburden well above the expected failure loads with the
correction (Bowels, 1988). Some designers also use concept of sufficiently mobilizing the end bearing.
an equivalent term of β instead of Ks.tan.δ It must Load displacement curves of the test piles are shown
be noted that the empirical parameters discussed in Figure 3 for comparison.
here are based on the effective overburden pressure
σv’ which is calculated with the assumption of
hydrostatic conditions from the ground level, 7 ACTUAL SHAFT FRICTION CAPACITIES
ignoring the piezometric draw down conditions
below BSC shown in Figure 1. Actually mobilized shaft friction capacities are
calculated from the VWSG load transfer results at a
level where load displacement curves show a
6 TEST PILES maximum curvature and are given in Table 2, col. G.
Test piles discussed in this paper were all constructed
following the procedure discussed in section 4. All 8 EFFECT OF SLURRY VISCOSITY
piles were instrumented with Vibrating Wire Strain
Gauges (VWSG) at five to seven different levels to Effect of slurry viscosity on the shaft load transfer
estimate the shaft friction transferred to different soil can best be seen from the plot between the ratio of
horizons. One set of telltale extensometer rods was Actual/Estimated shaft capacity versus slurry
also used near the tip of each test pile to measure the viscosity, Figure 4. Over all scatter of data do not
elastic shortening of the pile shaft and finally pile suggest any trend. Since the effect of slurry viscosity
base movement at different stages of load testing. is difficult to separate, without normalizing the
Static maintained load testing method was used for effect of construction time. Test piles TP-6, 7 and 9
all piles. Test loads were applied using a system of have a slight difference in construction time and pile
hydraulic jacks against the reaction frame of steel age at load testing but slurry viscosity is
girders fixed against anchored reaction piles. significantly different, 45, 55, and 37 sec.
Normally three cycles of loading are applied with respectively but the ratio of Actu./Esti. shaft
first cycle up to the design load Qd and maintained capacities for these piles are necessarily the same
174
34
Applied Load (Tons)
5, plot between Actu./Esti. shaft capacity versus
0 500 1000 1500 2000 2500 construction time. Though the data is widely
0 scattered, best fit curve shows the trend of
exponential decrease of shaft load transfer with
20
TP-1
increase in construction time. Best fit curve for the
TP-2 available data also reveals that the available shaft
capacity is more than the estimated capacity for a
TP-3
40 TP-4
Movement (mm
TP-5
TP-6
construction time as long as 40 hours. This confirms
60 TP-7 the validity of recommendations made by some
researchers like Fleming (1977), who suggest to
TP-8
TP-9
80 TP-10
TP-11
complete the concreting within 24 hours after
finishing the drilling, and seems to be more
100
reasonable because slurry is kept continuously
120
agitated during drilling operation and actual
exposure time for slurry starts, once the drilling is
Figure 3. Load settlement curves of the test piles. completed. Test pile TP-4 which exhibited the
lowest capacity and has the minimum slurry
which suggest that the sole effect of slurry viscosity viscosity (37 sec.), second longest construction time
on shaft capacity reduction is insignificant. (38.7 hours) and the minimum age (15 days) at the
Additionally, TP-1, 4 and 5 which exhibit a time of load testing, in the group reveals that the
maximum reduction in shaft capacity have almost shaft capacity reduction is contributed by increased
same slurry viscosity but different construction time construction time as well as reduced pile age while
which confirms that slurry viscosity within the data slurry viscosity have minor effect.
set available do not directly have any effect on the It can also be noted that the major part of the shaft
shaft load transfer of the piles. capacity reduction is contributed by the first 24
hours of construction time which normally
considered feasible and agreeable by most of the
9 EFFECT OF CONSTRUCTION TIME piling contractors under normal circumstances, and
further delays beyond 24 hours have a minor
Since ten out of eleven bored piles were constructed addition to the shaft load capacity degradation.
with slurry viscosity varying within narrow range These findings are in good agreement with the
effect of construction time can best be seen in Figure previous observations discussed in section 3.
1.60
25
TP-1
1.40 29(Pile age at load testing, days) TP-2
Actual / Estimated Shaft Capacity
18 27 TP-3
32 32
39 TP-4
1.20
24
TP-5
TP-6
1.00 19
32 TP-7
TP-8
TP-9
0.80 15
TP-10
TP-11
0.60
35 40 45 50 55 60
Bentonite Viscosity (sec)
Figure 4. Effect of slurry viscosity on shaft capacity of bored piles.

175
35
1.60
25
TP-1(Tip in SAND)
TP-2(Tip in SAND)
Actual / Estimated Shaft Capacity
1.40 29 (pile age at load testing, days)

TP-3(Tip in SAND)
18 27
32 32 TP-4(Tip in CLAY)
39
1.20 TP-5(Tip in SAND)
24
TP-6(Tip in SAND)
19 TP-7(Tip in CLAY)
1.00
32
TP-8(Tip in SAND)
TP-9(Tip in SAND)
0.80 15
TP-10(Tip in
SAND,Base Grouted)
TP-11(Tip in
SAND,Base Grouted)
0.60
5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0
Construction Time (hours)
Figure 5. Effect of construction time on shaft capacity of bored piles.
10 DISCUSSION or Maintained Load (ML). Reduction in construction

time, generally, lead to better performance of pile
From Figure 5 it can be noted that actually shafts but this reduced construction time could
mobilized shaft capacity is equal to the estimated never exclude the accidental occurrence of delays.
shaft capacity for a construction time of Any potential degradation in load carrying capacity
approximately 35 - 40 hours or in other words, the of such piles caused by uncontrollable delays must
empirical design parameters commonly used to be considered along with the conceivable trend in
estimate the shaft friction capacities discussed in reduction under these conditions.
section 5, are justifiable for a total construction time It must also be noted that because of high
of longer than 24 hours, especially, to account for differential pressures imposed due to depressurized
the unexpected delays during construction in the sand layers in Bangkok sub soils (differential
downtown areas of Bangkok which are notorious for pressures of the order of 30 psi will be imposed if
its traffic congestion and other constraints imposed slurry level is maintained near the ground level due
by local authorities. This also confirms that the to piezometric draw down of 20 m in the sand
allowable time of 24 hours after finishing the layers), formation of filter cakes which are
drilling operation (Fleming, 1977) is also justifiable. considered to degrade the shaft load transfer would
However, with no dispute, shorter the construction theoretically take fraction of seconds to develop, if
time better the shaft load transfer response. we calculate on the basis of procedure recommended
It must therefore be emphasized that the selection by Majano (1993). This means that the degradation
of empirical parameters like α, β, δ and Ks, to of pile shaft interface due to the formation of filter
estimate shaft load capacities at the design stages cake in permeable soil layers is unavoidable within
must also consider the practically possible the practical construction time limits.
construction times by suitably incorporating the
effects of any possible delays specific to the site and
working conditions. Empirical factors mentioned 11 CONCLUSIONS
above are not the fundamental properties of the soil-
pile system, and need to be accounted for the 1. Bentonite slurry viscosity within the data set
installation procedures or even the method of pile available do not have significant effect on the
load testing like Constant Rate of Penetration (CRP) shaft load transfer of bored cast in situ piles.
176
36
2. Longer construction time reduces the shaft Suchada, (1989), Bored Piles in Bangkok Subsoils, , M. Engg.
carrying capacity of bored piles constructed Thesis, AIT, Bangkok.
Tomlinson, M. J. (1957) The adhesion of Clay Piles Driven in
under slurry. However, the major portion of Clay Soils, Proc. 4th Intl. Conf. on Soil Mechanics and
degradation occurs within the first 24 hours Foundation Engineering, London.
period, which is unavoidable under normal O’ Neil et al.(1992), Behaviour of Axially Loaded Drilled
construction time. Shafts in Beaumont Clay, The University of Texas at
3. Degradation of shaft capacity with increase in Austin.
construction time beyond 24 hours is not Veder, C. (1963), An Investigation on The Electrical
Phenomena At The Area of Contact Between Bentonite
excessive and can be adjusted within the Mud And Cohesionless Material, 5th Intl. Conf. on Soil
empirical design parameters or factor of safeties Mechanics, Paris.
chosen at the design stages.
4. Piles constructed within 24 hours showed
generally higher shaft capacities than estimated
as per normal design practice adopting
empirical parameters suggested by different
researchers for Bangkok subsoils and can safely
be used for slurry viscosity up to 55 sec. and
total construction time up to 40 hours.
AKNOWLEDGEMENTS
Writers wish to acknowledge the co-operation of

Mr. Aung Win Maung and Mr. Pornpot Tanseng, in
the preparation of this paper.
REFERENCES
Baker, Jr. C. N. et al. (1970), Load Carrying Capacity

Characteristics of Drilled Shafts Constructed with the Aids
of Drilling Fluids, Research Report 89-6, Center for
Highway Research, The University of Texas at Austin, TX.
Balasubramanium, A. S. (1991), Inaugural Lecture on
Contributions in Geotechnical Engineering Soil Mechanics
and Foundation Engineering, AIT, Bangkok, March 14,
1991.
Bowles, J.E.(1988), Foundation Analysis and Design,
McGraw-Hill.
Cernak, B. (1976), The Time Effect of Suspension on the
Behaviour of Piers, 6th European Conf. On Soil Mechanics
and Foundation Engineering, Vienna.
Corbett, B. O. et al. (1975), A Load Bearing Diaphragm Wall
at Keinsington and Chelsa Town Hall, London, Conf. on
Diaphragm Walls and Anchorages, ICE, London.
Fleming, W. K. (1977), The Use and Influence of Bentonite in
Bored Pile Construction, CIRIA Report PG3.
Littlechild, B. and Plumbridge, G. (1998), Effects of
Construction Technique on the Behaviour of Plain Bored
Cast In Situ Piles Constructed Under Drilling Slurry, 7th
Intl. Conf. and Exhibition on Piling and Deep Foundations,
Vienna.
Majano, R. E. and O’Neil (1993), Effect of Slurry Dosage and
Exposure Time on Perimeter Load Transfer of Drilled
Shafts, ADSC Publication.
Nutalaya, P. (1981), Subsidence Studies of the Bangkok Plain,
AIT Research Reports.
Pitupakorn, W. (1985), Predictions of Prestressed concrete pile
capacity in Bangkok stiff clay and clayey sand, 8th
Southeast Asian Geotechnical conf., Kuala Lumpur.
Stas, C. V. and Kulhawy, F. H. (1984), Critical Evaluation of
Design Methods for Foundations Under Axial Uplift and
Compression Loading, Report EL-3771, EPRI, Palo Alto,
Calif.
177
37
PROCEEDINGS OF THE SYMPOSIUM IN HONOR OF
PROFESSOR SENG-LIP LEE
May 14-15, 1999
Asian Institute of Technology, Pathumthani, Thailand
Innovative Solutions in Structural and

Geotechnical Engineering
Edited by
Pisidhi Karasudhi and Anil C. Wijeyewickrema

Asian Institute of Technology, Thailand
K. Y. Yong and T. Balendra

The National Institute of Singapore
Performance comparison of bored and excavated piles

in the layered soils of Bangkok
N. Thasnanipan, M.A. Anwar, Maung A. W. and P. Tangseng
Asian Institute of Technology National University of Singapore
39
Symposium on Innovative Solutions in Structural and Geotechnical Engineering
in Honor of Professor Seng-Lip Lee, May 14-15, 1999
PERFORMANCE COMPARISON OF BORED AND EXCAVATED PILES, IN

THE LAYERED SOILS OF BANGKOK
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung, Pornpot Tanseng
SEAFCO Co., Ltd., Bangkok, Thailand.
ABSTRACT: Installation techniques are considered to effect the performance of cast in situ piles.
This paper compares the load transfer characteristics of a fully instrumented bored pile (Dia. 1.5 m)
and barrette (1.5 x 3.0 m) having same length equal to 57 m, embedded in identical ground
conditions (on same site, 30m apart). Total time consumed to construct the pile and barrette differ
considerably, 27 and 75 hours, respectively. No significant difference in load transfer has been
observed between the pile and the barrette. It has been concluded that, by observing good engineering
practices, any conjectured degradation effects on load capacity of such piles can be eliminated.
Additionally, the load applied on barrette was one of the highest test load in the region with a
reaction frame “tower” capacity of 6000 tons. Reaction frame and loading system arrangement, to
achieve such a high load capacity is also discussed.
1. INTRODUCTION
Barrettes become a common foundation element in combination with bored piles and diaphragm
walls in Bangkok (Refer: Thasnanipan et al, 1998-b). Foundations for the BECM tower project
located at Rama IX road Bangkok, Thailand, required 560 bored piles of 1.2 m and 1.5 m diameter
and 24 number of barrettes having cross section size of 1.5 m x 3.0 m. Foundation plan of the tower
is shown in Figure 1. Barrettes were necessary to transfer the very high loads coming from the
central 52 story tower’s lift shafts. Piles were approximately 57.0 m long with tips embedded into
the second sand layer. Depending upon the load requirements piles were base grouted in the central
tower area and at the locations of high column loads. Load testing on working piles was required to
verify the design parameters, and it was proposed to test one bored pile of 1.5 m diameter and one
barrette with full instrumentation. Since, pile and barrette were constructed using two different
techniques and the parameters like construction time, bentonite properties etc, were considerably
different, results from the load tests have provided a unique chance to assess the extent of difference
in behavior, which have been speculated to effect considerably in the past by some theoreticians.
2. SOIL PROFILE
Soil profile at the site is no different than the typical soil profiles of plain of Bangkok and is shown in
Figure 2. Below the top 2 m of weathered crust 11.5 m thick Bangkok Soft Clay is present. First stiff
clay starts from 13.5 m depth and extends to 26 m, below is the First sand layer and is about 10 m
thick. Below the First sand layer a thick layer of Hard clay which extends down to 54 m, is present,
and then starts the second sand layer which extends down to about 80 m where the next hard clay
layer starts. Undrained shear strength, Su obtained from unconfined compression test and Standard
Penetration Test (SPT) are shown in Figure 2. Ground water table depth at the site generally varies
from 1 to 1.5 m, piezometeric conditions of the site are also quite similar to the other parts of the
Bangkok with piezometric draw down of about 22 m near the bottom of the first stiff clay layer.
41
346
132.10
96.20
Test-Barrette
Test
Pile
132.68
94.00
Figure 1. Foundation footprint of BECM tower project, showing the location of test pile and barrette.
3. CONSTRUCTION PROCEDURE
3.1 Bored Pile
Wet process method with rotary drilling bucket was employed. Bentonite slurry conforming to the
widely accepted specifications were used as supporting fluid. Top 15 m soft clay was temporarily
cased to assure the stability of the borehole. Firstly, auger was used to drill within the temporary
casing, followed by rotary bucket with bentonite slurry down to final depth of excavation. Airlift
technique was applied to clean the borehole base of any congregated sediments. Before lowering the
reinforcement cage special cleaning bucket was used to scrap of the borehole walls and the base.
Reinforcement cages were lowered inside the borehole by simultaneously attaching the
instrumentation at specified locations. Soon after lowering the rebar cage tremie concreting was
started. Properties of the bentonite slurry used are given in Table 1. Polystyrene grains plug was
used before the first charge of concrete to avoid the mixing of bentonite with concrete. Time
consumed in different construction activities is plotted in Figure 3.
3.2 Barrette
Mechanical rope-grab in conjunction with bentonite slurry was used to excavate the trench. A guide
wall cast with inside clear dimensions slightly more than the nominal size of the barrette was used to
guide the grab during initial bites. Since time consumed in the preparation of instrumentation was
quite long, refer Figure 3, desanding was continuously done to keep the bentonite slurry agitated,
which also helps to minimize the growth of filter cake by actually reducing the exposure time.
42
347
Su (UC) t/m2 BUTT LOAD (TON) Max. Mobilized Skin Friction
(ton/m 2)
0 5.0 10.0
DEPTH (m) 0 1000 2000 3000 0 1000 2000 3000 4000 5000 0.0 10.0 20.0 30.0
0 0
Weathered
Crust
Bored Pile
Soft Barrette
Clay
-10
First Stiff
-20 Clay
First
-30 Sand
-40
Hard
Clay
-50
Second
Tip Level Tip Level
Sand
-57.5 m -57.5 m
-60
60
1.50 m
60
3.00 m
0 50 100
SPT-N (blows/30cm)
VWSG
Bored Pile 1.50 m Barrette
Mechanical Extensometer
Figure 2. Comparison of bored pile and barrette, with soil profile at the site.
As another measure, trench was once again occupied by grab to scour the trench walls and to
remove, if any, filter cake formed on the walls (Refer: Drilled Shaft Inspector’s Manual, 1989). It is
authors opinion that, if due to some unforeseen reasons, reinforcement cage lowering have to be
delayed for considerable period of time or for overnight. It is a good practice to use the grab again to
scrap the trench walls which eliminates any foreseeable bad effects due to such unexpected delays.
After lowering the rebar cage, tremie concreting was done.
Table 1. Comparison of bentonite slurry properties.

Pile Barrette
Before feeding to the borehole After Recycling & Before Before feeding to the After Recycling & Before concreting
concreting (near trench (near trench base)
borehole base)
Viscosity (sec) 33 36 36 49
Density (g/cc) 1.08 1.10 1.10 1.17
pH value 8 8 8 9
Sand Content (%) 0.1 0.8 1.0 1.1
43
348
B arrette
D rilling
D eSanding
C age Low ering
B ored P ile Trem ie Pipe Preparation
C oncreting
0 10 20 30 40 50 60 70 80
T im e Co n su m e d (h o u rs)
Figure 3. Comparison of time consumed in different construction activities.
4. INSTRUMENTATION
Both, pile and barrette were instrumented with Vibrating Wire Strain Gauges (VWSGs) and
Mechanical Extensometers (ME) at five levels along the pile shaft at the known interface boundaries
of different soil layers. Pile has four sets of VWSGs and one set of ME at each level. While the
Barrette has six sets of VWSGs and two sets of ME at each level, as depicted in Figure 2.
5. LOAD TEST ARRANGEMENT
Reaction frame anchored against the four working piles was used to load the piles. Reaction
frame used to load the barrette was one of the biggest of its kind used in the region with overall
height equal to about 8 m as shown in Figures. 4a and 4b. Five numbers of built-up steel girders
supported on each side by two 1st level cross girders were used to achieve the required maximum
frame capacity of 6000 tons. 1st level crossbeams were supported against the 2nd level cross girders.
2nd level cross girders were then anchored against the surrounding barrettes using anchor blocks at
the top. Especially fabricated rigid transfer girders as shown in Figure 5 were used to distribute the
tension force coming from the Tie-bars to the dowel bars above the anchor barrette heads. Sixteen
number of hydraulic jacks each having 500 ton capacity, were installed between the barrette cap and
main girder of the reaction frame. Jacks were arranged in a symmetrical fashion as shown in Figure
4b, to avoid any possible eccentricity of loading. Ball bearings were used between the main girders
and hydraulic jacks to keep the line of loading in the true vertical direction.
6. EFFECT OF SHAPE ON SHAFT LOAD TRANSFER
Assessment of shape effect on the shaft load transfer has been attempted by various researchers
in the past. Hosoi et al (1994), concluded that the earth pressure acting on the flat surface of a
rectangular diaphragm wall panel is larger than that of circular bored piles. Numerical analysis
performed by them showed that the earth pressure on the flat surface of a trench with L/B = 1000
(plain strain conditions), where, L and B are the length and width of the trench in plan respectively, is
higher than that of circular borehole. In order to assess the effect of different aspect ratios (L/B) on
the earth pressure developed around the trench, attempt has been made to model the behavior of
trench using finite element computer program. Figure 7 shows the orientation of principal stresses
and stress contours developed around the trenches with different L/B ratios. It has been observed that
sufficient hoop compression stresses develop around the trench and if L/B <3. But if L/B is
increased above 5, longer sides of the trench start yielding and it can be concluded that under such
conditions hoop stresses around the trench become less effective which give rise to the increased
44
349
Figure 4a. View of the test frame having maximum capacity of 6000 tons.
5.00 m 5.00 m
2nd Cross Beam

I 350 x 2000 x 8000 mm.
B20
1st Cross Beam (R)

2 x I 600 x 2700 x 10000 mm.
Main Beam
5.00 m
5 x I 400 x 2000 x 9000 mm.
B13
B14
B15
Hydraulic Jacks
5.00 m
Reference Beam
B9
2nd Cross Beam
I 350 x 2000 x 8000 mm.
1st Cross Beam

2 x I 600 x 2700 x 10000 mm.
2nd Cross Beam

I 350 x 400 x 7500 mm.
Figure 4b. Barrette load test plan.
45
350
Figure 5. Stiffening girder used to distribute force to dowel bars.
Figure 6. A view of the instrumented barrette cap with jacks.
46
351
L/B = 1
L/B = 3
L/B = 5
Figure 7. Orientation of principal stresses and stress contours around the excavated trenches with
different aspect ratios.
47
352
earth pressure on flat surface of trenches with large aspect ratios. Although the exact determination
of earth pressure developed on the surface of trench depends on the soil properties and the over
burden pressure at the level of consideration, it can be concluded that no significant difference in
earth pressure is observed between the bore hole (L/B=1) and the barrette (L/B =2).
7. RESULTS OF LOAD TEST
Values of unit skin friction developed at the interface of different soil horizons along the pile
and barrette shafts are given in Table 2 and are also plotted in Figure 2. Comparison of unit skin
friction values prove that there is no significant difference in shaft load transfer between pile and
barrette, even though the construction methodology adopted is different. These findings are in line
with the conclusions made by Thasnanipan et al (1998-a). Load settlement curves of the pile and
barrette are shown in Figure 8.
Table 2. Comparison of unit skin friction and end bearing mobilized for the test pile and barrette.
Mobilized Skin Friction (ton/m2)
Depth Soil Type Pile Barrette
(m) At Design Load Max. Mobilized At Design Load Max. Mobilized
0 - 13.5 Bangkok Soft Clay (CH) 0.20 2.63 0.26 2.24

13.5 - 25.0 Stiff - V. Stiff Clay (CH) 5.90 7.01 3.58 8.86
25.0 - 35.0 Med.-Dense Silty Sand (SM-SP) 11.46 22.76 13.25 25.28
35.0 - 50.0 Hard Silty Clay (CH) 3.09 11.11 3.55 9.61
50.0 - 55.0 3.5m Hard Clay + 1.5m Clayey Sand (SC) 3.43 14.84 3.19 7.49
Maximum Mobilized End Bearing (Tom/m2) 270 101
A p p l i e d L o a d (T o n s)
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
0
P ile H e a d M o ve m e n t (m m
20
B ar r ette
Pile
40
60
80
Figure 8. Comparison of Load Settlement Curves of Barrette and Pile.
48
353
8. CONCLUSIONS
1) No significant difference in shaft load transfer has been observed between the wet process bored
pile and barrette, in spite of the presence of considerable difference in construction techniques
followed and other relevant parameters like bentonite slurry viscosity and construction time.
2) All foreseeable negative effects attributed to construction parameters can be eliminated by

observing good engineering practices.
3) Reaction frame can be designed and erected to apply very high compression loads of the order of
6000 tons for pile load testing.
ACKNOWLEDGEMENTS
The authors wish to thank Dr. Wanchai Teparaksa and their colleague Mr. Kamol Singtogaw for
their invaluable suggestions. They also wish to thank Mr. Natamon of STS Engineering Consultants
Co., Ltd. for providing the necessary data.
REFERENCES
Drilled Shaft Inspector’s Manual (1989), Prepared by The Joint caisson – Drilled Shaft Committee of
the ADSC: The International Association of Foundation Drilling and DFI: Deep Foundation
Institute, NJ. USA.
Hosoi, T, Yagi N & Enoki, M. (1994), Consideration to the Skin Friction of Diaphragm Wall
Foundation, 3rd Intl. Conf. on Deep Foundation Practice incorporating PILETALK, Singapore.
Static Load Testing Report of Barrette Pile for BECM Tower Project, Rama IX Road, Prepared by
EDE-STS Engineering Joint Venture, Bangkok, Thailand.
Static Load Testing Report of Bored Pile for BECM Tower Project, Rama IX Road, Prepared by
EDE-STS Engineering Joint Venture, Bangkok, Thailand.
Thasnanipan, N, Baskaran G. & Anwar, M.A. (1998-a), Effect of Construction Time and Slurry
Viscosity on Shaft Capacity of Bored Piles, Proc. of the 3rd Intl. Geotechnical Seminar on Deep
Foundations on Bored and Auger Piles, Ghent, Belgium, Balkema.
Thasnanipan, N., Maung, A.W. & Tanseng, P (1998-b), Barrettes Founded in Bangkok Subsoils,
Construction and Performance, Proc. of the 13th Southeast Asian Geotechnical Conf. Taipie,
Taiwan, R.O.C.
49
PROCEEDINGS OF THE ELEVENTH ASIAN REGIONAL CONFERENCE ON SOIL MECHANICS
AND GEOTECHNICAL ENGINEERING : SEOUL/ KOREA/ 16-20 AUGUST 1999
Soil Mechanics and

Eleventh Asian Regional Conference
Edited by
S.W.Hong, M.M.Kim, G.S.Yang, S.R.Lee, S.S.Chung, C.C.Ihm, H.T.Kim
J.B.Park & B.S.Lee
Base grouting of wet process bored piles in Bangkok subsoil

Teparaksa W.
Chulalongkorn University, Thailand
Thasnanipan N., Anwar M. A.

51
Eleventh Asian Regional Conference on Soil Mechanics and Geotechnical Engineering, Hong et al. (eds)
 1999 Balkema, Rotterdam, ISBN 90 5809 053 1
Base grouting of wet process bored piles in Bangkok subsoils.
W. Teparaksa
Chulalongkorn University, Bangkok, Thailand.
N. Thasnanipan & M.A. Anwar
Seafco Company Limited. Bangkok, Thailand.
ABSTRACT: Base grouting techniques are widely used to rectify the soft toe problem of bored piles in Bangkok. Two different
techniques, commonly known as flat jack and tube-a-manchette are used. This research study is mainly focused on the tube-a-
manchette technique. Six piles were constructed with the provision of PVC-casing inside the shaft to allow the drilling through the
pile bases in order to collect the soil samples below the pile tip. Piles were base grouted by varying the controlling parameters such as
grout volumes, pressures and injection rates. Soil samples were collected from beneath and at some distance away from the pile tip,
and analyzed. It is observed that grout does not penetrate into the surrounding sand of the pile toe even at high pressures. It partially
replace and/or precompress the sediments present under the pile tip and most of the grout just rises up the soil/pile interface by
replacing the bentonite cake. Recommendations are made to eliminate the soft toe problem more effectively.
1 INTRODUCTION 0
FILL
X- Section through pile base
The method of constructing bored piles under bentonite
SOFT
CLAY
suspension has become well established in Thailand. It is -10 1

particularly used for large diameter piles, where powerful rigs 2
can be used with rotary tools to bore through otherwise unstable Manchettes
water bearing strata. It is almost always used with a temporary Grouting
FIRST STIFF
circuit
casing to support the top most Bangkok Soft Clay (BSC), -20
CLAY
(U-loop)
1
resulting in a hybrid of the casing and bentonite methods.
DEPTH
(m)
Construction problems are affecting integrity and

performance of such piles. A pile without a sound toe and -30
without a proper seat into the virgin ground should prove
FIRST
SAND
disastrous particularly if required to act in point bearing. Even PVC- 3

otherwise, such piles could lead to higher orders of detrimental CASING
settlements. The known causes of ending up with a defective -40

SECOND STIFF
30 cm
pile toe, are the loosened materials during boring operation and
CLAY
accumulation of sediments from the bentonite column at the base

of the borehole. The technology of flowable concrete and tremie -50
placing are well known to avoid this problem up to certain extent Pile Concrete
and it is already a standard procedure to use a cleaning bucket or tip coring 0.5-1.0m
SECOND SAND
Position of PVC-casing
an airlift or both to clear the borehole base of any soft or loose with test pile no.
-60
material. Manchettes SPT 2-3m
Washboring & soil
Another solution was found in pressure grouting the base -65 sampling location
after the completion of the pile. This improves the condition of
the soil and pre-stresses the pile. Base grouting is widely used in Figure 1. Schematic diagram of the positions of PVC casings in test piles
Thailand to rectify the soft toe problem. with typical soil profile at the site.
Pile base grouting was started in Bangkok in 1985, during the

2 BASE GROUTING IN BANGKOK construction of Rama IX cable stayed bridge (Morrison, 1987).
Here 172 bored piles, 2.0 m dia. and 35 m long seated in the 1st
Teparaksa (1991-1992) reported the construction and sand were base grouted. Tube-a-manchette technique was used
performance of base-grouted bored piles in Bangkok subsoils. It with grout volumes up to 4000 liters and maximum pressures of
was found that base grouting not only helps to improve the end 60 bar. With the passage of time engineers started applying this
bearing capacity but skin friction as well (ref: Table 1). technique for the foundations of expressways, high-rise
buildings and other heavy structures. In the absence of standard
guidelines, the criteria and the controlling parameters changed
Table 1. Improvements made by base grouting (Teparaksa, 1994). with time. In current practice, pressures ranging from 20 to 60
Layer of Increase Increase Increase Displacement at fully bar and grout volumes 500 to 1000 liters are normally applied
Pile tip in in skin in end mobilized skin friction with Injection Rate (IR) varying from 5 to 36 liters/min.
failure friction bearing
load (%) (%) (% of Pile Dia.)
(%) NGB GB
3 GROUTING PROCEDURE
Ist Sand 26-66 24-66 28-61 0.49 – 0.51 1.34-1.35
2nd Clay 27 51 1 0.37 0.74
2nd Sand 12-24 9-27 11-21 1.04 1.59-1.86
Two grouting circuits in the form of U-loops with manchettes
NGB: Non Grouted Base, GB: Grouted Base
placed at a level of 5 to 10 cm above the pile tip arranged in a
53
269
Portland cement Type I grout with w/c ratio equal to 0.55 was
used. Volume of grout was measured from the number of pump
A strokes and also crosschecked against the volume consumed
from the agitator tank. Nominal grout volume used for each pile
was 540 liters but a slight difference shown in Table 2 in actual
volume of grout measured might be due to the difference in IR
used during grouting and at the time of calibration.
All the three piles were grouted with different IR varying
from 3 to 36 liters/min. In each case, after injecting target grout
volume, grout circuit valve was closed to lock the maximum
pressure achieved. Residual pressures two minutes after closing
B the pump were also measured and are given in Table 2.
B After 15 to 20 days of base grouting, diamond bit rotary
drilling with double tube sampler was used to core through
concrete present below the base of the PVC-casing. Right after
reaching the pile tip, Standard Penetration Test (SPT) was
started. Steel liners were used inside the split spoon sampler to
collect soil samples. All the concrete and soil samples retrieved,
carefully examined on site to determine the grout content
present. Additionally, in order to determine the lateral spreading
A: Base of PVC-casing, B: Manchettes of grout into the surrounding soil of pile tip, additional boreholes
Figure 2. PVC casing attached to the rebar-cage, before lowering in to were drilled 30 cm away from the pile periphery. Washboring
the borehole. was used down to approximately 1 m above the pile tip level and
then SPT was employed to collect the samples.
symmetrical fashion are used (Figure 1). Normal practice is to
use PE tubes above the manchettes running along the rebar cage
to the top, but if specification require cross-hole sonic logging, 5 CONCRETE CORES RECOVERED
same sonic logging tubes are used to connect the manchettes.
Normal practice is to start grouting 24 hours after concreting From the concrete cores recovered no trace of grout was found
when high pressure criteria is not to be strictly followed. This is except in case of test pile TP-1 where a 6.5 cm thick grout layer
especially suitable for congested piling plans where subsequent was recovered right below the pile tip as shown in Figure 3.
maneuvering of equipment can damage the circuit tubes. If high
pressure is to be maintained and/or sonic logging is required,
grouting is started after seven to ten days. First the pile toe is
cracked with high-pressure water flow to flush and open the
grouting circuit to make way for forth coming grout and then
cement grouting is started. Grouting is stopped when the target
grout volume or maximum required pressure is achieved.
4 TESTING METHODOLOGY
B
In order to investigate the mechanism of grout spreading beneath
and in the surrounding soil of pile tip and the effect of different A
grouting parameters, an attempt was made to collect the soil
samples from below the tip of the base grouted piles (Anwar,
1997). PVC-casings were placed in the prototype bored piles as
shown in Figure 1 and 2, to allow for the passage of drilling
tools down to the pile base. Three piles, 1.0 m in diameter and
length in the range 50 to 55 m were constructed with the
provision of 7.5 cm diameter PVC-casings at a site in Bangkok A: Grout core, B: PVC-casing bottom cap.
metropolitan area. Unfortunately, only one pile out of this group Figure 3. Grout layer of 6.5 cm thickness recovered from TP-1.
could be drilled because of the failure to lower the drilling tools
inside the PVC-casings. For the second time another group of Below the grout core recovered no traces of grout were found in
three piles having 1.5 m diameter and length 56 - 57 m were the sand samples recovered by SPT. Similarly for the rest of two
constructed. This time diameter of PVC-casing was increased to test piles no traces of grout were seen in the soil samples
10 cm. Positions of the PVC-casings were staggered with recovered through SPT.
respect to the manchettes as depicted in Figure 1. All the three
piles were base grouted following the parameters shown in Table
2. 6 DEVELOPMENT OF GROUTING PRESSURE
Table 2. Grouting parameters of the test piles. Development of grout pressure with the increase in grout volume
Test Grout Injection Maximum Residual is shown in Figure 4. TP-1 was grouted with a high IR of 32
Pile Volume Rate Pressure Pressure 2 min liters/min for the first 140 liters and then IR reduced to 3
(Liters) (Liters/min) Achieved after closing the liters/min for the rest of the grouting operation. It must be noted
(bar) pump
(bar)
that as the IR is reduced, pressure dropped from 26 to 6 bar and
then increased gradually to 16 bar at the end of the grouting.
TP-1 500 32 – 3* 26 – 16* 14
TP-2 545 36 36 3
Two minutes after closing the pump, pressure dropped to 14
TP-3 520 8 24 8
bars. TP-2 was grouted at a constant and very high IR of 36
liters/min, which corresponds to the maximum available speed of
* in two phases, refer Figure 4 for clarification. the grout pump. It must also be noted that pressure development
54
270
in this case was very quick as it jumped to 15 bars within first SAND SILT CLAY
100
few strokes of grout and then rose to 36 bar at about 200 liters.
90
No further increase in pressure was observed till the target Type III MC-300
MC-500
volume of grout had been pumped. Pressure abruptly dropped to 80
about 3 bar two minutes after closing the pump. TP-3 was 70
Percent Finer (WT)

Portland Type I
grouted with a constant IR of 8 liters/min. Development of 60
pressure in this case was quite gradual and it approached 50
maximum pressure of 24 bar and dropped to 8 bar 2 min after 40
closing the pump. D15 - Range of

Sand Samples
30
20
10
40 40 0
1 0.1 0.01 0.001
Residual pressure after 2 min.

Grain Size ( mm)
Injection Rate (Liters/min)

30 30
Figure 5. Grain size distribution curves of samples collected from
Pressure (bar)
Pressure
Injection Rate
TP-1 Second Sand Layer around the pile tip.
20 TP-2 20
TP-3
In order to assess the possibility of permeation grouting of the

10 10
2nd sand layer present around the pile tip, Groutability Ratio
(GR) of the sand layers was determined. GR is a useful
0 0
parameter for checking the applicability of grout for use in sand
0 100 200 300 400 500 600
Volume of Grout (Liters)
and is given by the formula as follows (Mitchell, 1970):
Figure 4. Curves showing gradual increase in grout pressure. GR = D15/D95 (1.1)
Where D15 = The particle diameter of the soil to be grouted 15%

of which is finer by weight; and D95 = the particle diameter of
7 GROUT INJECTION RATE the grout 95% of which is finer by weight.
Weaver (1991) summarizes the possibility of grouting a soil
for GR ranges as:
GR > 24, usually;
7.1 High Injection Rate GR < 19, not likely; and
It has been observed that IR is a key factor in the process of GR < 11, not possible.
base grouting. If IR is kept very high, grout pressure rises GR range for the 2nd sand layer is found to be 1.2 to 4 for the
abruptly but can not be smoothly maintained. A clear fluctuation Portland cement type I which means that its not possible for the
in pressure is observed during grouting with high IR. Although cement grout to permeate into the sand layers. GR using
the apparent maximum pressure indicated by the dial gauge is microfine cements MC-300 (Henn, 1996) shown in Figure 5
high at the initial stages but can not be increased further. It comes out to be 12 and 30, so the silty sand layer encountered
seems that at high IR, hydraulic fracturing of the surrounding can be permeation grouted using microfine cements.
soil took place which is also evidenced by sharp fluctuations in
pressure during grouting Similar phenomenon have been
observed by Yeats, 1989. It must also be noted that maximum 9 INSTALLATION EFFECTS
pressure achieved in this case dissipates very quickly as the
grouting operation is terminated. This means that pressure SPT-N values obtained right below the pile tips and 30 cm away
developed in this case mainly consists of the circuit resistance from the pile are compared with the values from nearest
and is not the true pressure induced at the pile/soil interface. borehole (10 – 20 m) performed at the time of soil investigation
Since pressure ceases to increase beyond 400 liters it seems that in Figure 6. It is quite clear that the density of the sand
no further increase in pressure is possible even if large volumes surrounding the tip have been increased appreciably, especially,
of grout have been injected. below the tip of the piles.
7.2 Low Injection Rates SPT-N (Blows/30cm)
If IR is kept low, development in pressure took place quite 25 50 75 100 25 50 75 100 25 50 75 100
52
gradually and the pressure fluctuation is minimal. It must also be TP-1 TP-2 TP-3
noted that the pressures developement by keeping IR low is quite
steady and can also be maintained for certain duration of time, as 53
shown in Figure 4 for TP-1 where a residual pressure of 14 bars
has been observed two minutes after closing the pump with a 54
total drop of only 2 bar. It must also be noted that pressure keeps
on increasing gradually and higher pressures than that of TP-2
Depth (m)
Pile Tip
can be achieved with more grout volumes. 55

Pile Tip
Pile Tip
56
8 PENETRATION OF GROUT INTO THE SAND.
57
From the samples collected, no traces of grout have been
observed in the sand present below or surrounding the pile tip.
Grain size distribution curves of sand samples collected are 58
shown in Figure 5. It can be seen that the second sand layer falls Under the pile tip 30cm away from pile Nearest borehole
in the range of medium to fine sand range. Grain size
distribution for different types of available cements is also Figure 6. Comparison of SPT-N values around the grouted pile tip and
shown in Figure 5. the nearest borehole from soil investigation.
55
271
10 MANCHETTE BREAKING MODELS
Thickness of concrete cover present below the manchette is very

important and determines the flow path of the grout injected. If
concrete cover present below the manchette is thicker than the
side cover to the outlet points on manchette, it can be suspected
that grout directly percolates toward the pile shaft by bursting
the side cover. In order to monitor the bursting mechanism of
manchettes, 9 model piles 1.2 m diameter and 2.0 m deep were
constructed by varying manchette levels above the pile tip A
(Thasnanipan et al, 1998). Model piles were water burst in a
similar way comparable to the normal bored piles. Colored
grout was injected to differentiate it from normal pile concrete.
Suitable time was given to allow for the grout setting before
these piles were exhumed to determine the flow pattern of the
injected grout. Some of the unearthed model piles are shown in
Figure 7 and 8. A: colored grout zone
Figure 8. Model pile showing the grouted zone below the tip.
Low IR of the order of 3–5 liters/min are recommended for

base grouted piles in silty sand layers of Bangkok.
3. Significant improvement in the density of sand around the
pile shaft near the base and especially, below the grouted
pile tip has been observed. Although this is also
contributed by increased overburden pressure due to
concrete, base grouting helps to improve the soil around the
pile tip, disturbed during the drilling process.
4. Position of manchettes above the pile base is very important
to effectively grout the base of the pile. Care must be
exercised to place the manchettes as close to the pile tip as
possible but if due to some unforeseen reasons manchettes
are lifted up grout can be conjectured to percolate towards
the pile base if the concrete cover thickness below the
manchettes does not exceed 30 cm.
Figure 7. Typical model pile used to determine manchette breaking
mechanism.
REFERENCES
From the analysis of the model piles, it was revealed that: a) full
Anwar, M.A. 1997, base grouting to improve the performance of wet
pile base area coverage by grout can be achieved if manchette process bored piles in Bangkok sub-soils, M. Engg. Thesis, AIT,
level is less than 20 cm above the pile tip; b) a partial base area Bangkok, Thailand.
coverage was found up to 30 cm and if; c) manchette level is Francescon, M. & Solera, S.A. 1992, Base grouting to improve
more than 40 cm above the tip, grout directly percolates towards performance of piles bored under bentonite in the Thanet Sands, Proc.
the pile shaft by bursting the side concrete cover. Presence of of the Conf. on Grouting in the ground, ICE, London.
grout on the vertical pile/soil interface at the shaft was also Henn, R.W. 1996, Practical guide to grouting of underground structures,
confirmed in all cases. Although a direct comparison of grout ASCE publication, NY.
flow in model piles and prototype piles can not be made due to Mitchell, J.K. 1970, In-place treatment of foundation soils, Journal of the
the difference in overburden pressures but the upward flow of Soil Mechanics and Foundation Division, ASCE, 96(1), 73-110.
Morrison, I.M. 1987, Bored piled foundation for Chao Phya river
grout along the pile shaft can be the most likely phenomenon in th
crossing at Wat Sai, Bangkok, 9 SEAGS Conf. Bangkok, pp. 6-207
base grouting. Which, subsequently act as a ‘rock socket’ and to 6-218.
significantly increases the pile stiffness (Francescon, 1992). Teparaksa, W. 1991, Performance of base-grouted bored piles in
Bangkok subsoils, First Young Asian Geotechnical Engineers
Conference, AIT, Bangkok, Thailand, pp.117-128.
11 CONCLUSIONS Teparaksa, W. 1992, Behavior of base-grouted bored piles in Bangkok
subsoils, Piling Conference, European Practice and Worldwide
From the study carried out following conclusions can be drawn. Trends, Institution of Civil Engineers, London, pp.296-301.
Teparaksa, W. 1994, Newly developed toe-grouted bored piles in soft
Bangkok clay: performance and behavior, Proc. of the Intl. Conf. on
1. Most of the grout injected during base grouting rises up Design and Construction of Deep Foundation, FHWA, Orlando,
along the pile/soil interface, and does not permeate in to the USA, pp.1337-1351.
surrounding sand layer to any degree. So the improvement Thasnanipan, N. et al, 1998, Design, Construction and behavior of bored
made by base grouting is mainly contributed by increase in th
cast in-situ concrete piles in Bangkok subsoil, Proc. 4 Intl. Conf. on
skin friction in the form of rock-socket effect which Case Histories in Geotechnical Engineering, St. Louis, Missouri,
increases the pile stiffness significantly. Since the increase USA, paper no.1.20.
in skin friction depends only upon the volume of grout Thasnanipan, N. & Umesh, S. 1998, Pile base cracking models with
injected, pressures achieved during grouting are of different manchette levels from the pile tip, SEAFCO internal report.
Weaver, K.D. 1991, Dam Foundation Grouting, American Society of
negligible importance in this regard. Partial replacement
Civil Engineers, NY.
and/or precompression of the sediments below the pile tip is Yeats, J.A. & O’Riordan, N.J. 1989, The design and construction of large
possible by following low IR hence, smoothly increasing diameter base grouted piles in Thanet Sand at Blackwall Yard,
and maintaining the pressures achieved, if it is desired so. London, Proc. Intl. Conf. on Piling and Deep Foundations, London,
2. IR is the key factor in the base grouting process, especially Vol. 1, pp. 455-461, Balkema.
if, higher pressures are to be achieved and maintained.
56
272
Under the Royal Patronage of His Majesty King Bhumibol Adulyadej
Sixth Cycle Celebration of

His Majesty the King of Thailand and
th
40 Anniversary of the Asian Institute of Technology
Civil and Environmental Engineering Conference

New Frontiers and Challenges
8-12 November 1999
Bangkok, Thailand
Conference Proceedings
Volume 2 (Part II)
Editors:
D. G. Lin
D. T. Bergado
N. Phien-wej
P. Nutalaya
A. S. Balasubramaniam
Review of the shaft capacity degradation of bored piles

constructed with bentonite slurry
N. Thasnanipan, M.A. Anwar and Maung A.W.
57
New Frontiers and Challenges, 8-12 November 1999, Bangkok, Thailand
REVIEW OF THE SHAFT CAPACITY DEGRADATION OF BORED PILES

CONSTRUCTED WITH BENTONITE SLURRY
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung

ABSTRACT
Over the last few decades, different researchers have studied the influence of different
construction parameters on the shaft load transfer of bored piles constructed with slurries.
Still the issue is not fully resolved, different parties involved in the bored piling industry
believe on the degree to which these parameters effect the perimeter load transfer. This paper
attempts to summarize all the significant works previously done on this subject. It is
concluded that the realistic formulation of specifications and selection of design parameters
with the consideration of practical constraints at site is very important. Performance of the
bored piles is effected by installation procedures, all the ill effects attributed to the
construction parameters can be eliminated by following good engineering practices.
INTRODUCTION
It is commonly accepted that the construction parameters like slurry properties,

construction time, equipment used, etc., effect the performance of wet process bored cast in-
situ piles. But the extent to which these parameters effect the pile capacity is not clearly
understood due to the wide variation in the soil conditions and other influencing factors which
are simultaneously acting. It has also been recommended in some codes of practice that
observing good engineering practices can eliminate all the foreseeable ill effects attributed to
these parameters.
CONSTRUCTION PARAMETERS
Among the other various parameters, construction time and slurry properties are
particularly considered responsible for the degradation of the pile shaft capacity. In fact the
total construction time is not directly related to the pile performance. Critical time, which
might contribute to the degradation of the pile, starts after the borehole excavation is
completed till the completion of the concrete pouring. Time consumed in the excavation
operation is not generally considered to contribute any degradation since slurry is
continuously kept agitated during drilling operation. A maximum elapsed time of 24 hours
after the completion of the borehole excavation is generally recommended under normal
conditions. Slurry properties especially the viscosity is believed to be another important
factor related to the quality of the constructed pile, because it determines the degree of
replacement by concrete during tremie concreting.
Previous published researches on this issue along with the brief conclusions made are
given in the following in a chronological order.
Chaidesson (1961) observed a better load-settlement behavior for one pile drilled with
slurry than the other pile drilled by a casing method.
V ~ 59
59
Corbett et al. (1970) presented the results of load testing of test panel excavated under
bentonite slurry, excavation took 2 days and on completion the panel was left open overnight,
but bentonite was renewed prior to concrete. The authors inferred that the presence of
bentonite suspension even for such a long time was not adversely affected the development of
skin friction.
Farmer, Buckley & Sliwinski (1970) conducted the laboratory tests on model concrete
pile. The results indicated that presence of a bentonite filter cake at the concrete/sand
interface has little effect on load transfer. Full scale tests on three deep bored piles
constructed under bentonite, indicated that the load transfer was extremely higher than the
expected. They concluded that high load transfer level may be obtained from piles
constructed under bentonite, this may be due to uneven side-wall configuration.
Reese & Tauma (1972) found that the values of load transfer developed in clayey soils
on the sides of the shaft constructed by the slurry displacement method are comparable to
those in the shafts constructed in the dry.
Wates & Knight (1975) concluded from their investigation that a bentonite filtercake of
a significant thickness would develop in 24 hours. In order to prevent a build up of filtercake,
the slurry should be left in place much less than 24 hours.
Cernak (1976) performed full scale load test on three barrettes in sandy gravels. Two of
the barrettes were constructed using slurry with different exposure times (8 and 97 hours) and
their load test results indicated a decrease of skin friction capacity of 43% and 56%
respectively, as compared to the other barrette excavated dry and concreted immediately. It
must be noted that the major part of the reduction i.e. 43% took place in the first 8 hours of
construction time while only 13% (56-43) reduction took place in the rest of 89 hours (97-8)
for the second barrette. These findings support the conclusion of Wates & Knight (1975).
Sliwinski (1977) reported that displacement of suspension of slurry from the sides of the
bored hole does not constitute a major problem. The rising column of concrete will displace
the fluid suspension because of the considerable difference in unit weight and shear strength
of the materials. The author stated that the field and laboratory tests seem to indicate that the
influence of some bentonite in the parent soil has an insignificant influence on load support
providing that the properties of the slurry are within reasonable limit and the concreting is
done within a reasonable short time after the excavation is completed.
Fleming & Sliwinski (1977) reported on 49 filed tests from several countries. The test
results suggest that the use of bentonite suspension has no detrimental effect on shaft friction
while in granular soil, there is an indication that skin friction at high displacement may be
slightly be reduced, though there may be reasons unrelated to the bentonite process. They
suggested that concreting should be completed within 24 hours of the completion of boring.
Reese & Tucker (1985) concluded that the capacity of bored piles constructed under
slurry could be substantial with proper construction technique. The slurry will be ejected from
the excavation and substantial bond will develop at the interface of the concrete and the
supporting soil. However, the authors also suggested that the slurry must not be left more
than few hours in the excavation without operation of drilling tool; otherwise, a thick filter
cake will develop on the side of the excavation.
Reese & O’ Neil (1988) suggested that a solution to the problem of reduced skin
friction due to excessive filter cake is to maintain the properties of the slurry within tolerable
limits and to place the concrete the same day that the excavation is completed. If for some
reason, it is impossible to place the concrete without undue delay, the drilling machine must
be re-occupied the excavation to rework the borehole. If the slurry remains for a period of
time without agitation, the filter cake can become thick.
V ~ 60
60
Fleming et al. (1992) concluded that the process of forming a pile under bentonite
suspension does not materially reduce the shaft friction in both granular soil and cohesive
soil. The rising column of concrete with a slump in excess of 175 mm from tremie process
will largely remove the filter cake layer on the wall surface of shaft. However, if the hole is
left filled with slurry for long period (no agitation action), this will give rise to a significant
reduction in shaft friction on completed piles.
O’Neil & Reese (1992) reported the effect of slurry properties on two drilled shafts
constructed in similar ground conditions (alternating layer of very stiff clay and dense sand).
Shafts 1 and 2 were constructed with Marsh cone viscosity of 37 and 49 sec respectively.
Time between opening the borehole and completion of concreting for both piles was 5 to 7
hours. Minor difference in load transfer of both shafts was found and the authors concluded
that the difference in slurry properties used for the construction for both shafts had a little
effect on shaft load transfer.
Wardle et al. (1992) concluded from four piles tested in London that the axial capacity
of driven and jacked piles after one or two months were 14 to 28% greater than those
measured during or immediately following installation. CRP tests performed after about 3
years recorded further increase of between 14 to 20% (total became 28 to 48%). Although, it
was not possible to record the capacity of the bored pile immediately after installation, the
capacity of this pile increases by 47% between 2 months and three years.
Majano & O’Neil (1993) attempted to model the formation of filter cake in the
laboratory with different slurry dosages, differential pressures and exposure times. They
stated that perimeter load transfer is a complex function of the physical and chemical
characteristics of the slurry and the geomaterials, the roughness of the borehole, the fluid
pressures exerted by concrete, the shearing properties of the soil, and possibly the chemistry
of the fluid concrete. They attempted to correlate the potential degradation of the soil-pile
interface to the thickness and shear strength of the cake formed against the walls, and argued
that the shear strength and thickness can not be measured with only one or two parameters
like slurry dosage and differential pressure. With a bentonite concentration of 72 kg./m3 and
differential pressure of 0.5 psi, they achieved a filter cake thickness of 3.14 and 4.5 mm after
a contact time of 4 and 24 hours. Authors also compared the laboratory tests with the actual
field conditions and proved that due to the presence of enormously high differential pressures
in the field in some cases, thickness of filter cakes which would be considered detrimental for
shaft load transfer, require few hours to develop (Authors proved that a filter cake of 6 mm.
thickness would require only 2.3 hours to develop in the field under a differential pressure of
10 psi).
Hosoi et al. (1994) reported that when the diaphragm wall foundation is cast in slurry,
mud cake could not be fully scoured by concrete. So the evaluation of the shaft friction
mobilized on the wall foundation cannot overlook the effect of mud cake. They concluded
from laboratory tests that the friction resistance between concrete and the sample soil
(decomposed granite) through bentonite slurry cake is smaller than the shear strength of the
sample soil.
Ho & Lim (1998) reported that the barrette sized 2.8 m x 0.60 m x 47.4 m excavated
under bentonite slurry support at Singapore Post Center project with a total construction time
of about 110 hours has been tested. The test result showed that substantial shaft friction could
be mobilized to carry load satisfactory with a reasonable factor of safety and within a
tolerable displacement of 12 mm at working load.
Littlechild & Plumbridge (1998) presented that the values of actual shaft resistances
mobilized for in the five pile load test carried out at BERTS project and compared to the
calculated shaft resistances, where calculated shaft resistance is based on the design approach
adopted on the project. For cohesive soil fs = α Cu, α taken as 0.5 for stiff and hard clay
61
V ~ 61
layers and for sand layers shaft resistance of 120 KPa has been determined for all sand layers.
Construction time and slurry viscosity on this project was controlled to be less tan 24 hours
and 35 sec respectively. They concluded that shaft friction capacity of bored piles constructed
under bentonite slurry tends to decrease as the construction time and slurry viscosity increases.
Thasnanipan, Baskaran & Anwar (1998) presented the shaft friction values mobilized
from eleven instrumented bored piles constructed under slurry with different construction
times and bentonite viscosity from various sites in Bangkok area. They concluded that the
slurry viscosity up to 55 sec./Qt do not have significant effect on the shaft load and
degradation of shaft capacity with increase in construction time beyond 24 hours is not
excessive.
DESIGN CONSIDERATIONS AND PARAMETERS
Selection of design parameters, which suits the local construction practices and
procedures, is very important. Design parameters are normally selected based on the previous
researches of local institutions/organizations along with some special considerations specific
to the sites. Following are the recommendations and suggestions that need to be observed for
the estimation of shaft load capacities in Bangkok subsoils.
Sliwinski and Fleeming (1974) suggested that in clay soil, the bentonite has no effect on
friction and no reduction of the normal factors for deriving adhesion is justified. There may
indeed be a slight increase of adhesion. In granular soils, the bentonite has some effect. The
reduction of friction calculated from effective lateral pressure and coefficient of friction
should be considered in the region 10-30%.
Oonchittikul (1990) back analyzed a parameter β = (Ks.tan φ) from instrumented bored
pile load tests conducted with bentonite slurry in Bangkok subsoils by using the effective
stresses with both the hydrostatic pore pressure and the declined pore pressure. The resulted β
value in the first and second sand layers from effective overburden pressure based on
hydrostatic pore pressure falls in the range of 0.4 to 0.65 and 0.2 to 0.4 respectively while β
value in the first and second sand layers from effective overburden pressure based on declined
pore pressure falls in the range of 0.1 to 0.48 and 0.1 to 0.2 respectively. In both cases β
value in second sand layer is less than the value in the first sand layer. The author inferred
that the overburden pressure might not fully develop at great depth in sand layer as it has been
assumed in the calculations.
Hooley & Brooks (1992) have suggested that if the traditional pile design approach is
used, the proper interpretation must be given to the accuracy of the data employed by soil
investigation as the potential error in the calculation of ultimate base resistance increases
steadily with depth and shaft resistance error increases gradually with depth. The authors
reported that it is noticeable that the amount of scatter in the test result from soil investigation
samples increases with depth and thus the average line is more difficult to define at greater
depths. If the scatter is analyzed statistically, upper and lower bound of average line can be
identified on the basis of chance of error 0.01% probability.
Eide & Bellis (1992) extracted data from 13 numbers of loading tests taken to failure
from the test results of over 109 tests in Bangkok which are presented by Oonchittikul (1990).
The authors found average shaft friction in stiff clay and dense sand in the range 78 to 130
KPa and adopted 100 KPa as design value, which is commonly used for design in Bangkok
Subsoil.
Teparaksa (1992) reported the problem from bentonite sedimentation causes a soft base.
Unexpectedly, low skin friction can be mobilized at failure (design at yield point) in soft base
condition. The author also reported that the static load test results on 1.50 m dia. pile at Pok
62
V ~ 62
Klao bridge project in Bangkok indicates an excessive settlement at 4000 KN load which is
caused by deposition of bentonite at the base of the pile. Analysis of load transfer indicates
that the skin friction of first stiff clay, first sand, and second stiff clay are equal and about 40
KPa at 4000 KN load. After the base grouting with cement-mortar, the analysis shows the
skin friction of this soil is between 40 to 90 KPa at about 5 mm. movement. They are in the
expected range for each soil layer, which is not affected by grouting.
Meiβner et al. (1993) studied the influence on bearing behavior of bored piles when the
pile base is approaching a layer of soft soil layer and concluded that a significant punching
mechanism is obtained for value of t/d =2, where t is the distant from pile base and the clay
layer and, d is the pile diameter. This punching mechanism would decrease significantly both
shaft and base resistance.
Tomlinson (1996) suggested that if pile is excavated by grab under water, there is
considerable loosening of the soil. This causes a marked reduction in both end bearing and
skin friction. Then the design calculation will be calculated on the basis of a low relative
density of the granular soil.
Littlechild et al. (1998) reported that the shaft resistance of both the plain and shaft
grouted bored piles in sand appears to be mostly independent of the vertical effective stress
over the stress range encountered. From the test result of plain bored piles constructed in
accordance with a given specification of BERTS project in Bangkok i.e. bentonite viscosity
<35 sec and construction time of less than 24 hours, the shaft resistances fall in the range 65
to 170 KPa. In addition, they have determined a design line for shaft resistance of plain bored
piles in Bangkok sand of 120 KPa for all ranges of vertical effective stresses.
Thasnanipan et al. (1998) revealed that the end bearing capacity of bored pile where toe
is embedded in to the clay layer would be reduced and this end bearing capacity reduction
might influence the shaft resistance capacity too. This conclusion was reinforced by the pilot
pile load test result in the project located on Rama III road in Bangkok having diameter 1500
mm, toe founded at 60m in clay layer, failed at only about 17500 KN with total settlement of
over 90 mm. While the designed maximum test load was 25000 KN compared to the pile load
test from contract pile of same diameter toe depth founded in sand layer at depth 55 m and
base grouted. Result of test up to 2500 tons with total settlement of only 29.26 mm and elastic
recovery was 17.52 mm recorded.
CODES AND SPECIFICATIONS
Specifications are prepared for the good of the project. Specs drafted by ignoring the
practical site constraints and specific project needs can cause a job to become problem for all
parties. For example, some specs have been found to stressing on maximum allowable
construction time of 24 hours, which sometimes become impossible to adhere to because of
the prevailing traffic congestion in the city centers and local authority regulations in Bangkok,
and ultimately instigate the contractor to breach the law in order to comply with such
unpractical specs. A summary of the slurry specifications used worldwide is given in Table 1.
In the following are given some suggestions related to the excavation under bentonite slurry.
Fleming et al. (1975) suggested that the specifications should set out a cause of action to
perform the work, which is reasonable from the contractor’s point of view and gives a
reasonable expectation of achieving the desired end product. A specification loses
creditability on site if its requirements are outside the range of practical achievement and such
a specification can only be detrimental to the whole job. They further suggested that obscure
requirements will naturally be the first to be forgotten. Thus, for example, it is better to say
“after mixing concrete shall be placed before it lose the specific workability” than to say
“concrete shall be placed within X minute (s) from mixing”.
V 63
~ 63
Lttlechild & Plumbridge (1998) reported that the majority of the piling contractors state
that a maximum construction time for bored pile constructed to depth 50m under bentonite
slurry for BERTS project in Bangkok of less than 24 hours per pile would not be realistic. A
maximum construction time of 24 hours and slurry viscosity of less than 35 seconds (use of
Bentosund bentonite only) were therefore adopted for the project.
Chodorowski & Duffy (1998) presented the typical control and progress charts for 24
hours limit on construction time of bored piles at BERTS project, Bangkok. Charts prepared
by the authors show that during the 24 month period of bored pile construction, there was
noticeable percentage of pile constructed longer than 24 hours of approximately 15% which
in some months as much as 30%. They reported that longer construction time of more than
24 hours per pile was due to difficulties in securing continuous concrete supply as a result of
traffic congestion in the area of different piling sites.
Table 1: Summary of recommended values of slurry properties used for bored piles.
FP Hutchin Hodgso FDOT Rees

Authors S gso
- n
Holden
, e&
ICE ACI AASHTO Majano
&
ICE
n 1987, O' FPS
1975, et al., 1977 1984 1988 1989 1992
Parameter 1977 1975
1988 O'Neil Neil ARUP
1988 1993
1996
Sodium Bentonite Bentonite Bentonite Bentonite
Slurry Type Bentonite Calcium Bentonite or or Bentonite or or Bentonite Bentonite
Bentonite Bentonite
Attapulgite Attapulgite Polymer Attapulgite
Density
(gm/ 1.024 to 1.024 to 1.03 to
1.03 to 1.20 1.03 to 1.20 <1.10
<1.36
1.03 to 1.10 1.02 to 1.07 <1.10
<1.10 1.135 Friction <1.15
1.218 1.20 1.03 to 1.20 1.02 to 1.13
cc) Shaft
8 to 11 8 to 10 9.5 to 10.8
pH 9.5 to 12 <11.7 10.8 to 11.7 8 to 11 8 to 12 8 to 11 9.5 to 12 8 to 12
8 to 11 8 to 10 9.5 to 11.7
<14% <4% <25% <4%
<6% <35% 10% max. <4 % <2%
Sand Content by volume
(by volume) - - (by volume <4% <10% <2%
(by weight) (by weight) (5% max. (by volume) Friction
(by volume)
by weight) Shaft)
Marsh Funnel 28 to 45 32 to 60 30 to 70
- - 28 to 40 28 to 45 30 to 90 26 to 50
Viscosity
(sec/
30 to 90 30 to 40 28 to 45 32 to 60
< 90
Qt)
Plastic (cP
Viscosity <20 <20 3 to 20 <20 <20 - <20 - - 6 to 8.5 -
6 to 10.0
)
Yield(Pa
Point - - - 4.2 to 41.8 - - -
-
- 2 to 6.0 -
2 to 6.5
)
10 min. Gel Strength 4 to 40
(Pa 4 to 40 3.6 to 20 10 to 40 2 to 10 1.9 to 10 - 4 to 40 - - -
4 to 40
)
Differential Head
(m >1 - - 1 to 1.5 - - - 1.5 - - -
)
Fluid Loss - - - - - < 40
- - - - -
(30 minute test) <60
* Maximum Contact - - - - - - - - 4 -
-
Time (hr.)
Note: For two rows in each cell (last three cloumns)
* Without agitation and side-wall cleaning First row: fresh or recycled as supplied to borehole
Second row: prior to concrete pouring
ADSC and DFI drilled shaft inspector’s manual clause 4.5.1, advised that where
bentonite slurry remains in the shaft unagitated for 3 or 4 hours, or more than 24 hours even
though agitated, a mud cake may develop on the shaft walls which can reduce friction.
Unless, the designer has accounted for ‘caking’ and appropriately reduced design friction
value or the specification gives other limitation, portion of the shaft or socket designed for
side friction should either be excavated and concreted within a 24 hours period (and no more
than 3 hours when the slurry is not agitated) or re-reamed or re-scraped to remove the mud
64
V ~ 64
cake prior to concreting, or some method which is acceptable to the geotechnical engineer and
the inspector used to demonstrate that detrimental mud caking has not occurred.
ADSC Specification Clause 2.3.5.2 (g) specified that the drilled pier constructed under
bentonite slurry should be concreted and completed the same day that of excavation. If this is
not possible, the excavation shall be re-drilled, cleaned and slurry tested before concreting.
Shaft friction could be mobilized to carry load satisfactorily with a reasonable factor of safety
and within a tolerable displacement of 12 mm at working load.
FHWA guide for drilled shaft specification, stipulates that, that drilling slurry is an
effective method of stabilizing drilled shaft excavation until either a casing has been installed
or concrete has been placed. Primary concerns connected to slurry use are: the shape of the
borehole would be maintained during the excavation and concrete placement; the slurry does
not weaken the bond between the concrete and both the natural soil and rebar; all of the slurry
is displaced from the borehole by the rising column of fresh concrete; and any sediment
carried by the slurry is not deposited in the borehole. The engineer’s concerns regarding the
behavior and effectiveness of slurry project, can be satisfied by appropriate specification
requirements. These requirements include: specifying a suitable range of slurry properties
both prior to and during excavation and prior to concreting; performing slurry inspection test;
and construction of pre-production trial shaft by the slurry method.
FULL SCALE TEST TO ASSESS THE EFFECT OF CONSTRUCTION PARAMETERS
Since the degree of effect of different parameters greatly depends upon the soil
properties encountered, conclusions made for one type of soil are not directly applicable to
other locations with different geology. Static, instrumented load test program on bored and
barrette pile conducted at BECM tower project Rama IX road, Bangkok, provided the unique
chance to determine the effects of construction parameters on the shaft capacity (Thasnanipan
et al., 1999). Bored pile 1.5m diameter and a barrette 1.5x3.0m in cross-section had same
lengths of 57.5m founded in the second sand layer and located 30m apart having similar soil
embedment conditions. Properties of the bentonite slurry used are given in Table 2.
Table 2. Comparison of bentonite slurry properties for test pile and barrette.
Pile Barrette
Before feeding After Recycling & Before feeding After Recycling &
to the borehole Before concreting to the trench Before concreting
Properties (near borehole base) (near trench base)
Density (g/cc) 1.08 1.10 1.10 1.17
pH value 8 8 8 9
Sand Content (%) 0.1 0.8 1.0 1.1
Total construction time for the bored pile and barrette was 27 and 75 hours,
respectively. However, the total elapsed time between the final excavation (re-drilling &
cleaning) and concreting for the pile and barrette was 16.25 and 20 hours, respectively. In
spite of considerable difference in the total construction time and slurry viscosity, values of
unit shaft friction mobilized in different soil layers did not show any difference in the shaft
load transfer characteristics. This proves that the total construction time is not a crucial
parameter, which effects the shaft load capacity if good engineering practices like re-drilling
the borehole is applied.
V ~ 65 65
CONCLUSIONS
Conclusions of various researchers can be summarized as follows:
• Bentonite viscosity up to the normally recommended range i.e. 60 sec., given in Table 1
does not have any significant effect on the skin friction capacity of bored piles. In
authors’ point of view, filter loss is the more relevant property of the slurry, which
directly effects the formation of the filter cake and finally results in the shaft capacity
reduction rather than the viscosity.
• Selection of pile design parameters with due consideration of site constraints is equally
important. Adequate strength reduction factors based on the local experiences suggested
by local research institutions need to be considered.
• Concreting should be completed within 24 hours after the finishing excavation of
borehole. If due to some unavoidable reasons borehole is left filled with slurry beyond 24
hours with out agitation, re-drilling must be carried out to scarp off any filter-cake present
on the walls of the borehole.
• Performance of bored piles, to a large extent, depends upon the installation procedures
and, the skill of the operating crew, if proper construction procedures have been followed
all the negative effects can satisfactorily be eliminated.
• It must be emphasized that the specifications for each project must be tailored to suit the
actual site conditions keeping in view the constraints like possible delays in concrete
supply due traffic congestion, local authority regulations etc. Among the available control
limits for bentonite slurry shown in Table 1, this is authors opinion and experience that
AASHTO specs are practically more suitable for pile excavation works in Bangkok
subsoils.
REFERENCES
ADSC and DFI (1989), Drilled Shaft Inspector’s Manual prepared by The Joint Caisson-
Drilled Shaft Committee of the ADSC: The International Association of Foundation
Drilling and DFI: Deep Foundations Institute.
ADSC Standards and Specifications, Foundation Drilling Industry. (1995). Incorporations.
ACI 336.1 94.
American Concrete Institute (1989), Standard Specifications for the Construction of Drilled
Piers, ACI Structural Journal 89, 1, pp. 94-103.
Cernak, B. (1976), The Time Effect of Suspension on the Behavior of Piers, 6th European
Conf. on Soil Mechanics and Foundation Engineer, Vienna.
Chaidesson, R. (1961), Influence of the Boring Method on the Behavior of Cast-In-Place
Bored Piles (in French with English Summary). Proc. 5th Int. Conf. Soil Mech. Paris.
Chodorowsi A.R. & Duffy, M.R. (1998), Quality Assurance and Testing on a contract for
16,000 Large Diameter Bored Cast In-Situ Piles. 7th Int. Conf. and Exhibition on Piling
and Deep Foundations, Vienna, Austria.
Corbett, B.O., Davies, R.V. & Langford, A.D. (1970), A Load Bearing Diaphragm Wall at
Kensington and Chelsea Town Hall, London. Proc. of the Conf. on Diaphragm Wall &
Anchorage organized by ICE, London.
Eide, O. & Bellis, J.A. (1992), Foundation of Bridges in Bangkok. Int. Conf. on Deep
Foundation. London.
V ~ 66
66
Farmer, I.W., Buckley, P.J.C. & Sliwinski, Z. (1970), The Effect of Bentonite on the Skin
Friction of Cast-in-Place Piles. Proc. of the Conference on Behavior of Piles. organized
by ICE London.
Federation of Pilling Specialists (1975), Specifications for Cast-In-Place Piles Formed under
Bentonite Suspension, Ground Engineering, 8, No. 2, 50 pp.
Fleming, W.G.K. & Sliwinski, Z.J. (1977), The Use and Influence of Bentonite in Bored Pile
Construction, CIRIA REPORT PG3.
Fleming, W.G.K., Fuchsberger, M., Kipps, O. & Sliwinski, Z. (1975), Diaphragm Wall
Specification, Proc. of the Conference on Diaphragm Wall & Anchorage organized by
ICE, London.
Fleming, W.G.K., Weltman, A.J., Randolph, M.F. & Elson, W.K. (1992), Piling Engineering
(2nd edition). John Willy & Sons.
Florida Department of Transportation (1988), FDOT Standard Specification for Road and
Bridge Construction, Section 455: Foundations of Structures, Tallahassee, FL.
Guide Drilled Shaft Specification, U.S. Department of Transportation, Federal. Highway
Administration. Geo-Technical Engineering Note Book. Geotechnical Guideline no. 14.
Distributed by ADSC, Texas, USA.
Ho, C.E. & Lim, C.H. (1998), Barrettes Designed as Friction Foundations, Proc. of Fourth Int.
Conf. on Case Histories in Geotechnical Engineering. St. Louise, Missouri.
Hodgson, F.T. (1977), Design and Control of Bentonite/Clay Suspensions and Concrete in
Diaphragm Wall Construction, Seminar on A Review of Diaphragm Walls, ICE,
London.
Holden, J.C. (1984), Construction of Bored Piles in Weathered Rocks, Part 4: Bentonite
Construction Procedures, Technical Report No. 69, Road Construction Authority of
Victoria, Australia.
Hooley, P. & Brooks, S.R. (1992), Site Investigation of Bored Piling in London Clay. Piling
European Practice and Worldwide trends. Proc. of a Conf. organized by the ICE,
London.
Hosoi, T., Yaki, N., & Enoki, M. (1994), Consideration to the Skin Friction of Diaphragm
Wall Foundation. 3rd Int. Conf. on Deep Foundation Practice incorporating Piletalk
1994, Singapore.
Hutchinson, M.T., Daw, G.P., Shotton, P.G., & James, A.N. (1974), The Properties of
Bentonite Slurries Used in Diaphragm Walling and Their Control, Proc. of Conf. on
Diaphragm Walls and Anchorages, ICE, London.
Littlechild, B. & Plumbridge G. (1998), Effects of Construction Technique on the Behavior of
Plain Bored Cast in Situ Piles Constructed under Drilling Slurry, 7th International Conf.
and Exhibition on Piling and Deep Foundations Vienna, Austria.
Littlechild, B.D., Plumbridge , G.D. & Free, M.W. (1998), Shaft Grouted pile in sand and
clay in Bangkok .7th Int. Conf. and Exhibition on Piling and Deep Foundations, Vienna,
Austria.
Majano, R.E. & O’Neil, M.W. (1993), Effect of Mineral and Polymer Slurry on Parameter
Load Transfer in Drilled Shaft. A Report to ADSC; by University of Houston, USA.
V ~ 67
67
Meibner, H., Shen, Y.L., Van Impe, W.F & Vogt, C. (1993), Punching Effect for Bored Piles.
Proc. of the 2nd Int. Geotechnical Seminar on Deep Foundation on Bored and Auger
Piles; Ghent, Belgium.
O’Neil, M.W. & Reese, L.C. (1992), Behavior of Axially Loaded Drilled Shafts in Beaumont
Clay. Report No. 89-8, The University of Texas at Austin. U.S.A.
Oonchittikul, S. (1990), Performance of Bored Piles in Bangkok Subsoils. M. Engg Thesis,
V ~ Bangkok,
Asian Institute of Technology (AIT), 67 Thailand.
Reese, L.C. & Tauma, F.T. (1972), Load Test of Instrumented Drilled Shafts Constructed by
the Slurry Displacement Method. Center for Highway Research. The University of
Texas at Austin. U.S.A.
Reese, L.C. & Tucker, K.L. (1985), Bentonite Slurry in Construction Drill Piers. Proc. of a
Session sponsored by the Geotechnical Engineering Division of the American Society
of Civil Engineers in Conjunction with the ASCE Convention, Denver, Colorado.
Reese, L.C. & O’Neil, M.W. (1988), Drilled Shafts: Construction Procedure and Design
Methods prepared for U.S. Department of Transportation, Federal Highway
Administration office of Implementation, Mc.Lean, Virginia. in cooperation with
ADSC: The International of Foundation Drilling. Dallas, Texas.
Sliwinski, Z. & Fleming, W.G.K. (1974), Practical Consideration Affecting the Performance
of Diaphragm Walls, Proc. of the Conf. on Diaphragm Wall and Anchorages, ICE,
London.
Sliwinski, Z.J. (1977), The Effect of the Bentonite Displacement Method of Construction on
the bearing Capacity of piles and Diaphragm Wall Elements. Notes for the meeting of
the concrete Society. Croydon, October 12, 1977, (Unpublished.).
Teparaksa, W. (1992), Behavior of Base Grouted bored piles in Bangkok Subsoils. Piling
European Practice and Worldwide Trends. Proc. of the conf. organized by the ICE,
London.
Thasnanipan, N., Baskaran G. & Anwar, M.A. (1998), Effect of Construction Time and
Bentonite Viscosity on Shaft Capacity of Bored Piles, Proc. of the 3rd Int. Geotechnical
Seminar on Deep Foundations on Bored and Auger Piles, Ghent, Belgium.
Thasnanipan, N., Maung, A.W. & Tanseng, P. (1998), Barrettes Foundation in Bangkok
Subsoils, Construction and Performance. Thirteenth Southeast Asian Geotechnical
Conference, Taipei, Taiwan.
Thasnanipan, N., Teparaksa, W., Maung, A.W. & Baskaran, G. (1998), Design, Construction
and Behavior of Bored Cast In-Situ Concrete Piles in Bangkok Sub Soil. Proc. of
Fourth Int. Conf. on Case Histories in Geotechnic Engineering, St. Louise , Missouri.
Tomlinson, M.J. (1996). Pile Design and Construction Practices. (4th Edition).
Wardle, I.F., Price, G., & Freeman J., (1992), Effect of Time and Maintained Load on the
Ultimate Capacity of Pile in Stiff Clay. Piling European Practice and Worldwide trends.
Proc. of the Conf. Organized by the ICE, London.
Wates, J.A. & Knight, K. (1975), The Effect of Bentonite in the Skin Friction in Cast in-Place
Piles and Diaphragm Walls. Proc. of Sixth Regional Conf. for Africa on Soil Mechanics
and Foundation Engineering, Dorhan, South Africa.
V ~ 68
68

th

8-12 November 1999
Bangkok, Thailand
Editors:
W. Kanok-Nukulchai
C. Polprasert
P. Nutalaya
A. S. Balasubramanian
Failure mechanism of long bored piles

in layered soils of Bangkok
N. Thasnanipan, M.A. Anwar and Maung A.W.
69
FAILURE MECHANISM OF LONG BORED PILES IN LAYERED SOILS OF

BANGKOK
Narong Thasnanipan, Muhammad Ashfaq Anwar, Aung Win Maung

ABSTRACT
Simple superposition principle for the estimation of ultimate shaft friction capacity of
bored piles sometimes found to be not directly applicable for the piles embedded in the
multilayered soils of Bangkok. Because of the early yielding of the stiff clay layers, a
localized plunging of the load-settlement curve is often observed. For these layers, although
the maximum mobilized skin friction corresponds to the adhesion factor α equal to around
0.35, suggested by different researchers. But due to the movement softening behavior of
these layers, residual friction values near the ultimate failure of the piles drops to a
corresponding α values of the order of 0.15 or even smaller. It is recommended that due to
the progressive nature of failure, lower α values for stiff to hard clay layers should be
considered for the estimation of ultimate skin friction capacity under such conditions.
INTRODUCTION
Ultimate shaft friction capacity of piles embedded in the layered soils is calculated as
summation of the ultimate shear resistances offered by different soil horizons present along
the pile shaft. Load movement curves obtained from instrumented pile load tests have
revealed that the very stiff to hard clay layers present within the Bangkok aquifer sands
frequently exhibit a well defined movement softening behavior. Peak friction values in these
layers have been found to be mobilized at very low pile head movements and well before the
ultimate friction values of other layers present along the pile shaft. When rest of the layers
reach their ultimate values these layers have already reached their residual friction values,
which are as low as 50 percent of the peak friction values. So the simple superposition of
ultimate friction values give rise to overestimation of the total shaft friction capacity under
these conditions.
SOIL PROFILE
Typical soil profile of plain of Bangkok is shown in Figure 1. Thickness of Bangkok

Soft Clay (BSC) at the top varies generally between 15 to 18m. BSC changes to medium stiff
consistency before the first stiff clay, which is generally present at 20 to 25 m depth. First
sand layer is usually 5 to 10 m thick and found at 25 m to 30 m depth, below is a series of stiff
to hard clay and medium to very dense silty sand layers.
PILES EMBEDMENT CONDITIONS
Pile embedment conditions are designed depending upon the actual soil profiles
encountered at the sites, as depicted by three locations P-1, P-2 and P-3 in Figure 1.
Thasnanipan et al. (1998), presented the instrumented pile load test results of wet process
bored piles tested in Bangkok subsoils. The authors have revealed that in case of pile
V ~ 69
71
P-1 P-2 P-3
Depth (m)
Figure 1: Typical soil profile along a 4.5 km long stretch of an expressway in the plain of
Bangkok, with different pile embedment conditions.
embedment conditions like P-1 and P-3, 2nd stiff clay often exhibit a brittle type of failure and
peak skin friction values are mobilized well before the sand layers.
A family of t-z curves obtained from the instrumented pile load tests showing this type
of failure is presented in Figure 2. The peak friction values in these layers are mobilized at
very low pile head movements (10 – 15 mm depending upon the depth of stiff clay layer) and
well before the ultimate friction values of other soil layers present along the pile shaft. When
rest of the soil layers reach their peak friction values stiff clay layers have reached the residual
friction values.
10
9
8
Skin Friction Mobilized (ton/m )
2
7
6
5
4
3
2
1
0
0 200 400 600 800 1000 1200
Butt Load (ton)
Figure 2: Family of curves obtained from instrumented pile load tests showing typical brittle
failure in stiff clay layers.
V ~ 70
72
A typical load test showing this type of behavior is shown in Figure 3. It can be seen
that the stiff clay layers reach their ultimate friction values well before the sand layers and at
the maximum test load stiff clay layers reach their residual friction values. Design parameters
like α and β suggested by different researchers like Suchada (1995) for the estimation of
ultimate skin friction in Bangkok subsoils give a reasonable estimate for the embedment
conditions like at P-2 in Figure 1 or where the thickness of 2nd stiff clay layer is relatively
low. Exceptions have been observed in case of embedment conditions especially like P-1,
where a thick 2nd stiff clay layer is present. Owing to the difference in shaft friction
development with pile head movement in the stiff clay layers as shown in Figure 3, direct
summation of (simple superposition) ultimate shaft friction capacities in different soil
horizons leads to over estimation of shaft capacity of these piles. Due to the brittle nature of
failure mechanism in very stiff to hard clay layers peak friction resistance is mobilized at
comparatively very little movement at the pile/soil interface. Residual values of friction
resistance in these layers drop to 50 percent or even smaller of the peak friction values
mobilized.
Butt Load
0.00
14
Soft Clay Su = 1.5 t/m 2

Soft Clay
10.00 12
Medium Clay
1st stiff Clay

Medium Su = 4.5 t/m 2
Bored Pile
Clay 10 1st Sand

Mobilized Skin Friction (ton/m )
2
2nd Stiff Clay

20.00 1st Stiff
Su = 13.5 t/m 2 2nd Sand
Clay
8
Depth (m)
30.00 1st Sand 6

SPT-N = 35
40.00
2nd Stiff
Clay Su = 25.5 t/m 2 2
50.00 0
0 400 800 1200 1600 2000

2nd Sand SPT-N = 50
Butt Load (ton)
60.00
Figure 3: Skin friction development with the increase in butt load which clearly shows a
typical early yielding of stiff clay layers.
On the extreme, from some instrumented pile load test results it has been observed that
stiff clay layers reach their peak friction values near the design service loads. Rest of the soil
layers need further pile/soil interface movement for the full mobilization of skin friction, and
at this stage, stiff clay layers have already arrived at their residual friction values. Now
depending upon the relative thickness of stiff clay layers, overall shaft friction capacity
mobilization is affected differently. If the thickness of very stiff to hard clay layers is such
that their contribution in the total shaft friction capacity is proportionally high, when stiff clay
layer yields, load settlement curve shows a localized plunging behavior. The movement of
such plunging consists of few millimeters and when the other soil layers shares the deficit in
73
V ~ 71
Table 1: Comparison of maximum skin friction values mobilized and the values at maximum
test load.
Undrained Mobilized Unit Skin Friction Values

Clay Layer Shear (ton/m2)
Strength Max. Mobilized At Max. Test Load
Su (ton/m2) fs α fs α
Soft Clay 1.5 1.7 1.13 1.6 1.07
Medium Clay 4.5 2.2 0.49 1.9 0.42
1st Stiff Clay 13.5 5.8 0.42 5.1 0.38
2nd Stiff Clay 25.5 8.8 0.34 4.2 0.16
peak and residual friction capacity of these layers, once again pile starts taking load followed
by this local plunging, as shown in figure 4, till the final plunge exhibiting the failure of all
soil layers along the pile shaft reaches. It must be noted that this localized plunging was
earlier conjectured to be because of the soft toe of the piles. On the other hand if the
contribution of skin friction derived from very stiff to hard clay layers is proportionally low,
localized plunging due to earlier failure of very stiff to hard clay layer is not visible. Normal
practice to estimate the contribution of skin friction from clay layers is by using α−method,
with a value of α equal to around 0.3 for very stiff to hard clay layers. Instrumented pile load
test results have shown that skin friction values calculated using α value equal to 0.3
corresponds to the peak skin friction in these layers. Owing to the brittle type of failure
mechanism in these layers, α values at full mobilization of total shaft friction capacities drop
to as low as 0.15 in some cases and the ultimate shaft capacity calculated gives over estimates
as compared to the actually mobilized capacities. Skin friction values mobilized in the pile
load test shown in Figure 3 are given in Table 1.
Butt Load (Tons)
0 500 1000 1500 2000
0
20
TP-1
TP-2
40
60
80
100
120
Figure 4: Typical load-settlement curves showing localized plunging before the overall failure
of the pile.
74
V ~ 72
Peak and residual α values mobilized in the stiff clay layers from Table 1 are plotted in
figure 5 along with the suggested curves by different researchers. It can be noted that the
peak α value mobilized in 2nd stiff clay layer corresponds well to the value suggested by Stas
& Kulhawy (1984). But the residual α value at the maximum test load drops below the curve
suggested by Suchada (1989). So the α values proposed in figure 5 for the stiff to hard clay
layers gives overestimate of the ultimate shaft friction under theses conditions.
1.20
Bored Cast Insitu Piles
1.00
Tomlinson,1957
0.80
αAdhesion Factor ,
Suchada,1989
0.60 for bored piles in Bangkok soils
0.40
Peak
0.20
Residual
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
2
Undrained Shear Strength (ton/m )
Figure 5: Comparison of adhesion factor α , suggested by different researchers with the actual
mobilized in the stiff clay layers.
EFFECT OF LOWER END BEARING VALUES ON SKIN FRICTION
Meibner & Van Impe (1993) firstly reported the influence of lower end bearing values
on the shaft friction capacity of piles. Laboratory model tests as well as finite element
analysis performed by the authors showed that the presence of soft soil within a depth of two
to four times pile diameter below the pile tip, not only decrease the end bearing but also the
skin friction capacity. Thasnanipan et al. (1998) also revealed that the end bearing capacity of
bored pile where toe is embedded in to the clay layer would be reduced and this end bearing
capacity reduction might influence the shaft resistance capacity too. This conclusion was
reinforced by the pilot pile load test result in the project located on Rama III road in Bangkok
where bore pile having diameter 1500 mm, toe founded at 60m in clay layer, failed at only
about 1750 tons with total settlement of over 90 mm. While the designed maximum test load
was 2500 tons. An other contract pile of same diameter and founded in sand layer at depth 55
V ~7573
m and base grouted was successfully tested up to 2500 tons with total settlement of only
29.26 mm and elastic recovery was 17.52 mm recorded. Even though the second pile was
shorter in length than the first, more skin friction was mobilized due to the improved end
bearing conditions. A similar decrease in the shaft friction capacity in the sand layers has
been observed from the instrumented pile load tests at another project in Bangkok.
CONCLUSIONS
For the piles embedded in the multi layered soils of Bangkok, estimation of ultimate shaft
friction capacity needs to consider the brittle type of failure mechanism of stiff to hard clay
layers and the α values selected need to be adjusted accordingly. Where the pile tips are
embedded in the clay layer, effect of lower values of end bearing on shaft capacity should also
be considered.
ACKNOWLEDGEMENTS
The authors are grateful to their colleagues, especially Mr. Pornpot Tanseng for their
assistance in the preparation of this paper.
REFERENCES
Meibner, H., Shen, Y.L., Van Impe, W.F. & Vogt, C. (1993), Punching Effect for Bored
Piles. Proc. of the 2nd Int. Geotechnical Seminar on Deep Foundation on Bored and
Auger Piles; Ghent, Belgium.
Stas, C.V. & Kulhawy, F.H. (1984), Critical Evaluation of Design Methods for Foundations
Under Axial Uplift and Compression Loading, Report EL-3771, EPRI, Palo Alto, Calif.
Suchada (1989), Bored Piles in Bangkok Subsoils, M. Engg. Thesis, AIT, Bangkok.
Thasnanipan, N., Baskaran G. & Anwar, M.A. (1998), Large diameter bored piles in
multilayered soils of Bangkok, Proc. of the 3rd Intl. Geotechnical Seminar on Deep
Foundations on Bored and Auger Piles, Ghent, Belgium, Balkema.
Thasnanipan, N., Teparaksa, W., Maung, A.W. & Baskaran, G. (1998), Design, Construction
and Behavior of Bored Cast In-Situ Concrete Piles in Bangkok Sub Soil. Proc. of Fourth
Intl. Conf. on Case Histories in Geotechnic Engineering, St. Louise, Missouri.
Tomlinson, M. J. (1957), The adhesion of Clay Piles Driven in Clay Soils, Proc. 4th Intl.
Conf. on Soil Mechanics and Foundation Engineering, London.
76
V ~ 74
GeoEng2000
An International Conference on
Geotechnical & Geological Engineering
19-24 November 2000

Melbourne Exhibition and Convention Centre
Melbourne, Australia
Barrettes : A versatile foundation for transmission line towers

Narong Thasnanipan, Pornpot Tanseng, Aung W. Maung and Muhammad A. Anwar
ISBN No. 1-58716-068-4
77
GeoEng2000 An International Conference on Geotechnical and Geological Engineering 19-24 November 2000, Melbourne, Australia
BARRETTES; A VERSATILE FOUNDATION FOR

TRANSMISSION LINE TOWERS
Narong Thasnanipan1, Pornpot Tanseng2, Aung W. Maung3 and Muhammad A.

Anwar4
ABSTRACT
An efficient design of foundations for transmission line towers has always been a challenge for the
engineers due to the variety and cyclic nature of the loads. Foundations, especially for the four-legged
towers, are subjected to all types of loads (compression, tension, torsion and shear) in different combinations.
The cyclic nature of the loads further complicates the situation. Available design parameters proposed by
different researchers are also mostly based on the monotonic loading conditions and are not directly
applicable for tower foundations. This paper presents the analysis, design and construction practice of the
barrettes used for the transmission line towers in Thailand. Adaptability of the barrettes under different site
constraints like sensitive underground pipelines is also described.
INTRODUCTION
The 230kV-transmission line project presented in this paper involved construction of 40 transmission
towers, 400m apart. The project site was located about 200km, east of Bangkok, Thailand. The transmission
line was planned to convey electricity from the power plant to a sub-station. The alignment of 15.6km-long
transmission line partially runs along the boundary of an Industrial Estate and thus the foundations of these
towers are in close vicinity of existing underground gas pipelines and utilities above ground (Fig.1).
Foundation layouts, construction method and the work area were thus restricted by presence of these utilities.
LEGEND :
LATTICE TYPE TOWER
Initially, the structural designer proposed
large diameter bored piles of 2.0m in
ILAND
MONOPOLE TYPE TOWER

TRANMISSION LINE
POWER
diameter for tower foundations.
GULF OF THA
GENERATION GAS PIPELINE
Construction of such large bored piles

PLANT
would require heavy equipment and plant,

CHEMICAL
PLANT and would induce unacceptable vibration
level to the existing underground gas pipe
CHEMICAL CHEMICAL
PLANT PLANT
CHEMICAL
lines located within and along the

PLANT
GAS
CHEMICAL
SEPARATION
construction area. As an alternative

CHEMICAL PLANT
PLANT
PLANT
construction method and foundation type

to suit the site conditions, barrette
construction that utilizes relatively less
equipment and induces less vibration than
0km 1.5km 2.5km 5km SUB STATION
conventional bored pile construction was

adopted.
Figure 1. Layout of project site
1
Managing Director, Seafco Co., Ltd.. Bangkok, Thailand
2
Geotechnical Engineer, Seafco Co., Ltd.. Bangkok, Thailand
3
Project Manager, Seafco Co., Ltd.. Bangkok, Thailand
4
Project Manager, Middle East Foundations Group, Dubai, UAE
79
Construction of barrettes in a site with limited headroom and constraints and performance of barettes in
Bangkok soil has been reported by Thasnanipan et. al. (1998) and (1999). The past experience indicated the
feasibility of this project.
SOIL CONDITIONS
Initially, only a few boreholes were made to investigate subsoil conditions. However as the great distance
between towers meant that substantial variations in soil conditions could occur. An extensive soil
investigation was later carried out at all tower locations by drilling to verify subsoil condition to optimize
barrette design and for excavation feasibility. The general subsoil profile revealed fine to coarse sand at the
top, overlaying silty clay (lateritic soil) with varying thickness. The lateritic soil was not present at some
locations. Below the sand and lateritic soil was completely decomposed granite at varying depths. Figure 2
shows cross sections of subsoil at tower locations along the transmission line. The groundwater level found
also varied from about 0.6m to 9.2m below ground surface.
13/A
No.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
BH
7
8
0.00
5.00
10.00
DEPTH (M)
15.00
20.00
25.00
LEGEND
30.00
CLAYEY SAND
SANDY CLAY (LATERITIC SOIL)
SANDY CLAY (COMPLETELY DECOMPOSED GRANITE)
SILTY SAND (COMPLETELY DECOMPOSED GRANITE)
Figure 2. Soil profile along the transmission line

BH-3 BH-11 BH-20 BH-32 BH-39
SPT-N SPT-N SPT-N SPT-N SPT-N
0 50 100 0 50 100 0 50 100 0 50 100 0 50 100
0.00
5.00
10.00
DEPTH (M)
15.00
20.00
25.00
30.00
Figure 3. SPT test results at some tower locations.
SPT N values of soil layers from different boreholes indicated that the depth of competent layer varied
significantly (Fig. 3). This meant that penetration depth of barrettes would vary with soil conditions and
barrettes need to be designed individually.
80
BARRETTE DESIGN
The design of barrettes mainly involved (1) layout arrangement and (2) structural and geotechnical
capacity of barrettes in connection with load, site and soil conditions and construction practicality.
Regarding the layout, a four-barrette group (Fig. 5) was selected to support 38 lattice towers (up to 61.4m
high) individually where a large area was available. Otherwise, a cruciform barrette (Fig. 6) was used for
monopole towers (51.4m high). Compression, uplift and lateral forces were considered the significant loads
to barrettes and they depended on the position of the tower in the line, the natural forces (wind) exerted on
towers and cables and the weight of tower and cables. The towers were generally classified according to
their positions on the line; end towers, edge towers and towers on a straight line between two other towers.
The end towers were longitudinally loaded from one side by the cables and the overturning moment on such
towers was large. The edge towers were loaded both vertically and transversally while the towers on straight
line mainly carried a vertical load. A summary of estimated loads on the individual barrettes of the towers is
presented in Table 1. To achieve load carrying and uplift capacity based on the imposed load and soil
conditions, barrette sizes of 1.0x2.7m and 0.8x2.7m with depths from 11.0m to 22.0m were selected.
Table 1. Summary of Loads on individual barrettes

Tower Classification Compression (kN) Uplift (kN) Horizontal (kN)
End Tower 3792 3078 1087
Edge Tower 3479-3792 2804-3078 898-1087
Tower on straight line 1370-2039 1032-1565 309-505
Remark
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower
1.0x2.7 Edge tower

1.0x2.7 End tower
1.0x2.7 End tower

1.0x2.7 Monopole
1.0x2.7 Monopole
1.0x2.7 Monopole
Tower on a straight line Tower on a straight line

Tower Barrette
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
0.8x2.7
Size
13/A
No.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
1
7
8
0.00
5.00
10.00
DEPTH (M)
15.00
20.00
25.00
30.00
Figure 4. Dimensions and type of barrettes for the towers

15.90
GAS PIPE
15.90
2.47
2.67
LINE
φ 34" BARRETTE
0.80X2.70m 3.48 4.47
A GAS PIPE LINE

- φ 34"
15.90
2.
70
80
0.
SECTION A-A
Figure 5. Typical barrette layout for a transmission tower (Lattice Type)
81
Generally, a safety factor higher than 2.3 and 1.2 was adopted for compression and uplift load capacity
respectively. For uplift capacity, reduced friction capacity (up to 50%) in the cohesionless soil above fixity
level was used in calculation since barrettes were subjected to horizontal cyclic loading. Alexander J.
Verstraeten. (1987) has described decrease of friction by dynamic loading patterns in details. Finite element
method (FEM) was used for calculating bending stesses in the barrettes and prediction of lateral movements.
A total of 6 pile types was classified according to imposed loads, pile dimensions and soil conditions, and
then analysed by FEM. Reinforcement of barrettes was designed using strength design method. A concrete
cover 75mm was used for all barrette types.
The lattice towers were anchored in the barrettes by stubs and cleats which were cast into the extended
barrette section after trimming the barrette top. For monopole towers, anchor bolts and steel flanges were
used.
Three monopole towers were used where the lattice type tower could not be allocated. Due to the
presence of underground gas pile lines and the narrow area available, it was difficult to allocate barrette
groups with adequate alignment to resist the maximum bending moment and horizontal force up to
25,840kN-m and 635kN respectively. Cruciform barrettes were then designed to resist such high bending
moments and forces acting on the base of the monopole.
MONOPOLE
TOWER
A 2.00
-
1.00 2.70 3.40 PILE CAP
CRUCIFORM
BARRETTE
1.0X2.70m
1.00
2.70
3.40
SECTION A-A
Figure 6. Cruciform barrette for a monopole type tower
BARRETTE CONSTRUCTION
A mechanical rope grab was used to excavate the trench under bentonite slurry. The same crane was used
for excavation and lifting reinforcement cages as well as construction plant and facilities. A guide wall with
inside clear dimensions slightly more than the nominal size of the barrette was used to guide the grab.
Circulation of slurry was continuously done to keep the bentonite slurry agitated, to minimize the building up
of filter cake on trench wall surfaces. Properties of bentonite slurry were maintained within the specified
ranges in wide use. Construction sequences are shown schematically in Figure 7.
GUIDE WALL
TREMIE PIPE
REINFORCEMENT
CAGE
BENTONITE
SLURRY
EXCAVATION INSTALLATION OF CONCRETING

REINFORCEMENT
Figure 7. Barrette construction sequences (schematic)
82
Figure 8. Guide wall for a cruciform barrette Figure 9. Reinforcement being installed in the trench
Sediment or loose materials at the bottom of the trench were removed and any built-up filter cakes were
scraped off by the grab before reinforcement cage installation. Entire barrette length was fully reinforced
and up to 24m-long reinforcement cages were fabricated in one complete section, for installation into the
trench. The reinforcement cage of the cruciform barrette was as the same shape as the barrette. For the
purpose of lifting and handling the cage, temporary stiffeners, lacing and tie bars were necessary. These
temporary bars were removed section by section while the cage was lowered into the trench.
After excavation, the trench profile was checked with Koden drilling monitoring equipment. Tremie
concreting method was used for casting the barrettes. Since the cutoff level of barrettes was generally at
ground level, ready-mixed concrete was poured until all slurry and slime in the trench was completely
displaced and fresh concrete could be seen.
Up to 172kg/m3 and 381kg/m3 of SD40 steel bars with maximum steel percentage at barrette top section
(about 12.0m) were used to reinforce the group and cruciform barrettes respectively. Ready-mixed concrete
with cube strength of 38MPa at 28days was used for casting barrettes.
CONSTRUCTION PROBLEM AND CORRECTION
A trench collapse was reported during desanding operation after trenching was completed for a barrette
of lattice tower. A sudden fall of bentonite slurry level by a few meters in the trench was observed and then
the trench was found filled with soil up to 10m depth. Mode of failure of the trench was found to be in the
form of a localized cavity failure in the loose sand present at 2.0m to 6.0m depth. Failure surface did not
extend up to the ground surface and the guide wall was still intact without any distortion or damage. The
most probable cause of the trench collapse was due to heavy rainfall during construction. Surface water was
seen in the vicinity of the construction area. The rain had raised the groundwater level to near the ground
surface and caused a groundwater flow reversal towards the trench, triggering cavity formation in the loose
sand stratum. The cavity progressively increased in size, leading to the collapse of the trench during the
desanding operation. The collapse trench was backfilled with cement-bentonite mix, and wooden piles were
driven into the collapse area outside of the barrette outline. The guide wall was then raised up to 1.5m above
ground level. The trench was re-excavated under the slurry head 1.5m above the expected groundwater level
and the barrette was completed.
QUALITY CONTROLS
All trenches were checked for verticality and dimensions prior to reinforcement cage installation to avoid
any obstruction associated with trench inclination. In particular, all sides of the cruciform trench were
checked. If necessary, verticality of trench was improved by careful chiselling with the grab. Profiles of the
trench for a cruciform barrette is shown in Figure 10.
As the subsoil conditions varied from one location to another, they were observed during trenching
process and verified with the soil conditions assumed in design, especially in the lower section of the
barrette. Slurry quality was regularly tested and maintained within the specified ranges. Since barrettes
83
were highly reinforced, in order to achieve a good flow of concrete, the concrete mix was checked for
appropriateness of slump and cohesiveness prior to casting.
Figure 10. Trench profile of a cruciform barrette Figure 11. Test signal showing sound integrity and
recorded with Koden equipment - Tower no. 5 toe reflection of a barrette - Tower no. 2
Sonic integrity/seismic test was carried out on all barrettes about 5 days after casting for checking barrette
integrity. The test indicated that integrity of all barrettes were sound. Figure 11 shows a signal acquired by
sonic integrity test on the cruciform barrette.
DISCUSSION AND CONCLUSION
For barrette piling in loose cohesionless soil, chances of rapid groundwater level rise due to the rain, tide,
etc., need to be considered during trenching. If necessary, guide wall with appropriate height above the
possible highest groundwater level must be used.
Barrettes are versatile and can be used efficiently in the areas where conventional bored piling cannot be
done. Moreover, barrette construction uses less quantity of equipment than bored piling and thus it is
suitable for the piling work requiring shifting from one location to another.
Verticality of trench is critical for barrettes with complex cross-sectional shape, such as cruciform, T and
H for reinforcement cage installation. A thick concrete cover up to 100mm is thus recommended, in
particular for cruciform barrettes for ease of reinforcement cage installation and to maintain the adequate
concrete cover, considering trench verticality against high stiffness of rebar cage.
REFERENCES
Thasananipan N., Singtowkaew K. and Tanseng P. (1999), "Experiences in Construction of Barrette (in
Thai)", National Convention on Civil Engineering, Pattaya, Thailand, May 24-26, 1999. pp. GTE-144-
149.
Thasnanipan, N. Maung, A. W. and Tanseng, P. (1998), "Barrettes Founded in Bangkok Subsoils,
Construction and Performance", Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan,
ROC, November 16-20, 1998. pp. 573-578.
Alexander J. Verstraeten. (1987), "Steel Pile Foundations for Transmission Line Towers, as Used in Western
Europe", Proceeding of the 2nd International Conference of Deep Foundation Institute, Luxembourg, May
4-7, 1987. pp. 297-325.
84
Proceedings of the
GEOTECH – YEAR 2000
Development in Geotechnical Engineering
Editors:
Dennes T. Bergado
Lin Der-Guey
Tian Ho Seah
Kinya Miura
Noppadol Phien-wej
Prinya Nutalaya
Practical installation of stanchions for top-down

construction in Bangkok subsoil
Narong Thasnanipan, Aung Win Maung & Zaw Zaw Aye
27-30 November 2000

Bangkok, Thailand
85
GEOTECH-YEAR 2000
Developments in Geotechnical Engineering
27-30 November 2000, Bangkok, Thailand
PRACTICAL INSTALLATION OF STANCHIONS FOR TOP-DOWN

CONSTRUCTION IN BANGKOK SUBSOIL
Narong Thasnanipan, Aung Win Maung & Zaw Zaw Aye

ABSTRACT
Top-down construction method has been used for construction of deep basement at
major projects in Bangkok, Thailand. To be able to apply top-down construction method,
structure elements are required to sustain the construction load and to utilize as a part of
bracing system. Prefabricated steel columns known as stanchions embedded in bored piles or
barrettes are commonly used for this requirement. This paper presents some experiences in
the practical installation of the different stanchion types in large diameter deep-seated bored
piles and barrettes in Bangkok. The problems encountered during the actual installation of the
stanchions and solutions to them are also discussed.
INTRODUCTION
Conventionally, buildings with underground basements are built by bottom-up method

where sub-structure and super-structure floors are constructed sequentially from the bottom of
the sub-structure or lowest level of basement to the top of the super-structure. Though this
conventional method, also called as bottom-up method, is simple in both design and
construction, it is not feasible for the gigantic projects with limited construction time and/or
with site constraints. Top-down construction method which provides the significant saving of
the overall construction time has been adopted for some major projects where time factor is of
primary importance. To be able to apply top-down construction method, rigid permanent
retaining wall and pre-founded structural columns are required. Diaphragm wall is normally
used as basement retaining wall whereas prefabricated steel stanchions embedded in bored or
barrette piles are utilized to sustain the construction load as well as to form as a permanent
column at later stage. The common length of the stanchions used in Bangkok ranges from 15
to 30m for excavation depth of 12 to 25m. Installation of heavy and lengthy stanchion with
stringent positional tolerance in the boreholes or excavated trenches filled with drilling fluid
in prevailing subsoil condition of Bangkok is a challenging job for the foundation contractors.
SUBSOIL CONDITION AND PILE CONSTRUCTION METHOD
Bangkok subsoil is featured by alternating clay and sand layers of thick Quaternary
deposits. As characteristics and properties of Bangkok subsoil layers have been well
documented in many literatures, only a general description of Bangkok subsoil is presented in
this article as follows.
Made-Ground consists predominantly of Fill-Materials, Clayey Sand or Silty Clay with
some cement block rubble and rock fragments, is commonly found up to 4m depth. Soft to
very soft, highly compressible dark gray marine clay lies beneath Made-Ground and in some
337
87
areas it lies under weathered crust layers of 2m thick. Depending on the location, this layer is
extended up to 12-18m. About 2m thick Medium Clay layer can be observed between Soft
Clay and underlying Stiff Clay. Generally Stiff Clay layer occurs directly underneath
Medium Clay and its depth goes up to 22m. Below Stiff Clay layer, First Sand layer 5-8m in
thickness can be found. This First Sand layer, however, is absent in some areas. Stiff to Hard
Clay layer underlies First Sand and it is found to be about 5m thick. Second Sand layer
generally occurs at depths between 45 to 65m.
Stanchions are mainly installed in the large diameter deep-seated bored piles and
barrettes founded in first or second sand layer. Bored piles of diameter 1.50m to 1.8m and
rectangular-shape barrettes of 0.80x2.8m to 1.5x3.0m (thickness x width) founding at depth
between 50m and 60m are commonly used to accommodate the stanchions. Wet process
method is always used for construction of these large bored piles and barrettes in Bangkok.
For bored piles, temporary casing of 14-15m in length is used as a support in soft clay layer.
Soil inside the casing is normally excavated by auger applying rotary drilling method and
drilling is continued with the bucket under the slurry from the top of sand layer to the final
depth. A cable-hung-grab mounted on crawler crane is used for excavation of trench with
slurry support in barrette construction. Guide walls having a minimum depth of 1.2m are
commonly used for initial guiding of grab excavation at the top of the trench. Tremie
concreting is necessary for casting both bored piles and barrettes.
APPLICATION OF TOP-DOWN CONSTRUCTION METHOD
Top-down construction method as the name implies, is a construction method, which

builds the permanent structure members of the basement along with the excavation from the
top to the bottom. Top-down method is mainly used for two types of urban structures, tall
buildings with deep basements and underground structures such as car parks, underpasses and
subway stations. For tall buildings with deep basements, a full application of top-down
method enables superstructure to build concurrently with excavation and construction of the
basement giving a significant advantage in reducing the overall construction time. Top-down
method is the most appropriate solution in minimizing the duration and disruption to surface
traffic and other urban activities, for the construction of the underground car parks,
underpasses and subway stations where the structure is located directly underneath the
existing roads.
Application of top-down method offers three main advantages as outlined below.
• It allows early commencement of super-structure construction without the need to
wait for excavation reaching the bottom level. Overall construction duration can be
significantly less than that of conventional method thereby saving the cost.
• It does not require temporary bracing system so that time requirement and cost for
temporary works (both material and labor costs) are eliminated, in turn significantly
saving cost.
• Minimize soil movement induced by excavation works which is an important factor
for some sensitive locations.
STRUCTURAL MEMBERS REQUIRED FOR TOP-DOWN CONSTRUCTION
Design and construction principles for top-down method primarily call for two major
structural elements.
• Columns with sufficient capacity must be pre-founded in bored piles or barrettes to
sustain the construction load and to utilize as part of bracing system.
88
338
• Excavation for basement must be carried out with the support of permanent
retaining wall so that basement floor slabs can be utilized as lateral bracing.
Diaphragm wall of 0.8m to 1.2m in thickness with sufficient embedment in firm soil
layers is commonly used as a retaining wall whereas prefabricated steel columns known as
stanchions embedded in either large diameter deep-seated bored piles or barrettes are utilized
as structural columns. Figure 1 illustrates the top-down construction method with utilization
of stanchions and diaphragm wall.
F3
Top
F2
F1
GF
B1
Stanchion
B2
Down
B3
B4
D-wall
Bored Pile
Figure 1. Top-down construction with stanchion and diaphragm wall
TYPES OF STANCHION AND THEIR APPLICATION IN BANGKOK
Pre-fabricated “H” section steel columns and steel built-up-section columns are mainly
used as pre-founded structural columns or stanchions. Pre-cast reinforce concrete columns are
very seldom used (Manoharn S. & Aye Z. Z. , 1994). For the purpose of convenient reference
throughout the paper, the type of stanchion is categorized by size and capacity as presented in
the table below.
Table 1. Technical data, application and limitation of three types of stanchion

Type of General Information
Stanchion Material & Example Size Application Limitation
Light Steel H-beams • For semi top-down Limited capacity of light
stanchion 350x350x137kg/m construction stanchion does not allow for
• For temporary decking construction of super
structure until completion of
basement construction
Medium- Steel H-beams For semi and full top-down Limited number of super
sized 350x350x390 kg/m construction of shallow to structure floors construction
stanchion medium deep excavation
Heavy • Steel H-beams Full top-down construction in Depending on the loading
stanchion 508x457x738kg/m deep excavation condition, numbers of
• Composite steel columns superstructure floors can be
built up by 2 or more small constructed before
to medium size H-beams completing basement
• Large section pre-cast RC excavation
column (seldom use)
89
339
Embedded length of stanchion usually ranges from1.5 to 5m depending on the loading
condition applied in design and construction.
ALLOWABLE TOLERANCE FOR POSITION OF STANCHIONS
Allowable tolerance for stanchions is usually called by the designer and it is mainly
governed by the structural tolerance of the steel as well as required position of the finished
columns. In most projects, allowable vertical and horizontal deviation of stanchion are
specified as 25 to 50mm whereas verticality is required between 1:200 and 1:400. There are a
number of constraints to get highly accurate position of heavy and lengthy stanchions which
have to be installed in deep-seated foundation piles constructed by wet process in Bangkok
subsoil as described in the following sections.
STANCHION INSTALLATION METHODS
Stanchion installation method is usually selected by the piling contractor who take into
consideration three main factors such as installation depth, size of stanchion and size of bored
or barrette piles. Though installation details may be different from one contractor to another,
stanchion installation can be categorized under two main methods, post-concreting or
plunging installation and pre-concreting installation or placing stanchion prior to concreting.
Post-concreting installation or plunging method
In this method, stanchion is installed immediately after completion of bored pile

concreting process. General construction sequence involved in this method is demonstrated in
Figure 2. Guide frame is used to install the stanchion at the correct position.
Concrete
Guide frame
Tremie pipe Stanchion Stanchion

Slurry Slurry Slurry Backfill
Temporary Temporary Temporary
casing casing casing
Pile cut-off Pile cut-off Pile cut-off

level level level
Concrete
Concrete
Complete drilling and Concreting by tremie Installing stanchion by Complete casing extraction
reinforcement installation method plunging into concreted pile and backfilling
Figure 2. General construction sequence of pre-concreting installation method
Pre-concreting or pre-placing installation method
In this method, stanchion is installed immediately after completion of drilling and

reinforcement lowering prior to concreting process. In some projects stanchion is attached to
the last section of reinforcement and installed together with the reinforcement. General
construction steps involved in this method are demonstrated in Figure 3.
90
340
Concrete
Guide frame
Stanchion Stanchion Stanchion

Slurry Slurry Tremie pipe
Backfill
Pile cut-off Pile cut-off Pile cut-off Pile cut-off

level level level level
concrete concrete
Concreting
Complete drilling and Stanchion installation Complete casing extraction
(1 tremie for bored pile and
reinforcement installation and backfilling
2 tremie for barrette)
Figure 4 shows the installation of a 24m long built-up steel stanchion into the pile
borehole prior to concreting.
(a) (b)
Figure 4. (a,b) Installation of 24m long stanchion into borehole prior to concreting
Guide Frame for stanchion positioning
Guide frame is used for positioning of stanchion in both installation methods.

Effectiveness of guide frame plays one of the important roles in achieving positional accuracy
of stanchion. Figure 5 (a,b) shows the guide frame used for pre-concreting installation method
in one major project constructed by Seafco Co., Ltd.
91 341
(a) (b)
Figure 5 (a,b) Guide frame used by Seafco for pre-concreting installation in one major project
REVIEW OF THE STANCHION INSTALLATION METHODS USED IN OTHER PARTS

OF THE WORLD
The stanchion installation methods used in other parts of the world have been presented
in some published papers. Among these, Findlay (1989) reviewed a number of stanchion
installation methods used in construction of large diameter bored piles for top-down
construction particularly in UK. The author reported that steel columns can be placed with
better accuracy by dry process bored piling method than wet process method (with support
fluid).
Hollingsworth (1991) demonstrated the installation of stanchions by plunging into
concrete with the aid of an adjustable guiding frame called Cemloc (European Patent No.
0302717) for seven levels basement of 25m deep constructed by top-down technique in
London.
Ressi di Cervia and Tamaro (1991) briefly explained the stanchion installation method
used for construction of an underground garage with top-down technique in Boston, USA.
Steel columns were installed prior to concreting in load bearing elements (LBE) or barrette
excavated by cable-hung mechanical clamshell buckets. The authors cited that using the same
equipment for construction of diaphragm wall and load bearing elements minimized the
congestion on site and streamlined the schedule.
Crawley and Stones (1996) presented the installation of 30m long composite steel
columns for top-down construction applied for Westminster Station in Central London.
Columns were placed with the aid of guide frame followed by concrete pouring. Accuracy of
±25mm at the top of column with verticality of 1:200 was achieved as stated by the authors.
Arz (1989) presented an installation of heavy composite columns fabricated from six
steel H-beams encased by concrete (cast during fabrication) for construction of six storey
building with five basement using top-down method in Germany. Due to the high accuracy
requirement of 25mm, adjustable guide frame equipped with three hydraulically control arms
were used. Stanchions were installed prior to concreting as reported by the author.
From the available literatures it is noted that pre-concreting method is more common in
use than post-concreting method.
PROBLEMS ENCOUNTERED IN POST-CONCRETING INSTALLATION METHOD
The major problem commonly encountered in this installation method is inability to

install the stanchion at the design position due to one or combination of the following causes.
• Inclination of the borehole / trench
92
342
• Stanchion can not be inserted up to the required depth as concrete becomes hard due
to the premature setting
• During lowering, stanchion is stuck by reinforcement cage and due to the hardening
of the concrete, extraction of stanchion becomes impossible for reinstallation
As the installation of the stanchion is often associated with many unforeseen problems,
it is likely in many cases that concrete becomes stiff or prematurely set during the installation.
In Bangkok, premature setting of concrete is usually found to be attributed by;
• long delivery time due to traffic conjunction
• inappropriate mix
• severe weather
• disruption in concreting / equipment breakdown
PROBLEMS ENCOUNTERED IN PRE-CONCRETING INSTALLATION METHOD
Using this method can minimize some problems encountered during installation
provided that adequate clearance is available between pile reinforcement and the stanchion for
tremie pipes lowering. Inclination of borehole is also important for this installation method.
Proper planning in concrete ordering is required to avoid unnecessary waiting of concrete
trucks for stanchion placing which may affect the total batching time limitation. As stanchion
has to be hung in the position until completion of concrete placement, appropriate and
sufficient capacity of the hanging and positioning device should be provided to avoid falling
and deviation of stanchion in the borehole. Temporary casing should not be used as a hanging
system without additional support for the large stanchion.
FACTORS EFFECTING THE POSITIONAL ACCURACY OF THE STANCHION
It is hardly possible to install a stanchion so that its position and verticality is always as
designed. As-built position of stanchion is found to be generally influenced by the installation
method. The factors affecting the positional accuracy for two different methods are
summarized in the Table 2 and 3.
Table 2. Post-concreting installation method – factors effecting positional accuracy

Factors effecting Recommended measures to improve
Discussion
accuracy accuracy
Verticality of piles Due to the use of fixed guide arm for guiding • Use adjustable guide
stanchion against borehole wall, its verticality • Maintain good horizontal position of
is solely influenced by the verticality of drilling rig and verticality of
borehole. temporary casing
• Perform drilling monitor test after
completion to check verticality prior
to stanchion installation
Concrete stiffening Due to the stiffening or hardening of concrete, • Use concrete with appropriate
or hardening stanchion may not be able to place at the right admixture to achieve longer setting
during installation elevation. Using vibro-hammer to force down time
stanchion can cause stanchion to deviate from
the position.
Inappropriate Large aggregates backfill can cause locking • Use appropriate backfill such as sand
backfill material up casing with stanchion upon casing / fine aggregates.
extraction which create stanchion to deviate
Improper back Back filling from one side of stanchion can • Maintain equal distribution of back
filling cause deviation of stanchion. fill around stanchion
Improper Stanchion deviation in both horizontal and • Apply proper equipment and method
temporary casing vertical can be caused by improper temporary in casing extraction to suit the site
extraction casing extraction. condition and design requirement
93
343
Table 3. Pre-concreting installation method – factors effecting positional accuracy
Factors effecting Recommended measures to improve
Discussion
accuracy accuracy
Verticality of piles In this method stanchion position may be • Maintain good horizontal position of
independent from borehole verticality prior to drilling rig and verticality of
concreting. However due to poorer borehole temporary casing
verticality in comparison with that of • Perform drilling monitor test after
stanchion, clearance between stanchion and completion of drilling to check
borehole wall becomes smaller with depth at verticality prior to stanchion
one side of stanchion. Thus during the tremie installation and adjust the tremie pipe
pipe lowering and extraction, stanchion is position according to borehole
pushed by tremie causing stanchion to move verticality
horizontally.
Concrete flows During pouring of concrete from tremie pipe, • Use appropriate fixing (lock)at top
pushing at force induced by concrete flow tends to move • Use 2 sets of tremie pipes placing at
stanchion during stanchion in both horizontal and vertical other side of stanchion if possible
tremie concreting direction
Improper back Same as post-concreting installation method Same as post-concreting installation
filling presented in above Table 2 method presented in above Table 2
Improper temporary Same as post-concreting installation method Same as post-concreting installation
casing extraction presented in above Table 2 method presented in above Table 2
Skill and experience of the contractor plays a major role in achieving positional
accuracy of the stanchion provided that the installation is practical with the design elements
such as size of foundation piles in relation to size of stanchion.
Figure 6. View of stanchions embedded in bored piles at final excavation level
STANCHION POSITION EFFECTED BY EXCAVATION
In Bangkok, post-installation movement (horizontal movement) of the stanchion is

frequently encountered and it is mainly caused by excavation induced soil displacement
particularly at initial stage of excavation where diaphragm wall deflection is characterized by
large cantilever rotation within Soft Clay layer. In some projects, deviated stanchions had to
be pushed or jacked back to the original position and restrained by means of temporary
support until permanent slab was cast.
REVIEW OF THE COMPLETED PROJECTS
The authors were involved in a number of projects constructed by Top-down method in

Thailand particularly in Bangkok.
94
344
Table 4 shows the summarized technical information of some major projects with
different stanchion type, size and installation depth.
Table 4. Summarized technical information of stanchion application in some major projects in

Thailand
Project Stanchion Type and Size Pile Type and Installation Basement Location
No. Size Method Excavation
Depth (m)
1 Built-up steel column Bored pile Pre-concreting -20.0m Silom Road
850x850mm, L=24m φ1.5x 60m Bangkok
built-up by 4 H-beams
270x248x157kg/m
2 H-section steel column Bored pile Post- -29.5m Rama IV Road
508x457x 738 kg/m, L=30m* φ1.8x65m concreting Bangkok
3 H-section steel column Barrette pile Pre-concreting -24m Ratchadapisek
419x407x390kg/m, L=24.7m* 1.2x3.0x47m Road Bangkok
4 H-section steel column Bored pile Pre-concreting -14.5m Payathai Road
414x405x232 kg/m, L=18m φ1.8x65m Bangkok
5 Pre-cast RC column Bored pile Pre-concreting -11.5m Haadyai
500x1000mm, L=15m φ1.5x35m
500x1500mm. L=15m φ1.8x35m
Note: * Top levels of stanchion in Project 2 and 3 are 3m and 1.8m below ground level respectively.
A view of heavy stanchions (built-up sections) exposed after excavating to the first
basement level is presented in Figure 7. Pre-concreting installation was used to install the
stanchions in this project.
Figure 7. View of built-up stanchions exposed after excavation to the first level of basement
COMPARISON OF AS-BUILT POSITIONAL TOLERANCE ACHIEVED BY TWO

INSTALLATION METHODS
Although the as-built position of stanchions installed by two different methods have yet
to be statistically analyzed, field survey data of Projects 2 and 3 indicated in above Table 4
suggested that stanchions could be more accurately installed by pre-concreting method.
Stanchions with as-built horizontal deviation greater than 100mm installed by two methods
(expressed by percentage) are presented in the table below.
95
345
Table 5. Comparison of as-built position of stanchions by two installation methods
Project Installation Method Depth of As-built Position of Stanchions with
Measurement Horizontal Deviation >100mm
2 Post-concreting Method 15.5m B.G.L 20 %

3 Pre-concreting Method 13.5m B.G.L 4%
CONCLUSION
Two common stanchion installation methods have been presented with particular
emphasis on the common problems encountered by each method and recommended measures
to alleviate them. According to the author’s experience, pre-concreting installation method
provided fewer problems in practical installation and achieved better positional accuracy. A
comparison of as-built position of stanchion between two projects also suggested that
stanchions could be more accurately installed by pre-concreting installation method.
The designers such as structural engineers and architects should be aware of the
accuracy achievable by practical installation method and take it into consideration at the
design stage.
Stanchion installation contractor should select the appropriate method, equipment as
well as experienced personnel plus a well-formed plan with the consideration of all potential
problems to achieve the successful construction of foundation structure, which is of primary
importance for top-down technique.
REFERENCES
Arz P. 1989. Foundations of the Ko-Galerie in Dusseldorf Using Top-down Building

Technologies, Proceedings of the International Conference on Piling and Deep
Foundations, London, 15-18 May 1989 : 185-192. Rotterdam: Balkema.
Crawley J.D. & Stones C. S. 1996. Westminister Station – Deep Foundation and Top-Down
Construction in Central London, Geotechnical Aspects of Underground Construction in
Soft Ground, London : 93-97. Rotterdam : Balkema.
Findlay J. D., Wren G.E. 1989. Review of the Methods Used to Construct Large Diameter
Bored Piles for Top Down Construction. Proceedings of the International Conference
on Piling and Deep Foundations, London, 15-18 May 1989 : 199-210. Rotterdam:
Balkema.
Finsted J. A. 1991. Royal Christiana Hotel : Basement with Permanent Sheet Pile Wall, “Up-
and-down” Method, 4th International DFI Conference : 387-392. Rotterdam: Balkema.
Hollingsworth J.R. 1991. Diaphragm Walls, Load Bearing Piles and Piled Soil
Reinforcement for a Deep Top-down Basement Construction, 4th International DFI
Conference : 79-84. Rotterdam : Balkema.
Manoharn S. & Aye Z. Z. (1994) , Bored Pile with King-post and Secant Pile Wall
Construction for Top-Down Method – Central Sukhothai Hotel Haddyai, Latest
Technology in Construction Technology Conference, Bangkok, Thailand
Yong K.S., Tsen C.N. & Tang S.K. 1998. Performance of a Deep Excavation with Top-down
Construction, 13th Southeast Asian Geotechnical Conference, 16-20 November 1998.
Taipei, Taiwan ROC : 587-593
96
346
Proceedings of the
GEOTECH – YEAR 2000
Development in Geotechnical Engineering
Editors:
Dennes T. Bergado
Lin Der-Guey
Tian Ho Seah
Kinya Miura
Noppadol Phien-wej
Prinya Nutalaya
Concrete for wet processed bored piles

Thasnanipan N., Maung A. W., Tanseng P. & Navaneethan T.
27-30 November 2000

Bangkok, Thailand
97
GEOTECH-YEAR 2000
Developments and Geotechnical Engineering
CONCRETE FOR WET PROCESS BORED PILES
Thasnanipan Narong, Maung Aung Win, Tanseng Pornpot & Navaneethan Thiruchelvam
ABSTRACT
The final quality of wet process bored piles is not only depending on the good
construction practice of piling contractors but also on the characteristics of concrete supplied.
The concrete characteristics are of the major items that influence the performance of piles
after casting. Due to the nature and environment of wet process bored piling, both proper
drilling and pouring of concrete would not produce the good quality piles, if the quality of
concrete did not meet the basic characteristics required for bored piling process. This paper
reviews the basic characteristics of concrete required in wet process bored pile and presents
some example problems caused by improper concrete mix. Additionally, the problems caused
by improper concrete pouring process and preventive measures are also described.
INTRODUCTION
The quality of piles depends on a good construction process including drilling,

reinforcement installation and concrete pouring as well as on good quality of concrete.
Quality of concrete has some influence on the workmanship with interrelated performances.
For wet-process bored piles, concrete is cast under drilling slurry using tremie pipes. Good
quality concrete in bored piling sense means that the properties and characteristics of the
concrete are suitable for the process of work and subsequently meet requirements of the
finished product. Continuous concrete pouring which is mandatory in piling and it is
sometime disrupted by blockage of segregated or prematurely set concrete mix in the tremie
pipe. Early setting of concrete after pouring in bored hole can also cause discontinuities in
pile by accidental lifting of set concrete during extraction of the temporary casing. Dampness
is sometimes found in top section of piles constructed in water-bearing permeable soil layer.
The dampness was found to be caused by capillary action of ground water through
interconnecting voids formed in the improperly mixed concrete.
Ready mixed concrete for bored piles usually specified as self-compacting concrete and
its “compacting factor” is very important for achieving required strength, especially at the top
section of pile which usually carries the maximum portion of transferred load. However, the
compacting factor is seldom mentioned in piling practice. Besides, no vibration is allowed for
tremie concrete while structural concrete is compacted by vibrator during casting. The
concrete strength test is usually carried out on the compacted test samples and thus the actual
strength of pile can be different from sample strength according to the compacting factor.
WET PROCESS BORED PILING METHOD
In wet process bored piling, bentonite or other type of suitable drilling slurry is used as
drilling fluid to support the borehole during construction. A steel temporary casing is usually
347
99
used to case the top weak soil subjected to heavy construction loads. Drilling and
reinforcement cage installation and concrete placing are successively executed under drilling
slurry. The concret is poured with tremie pipes, displacing the slurry well above the cutoff
level. The temporary casing is then extracted immediately after concreting.
CHARACTERISTICS OF CONCRETE AND CONCRETE MIX
Concrete mix for bored piles is designed according to concrete pouring process and
mechanical properties required. Concrete for wet process piles needs to be specially mixed
having cohesiveness with high workability (high slump/excellent fluidity) which is not prone
to segregation and retain it workability as far as possible throughout the tremie placing
operation for the complete pour. Addition to those characteristics, compaction under self-
weight, resistance to harsh environment, resistance to leaching, and appropriate strength are
essential.
Workability
Excellent fluidity is essential that the concrete has the ability to flow readily through the
tremie pipe, to flow laterally through a reinforcement cage and a high lateral stress against the
sides of borehole. High workability is best achieved with rounded natural aggregates and
natural sand in the mix.
Self Compaction
Compaction under self-weight is essential as vibration of concrete is impractical, except

near the surface. The degree of compaction achieved is determined by the density ratio (the
ratio of density actually achieved to the density of the same concrete fully compacted). The
recommended compacting factor for the required workability of tremie concrete is 0.95 to
0.96 (Xanthakos 1994). Fresh concrete is usually placed through tremie pipes and displaces
the slurry by gravity action only. In some cases, lack of self-compaction in the concrete will
lead to defects, such as reversed “hanging up”, and “whirls” in the completed pile. If the
initial shear of the concrete is very high, the flow is likely to restrain, resulting in bentonite
trapped in areas not reached by the concrete (Xanthakos 1994).
Resistance to Segregation
The concrete mix should be cohesive and resistant to segregation, as improperly

designed mixes will segregate during placement, resulting in inferior concrete containing
honeycombs and high permeable zones within the pile shaft. Concrete that bleeds or
disintegrates under the pressures of its own weight can also block the tremie pipe or accept
bentonite.
Controlled Setting
The concrete must retain it fluidity thorough the depth of borehole during complete
placement of the concrete in the borehole and attain an appropriate strength within a
reasonable time after placement. Retarders are used to prevent premature stiffening of some
cements or to delay stiffening under difficult placing conditions. The setting time must be
checked against the time necessary to complete the placement. The retarders should be used
under competent technical advice and after adequate testing.
100
348
Resistance to Aggressive conditions
The concrete should have high density and low permeability to resist the possible
(chemical and physical) attack of an aggressive sub-subsurface condition. In some instances
there is an underground flow of water that can cause a weakening of the concrete after it is
placed, and a properly designed mix should be resistant to such flow. However, if the rate of
ground water flow is substantial, a permanent casing will be necessary.
Good Mechanical Performance
The mechanical properties of hardened concrete can be satisfied in most instances.

However, appropriate tensile strength for the concrete without reinforcement in piles and high
level of bending and axial stress must be considered in some cases.
Reese and O’Neil (1988) emphasis that the design of the concrete mix must be given
appropriate attention and the design of the mix is dependent strongly on the particular job, and
the cement will be selected to be consistent with the design requirement. They observed that
bleeding is not a problem for concrete mixes that are properly designed. The trial mix method
is usually used in the laboratory. It is necessary to follow-up to see that the materials and
proportions used by the batch plant are those of that are recommended. Inspection at the
batch plant should include checking the nature, quantity and temperature of the components
of the mix, the aggregates, cement, water, admixtures and of the completed mix for
conformance with the specifications. For testing at the job site, the organization of the job
must be such that time required to perform tests at the job site is kept to a minimum.
Excessive job-site testing can lead to harmful effects. No delay in pouring should occur due
to field test.
Adding of water to the concrete with very low slump on site to increase the workability
can have detrimental effect of reducing the strength, compactability and impermeability of the
concrete. The results of adding water could be a significant change in the characteristics of
mix and the possibility of segregation as the pour is made. Segregation of concrete during
pouring can also lead to increase in permeability of concrete, especially at the top section of
piles due to upward migration of water in the concrete mix. Adding of water to the concrete
on site must not be allowed unless specified.
Suggested concrete mix by Fleming & Sliwinski (1977) for bored piles cast under
Bentonite was shown in Table 1 while range of cement content and water-cement ratio in
general use for concrete mixes by Bartholomew (1979) is shown in Table 2.
Table 1. Suggested Concrete Mix for Bored Piles Cast under Bentonite (After Fleming &
Sliwinski 1977)
Slump >175 mm.
Water/Cement Below 0.6
Aggregate Type Natural round stone if possible, 20 mm. max. size
Sand Type Natural and complying with zone 2 or 3 grading
Sand Content 30% to 45% of total aggregate weight
Cement Content Not less than 400 kg./m3
Admixture The use suitable admixture which will improve the workability and
extend the period during which such workability is maintained are to be
advocated.
101
349
Table 2. Range of Cement Content in kg/m3 and Water Cement Ratio in General use for
Concrete Mixes (After Bartholomew 1979)
Conditions
Normal Moderately Highly and Very
Pile Type
Aggressive Highly
Aggressive
450-475 450-475 450-475
Precast
0.4 – 0.5 0.4 0.4
300-450 350-450 380-500
Bored Piles Dry Process
0.5 – 0.55 0.475 - 0.5 0.45 - 0.5
Bored Piles with Tremie Process 350-450 350-450 400-500
0.5 – 0.6 0.475 - 0.5 0.43 - 0.45

280-370 330-450 370-500
Driven Cast-In-Situ Piles
0.25 – 0.6 0.3 – 0.55 0.3 - 0.45
Concrete for bored piles compared to that for pre-cast drive piles is the least dense
concrete due to pouring and casting process. Bored pile concrete was cast in the aggressive
subsurface conditions such as, high salinity, ground water fluctuation or in the vicinity of sea
and river. In such environment, the concrete can easily be leached out by ground water. In
this case cement content and water cement ratio given in the Table 2 needs to be reviewed
considering the site conditions.
Fleming et al (1977) pointed out that the high cement content favored for in-situ pile
construction enable the necessary workability mixes to be used with adequate margins of
safety against the inevitable variations in strength and workability. It also compensates for
some reduction in strength, which may occur on interfaces during displacement.
Concrete is permeable to water to the extent that it has interconnecting void spaces
through which water can move. Calcium hydroxide liberated by hydrating cement is water-
soluble and may leach out of harden concrete, leaving voids for the ingress of water.
Permeability of concrete is governed by amount of cementitious material, water content,
aggregate grading, consolidation and curing efficiency.
MEHTOD OF CONCRETE POURING
The quality of piles depends on a good pouring procedure as well as on good quality of
concrete. In wet process, the concrete is usually placed by a steel tremie pipe of 20-25cm in
diameter (minimum 6 times of coarse aggregate size).
Prior to charging the tremie pipe with concrete, the bottom of tremie needs to be sealed
by a plug of some descriptions may be inserted at the top of tremie before or after the tremie
is placed in the bored hole as appropriate. There are two potential problems that are associate
with the initial charging of tremie with concrete; the concrete can segregate during placement
and the air in tremie will prevent the complete filling of the tremie. These problems can be
avoided if the tremie is filled slowly. Faulty initial charging of tremie during concreting can
cause entrapment of mud within the concrete.
Excessive initial lifting of tremie can result in possible distribution of leached concrete
caused by concrete falling through the slurry. The bottom of tremie must stay well below the
top of the column of fresh concrete all the time. Moreover, the tremie must not be lifted and
102
350
lowered rapidly to avoid the cause of contamination of concrete with slurry. It is suggested
that the tremie pipe must not be lift and lowered rapidly to start or restart the flow of the
concrete (Reese & Neil 1988). However Xanthakos (1994) suggested that if the concrete is
not deposited easily the triemie pipes may be moved up and down with movement not
exceeding 30cm. Moreover, the tremie pipes should not be moved horizontally.
In the area of high ground water level, the concrete must be deposited above the
external water table before the casing is withdrawn. The hydrostatic pressure in the concrete
column should be greater at all time than the pressure in any column of fluid outside the
casing.
COMMON DEFECTS
The common defects of piles are cold joints, zone of segregated or contaminated
concrete, trapping of bentonite mud and cavities. The first two types of defect result from
interruptions in the concrete placement or premature extracting of the tremie pipe either
partially or completely above the concrete-slurry interface. Mud trappings are caused by
concrete of low workability and impediment to the flow of concrete due to closely spaced
bars. Discontinuities or partial separation in the piles at the bottom edge of temporary casing
can be caused by accidental lifting of low workability concrete or concrete without controlled
setting during casing extraction. If the concrete in the casing is too stiff and has considerable
frictional resistance against the casing, a column of concrete can be pulled up with the casing.
Permeability of concrete depends on the capillary porosity, water-cement ratio and
degree of hydration. High permeability of concrete can be contributed by presence of
capillary pores that are interconnecting voids in the concrete (Figs 1 & 2). The
interconnecting voids are caused by bleeding of concrete due to excessive water used in the
mix. Bleeding raises the water and air bubbles to top surface of fresh concrete and raises the
water-cement ratio of concrete upper part of the forms, thereby reducing the strength and
increasing porosity of concrete. Concrete mixed with a water-cement ratio higher than 0.6
can be more permeable. Function of concrete self-compaction can also reduce the
permeability and increase the density. In piles, usually the fresh concrete compacts under its
own weight, resulting in an increasing density with depth. Besides, concrete in a great depth
of pile is generally cured in stable temperature and moisture.
Figure 1. High permeability of concrete Figure 2. Low permeability of concrete

caused by interconnecting voids caused by less interconnecting voids
Usually one type of concrete mix is used for piling in a particular project in most cases.
However, the project in the vicinity of the river, some bored piles are to be installed closer to
103
351
the river than the remaining piles and thus the concrete mixes need to be designed
accordingly. For the piles installed in the deposition side of the river where sand deposits
occur, these piles are usually subjected to be effected by the ground water flow (Fig. 3). If
the cement content is not high enough in the concrete mix and bleeding or segregation occurs,
dampness or wet patches caused by capillary suction to ground water through the previous
tremie location and vertical steels can be found on the pile head (Figs 4 & 5). Figure 6 is a
photo of core sample obtained from the pile that exhibits the dampness on top, showing
segregated concrete caused by bleeding.
Crust Soil
Capillary Suction
Thru Previous
Tremie Location Hydraulic pressure
and Capillary suction

Permeable
Soil
Water Flow into

High Permeable
Concrete
Soft to Stiff
Clay
Figure 3. Capillary flow of Figure 4. Capillary suction and hydraulic pressure

ground water in bored pile in concrete
Figure 5. Wet pile head caused by capillary Figure 6. Core samples showing
flow of ground water through segregated segregated concrete in the pile
concrete in the pile
Figure (7) show the some segregated concrete extracted from the tremie pipe, which
was blocked with such concrete. Settlement of solid particles or aggregates in the concrete
mix cause bleeding by migration of the water to the top surface of fresh concrete, reducing
water-cement ratio of the lower part of the mix. As a result, lower part of the concrete mix
stiffens rapidly with aggregates. In such case stiffening concrete can also block the tremie
104
352
pipe during concrete pouring (Fig. 8) causing disruption in casting process and thus effects
the quality of pile. Equipment for concrete pouring must also be adequate and reliable to
avoid any interruption due to a breakdown in continuous tremie concrete pouring operation.
Figure 7. Tremie blocked by segregated Figure 8. Tremie blocked by stiffening

concrete. concrete
Due to a congested bar arrangement, concrete cannot flow through the bars and as a
result concrete cover can be lost. Good arrangement of reinforcing bars is thus necessary.
The horizontal spacing of main bars should be at least 10cm and 15cm (minimum 4 times of
the maximum size of aggregate) for small and large diameter bars respectively.
RECOMMENDATIONS AND CONCLUSIONS
Foundation designers and concrete suppliers should pay more attention to workability
and many other important factors required for cast in-situ tremie concrete than strength alone.
It is essential to design the better quality mix than the concrete for other structural works in
some aspects, considering the process of work.
For the projects, such as elevated highways, viaducts and bridges, variable soil and
ground water conditions can be encountered along the project area. In such cases, concrete
mixes need to be designed to suit these conditions and used accordingly in bored pile
construction.
An adequate cement content and water cement ratio is necessary to have good
impermeability of concrete which is one of factors influencing the durability of concrete,
especially for bored piles in water bearing subsoil.
It is concluded that appropriate concrete mix and casting practice is essential in bored
piling work to achieve good quality piles.
REFERENCES
Bartholomew, R. F. (1979), The Protection of Concrete Piles in Aggressive Ground

Conditions: An International Appreciation”, Proceedings of the Conference on Recent
Developments in the design and Construction of Piles, ICE London, 21-22 March 1979,
pp. 131-141
Fleming, W. K. and Sliwinski, Z. J. (1977), “The Use and Influence of Bentonite in Bored
Pile Construction”, CIRIA Report . PG. 3.
105
353
Reese, L. C. & O’ Neil, M. W. (1988), Drilled Shafts: Construction Procedures and Design
Methods, U.S. Department of Transportation, Federal Highway Adminstration, Virginia
incorporation with ADSC., Dallas, Texas, USA.
Xanthakos, P. P (1994), Slurry Walls as Structural systems, 2nd Edition, McGraw-Hill Inc.
106 354
5th International Conference on
DEEP FOUNDATION PRACTICE

incorporating PILETALK International 2001
4 – 6 April 2001, Singapore
RECORD LOAD TEST ON A LARGE BARRETTE AND ITS

PERFORMANCE IN THE LAYERED SOILS OF BANGKOK
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye
Pornpot Tanseng
Suranaree University of Technology, Nakornratchasima, Thailand
Organized by
• CI-PREMIER CONFERENCE ORGANIZATION
ISBN: 981-04-2512-0
107
th
5 International Conference on Deep Foundation Practice incorporating Piletalk : 4 – 6 April 2001, Singapore
Record load test on a large barrette and its performance

in the layered soils of Bangkok
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye

Pornpot Tanseng
ABSTRACT
This paper presents the load transfer characteristics of fully instrumented barrette of 1.5x3.0m in size
seated about 57m below the ground level. The test results were compared with those from the
instrumented load test on bored pile of diameter 1.5m with the same length, located 30m away. No
significant difference in load transfer has been observed between barrette and pile despite the
considerable difference in construction method applied and time consumed.
Keywords : Barrette, static load test, load transfer characteristics
1. INTRODUCTION
Barrettes have been used as deep foundations for various structures in Bangkok for a number of
years. The large load carrying capacity achievable by flexible dimension and length of barrettes
provides a major advantage in prevailing subsoil condition of Bangkok. In addition to extensive bearing
capacity requirement, the demand for barrettes is necessitated mainly by site constraints, applicable
construction method and equipment. Barrettes with dimension ranging from 0.80mx2.7m to
1.5mx3.0m for safe working load capacity from 1100 to 2300 ton have been used in some major
projects. To assess the performance of large-capacity foundation element in layered subsoil of
Bangkok, instrumented load testing is compulsory. This paper presents the static load test results of
instrumented barrette tested up to 5290 ton for foundation of a fifty-storey building in Bangkok. The
test results, particularly load transfer characteristics, and shaft friction capacity were compared with
those of bored pile diameter 1.50m with the same length located 30m away.
2. OVERVIEW OF THE PROJECT
Foundation of the fifty-storey tower called for 560 bored piles of 1.2m and 1.5m diameter and 24
number of barrettes having cross section size of 1.5m x 3.0m. Bored piles and barrettes were seated
at approximate depth of 57m in the second sand layer. Barrettes were designed to support the large
load for tower’s lift shafts as bored piles were not feasible to utilize in such case. Base grouting was
applied for barrettes and bored piles at the locations of high column load mainly in the central tower
area. Instrumented static pile load test was proposed for one barrette and one bored pile of diameter
1.5m. The design safe working loads for the base grouted barrettes and bored piles are 2,300 ton and
1,175 ton respectively. Foundation plan of the project is shown in Figure 1.
363
109
132.10
96.20
Test-Barrette
Test
Pile
132.68
94.00
Figure 1 Layout of foundation showing location of test barrette and bored pile
3. SUBSOIL CONDITION
Soil investigation from five boreholes at different location reveals that subsoil layers along the site
are relatively consistent. Similar to other localities in Bangkok a typical subsoil profile at the site is
characterized by the alternating layers of clay and sand deposits as soil succession shown in Figure 7.
Soft, highly compressible dark gray marine clay lies beneath weathered crust layers of 2m thick and
extends up to 13.5m. Stiff Clay layer occurs directly underneath Soft Clay and its depth goes up to
26m. Below Stiff Clay layer, First Sand layer of 10m in thickness can be found. Hard Clay layer
underlies First Sand and it is found to be about 12m thick. Second Sand layer occurs at depths
between 50 to 72m. Undrained shear strength (Su) obtained from unconfined compression test and
Standard Penetration Test (SPT) are shown in Figure 7.
4. CONSTRUCTION METHODS
Both barrette and bored pile were constructed by wet process under bentonite slurry. Bentonite
slurry conforming to the widely accepted specification was used. Properties of the bentonite slurry
used are given in Table 1. Single stage base grouting was applied 24 hours after concreting for both
barrette and bored pile. The detailed procedures utilized in the construction of two different foundation
structures are outlined below.
Table 1. Comparison of bentonite slurry properties

Barrette Bored Pile
Properties Before feeding After Recycling & Before feeding to After Recycling & Before
to the borehole Before Concreting the borehole concreting
(near trench base) (near borehole base)
Density (g/cc) 1.10 1.17 1.08 1.10
Sand Content (%) 1.0 1.1 0.1 0.8
pH value 8 9 8 8
Mechanical rope-grab was used to excavate the trench. A guide wall cast with inside clear
dimensions slightly larger than the nominal size of the barrette was used to guide the grab during
364
110
initial bites. Since time consumed in the preparation of instrumentation was relatively long, desanding
was continuously done to keep the bentonite slurry agitated, which also helps to alleviate the growth of
filter cake by minimizing the actual exposure time. As another measure, trench was once again
occupied by grab to scrap the trench walls to remove, if any, filter cake formed on the walls. This
attempt is in line with the recommendation made by Reese and O’Neill (1988) [1]. It is authors opinion
that, if due to some unforeseen reasons, reinforcement cage lowering have to be delayed for
considerable period of time, it is a good practice to use the grab again to scrap the trench walls. This
measure eliminates any foreseeable negative impacts caused by unexpected delays. After lowering
the rebar cage, tremie concreting was done.
Rotary drilling was employed for bored pile excavation. Different from barrette excavation,
temporary casing of 15 m length was used as a support in Soft Clay layer for bored pile drilling to
assure the stability of the borehole. Firstly, auger was used to drill within the temporary casing,
followed by rotary bucket with bentonite slurry down to final depth of excavation. The base of the
borehole was cleaned by recycling technique to minimize any congregated sediments. Before lowering
the reinforcement cage special cleaning bucket was used to scrap of the borehole walls and the base.
Reinforcement cages were then lowered inside the borehole while attaching the instrumentation
simultaneously at specified locations. Soon after lowering the rebar cage tremie concreting was
commenced. Polystyrene grains plug was used before the first charge of concrete to avoid the mixing
of bentonite with concrete. Time consumed in different construction activities is plotted in Figure 2.
B arrette
D rilling
D eSanding
C age Low ering
B ored P ile Trem ie Pipe Preparation
C oncreting
0 10 20 30 40 50 60 70 80
Tim e Co nsum e d (ho urs)
Figure 2 Time consumed in different construction activities of barrette and bored pile
( after Thasnanipan, 1999 [2] )
5. LOAD TEST PROGRAM AND INSTRUMENTATION
5.1 Test Pile Layout
Test pile layout of barrette is presented in Figure 3. With overall height of 8m, reaction frame
utilized for static load testing on barrette in this project was claimed to be one of the biggest of its kind
in the region. Four barrettes were used as anchoring system. Five numbers of built-up steel girders
st
supported on each side by two 1 level cross beams were used as main beams to achieve the
maximum capacity of 6000 tons. First level beams were supported against the second level cross
beams. Second level cross beams were anchored against surrounding barrettes using anchor blocks
at the top. Specially fabricated rigid transfer girders were used to distribute the tension force coming
from the tie-bars to dowel bars above the anchor barrette heads. Sixteen numbers of hydraulic jacks
each having 500 ton capacity were placed between the test barrette cap and the main beams of the
reaction frame. General view of the barrette load test set up is presented in Figure 4.
Test pile layout of bored pile was similar to those of other static bored pile load tests in Thailand.
Steel test frame anchored against four bored piles was used in bored pile load test. Different from the
test frame of barrette, only 1 layer of cross beams was required for that of bored pile.
111365
5.00 5.00
2n d C ro ss
B eam
1st C ro ss
B eam
AN C H O R
B AR R E TTE
5.00
M ain B eam
H ydrau lic
Jacks
TE S T
5.00
B AR R E TTE
2n d C ro ss
B eam
1st C ro ss B eam
2n d C ro ss
B eam
Figure 3 Barrette test pile layout showing the configuration of 6000 ton capacity reaction frame
5.2 Monitoring System and Instrumentation
Direct measurement from four dial gauges placed in diametrically opposite positions having
equidistance from the test pile axis was used as a main monitoring system of test pile head
movement. Precise leveling and piano wire were also utilized as backup for pile head movement
measurement. Additional two dial gauges were also used to monitor the lateral movement of the pile.
Eight number of SINCO load cells of 500 ton capacity each were installed on top of the hydraulic jacks
to evaluate the actual applied load of the first load cycle . Vibrating wire strain gauges (VWSGs) and
Mechanical Extensometers (ME) were fixed at five levels along the shafts at the known interface
boundaries of different soil layers. At each level, four sets of VWSGs and one set of ME were installed
for the pile whereas six sets of VWSGs and two sets of ME were installed for the barrette.
Figure 4 General view of the reaction frame for static load test on barrette
112 366
5.3 Load Test
Pile load tests were carried out in accordance with ASTM D 1143-81. Pile and barrette were
tested under three and four cycles of loading and unloading respectively.
6. LOAD TEST RESULTS AND INTERPRETATIONS
6.1 Pile Head Movement at Applied Load
Load vs pile head movement graph of barrette and bored pile is illustrated in Figure 5. Measured
pile head movements at design load, double of design load and maximum test load are presented in
Table 2. As can be seen in the table, both barrette and bored pile experienced only negligible pile
head movement at relevant design load.
Table 2 Predicted and measured pile head movement at specific applied load
Measured gross pile head movement at applied load
Test Pile Design Load (DL) Max. Test Load
At DL At 2 x DL At Max. Test Load
Barrette 2300 ton 5290 ton -5 mm -12 mm -61 mm
Bored Pile 1175 ton 2700 ton -6 mm -19 mm -67 mm
Table 3 shows the estimated ultimate load capacity of barrette and bored pile from the load vs pile
head movement using different method.
Table 3 Estimated ultimate capacity of barrette and bored pile

Ultimate capacity (Qult) and pile head movement
Test Pile Parameter at Qult estimated by different method
Davisson’s Mazurkiewicz’s Butler and Hoy’s
Qult (ton) 5180 ton 5156 ton 5100 ton
Barrette
Pile head movement at Qult 18 mm 18 mm 17 mm
Qult (ton) 2675 ton 2734 ton 2620 ton
Bored Pile
Pile head movement at Qult 30 mm 38 mm 22 mm
Applied Load (Tons)

0 1000 2000 3000 4000 5000 6000
0
20
Barrette
Pile
40
60
80
Figure 5 Pile head movement at the last load cycle (maximum loading) of barrette and bored pile
367
113
6.2 Mobilized Skin Friction vs Pile Head Movement
The graphs showing the ratio of mobilized skin friction to maximum mobilized skin friction against
ratio of pile head movement to pile diameter (s/D) are presented in Figure 6a and 6b respectively.
Symbol ‘D’ shown in Figure 6a represents the barrette dimension in equivalent diameter. According to
the figures, in general, skin friction mobilized to the maximum values at pile head movement of 0.6%
and 1.6% of equivalent shaft diameter for barrette and diameter of bored pile respectively.
1 .2
1 .2
(fs As)m obiliz e d / (fs As) m a x im um
(fs As) m obiliz e d / (fs As) m a x im um

1 .0
1
0 .8 0 .8
0 .6 0 .6
0 .4 0 .4
0 .2 0 .2
0 .0 0
0 .0 0 .5 1 .0 1 .5 2 .0 2 .5 0 .0 1 .0 2 .0 3 .0 4 .0 5 .0
s / D (% ) s / D (% )
6a (Barrette) 6b (Bored Pile)
Figure 6 Ratio of mobilized skin friction to mobilized maximum skin friction against ratio of
pile head movement to equivalent pile diameter
6.3 Load Transfer Characteristics of Barrette and Bored Pile
Load transfer curves along the shaft of barrette and bored pile at various applied load are shown
in Figure 7. The values of unit skin friction developed at the different soil layers along the shaft of
barrette and bored pile in comparison with those of calculated ultimate unit skin friction are
demonstrated in Figure 8. This comparison suggests that mobilized skin frictions of barrette and bored
pile are higher than those of calculated values using empirical formulas. It proves the findings of
various researchers on shaft friction improvement of base grouted piles ( eg. Teparaksa et al 1999,
[3] ). It is also evident that overall shaft resistance of both pile was not fully developed at design load.
S u ( U C ) t/ m2 BUTT LOAD (TON)
0 5 10
DEPTH (m) 0 1000 2000 3000 0 1000 2000 3000 4000 5000
0 0
Weathered
Crust
Soft
Clay
-10
First Stiff
-20
Clay
Fi r st
-30
Sand
-40
Hard
Clay
-50
Second
T i p Level T i p Level
Sand
-57.5 m -57.5 m
-60
60
1.50 m 3.00 m
0 50 100
S P T - N ( b l ows / 3 0 c m)
VW SG
1.50 m
B ored
B arrette M echanical
P ile
Extensom eter
Figure 7 Load transfer curves of test barrette and bored pile with typical soil profile at the site
114 368
Mobilized Unit Skin Friction Max. Mobilized Unit
S u ( U C ) t / m2
at Design Load (ton/m2) Skin Friction (ton/m2)
0 5 10
DEPTH (m) 0 10 20 30 0 10 20 30
0 0 0
Weathered Bored Pile
Crust Bored Pile Barrette
Barrette Calculated
Soft Calculated
Clay
-10 10 10
First Stiff
-20 20 20
Clay
Depth (m)
Depth (m)
F i r st
-30 30 30
Sand
-40 40 40
Hard
Clay
-50 50 50
Second
Sand
-60 60
60
0 50 100
SPT-N (blows/30cm)
Figure 8 Comparison of calculated and mobilized unit skin friction of barrette and bored pile
6. 4 Effect of Construction Time on Shaft Capacity Reduction
Excessive construction time is one of the main parameters claimed to be responsible for shaft friction
capacity reduction of drilled shaft foundation constructed with wet process under bentonite slurry as
reported by various researchers. Total construction time of barrette and pile were 75 hours and 27
hours respectively. Though total construction time consumed for test barrette was almost 3 times more
than that of bored pile, there is no significant difference in shaft load transfer between them. The
measures adopted in construction of barrette to minimize the excessive filter cake formation along the
sand layers by proper desanding (continuous agitation), and retrenching with grab are considered the
main reasons contributed to this achievement. Comparison of developed unit skin friction values of
barrette and pile proves that difference in utilized bentonite viscosity as shown in Table 1 does not
have significant effect on the shaft capacity. These findings are in line with conclusions made by
Thasnanipan et al (1999) [4].
6.5 Effect of Shape on Shaft Load Transfer
Hosoi et al (1994) [5] concluded, from the results of numerical analysis that earth pressure acting on
the flat surface of diaphragm wall panel is larger than that of circular bored pile. Thasnanipan et al
(1999) [2] attempted to assess the effect of different aspect ratios (L/B) on the earth pressure
developed around the trenches by using finite element program and reported that no significant
difference in earth pressure was observed between the borehole (L/B=1) and the barrette (L/B=2).
Further research is necessary to evaluate this finding. No significant difference is found between the
maximum mobilized skin friction against displacement of rectangular-shape barrette and circular-
shape bored pile according to Figure 6a and 6b respectively. Displacements at maximum mobilized
skin frictions of barrette and bored pile fall within the range (0.5 % - 2 % of shaft diameter) reported by
Reese (1978) [6].
369
115
7. CONCLUSIONS
(1) Elastic deformation of both barrette and bored pile at relevant design load was found to be
negligible. Shaft resistance of barrette and bored pile was not fully mobilized at relevant design
load. These observations suggested that the selected size and length of barrette and bored pile
for specified design loads are sufficient for acting as friction piles.
(2) Mobilized shaft frictions of barrette and bored pile at maximum test load particularly in the first
sand layer are considerably higher than calculated values. Shaft friction capacity improvement
was considered to be contributed by base grouting which is in line with the findings reported by
various researchers.
(3) Despite the considerable difference in construction method applied and construction time
consumed (75 and 27 hours for barrette and bored pile respectively), similar characteristics of
shaft load transfer were observed between barrette and bored pile. This achievement indicated
that adopted measures (proper desanding and scraping of trench wall by grab prior to
reinforcement lowering) to alleviate the excessive growth of bentonite filter cake on the barrette
wall proved to be effective.
(4) Static load test up to 5290 ton conducted on barrette set the record as the highest load ever
tested for a single cast-in-situ deep foundation in Thailand.
ACKNOWLEDGEMENTS
The authors wish to express their appreciation to Dr. Wanchai Teparaksa and Mr. Kamol Singtogaw
for providing invaluable suggestions in the initial stage of the load test program. The authors also wish
to thank Mr. Chanchai Submaneewong for his assistance in preparing this paper.
REFERENCES
[1] Reese L.C., & O’Neill M. W., Drilled Shafts : Construction Procedures and Design Methods,
ADSC: The International Association of Foundation Drilling, Dallas, Texas, USA, 1988
[2] Thasnanipan N., Anwar M. A., Maung A.W., Tanseng P., Performance Comparison of Bored and
Excavated Piles in the Layered Soils of Bangkok, Symposium on Innovative Solutions in
Structural and Geotechnical Engineering in Honor of Professor Seng-Lip Lee, Asian Institute of
Technology & National University of Singapore, 1999, p.p. 345-353.
[3] Teparaksa W., Thasananipan N., and Anwar M. A., Base Grouting of Wet Process Bored Piles in
Bangkok Subsoils, The Eleventh Asian Regional Conference on Soil Mechanics and
Geotechnical Engineering, Seoul, Korea, 1999, p.p. 269-272.
[4] Thasnanipan N., Anwar M. A. & Maung A. W., Review of the Shaft Capacity Degradation of
Bored Piles Constructed with Bentonite Slurry, Civil and Environmental Engineering Conference,
Asian Institute of Technology, Bangkok, Thailand, 1999, p.p. V-59 to V-68.
[5] Hosoi T., Yagi N., & Enoki M., Consideration to the Skin Friction of Diaphragm Wall Foundation,
rd
3 Intl. Conf. on Deep Foundation Practice incorporating PILETALK, Singapore, 1994
[6] Reese L. C., Design and Construction of Drilled Shafts, 12th Terzaghi Lecture, Proc. ASCE, Vol.
104, No. GT1, 1978, p.p. 95-116
116
370
GEOTECHNICAL SPECIAL PUBLICATION NO. 116
DEEP FOUNDATIONS 2002

An International Perspective on Theory,
Design, Construction and Performance
Volume One
PROCEEDINGS OF THE INTERNATIONAL

DEEP FOUNDATIONS CONGRESS 2002
SPONSORED BY
The Geo-Institute of the American Society of Civil Engineers
February 14-16, 2002

Orlando, Florida
EDITED BY
Michael W. O’Neill
Frank C. Townsend
117
Performance of Wet-Process Bored Piles Constructed with
Polymer-Based Slurry in Bangkok Subsoil
Narong Thasnanipan*, Zaw Zaw Aye**, Chanchai Submaneewong*** and

Wanchai Teparaksa, Ph.D. ****
*Managing Director, **Project Manager, ***Geotechnical Engineer, SEAFCO Co.,

Ltd., 26/10 Rarm Intra 109 Rd., Bangchan, Klong Sam Wah, Bangkok 10510, Thailand,
Tel. (662) 9190090, Fax. (662) 9190098, seafco@seafco.co.th, www.seafco.co.th
****Professor, Department of Civil Engineering, Chulalongkorn University, Bangkok,
Thailand, fcewtp@eng/chula.ac.th
Abstract
Conventionally, medium to large bored piles deep-seated in the multi-layered soils of

Bangkok were constructed by rotary drilling under wet process or slurry replacement
method, using betonite slurry as support fluid in boring process. As an alternative to
bentonite slurry, polymer-based slurry has been increasingly used in Bangkok recently.
Polymer-based slurry having environmentally friendly properties, is found to be a good
solution in minimizing the problems associated with the disposal of contaminated soils
as well as the site pollution. This paper presents the performance of medium to large
diameter bored piles constructed with polymer-based slurry in Bangkok. The static pile
load test results of polymer-based slurry bored piles are compared with the predicted
values using bentonite piles parameters. The measured frictional resistance of pile shaft
from instrumented pile load tests are also reported. Shaft resistance of bored piles
constructed with polymer-based slurry is found to be better than that of bored piles
constructed with bentonite slurry.
Introduction
Medium to large diameter bored piles with length ranging from 24 m to 60 m have been
used as foundations of the various structures in Thailand for over 3 decades. Rotary
drilling technique coupled with wet process method using bentonite slurry as support
fluid is commonly applied to construct these piles. As bentonite slurry had proven
acceptable to make qualified bored piles in the past, most of them were constructed only
by using this mineral based slurry. Recently, polymer-based slurry has been
increasingly used in Bangkok as an alternative supporting fluid in bored pile drilling. In
the initial stage, the primary reason for using this alternative slurry was to minimize the
problems associated with using classic bentonite such as site pollution and the disposal
of excavated soils which are usually contaminated. Construction records and static pile
load test conducted on bored piles constructed with polymer-based slurry proved that
there is no adverse effect on stability of borehole during drilling as well as capacity of
the pile from using this alternative supporting fluid in Bangkok subsoil. Dry anionic
partially hydrolyzed polyacrylamide (PHPA) polymer powder premixed with fresh
water and a small percentage of bentonite was used in the construction of bored piles
presented in this paper.
119
143
144
Geotechnical Pore Pressure (kN/m 2)

Parameters 0 200 400 600
0 Weathered Crust
0
Bangkok Soft Su= 10-25 KPa

Clay γ t = 14-16kN/m 3
10 -
10
Medium to
Depth (m)
20 Stiff Clay Su= 50-14020- KPa
γ t = 17-21kN/m 3
Act
ual
Medium
30 Dense Sand SPT-N= 20-40
-
Pie
30
Hyd
(First Sand)
γ t = 20kN/m 3
zom
ros
etri
tatic
40 Hard Clay Su> 200 KPa
-
cD
40
γ t = 21KN/m 3
Line
raw
dow
50 Dense to Very -
nL
50
Dense Sand SPT-N= 50-100
ine
(Second Sand) γ t = 21KN/m 3
60 -
60
Figure 1. Typical soil profile of Bangkok plain with piezometric drawdown conditions.
Subsoil Profile and Existing Pore Pressure Condition
Subsoil profile is relatively consistent at different localities in Bangkok. A typical

subsoil profile is characterized by the alternating layers of clay and sand deposits as
shown in Figure 1 and is briefly described below.
Weathered crust of 2 m thick is commonly found as the top layer. In urban areas
of Bangkok, this layer is covered by filled material. Soft to very soft, highly
compressible dark gray marine clay lies beneath weathered crust. Depending on the
location, this layer extends up to 12-18 m. About 2 m thick medium clay layer can be
observed between soft clay and underlying stiff clay. Generally stiff Clay layer occurs
directly underneath medium clay and its depth goes up to 22 m. Below stiff clay layer,
first sand layer 5-8 m in thickness can be found. This first sand layer, however, is
absent in some areas. Stiff to hard clay layer underlies first sand and it is found to be
about 5 m thick. Second sand layer generally occurs at depths between 45 to 65 m.
Existing pore water pressure conditions in upper part of Bangkok soft clay are
hydrostatic from nearly 1m below the ground level. Then the hydrostatic condition
changed to piezometric drawdown near bottom level of Bangkok soft clay as shown in
Figure 1. Excessive ground water abstraction from the deep aquifers in the past is
responsible for this significant drawdown of piezometric pressure (Nutalaya, 1981,
Thasnanipan et al., 1998).
Wet-Process Bored Pile Construction Method
In Bangkok, medium to large diameter (0.6 m to 1.8 m diameter) bored piles with depth
ranging from 24 m to over 60 m are constructed by wet process due to the prevailing
subsoil condition and limited application of dry process. Wet process method as its
120
145
name implies makes the pile under wet condition by using drilling slurry. Temporary
casing of 14-15 m in length is also used as a support in soft clay layer. Drilling is
commonly commenced by dry process with auger applying rotary drilling action in the
soft clay layer and stiff clay layer. Before reaching to the first sand layer slurry is fed to
the borehole and drilling is continued with the bucket to the final depth. A cleaning
bucket is commonly used to clean the bottom of the borehole prior to reinforcement
cage installation. Tremie concreting is necessary for the piles installed under wet
process.
Polymer-Based Drilling Slurry
The primary function of drilling slurry is to prevent the borehole instability as deep-
seated bored piles are required to install through water bearing and caving soils.
Bentonite slurry was mainly used in the past as it adequately fulfilled the requirement.
Polymer-based slurry has become increasingly popular as an alternative supporting fluid
in recent years. The main reason for switching to use polymer-based slurry is to
minimize the problems associated with environmental issues caused by the use of
bentonite slurry.
In the initial stage of using polymer-based slurry, the major concerns were the
potential negative impact on stability of the borehole and the shaft friction capacity as
this alternative slurry was not yet accustomed to the bored piling industry in Bangkok.
Trial boreholes conducted in the initial stage of introducing polymer-based slurry
revealed that borehole is well stablized if the right dosage is used. Dry anionic partially
hydrolyzed polyacrylamide (PHPA) polymer powder premixed with fresh water and a
small percentage of bentonite is mainly used for long bored piles in Bangkok. First, a
small dosage of bentonite powder is premixed with fresh water prior to polymer powder
introduction. The predetermined dosage of dry polymer is then added and mixed with
the betonite slurry. Premixed polymer-based slurry is stored in the steel tanks before
supplying to the bored pile drilling. The reason for adding bentonite for long bored piles
is to minimize the fluid loss which was observed from the trial boreholes drilled only
with pure polymer slurry. Depending on the thickness of sand layers present at site and
the length of pile, addition of bentonite may not be necessary. Close monitoring was
made by sonic caliper measurement (KODEN drilling monitoring equipment) to check
the stability of the borehole. The results of the sonic caliper measurement proved that
borehole can be well stabilized with the use of polymer-based slurry at total
construction time prior to concreting even over 24 hours. According to the cross-hole
sonic logging or sonic coring tests conducted on bored piles, there is no evidence that
degraded concrete remains along the pile shaft due to the use of polymer-base slurry.
Load tests were carried out to confirm the capacity of bored piles constructed
with the alternative drilling slurry. As most of the test piles constructed with polymer-
based slurry gave smaller settlement values, particularly at maximum test load than
those obtained from piles constructed with bentonite slurry, confidence in using
polymer-based slurry was raised among the designers and the contractors.
Review of Shaft Friction Capacities Affected by Drilling Slurry
Shaft capacity degradation of wet-process bored piles may be attributed to three main
factors : reduction of effective stress at borehole interface, properties of slurry and
121
146
excessive filter cake formation. Reduction of effective stress at borehole interface

caused by installation process of wet-process bored pile is theoretically explicable and
understandable but to the authors’ knowledge, detailed study on the issue has not been
well documented.
Properties of bentonite influence on the shaft capacity degradation has been
studied by many researchers. Thasnanipan et al. (1998) concluded from the case study
conducted on 11 bored piles in Bangkok subsoil that viscosity of bentonite slurry not
more than 55 sec/qt (Marsh Funnel Viscosity) does not have significant effect on the
shaft load transfer.
A number of researchers reported the reduction of shaft capacity caused by
excessive formation of the filter cake on borehole wall of bored piles constructed with
bentonite slurry. It is mainly contributed by long construction time without slurry
agitation as cited by them.
Few researches are available for the effect of polymer-based slurry on shaft
friction capacity of bored pile. Ata and O’Neill (1998) concluded that unit side shear
resistances in both clay and sand layers of polymer test piles are higher than those of
bentonite piles. Bustamante and Gianeselli (1998) carried out the comprehensive study
on the performance of polymer bored piles. The authors reported that no reduction of
shaft friction was observed from load test results of polymer bored piles in comparison
with bentonite bored piles.
Littlechild and Plumbridge (1998) concluded from test bored piles conducted in
Bangkok, that the shaft friction resistance of bored piles constructed under polymer
slurry shows no significance change with construction times up to 42 hours. Based on
the test results of bored piles they also reported a slight reduction in shaft resistance
with decreasing polymer viscosity.
Design Parameters for Bored Piles
Following empirical formula are normally adopted to estimate the ultimate skin friction
capacity of bored piles.
Total stress approach is used to calculate the shaft friction capacity of clay
layers.
fs = α.Cu
where, α = adhesion factor
Cu = undrained shear strength
Undrained shear strength, Cu is usually obtained directly from laboratory test

results for soft clay layer and indirectly estimated from SPT “N” values for deeper stiff
to hard clay layers.
Effective stress approach is applied to calculate the shaft friction capacity of

sand layers.
fs = β.σv’
where, σv ’ = effective overburden pressure
β = shaft friction factor = Ks. tanδ
Ks = coefficient of horizontal earth pressure
δ = coefficient derived from internal friction angle φ of sand layer
122
DEEP FOUNDATIONS 2002 147
Effective overburden pressure, σv’ at existing piezometric drawn down condition

is usually used to calculate shaft friction capacity of sand layers. The design lines of α
and β of the previous research data which were developed from the instrumented load
test results of bored piles constructed with bentonite slurry are commonly used in
estimation of pile capacity. Hence, it is important to justify whether these design lines
are applicable for bored piles constructed with polymer- based slurry. Assessment is
made in the later section of this paper to justify this issue.
Load Test on Polymer-Based Slurry Bored Piles
Static pile load test on eleven test piles constructed by wet process with
polymer-based slurry are presented in this paper. Out of these, three piles were
instrumented with vibrating wire strain gauges and mechanical extensometers at
interface of different soil layers. All test piles except TP-8 were tested to minimum 2.5
times of pre-defined design load, which is commonly estimated in the preliminary
design stage.
Slurry Properties and Construction Records of Test Piles. Table 1 shows the
properties of polymer-based slurry and construction records of test piles. It is to be
noted that all these test piles were constructed by the same contractor. Hence it is
assumed that influence of other variables such as equipment as well as construction
procedure used in drilling, slurry mixing and quality control measures applied for all
piles would be minimized in evaluation of the load test results.
Table 1. Technical data of polymer-based slurry used for test piles

Mixed Ratio
Properties of Slurry
Pile per m3 of water
Total Time Prior Prior to Concreting
Pile Dimension
to Concreting
ID Dia. x Depth Polymer Bentonite Sand
(Hrs : Mins) Density Viscosity*
(m) (kg) (kg) pH Content**
(g/ml) (sec/qt)
(%)
TP-1 0.80 x 49 0.50 12 17 : 20 1.06 40 8 0
TP-2 0.80 x 50 0.50 12 12 : 20 1.03 45 8 0.5
TP-3 1.20 x 51 0.50 10 14 : 00 1.01 44 8 0
TP-4 1.80 x 62 0.50 10 23 : 00 1.04 51 8 0
TP-5 1.50 x 54 0.50 12 15 : 30 1.04 44 8 0.5
TP-6 1.00 x 51 0.22 5 7 : 00 1.02 36 8 0
TP-7 1.00 x 30 0.50 5 3 : 45 1.03 45 8 0
TP-8 0.80 x 41 0.50 10 6 : 55 1.01 44 9 0.1
TP-9 0.80 x 43 0.50 12 8 : 00 1.03 45 8 0.5
TP-10 0.60 x 26 0.50 5 2 : 30 1.02 43 8 0
TP-11 0.60 x 24 0.70 - 0 : 45 1.02 42 8 0

Note : *Marsh Funnel Viscosity, **API RP 13-B
123
148
Prediction vs Performance of Test Piles. As none of the test piles except TP-5 was
tested to failure load, no attempt is made to determine the capacity of the piles from the
failure load criteria for further assessment. Instead the performance of test piles is
evaluated from load-settlement characteristics as described below.
As large quantities of tested-to-failure instrumented load test results of bored
piles constructed with bentonite slurry in Bangkok are available, load-settlement
behavior of this pile type can be accurately predicted using back-analyzed data. Load-
settlement behavior of all test piles constructed with polymer-based slurry presented in
this paper were predicted by taking the parameters obtained from back-analyzed data of
bentonite bored piles.
Predicted load-settlement curves are compared with those of actual test results as
shown in Figure 2 to 12. It is to be noted that the term “gross settlement” indicated in
the graphs represents the total pile head movement at applied load. The objective of
these comparisons is to assess from the actual load test results, whether or not use of
polymer-based slurry in bored pile construction affects the capacity of bored piles.
The summary of predicted and measured load-settlement data is presented in
Table 2. At the design load, measured pile head movement of most of the piles are
within 5 mm as can be seen in the table. It is also evident that movement measured from
the field load test is smaller than that of prediction for most of the test piles, and is more
clear at the applied load ranging from 1.5 to 3 times of pre-defined design load. This
finding pointed out that better capacity of bored pile constructed in Bangkok can be
achieved with use of polymer-based slurry. The pattern of the load-settlement curves
obtained from the filed tests suggested that majority of the applied load is resisted by
the shaft friction. Therefore, it is reasonable to conclude that overall shaft frictional
resistance of bored pile is improved with the use of polymer-based slurry.
SPT-N (Blows/ft)
1 0 50 100
0 0
Soft to
Applied Load (kN)
Medium
10 Clay 10 0 2000 4000 6000 8000 10000 12000
0
Gross Settlement (mm)
Stiff to Very
20 Stiff Clay 20 -5
-10
Depth (m.)
Medium to
30 Very Dense 30 -15
Sand Measurement
Hard clay
-20
40 40 Prediction
Medium to -25
Very Dense
50
-30
50 Sand
Very Dense
Sand
60 60
Figure 2. Load-Settlement curves of TP-1 (Dia. 0.80 x 49 m) : Prediction vs. Measurement.
124
149
SPT-N (Blows/ft)
1 Applied Load (kN)
0 50 100
0 0 0 2000 4000 6000 8000 10000 12000
Soft Clay
0
10 10
Loose Sand
-10

Stiff to Very
Depth (m.)
20 20
Stiff Clay -20
30 30 -30
Dense to Very
Dense Sand
-40
40 Hard Clay 40 Measurement
-50 Prediction
Dense to
Very 50
50
Dense Sand -60
60
60
Figure 3. Load-Settlement curves of TP-2 (Dia. 0.80 x 50 m) : Prediction vs. Measurement.

SPT -N (Blows /ft)
1
Applied Load (kN)
0 50 100
0
0
Soft Clay
0 4000 8000 12000 16000 20000
0
10 10
Stiff Clay
20
-10
Depth (m.)
Hard Clay 20
Dense Sand -20

30 30
Hard Clay Measurement

40 -30 Prediction
40
Dense to
Very Dense 50
50 Sand
-40
60
60
Figure 4. Load Settlement curves of TP-3 (Dia. 1.20 x 51 m) : Prediction vs. Measurement.
STP-N (Blows/ft)
0 50 100 Applied Load (kN)

0 0 5000 10000 15000 20000 25000 30000 35000
Soft to
10 Med. Clay
-5
Stiff Clay
20
Dense
-15
30 Sand
Depth (m.)
Measurement
40 Prediction
Hard Clay
-25
50 Very
Dense
Sand
60 -35
Hard Clay
70
Figure 5. Load-Settlement curves of TP-4 (Dia. 1.80x62 m) : Prediction vs. Measurement.
125
150 DEEP FOUNDATIONS 2002
STP-N (Blows/ft)
0 0 0 5000 10000 15000 20000 25000 30000 35000
Very Soft to 0
10 Soft clay

10
-25
20
Stiff to Very -50
20
Stiff Clay
-75
Depth(m.)
30
Measurement
30
-100 Prediction
V. Dense Sand
Hard Clay -125
40 40
Medium
Dense to -150
Very Dense
50 50
Sand
Hard clay
60 60
SPT-N (Blow/ft)

0
Soft Clay 0 1500 3000 4500 6000 7500 9000 10500 12000
10 0
Stiff Clay
-5
20
-10
Dense Sand
Depth (m.)
30 -15
V. Stiff Clay
-20
40 Measurement
Dense Sand -25 Prediction
50 -30
-35
Very Dense
60 Sand
70
SPT-N (Blows/ft)

0
Soft Clay 0 1000 2000 3000 4000 5000 6000 7000
5 0
Soft to
-5
10 Med. Clay
-10
Stiff Clay
Depth ( m.)
15
-15
Measurement
20 Hard Clay -20 Prediction
-25
25
Very -30
Dense Sand
30 -35
35
126
STP-N (Blows/ft)

0
0 2000 4000 6000 8000 10000 12000
Soft - Med.
Stiff Clay
0
10

Medium to -20
20 Dense Sand
-40
Depth (m.)
Stiff Clay
30
Measurement
-60
Dense to
Prediction
40 Very -80
Dense Sand
50 -100
60
STP-N (Blows/ft)
0 0
0 1000 2000 3000 4000 5000 6000 7000
Soft to 5
Medium Clay 0
10 10
15 -5
Stiff to Very
20 20
Depth (m.)
Stiff Clay
25 -10
Medium to Measurement
30 30
Dense Sand Prediction
-15
35
40 Hard Clay
40
Dense to -20
V. Dense Sand 45
50
50
Figure 10. Load-Settlement Curves of TP-9 (Dia. 0.80x43 m) : Prediction vs. Measurement
SPT-N (Blows/ft)
0 50 100
Applied Load (kN)
0
0 500 1000 1500 2000 2500 3000
Very Soft 0
5
to Soft Clay
-2
10
-4
Stiff to Very
Depth (m.)
15 -6
Stiff Clay
Measurement
20 -8 Prediction
Dense to Very -10

25 Dense Sand
-12
30
35
Figure 11. Load-Settlement curves of TP-10 (Dia. 0.60x26 m) : Prediction vs. Measurement
127
SPT-N (Blow/ft)
0 50 100
0
Applied Load (kN)
5
Soft Clay 0 500 1000 1500 2000 2500
0

10
Medium Clay
Depth (m.)
15 -5
Stiff Clay
20 -10 Measurement
25
Prediction
-15
Dense Sand
30
-20
35 Hard Clay
40
Figure 12. Load-Settlement curves of TP-11 (Dia. 0.60x24 m) : Prediction vs. Measurement
Table 2. Summary of predicted and measured gross settlement

Pile Design Gross Settlement at Applied Load (mm)
Maximum
Dimension Load, Prediction Measured
Pile ID Test Load
Dia. x Depth DL
(kN)
(m) (kN) DL 2xDL 2.5xDL DL 2xDL 2.5xDL
TP-1 0.80 x 49 3300 9900 5.50 12.70 17.50 3.90 9.80 14.95
TP-2 0.80 x 50 3300 9900 6.50 16.52 24.20 5.20 14.74 21.05
TP-3 1.20 x 51 7200 18000 6.62 17.50 29.80 4.61 14.19 19.76
TP-4 1.80 x 62 11000 33000 5.62 14.07 20.55 4.00 9.88 12.96
TP-5 1.50 x 54 12000 32000 9.31 26.29 127.10 8.60 12.79 23.03
TP-6 1.00 x 51 4500 11250 5.50 13.91 19.80 2.49 7.10 9.15
TP-7 1.00 x 30 2600 6500 3.50 10.65 20.80 1.67 3.96 5.38
TP-8 0.80 x 41 3300 9900 5.72 15.90 30.20 2.75 8.05 11.60
TP-9 0.80 x 43 3300 6600 6.20 15.10 - 4.52 11.20 -
TP-10 0.60 x 26 1000 2500 2.56 6.35 9.50 1.61 3.76 4.55
TP-11 0.60 x 24 800 2000 2.59 7.84 14.30 1.09 3.19 4.22
Instrumented Pile Load Test. Shaft frictional resistance of polymer-based bored piles
in sand layers were of major interest to the authors as it is considered that drilling slurry
has less influence on clay layers according to the previous studies on bentonite bored
piles.
Three instrumented pile load tests were available at the time of this paper
preparation. Vibrating wire strain gauges (VWSG) and mechanical extensometers (ME)
were installed at the major soil boundaries for all 3 piles. It is to be noted that the lowest
level of strain gauges and extensometer of test piles TP-1, TP-2 and TP-3 are located at
1.5, 1.0 and 1.0 m above the relevant pile tip level respectively.
128
Load distribution curves along the test pile shaft at various applied load are
shown in Figure 13 to 15. The values of unit skin friction developed by the maximum
test load at the different soil layers along the shaft of test piles in comparison with those
of ultimate unit skin friction calculated using the formula shown in the earlier section
are also illustrated in these figures.
SPT-N (Blows/ft) Load (kN) Unit Skin Friction (kN/m2)

1 0 4000 8000 12000
0 50 100 0 50 100 150
0 0 0.0 0
Soft to
Medium Clay
10 10 10.0 10
Stiff to
Very Stiff 20.0
20 20 20
Clay
Depth (m.)
Medium Dense
30 to V.Dense 30 30.0 30
Sand
40 Hard Clay 40 40.0 40

Medium to
V.Dense Sand
50 50 50.0 50
Mobilized at 9900 kN
Very Dense Sand Qult (calculated)
60 60 60.0 60
Figure 13. Load distribution curves and unit skin friction of TP-1 (Dia. 0.80x49 m).
SPT-N (Blows/ft) Load (kN) Unit Skin Friction (kN/m2)

1
0 50 100 0 4000 8000 12000 0 50 100 150
0 0 0 0
Soft clay
10 10 10 10
Loose Sand
Depth (m.)
20 Stiff to Very 20 20 20
Stiff Clay
30 Dense to Very 30 30 30
Dense Sand
Hard Clay
40 40 40 40
Dense to Very
Dense Sand
50 50 50 50
Qult (calculated)
60 60 60 60
129
154
1 SPT-N (Blows/ft) Load (kN) Unit Skin Friction (kN/m2)

0 50 100 150 200 250
0 0 50 100 0 6000 12000 18000
0 0
0
Soft Clay
10 10
10 10
Stiff Clay
20 20
20
Depth (m.)
V. Stiff to 20
Hard clay
Dense Sand
30 30 30
30
Hard Clay
40 Dense to 40 40
40
Very Dense
Sand
50 50 50 50
Qult (calculated)
60 60 60 60
A summarized comparison between measured total shaft friction resistance at

maximum test load and that of theoretically calculated values of 3 test piles is presented
in Table 3.
Table 3. Calculated and measured shaft resistance

Pile Total Shaft Resistance of Pile (kN)
Design Maximum
Pile Dimension Measured /
Load Test Load Measured at
ID Dia. x Depth Calculated Calculated
(kN) (kN) Maximum Test Load
(m)
TP-1 0.80 x 49 3300 9900 6900 9710 1.41
TP-2 0.80 x 50 3300 9900 6200 9670 1.56
TP-3 1.20 x 51 7200 18000 10820 17060 1.58
Displacements required to mobilize ultimate shaft friction of piles are 0.5 - 2 %

of the shaft diameter (Reese, 1978). As the total pile head displacements of TP-1, TP-2
and TP-3 are 2.8, 3.9 and 1.6 % of the shaft diameter respectively, it is likely that the
shaft friction of these piles were fully mobilized at relevant maximum test load.
However, the pile displacement value required to reach peak shaft resistance depends
upon many conditions of the pile-soil interface and the stress strain behavior of
surrounding soils : soil characteristics, stresses of the ground surrounding the pile shaft,
construction methods of bored piles, time from construction to a loading test, loading
procedure and speed, etc. (Hirayama, 1990). Moreover, shaft friction development
against pile displacement is also influenced by the pile length. According to
Thasnanipan et al. (1998), based on the various instrumented static pile load test
conducted on long bored piles in multi-layered soil of Bangkok, shaft frictions of deeper
soil layers were not fully mobilized though the piles were declared to have failed due to
the excessive pile head movement. This finding may also be valid for bored piles
constructed with polymer-based slurry. Figure 16 illustrates the shaft friction versus pile
head movement of different soil layers obtained from test pile TP-1. As can be seen in
the figure, at the maximum test load where pile head movement reached 2.8 % of the
130
shaft diameter, the shaft frictions of upper soil layers (soft clay and stiff to very stiff
clay) were fully mobilized and it is likely that medium dense sand (first sand) layer
almost reached to ultimate shaft friction resistance. However, it is most likely that very
dense sand (second sand) was not fully developed to the ultimate stage. The same
behavior is observed for test piles TP-2 and TP-3.
Shaft friction factors β of the first and second sand layer of TP-1, TP-2 and TP-3
are back calculated from the load transfer values of relevant layers obtained from the
VWSG data. However, the back-calculated β factors (named as pseudo-β for purpose of
convenient reference in this paper) particularly for the second sand layer are the values
at applied maximum test loads and not at ultimate test loads. In other words, it is likely
that the back-calculated β factors for the second sand layer are the values at partially
mobilized shaft friction of this layer. The back-calculation of β value is mainly aimed to
get an idea on the magnitude of β factors for polymer-based bored piles from the
currently available data in comparison with the design values commonly used for
bentonite bored piles constructed in Bangkok.
Figure 17 shows the values of pseudo-β of polymer-based slurry piles obtained
from 3 instrumented pile load tests plotted on the graph of β values derived from the
full-scale static pile load tests of bentonite bored piles in Bangkok subsoil from the
previous research works (Submaneewong, 1999). Submaneewong (1999) recommended
the design line of β factors for bentonite bored piles in the first and second sand as
indicated in Figure 17. It is evident that pseudo-β for the first sand layer of all 3 test
piles constructed under polymer-based slurry are significantly higher than those of
design values. The pseudo-β values of the second sand layer are also slightly higher
than those of design line. It is most likely that β values of second sand layers would be
higher than pseudo-β values indicated in Figure 17 if the piles were tested to higher
loads or to failure, since shaft friction at this layer might not have been fully mobilized
at applied maximum test load as described in above paragraphs. These findings
suggested that polymer-based slurry bored piles produced better shaft friction resistance
in sand layers than that of bentonite bored piles. Further research is needed to determine
the β values of the first and second sand layers at fully mobilized shaft friction
resistance.
200
Soft clay
Stiff to very stiff clay
Medium dense sand
Stiff to very stiff clay Very dense sand
150
Unit Skin Friction (kN/m 2)
Medium dense sand

(First sand)
100
Very dense sand

(Second sand)
50
Soft clay
0
0 4 8 12 16 20 24
Pile Head Movem ent (m m )
Figure 16. Unit skin friction versus pile head movement of different soil layers (TP-1).
131
1 .0
Submaneewong (1999) : Bored Piles (Tip in 2nd Sand)
Submaneewong (1999) : Bored Piles (Tip in 2nd Hard Clay)
Chiewcharnsilp (1988)
0 .8
Polymer Bored Pile TP-1 (1st Sand)
0 .6 Polymer Bored Pile TP-1 (2nd Sand)
β (Ks tan δ)
Polymer Bored Pile TP-2 (2nd Sand) Submaneewong

Polymer Bored Pile TP-3 (2nd Sand)
( 1999 )
0 .4
0 .2
0 .0
28 .0 30 .0 32 .0 34 .0 36 .0 38 .0
Angle of Internal Friction, φ '
Figure 17. Back-calculated β values (pseudo-β factor) of polymer bored piles at

maximum test load plotted on design line of bentonite bored piles constructed in
Bangkok subsoil .
Conclusions
(1) Performance of bored piles constructed with polymer-based slurry in Bangkok

subsoil have been presented using the results of eleven static pile load tests.
(2) The measurement of borehole profile by sonic caliper and construction records
proved that there is no adverse effect of using polymer-based slurry in Bangkok
subsoil as far as the stability of borehole is concerned.
(3) Pile head settlements at applied loads obtained from the field load test results of all
test piles constructed with polymer-based slurry are smaller than those of prediction
using bentonite piles parameters.
(4) Overall shaft resistance of bored piles constructed with polymer-based slurry is
higher than that of theoretically calculated values in the order of 1.5.
(5) The back-calculated β factors of current polymer-based slurry test bored piles are
higher than those of design values commonly used in Bangkok for bentonite bored
piles.
(6) Above findings suggested that higher shaft friction resistance can be obtained if pile
is constructed with polymer-based slurry.
(7) Further research is needed to establish the design parameters for bored piles
constructed with polymer-based slurry in Bangkok subsoil.
Acknowledgement
The authors wish to express their appreciation to the project directors and engineers of
SEAFCO Co., Ltd. for providing the bored pile construction records. Thanks are also
132
extended to EDE Co., Ltd. for providing the load test data. The authors also wish to
thank Miss. Siriphen Chaiyapak (former office engineer, SEAFCO Co., Ltd.) and Mr.
Artorn Kantawateera (trainee student, Khonkaen University) for their assistance in
preparing this paper.
References
Ata A. A. and O’Neill M. (1998). “Side-wall stability and side-shear resistance in

bored piles constructed with high-molecular-weight polymer slurry,” Third
International Geotechnical Seminar on Deep Foundation on Bored and Auger Piles,
Ghent, Belgium, 19-21 October 1998, Balkema, Rotterdam, 111-117.
Bustamante M. and Gianeselli L. (1998). “Performance of polymer slurries in large

diameter bored pile,” Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium, 19-21 October 1998, Balkema, Rotterdam
119-127.
Harayama H. (1990). “Load settlement analysis for bored piles using hyperbolic transfer
functions,” Soil and Foundations, Vol. 30, No. 1, Japanese Society of Soil Mechanics
and Foundation Engineering, 55-64.
Littlechild B. and Plumbridge G. (1998). “Effects of construction technique on the

behavior of plain bored piles constructed under drilling slurry,” Seventh International
Conference and Exhibition on Piling and Deep Foundations, Deep Foundation Institute,
Vienna, Austria, 15-17 June 1998, 1.6.1-1.6.8.
Nutalaya, P. (1981). “Subsidence Studies of the Bangkok Plain,” AIT Research Report,
Asian Institute of Technology, Bangkok, Thailand.
Reese L. C., (1978). “Design and Construction of Drilled Shafts,” Twelfth Terzaghi
Lecture, Proc. ASCE, Vol. 104, No. GT1, 1978, 95-116.
Submaneewong C. (1999). “Behavior of instrumented barrette and bored piles in

Bangkok subsoils,” (in Thai), M.Eng. Thesis, Department of Civil Engineering, Faculty
of Engineering Chulalongkorn University, Bangkok, Thailand.
Thasnanipan N., Ganeshan B. and Anwar M.A. (1998). “Effect of construction time
and bentonite viscosity on shaft capacity of bored piles,” Third International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent, Belgium,
19-21 October 1998, Balkema, Rotterdam, 171-177.
Thasnanipan N., Tanseng P. and Anwar M. A. (1998). “Large diameter bored piles in
multi-layered soils of Bangkok,” Third International Geotechnical Seminar on Deep
Foundation on Bored and Auger Piles, Ghent, Belgium, 19-21 October 1998, Balkema,
Rotterdam, 511-518.
133
GEOTECHNICAL SPECIAL PUBLICATION NO. 116

An International Perspective on Theory,
Design, Construction and Performance
Volume Two
PROCEEDINGS OF THE INTERNATIONAL

DEEP FOUNDATIONS CONGRESS 2002
SPONSORED BY
February 14-16, 2002

Orlando, Florida
EDITED BY
Michael W. O’Neill
Frank C. Townsend
135
Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok
Narong Thasnanipan*, Zaw Zaw Aye** and Wanchai Teparaksa, Ph. D.***
*Managing Director, **Project Manager, SEAFCO Co., Ltd., 26/10 Rarm Intra 109
Rd., Bangchan, Klong Sam Wah, Bangkok 10510, Thailand, Tel. (662) 9190090, Fax.
(662) 9190098, seafco@seafco.co.th, www.seafco.co.th
***Professor, Department of Civil Engineering, Chulalongkorn University, Bangkok,
Thailand, fcewtp@eng/chula.ac.th
Abstract
A static load test carried out on barrette having cross sectional size of 1.5 m x 3.0 m
seated at 57 m below ground level is presented in this paper. Over 50,000 kN load was
applied to evaluate the performance of this fully instrumented barrette. The load applied
on barrette with a reaction frame “tower” capacity of 60,000 kN was claimed to be one
of the highest test loads in the region. The test results were compared with those from
the instrumented load test on bored pile of diameter 1.5 m with the same length, being
located 30 m away. Since bored pile and barrette were constructed by different method
and the parameters such as construction time, pile shape, slurry properties etc, were
considerably different, results from the load tests have provided a unique opportunity to
assess the extent of difference in behavior.
Introduction
In the area of foundation engineering, Bangkok, the capital of Thailand, is well known
for its weak soils. As such, most of the structures are supported by piled foundations.
Construction of heavy structures such as sky scrapers, elevated expressways, overpass
bridges, underground car parks and subway stations commonly require the high
capacity foundations. Bored cast-in-place piles are mainly used to fulfill this
requirement. However, in some projects bored piles are not feasible due to the site
constraints, applicable construction method and/or extensive bearing capacity
requirements. In such cases, the use of barrette foundations would make a suitable
alternative. Barrettes with dimension ranging from 0.80 m x 2.7 m to 1.5 m x 3.0 m for
safe working load capacity from 11,000 to 23,000 kN have been used in some major
projects. To assess the performance of large-capacity foundation elements in layered
subsoil of Bangkok, instrumented load testing is compulsory. This paper presents the
static load test results of an instrumented barrette tested up to 52,900 kN for the
foundation of a fifty-storey building in Bangkok. The test results, particularly load
transfer characteristics, and shaft friction capacity were compared with those of bored
pile diameter 1.50 m with the same length located 30 m away.
1073
137
Project Summary
The project is located in the prime area of Bangkok on Rama IX road. The design of
fifty-storey office complex called for the high-capacity piled foundation : 560 bored
piles of 1.2 m and 1.5 m diameter and 24 barrettes having cross section size of 1.5 m x
3.0 m. Bored piles and barrettes were seated at an approximate depth of 57 m in the
sand layer. Barrettes were designed to support the large load for tower’s elevator shafts
as bored piles were not feasible in this case. Base grouting was applied for barrettes and
bored piles at the locations of high column load mainly in the central tower area.
Instrumented static pile load tests were carried out for one barrette and one 1.5 m
diameter bored pile. The design safe working loads for the base grouted barrettes and
bored piles are 23,000 kN and 11,750 kN respectively. The barrette load test was
conducted on July 8, 1998 with maximum applied load of 52,900 kN.
Subsoil Profile and Ground Water Condition
Similar to other localities in Bangkok a typical subsoil profile at the site is characterized
by the alternating layers of clay and sand deposits as shown in Figure 1. Undrained
shear strength (Su) from unconfined compression tests and standard penetration tests
(SPT) are indicated with depth in Figure 1. It is to be noted that unconfined
compression tests were conducted only for soft clay layer whereas SPT tests were
carried out for all the layers below soft clay. Soil investigation from five boreholes at
different location reveals that subsoil layers along the site are relatively consistent. The
soil layers present at site are briefly described below.
S u ( U C ) kN/m2
Pore Pressure (kN/m 2 )
0 50 100
DEPTH (m) 0 200 400 600
0 Weathered 0
Crust
S oft Clay -
-10 1
0
-20 Stiff Clay

Ac
tu a
Medium to
-30
lP
Dense Sand
Hy
ie
d ro
zom
s ta
-
e tr
tic
-40 4
ic D
Hard
L in
0
ra w
Clay
e
dow
-
-50 5
nL
Dense Sand 0
in
H ard C lay
e
Dens e to Very -
-60 Dense Sand 6
0 50 100 0
S P T - N ( b l o w s / 3 0 cm)
Figure 1. Soil profile at site and the typical ground water condition of Bangkok
138
Soft, highly compressible dark gray marine clay lies beneath a two meter thick
weathered crust and extends up to 13.5 m. A stiff clay layer occurs directly underneath
soft clay and its depth goes up to 26 m. Below the stiff clay layer, 10 m thick medium to
dense sand layer also known as the first sand layer can be found. Hard clay underlies
the first sand and it is found to be about 16 m thick. Dense to very dense sand (the
second sand layer) occurs at depths between 51 to 72 m. Within the second sand layer, a
thin layer of hard clay is found.
Figure 1 depicts the present typical piezometric profile of Bangkok. As can be
seen in the figure, existing pore water pressure conditions in upper part of Bangkok soft
clay are hydrostatic from nearly 1m below the ground level and then the hydrostatic
condition changed to piezometric drawdown below the bottom level of Bangkok soft
clay. Excessive ground water abstraction from the deep aquifers in the past is
responsible for this significant drawdown of piezometric pressure (Nutalaya, 1981,
Thasnanipan et al., 1998).
Installation Method and Process
Both the barrette and bored pile were constructed by wet process under bentonite slurry.
Bentonite slurry conformed to the locally accepted specification (AASHTO 1992) was
used. The actual properties of the bentonite slurry used are given in Table 1 along with
those specified by the project. Single stage base grouting was applied 24 hours after
concreting for both barrette and bored pile. The detailed procedures utilized in the
construction of two different foundation structures are outlined below.
Table 1. Comparison of bentonite slurry properties

Measured Actual Properties
Barrette Bored Pile
Project Before feeding After Recycling Before feeding After Recycling
Properties Specified to the borehole & Before to the borehole & Before
Values Concreting concreting
(near trench (near borehole
base) base)
Density
1.03-1.20 1.10 1.17 1.08 1.10
(g/cc)
Marsh
Funnel
28-45 36 49 33 36
Viscosity
(sec/qt)
API Sand
Content < 4.0 1.0 1.1 0.1 0.8
(%)
pH value 8-12 8 9 8 8
A mechanical cable-suspended grab was used for barrette construction.

Excavation of the trench was carried out by the cyclic process of lifting and lowering of
the grab under gravity and tangential force of the clamshell operated by cables. A guide
wall with inside clear dimensions slightly larger than the nominal size of the barrette
was used to guide the grab during initial bites. Bentonite slurry was introduced to the
trench as soon as initial excavation commenced. The excavation was continued under
139
the bentonite slurry to the final depth. Since time consumed in the preparation of
instrumentation was relatively long, circulation of the bentonite slurry was continuously
done to keep the slurry agitated, which also helped to alleviate the growth of filter cake
by minimizing the actual exposure time. Prior to lowering the reinforcement cage the
trench was once again occupied by grab to scrape the trench walls to remove, if any,
filter cake formed on the walls. This attempt is in line with the recommendation made
by Reese and O’Neill (1988). It is authors opinion that, if due to some unforeseen
reasons, reinforcement cage lowering have to be delayed for considerable period of
time, it is a good practice to use the grab again to scrape the trench walls. This measure
eliminates any foreseeable negative impacts caused by unexpected delays. After
lowering the rebar cage, tremie concreting was done.
Hydraulic rotary drilling was employed for bored pile excavation. Different
from barrette excavation, temporary casing of 15 m length was used as a support in Soft
Clay layer for bored pile drilling to assure the stability of the borehole. Firstly, an auger
was used to drill within the temporary casing and to the bottom of the stiff clay layer.
Bentonite slurry was then supplied to the borehole and drilling was continued with a
bucket down to final depth of excavation. Before lowering the reinforcement cage a
special cleaning bucket was used to scrape the borehole walls and to clean the base.
Reinforcement cages were then lowered inside the borehole and concreting was carried
out by tremie method. Time consumed in different construction activities of test
barrette and bored pile is shown in Table 2. Equipment used in the excavation of
barrette and bored pile are shown in Figure 3.
Table 2. Time consumed in different construction activities of test barrette and bored pile
Time Consumed (hours)
Activities
Barrette Bored Pile
Drilling 19 7
Recycling and Desanding 36 4
Reinforcement Cage Lowering 10 7
Delay 2.5 4.5
Concreting 7.5 4.5
Total Time Consumed 75 27
Pile Toe Grouting Process
Pile toe grouting was conducted on both test piles. A special grouting pipe system (a
grouting circuit) in the form of U-loop consisting of 0.5” PE pipes connected with 0.5”
steel pipe at the bottom was used. The steel pipe horizontally placed at the bottom of the
U-loop is perforated and wrapped by a rubber sleeve "Manchette". Three grouting
circuits (3 loops) were used for barrette whereas two grouting circuits (2 loops) were
employed for bored pile. The grouting circuits were attached to the reinforcement cage.
Twenty-four hours after concreting, the pile toe was cracked by pressurized
water to flush and open the grouting circuit so as to make a way for forth coming grout.
Cement grouting was commenced soon after completion of the water cracking. The
preset grouting criteria was to stop the grout injection when the grout volume reached to
140
the maximum target volume (1500 and 750 liter for barrette and bored pile respectively)
or the maximum target grouting pressure (40 kN/m2) was achieved, whichever came
first. The grouting for test barrette and bored pile was stopped by the volume criteria
since the maximum grouting pressure achieved for both piles was only 20 kN/m2.
BARRETTE BORED PILE
3.0 m.
DIA. 1.50 m.
1.5 m.
Po re P re ss ure (kN/m 2 )
Depth (m) 0 200 400 600
0 0
Weathered Crust
GUIDE WALL TEMPORARY
Soft Clay STEEL CASING -
-10 1
0
-20 S t iff Clay
BENTONITE SLURRY
Ac
tu a
-30 Medium to
lP
Hy
Dense Sand
ie
dro
z om
st a
-
e tr
t ic
-40 4
ic D
Hard Clay 0
Li n
r aw
e
dow
-
-50 5
n L
Dense Sand 0
i ne
Hard Clay -57.0 m. -57.0 m.
Dense to Very
-
-60 Dense Sand
6
0
Figure 2. General configuration of test barrette and bored pile
Mechanical cable-suspended Auger used for upper part of bored Bucket used for bored pile
grab used for barrette excavation pile drilling drilling under slurry
Figure 3. Excavation and drilling tools of barrette and bored pile
141
Major Differences of Two Test Piles
Basic installation parameters of the test barrette and bored pile are considerably
different as shown in Table 3. The general configuration of test barrette and bored pile
with soil profile at site is illustrated in Figure 2. The influence of the major differences
indicated in Table 3 on capacity of piles are discussed in later sections of this paper.
Table 3. Major differences in installation process of barrette and bored pile
Major Differences
Test Pile Equipment and Construction Total Shape Properties

Type Drilling Process Method / Use of Construction of
Temporary Time Bentonite
Casing
Barrette Mechanical grab None 75 hours Rectangular Different
with excavation as
under gravity and indicated
grab action in Table 1
Bored Hydraulic rotary 15 m Steel 27 hours Circular Different
Pile piling rig with Casing as
rotary drilling indicated
action in Table 1
Load Test Program
Pile load tests were carried out in accordance with ASTM D 1143-81 Item 5.1
“Standard Loading Procedure”. The pile and barrette were tested under three and four
cycles of loading and unloading respectively.
Test Frame Configuration. Test barrette and bored pile layout is presented in Figure 4.
General view of the barrette and bored pile test frame set up is presented in Figure 5.
With overall height of 8 m, reaction frame utilized for static load testing on barrette in
this project was claimed to be one of the biggest of its kind in the region. Four barrettes
were used as anchoring system. Four numbers of built-up steel girders supported on
each side by two 1st level cross beams were used as main beams to achieve the
maximum capacity of 60,000 kN. First level beams were supported against the second
level cross beams. Second level cross beams were anchored against surrounding
barrettes using anchor blocks at the top. Specially fabricated transfer-girders were used
to distribute the tension force transferred from the tie-bars to dowel bars above the
anchor barrette heads. Sixteen hydraulic jacks each having 5,000 kN capacity were
placed between the test barrette cap and the main beams of the reaction frame.
Test pile layout of bored pile was similar to those of other static bored pile load
tests in Thailand as it was considered that preset maximum test load would not require
special test frame set up. Steel test frame anchored against four bored piles was used in
bored pile load test. Different from the test frame configuration of the barrette, only 1
layer of cross beams was required for that of bored pile.
142
5.00 5.00
2n d C ro ss
B eam
8.00
1st C ro ss 4.00 4.00

B eam
AN C H O R
B AR R E TTE
5.00
An ch o r P ile
M ain B eam
M ain B eam 4.00
P ile C ap
H yd rau lic
Jacks 8.00 Test P ile
TE S T
5.00
H yd rau lic Jacks

B AR R E TTE 4.00
C ro ss B eam
2n d C ro ss
B eam
1st C ro ss B eam
2n d C ro ss
B eam
Barrette Test Frame Layout Bored Pile Test Frame Layout
Figure 4. Barrette and bored pile test frame configuration
View of Barrette Test Frame View of Bored Pile Test Frame
Figure 5. General view of the reaction frame for static load test on barrette and bored pile
Instrumentation and Monitoring System. Vibrating wire strain gauges (VWSGs) and
Mechanical Extensometers (ME) were fixed at five levels along the shafts at the known
interface boundaries of different soil layers. At each level, four sets of VWSGs and one
set of ME were installed for the pile whereas six sets of VWSGs and two sets of ME
were installed for the barrette.
Direct measurement from four dial gauges placed in diametrically opposite
positions having equidistance from the test pile axis was used as a main monitoring
system of test pile head movement. Precise leveling and piano wire were also utilized as
backup for pile head movement measurement. Additional two dial gauges were also
143
Conference Papers
NINTH INTERNATIONAL
CONFERENCE AND EXHIBITION
ON PILING AND DEEP
FOUNDATIONS
3 – 5 June 2002
NICE Acropolis
France
Performance of Toe-Grouted Large Diameter Bored Pile in Sand-Gravel

Formation of Chiang Mai City, Northern Thailand
Narong Thasnanipan, Zaw Z. Aye, Chanchai Submaneewong & Dr. Wanchai Teparaksa
Sponsored by
Deep Foundations
Institute
145
PERFORMANCE OF TOE-GROUTED LARGE BORED PILE IN SAND-GRAVEL FORMATION OF CHIANG
MAI CITY, NORTHERN THAILAND
Narong Thasnanipan, Zaw Z. Aye, Chanchai Submaneewong, Seafco Company Limited, Bangkok, Thailand
Dr. Wanchai Teparaksa, Department of Civil Engineering, Chulalongkorn University, Bangkok, Thailand
Performance of diameter 1.5 m toe-grouted bored pile founded at depth 44 m in sand-gravel

formation of Chiang Mai City is presented in this paper. Toe grouting was applied by using
tube-a-manchette method. Effectiveness of toe grouting was observed from the standard
penetration tests conducted immediately below pile toe through grouting pipes at two stages,
prior to grouting and three weeks after completion of grouting. Improvement of pile end
resistance by toe-grouting is demonstrated with a series of back analyses carried out by load-
transfer method employing hyperbolic transfer function. Construction experience of large
diameter deep-seated bored piles in multi-layered subsoil and practical aspects of toe-grouting
are also briefly discussed.
INTRODUCTION of Precambrian age (BAUM et al. 1970). Rhodes et al.

(1997) named the fluvial deposits poorly exposed along
Chiang Mai founded 800 years ago is now known to be the western edge of Chiang Mai Basin as Mae Rim
the tourist city of the northern Thailand. It is located Formation and stated that this formation is folded into
about 700 km north of Bangkok at latitude 18.75 N and gentle domes and basins. Based on the discovery of a
longitude 99 E. Bored piles have been used as distinct intraformational angular unconformity, the
foundations of heavy structures in Chiang Mai. Most of authors confirmed that the Mae Rim Formation
the bored piles constructed in Chiang Mai are of small to accumulated during the opening of the Chiang Mai
medium diameter short bored piles : diameter ranging Basin. According to the record of deep borehole drilled
from 0.60 to 0.80 m with toe depths of about 20 m. up to 1184 m, the basin deposits comprise claystone of
Large diameter deep-seated bored piles were rarely Tertiary age overlain by thick layers of unconsolidated
constructed in the city as high-capacity foundations were sand, clay and gravel of Quaternary age.
not in great demand in the past. Hence, behavior and
performance of large diameter deep-seated bored piles Intrasuta (1983) presented the geologic setting of
in this city have not yet been studied. In recent years, Chiang Mai area based on the lithology and structural
elevated bridges and underpasses are required in the features of each unit. The geological map of Chiang Mai
urban area and the vicinity of Chiang Mai to alleviate the and the surrounding area modified from the work of
increasingly congested traffic condition. In some Intrasuta (1983) is shown in Figure 1.
projects, design of bridge pier called for a single pile per
column concept to facilitate the available space. Large SITE GEOLOGY AND SUBSOIL PROFILE
diameter deep-seated bored piles are necessary in such
cases. Construction experience of diameter 1.5 m toe- The project is located in the inner part of the lowland
grouted bored piles founded at depth 44 m in sand– floodplain area. The site subsurface geology is
gravel formation of Chiang Mai City is presented in this characterized by thick alluvial deposits composed of
paper. A total of 33 bored piles were constructed for alternating series of sand and clay layers overlaying
bridge foundation of elevated interchange. The gravel beds. A relatively thin layer of gravel formation is
instrumented static pile load test was carried out up to observed between sand and clay layer at depth about 14
25,000 KN on one of the working piles as a proof- -17 m.
loading test, which was claimed to be the highest static
pile load test ever tested in the city of Chiang Mai. Figure 2 illustrates the generalized soil profile along the
site derived from 5 boring logs with indication of SPT-N
REGIONAL GEOLOGY AND MORPHOLOGY values obtained from each borehole. As can be seen in
the figure a large variation in soil profile along the site is
Geomorphologically, Chiang Mai is characterized by flat observed. Above 35 m depth, the thickness and
floodplain occupying the most part of the city and the occurrence of strata radically varies from one borehole
eastern area bordered by the mountain ranges in the to another. However, below this depth the top and
west. The main river, Mae Ping flows through the city bottom boundary elevation of stratum 11 (sand and
center from north to south. Geologically, the lowland gravel mixture formation) is relatively uniform across the
floodplain of the city is situated in the Chiang Mai Basin site. Pile toe was designed to be founded in this layer.
which was believed to be a horst-basin type structural The simplified description of soil stratum along with SPT-
basin formed by tectonic activities during Tertiary period. N values and undrained shear strength, Su (for clay
Western mountain ranges are formed by so-called layers) are presented in Table 1. Ground water level
Chiang Mai-Tak gneiss belt composed of medium to observed from the soil investigation boreholes was about
high grade gneiss, schist, calc silicate rocks and granite 1 m below the existing ground.
147
Figure 1 Geological map of Chiang Mai and adjacent areas (modified from Intrasuta, 1983)
BH-1 B H -2 B H -3 B H -4 TE S T P ILE B H -5
0
6 15 18
STR ATUM 2 9 STR ATUM 1
11 10 4 13
M ED IU M TO 18 9 LOOSE CLAYEY 12
14 STIFF S ILTY C LAY TO SILTY SAND 8
4 5
5 11 13 16
3 16
9 11 STR ATUM 3 15
5
16 12 LO O SE SAN D 10 8
19 STR ATUM 4
16 2 3 STR ATUM 5 6 SO FT TO M ED IU M STIFF 7
10 19 3
1 8 M ED IU M TO D AR K G R EY C LAY 27
13
1 5 VER Y D EN SE SAN D 17 16 23
17
15 20
10 24 15
46 UM 6 EL
STR AT G R AV 58
DENSE
15 23 56 24
3 7 M E D IU M T O 42
17 51 64
24 21
15 26 56
26
51 UM 8 58
DEPTH (m.)
24 E 18
20 STR AT M TO D EN S S AN D 29
53 40 M E D IU Y T O S IL T Y
3 5 STR ATUM 7 31
A Y E VER Y S TIFF SILTY CLAY
40 CL 40 68 49
57
33 51
33 32 STR ATUM 9 25
25 VER Y S TIFF TO H AR D 55 46
27 35 49
SILTY C LAY 20 47
29 43 42
32 42
50 37 40
STR ATUM 10 2 6 STR ATUM 10 C LAYEY SAN D 44
30 4 9 C LAYEY S AN D 33 28
33 40 22 55
30
52 53
50 41 50
58 39
62 ST R AT U 47 36
35 M 10 53
33 C LA YE Y 37 44 57
SA ND 67 55
56 69 41
31 40 55 50/5"
40
40 50 95/11" 31 50/6"
87 STR ATUM 11
50 54
90 93/11" SAND AND 50/3"
50/0" 78 G R AV EL M IXTUR E 50/5" 50/1"
50/5"
50/0.5" 50/5" 85
45 50/0"
STR ATUM 12 VER Y D EN SE G R AVE L 50/5"
50/0"
50
Figure 2 Generalized soil profile along the project (the numbers denoted behind the borehole are SPT-N values)
148
Table 1 Summarized description of soil stratum manchette were fixed in the reinforcement cage as all
Stratum Description Unit SPT Su bored piles were required to have sonic logging test.
No. Weight “N” 2
(KN/m ) The sonic logging tubes were also utilized as toe-
(KN/m3) grouting tubes. The arrangement and configuration of
1 Clayey sand 18.0 4 to 16 - grouting circuits are presented in the later section. Once
(brownish grey)
the reinforcement cage installation was done, tremie
2 Medium to very 19.0 8 to 18 90-120
stiff silty clay pipes complete with air lifting accessories were lowered
3 Loose sand 18.5 3 to 5 - and air lifting operation commenced. Air lift was
4 Soft to medium 18.0 3 to 7 35-60 continuously perform until the arrival of the first concrete
stiff silty clay truck to minimize the accumulation of sediments at the
5 Medium to very 19.5 15 to 46 - pile base if not entirely eliminated. The depth of the pile
dense sand was re-measured to ensure that the base of the pile was
6 Gravel 21.0 21 to 68 - reasonably cleaned. Hundred percent cleaning or
7 Very stiff clay 20.0 21 to 38 150-200 removal of undesired sediments was most unlikely
8 Medium to dense 21.0 26 to 40 - particularly for the long bored piles in the presence of
clayey sand thick sand formation and layers of gravel. Concreting
9 Very stiff to hard 22.0 22 to 69 > 200
was done by tremie method. All the contract piles
silty clay
including the test and anchor piles were constructed by
10 Clayey sand 22.0 26 to 49 -
11 Very dense sand- 22.5 40 to 90 - the procedure described above. The properties of
gravel formation bentonite slurry measured from the test pile at different
12 Very dense gravel - 50 / 0” - stages are shown in Table 2.
Y'
BORED PILE CONSTRUCTION METHOD
x x'
All the contract bored piles including the test pile were
constructed by rotary drilling under wet process or slurry- Y
replacement method, using bentonite slurry as support Depth (m) x x' Y Y' Top of
fluid in boring process. Depending on the presence of slurry
0.0
loose material, temporary steel casings of length ranging Clayey
from 3 to 11 m were used to stabilize the borehole in top sand 5.0
soil layer. After installing the temporary casing by Stiff clay
hydraulic vibrator (Vibro-hammer) soil inside the casing Fine sand 10.0
was drilled out by an auger. The borehole was filled with
bentonite slurry prior to advancing the drilling below the Coarse sand
15.0
casing. Drilling bucket was then used to excavate the Gravel with
soil under the slurry. Upon reaching to 15-19 m depth, sand
20.0
verticality of the borehole was checked by Koden drilling
monitor (sonic caliper method). Koden test at this depth Very stiff 25.0
also helped to assess the stability of borehole in gravel clay with
layer (stratum 6). After the careful examination of the trace of 30.0
borehole verticality monitoring results, drilling was sand
continued to the final depth. Additional Koden tests 35.0
were performed to ensure the verticality of the borehole Dense sand
at regular interval before reaching to the final depth if 40.0
excessive inclination was found in the first test. Sand-gravel
formation
Since the sand content in the slurry after drilling to the

final depth was higher than that of specified (API 4%), Figure 3 Borehole profile of test pile from Koden Test
recycling and desanding were carried out to screen out (soil profile shown in the figure is based on the record of
the excessive sand in the bentonite and to maintain the a visual inspection of soil from pile drilling)
properties of slurry. Up to 1m3 of sand in the slurry was
removed by the recycling process. This represented
about 0.5 m height of sediment at the base of pile. After Table 2 Properties of bentonite at different stages
completion of recycling, a final Koden test was Marsh Funnel API Sand
Density
conducted. Figure 3 shows the borehole profile obtained Stage
(g/ml)
Viscosity Content pH
from the final Kodent test conducted on test pile. Pile (sec/qt) (%)
base was then cleaned by cleaning bucket to remove the
Before drilling 1.05 31 0 8
accumulated loose material deposited on the base of the
pile during drilling process. After the cleaning, the depth After recycling 1.07 33 0.5 8
of the pile was measured and recorded. Reinforcement
steel cages were then installed section by section in the After cage
1.10 30 1 8
borehole. Four sonic logging access tubes (internal lowering
diameter 54 mm black steel pipes) complete with tube-a-
149
SONIC LOGGING TEST were fixed to the reinforcement cage and installed prior
to concreting as described in the earlier section. In the
The construction specification called for the integrity initial stage of piling, manchette opening and water
assessment of all the piles by sonic logging test. Sonic cracking was done 24 hours after casting of piles and
logging test was normally carried out at a minimum of 7 cement-grout injection was performed after 7 days or
days after completion of pile concreting. Figure 4 shows more upon completion of sonic logging test. However,
the sonic logging result of the lower part of the test pile with this process difficulty was found in cement grout
tested 7 days after pile concreting. filling process as grout circuit was blocked by intrusion of
sand and very small gravels into the pipes through
perforations since manchettes were broken out during
earlier water cracking.
1500
75 75
Pile W all
Main Reinforcem ent
Sonic Tube (ID =54 m m )
Rubber Sleeve Horizontal Pipe (ID. =40 m m )
(Manchette) with Manchette
Pile Base
S EC T IO N O F G R O U T C IR C U IT
Perim eter of Horizontal Pipe (ID.=40 m m )

Bored Pile Sonic Tube (ID.=54 m m )
Rubber Sleeve
(Manchette)
75 75
1500
Figure 4 Sonic logging test profile of lower part of the P LAN O F PILE B ASE
test pile (depth indicates in the figure are measured from
the top of sonic tubes) Figure 5 Grout circuit arrangement and layout (spiral
reinforcement is not shown in the figure for clarity)
PILE TOE-GROUTING
Original Specification Modification of Grouting Process
The original project specification called for the use of drill Therefore it was decided to change the sequence by
–and-grout method for toe-grouting. Drill-and-grout performing water cracking immediately prior to the actual
method involves coring through the pile base up to cement grouting. To facilitate this change, the level of
20 cm into the soil beneath the pile base. Coring is to be manchette was modified by placing just above pile tip
carried out by drilling tools using sonic tubes as access level (in the initial stage the manchettes were placed
to the base of the pile. After completion of drilling, sonic about 10 to 20 cm above the pile tip level). After
tubes are to be cleaned and filled with grout followed by completion of the sonic logging test, manchette opening
grout injection from top of the tubes. This method was and water cracking was carried out followed by the filling
considered inappropriate as it is likely that uplift pressure of grout in the grout tubes. Each grout loop was filled
at pile toe level blow out sand gravel mixture into the with cement grout from the top by using one pipe as inlet
sonic tubes which in turn would cause the undesired and another as outlet. Cement grout was injected at
consequences ; blockage of the sonic tubes and each loop with low pressure from inlet pipe until the
progressive collapse of the soil below pile base. cement grout comes out from the outlet pipe. Once both
outlet pipes were observed with discharge of cement
Alternative Method (Tube-a-manchette Method) grout, two outlet grout pipes were connected to the grout
pump and pressurized grout injection started. Toe-
The piling contractor therefore proposed tube-a- grouting was performed through all 4 pipes
manchette technique as an alternative toe-grouting simultaneously. Grouting was stopped when the injection
method. In this method a system of grouting consisted pressure reached to 20 bar. The target grouting
of two U-shape loops formed by two pairs of inside pressure was set at 20 bar (2,000 KN/m2) for two main
diameter 54 mm main vertical steel pipes (sonic tubes) reasons ; (i) to avoid excessive hydrofracturing of
attached with horizontal steel pipe of inside diameter surrounding soil as pile base was embedded in highly
40 mm at bottom of each pair. The horizontal steel permeable sand- gravel formation, (ii) the main purpose
pipes were perforated and wrapped with rubber sleeves of the toe-grouting in this project was to improve the soft
(manchettes). The bottom of the sonic tubes were toe condition of the pile. Compacting the remaining
securely closed with a steel cap. The position loose materials at the base of the pile and compressing
arrangement and layout of grouting circuit is shown in the distressed soil immediately below the pile base with
Figure 5. The grouting circuits completed with manchette the target pressure was considered sufficient to improve
fabricated together with the lowest section of sonic tubes the soft toe condition. Total volume of grout required per
150
pile was generally 350-500 liter, but up to 750 liter was As grout injection was applied under relatively low
consumed for a few piles. No pile uplift was observed pressure, upward migration of grout along the pile shaft
during pressure grouting. was unlikely. Following are the reasons.
• Toe-grouting was carried out minimum 7 days after pile
ASSESSMENT OF TOE-GROUTING BY SPT concreting. Time gap between concreting and grouting
would have allowed some positive improvement of
In order to evaluate the effect of toe grouting, standard bentonite skin (if any) along the pile shaft.
penetration tests (SPT) were conducted immediately • Sediments of approximately 5 cm were recorded for
below pile toe through sonic tubes at two stages, prior to most of the piles prior to concreting though intensive
grouting and after completion of grouting. The tests air lifting was carried out. Possibly loose gravels were
were performed on one working pile adjacent to the test also present at pile base since removal of relatively
pile. Toe-grouting was carried out from one loop so that heavy material by air lift is unlikely. Hence, it is most
the other loop (two sonic tubes) can be used for the likely that the base of the pile was loose and weaker
access of SPT sampler. The first SPT was conducted 7 than not only the surrounding soil but also pile / soil
days after casting of the pile whereas the second test interface along the lower portion of pile shaft.
was performed 3 weeks after toe-grouting. Prior to • At 20 bar, a relatively low pressure, the grout would
conducting the SPT at both stages, a steel cap at the first compact, compress and permeate the weakest
bottom of the sonic tubes were drilled through by zone around the applied area (pile base as described
diamond core bit coupled with core barrel. Figure 6 above). Grout would also bounded and cemented
shows the SPT-N values obtained from two different loose sand and gravel at pile base.
stages in comparison with that of original obtained from
nearest borehole (BH-4). As can be seen in the figure, it Cement grout mixed with sand and some gravels were
is clear that the density of the soil below pile toe has observed from the first 5 cm of the collected sample
been increased appreciably from soft toe condition by followed by pure sand-gravel mixture during standard
pile-toe grouting. After completion of second SPT, an penetration test at pile base, which favored the assumed
additional toe-grouting was carried out from two sonic mechanism described above. More investigations are
tubes which were used for the access of SPT sampler as required to establish the exact mechanism of grout at
described earlier. pile toe. Simplified version of the most likely geometry of
100 the pile base at different construction stages is
demonstrated in Figure 7.
80
Original Value from BH - 4
SPT
2nd SP T 1s t SPT
60
SPT-N
nd
2
S c hem atic sk etch o f pile b as e
40 Original SPT-N of BH- 4 P lan of P ile B ase

SPT
(nearest soil investigation

Before toe grouting ( 1 st SPT)
st
Borehole W all
1
20
Main Reinforcement
After toe grouting ( 2 nd SPT) Sonic Tube (ID =54 m m )
0
Horizontal Pipe with
Manchette
Figure 6 Comparison of SPT-N values Loose Material Borehole Base
(Sedim ents)
S TAG E 1 (After B orin g)
MECHANISM OF TOE-GROUTING
P ossib le m ech an ism
(i) Concrete casting pressure
Teparaksa et al. (1999) investigated the penetration of Pile
pushed loose m aterial
grout into the sand layer from the toe-grouting of bored Convex Concrete toward pile wall and
piles in Bangkok. Toe-grouting was performed 24 hours Concrete form ed convex surface
Surface (ii) Loose material trapped
after concreting with maximum pressure 40 bar for
Loose Material between bottom soil and
grouting the base of large diameter deep-seated bored (Sedim ents) concrete
piles in medium to fine sand layer. From the series of S TAG E 2 (After C on cretin g)
SPT and soil sampling performed below the pile toe, the
authors concluded that grout did not permeate into the P ossib le m ech an ism
sand layer at pile base. Upward flow of grout along the (i) Pressurized cem ent grout
pile / soil interface is the most likely phenomenon of toe- com pacted loose m aterial
Convex Pile
grouting in Bangkok as stated by them. Concrete
Concrete (ii) Grout cemented with
loose material and
Surface
form ed stiffer layer
In this case study, likely mechanism of toe-grouting Cem ent Grout
under maximum pressure 20 bar in sand-gravel (cem ented with loose m aterial)
formation are; S TAG E 3 (After G rou tin g )
• Compaction of loose material at the pile base
• Bounding and cementing the loose gravel at pile base Figure 7 Possible geometry of pile base at different
• Compressing of distressed soil at pile base construction stage
151
STATIC PILE LOAD TEST Load transferred to the pile toe at maximum test load
was about 5500 KN (accounted about 22 percent of the
Static pile load test (Maintain Load) was commenced 36 total applied load) as can be seen in Figure 9. At
and 29 days after concreting and toe-grouting maximum applied load, displacement at pile tip
respectively. Additional 4 bored piles constructed measured by extensometer was 8.29 mm (0.5 % of pile
without toe-grouting were used as anchor piles. Vibrating diameter).
wire strain gauges (VWSGs) and Mechanical
Extensometers (ME) were fixed at five levels along the LOAD-DISPLACEMENT ANALYSIS OF THE PILE
shafts at the known interface boundaries of different soil
layers. At each level, two sets of VWSGs and one set of Load-displacement behavior of pile was analyzed using
ME were installed. The lowest level of instrument was load-transfer method employing hyperbolic transfer
set at 70 cm above the pile toe level. The pile was function proposed by Hirayama (1990). A computer
tested to maximum 25,000 KN, which is 2 times of the program written based on this method was used to
design load (12,500 KN). analyze the load-displacement behavior of the pile.
Figure 10 illustrates the hyperbolic stress-displacement
Load vs. Pile Displacement curves of bored cast-in-place piles presented by
Hirayama (1990). It should be noted that end resistance
The total applied load vs. pile head displacement of the “qult” indicated in Figure 10 represents physical ultimate
test pile is shown in Figure 8. Pile shaft and end end resistance which is related to pile tip displacement in
resistance in relation to the pile head displacement are hyperbolic function. Conventional definitions of ultimate
also included in the figure. end bearing capacity for bored piles therefore do not
Load (KN) correspond to the “qult” indicates in this paper.
0 5000 10000 15000 20000 25000
0
1
4
1
Pile Head Displacement
f u lt af
S kin friction : f bf
8
(mm)
12 fu lt Z
2 f=
a f + bf Z
16
Zref, f
20 0
Total Load Resistance vs. Pile Head Displacement
Shaft Resistance vs. Pile Head Displacement P ile s h a ft d is p la c e m e n t : Z
End Resistance vs. Pile Head Displacement
Figure 8 Load-displacement curves of test pile 1

qu lt be
ae
End resistance : q
Load Distribution Mechanism

1
Load distribution along pile shaft at various applied loads
q u lt
derived from the strain gauge data is shown in Figure 9. Ze
2 q =
ae + beZe
Applied Load Maxim um Mobilized
(KN) U nit Shaft Friction
(KPa) Z ref, e
0
0 15000 30000 0 50 100 150
0
P ile tip d is p la c e m e n t : Z
0
Figure 10 Hyperbolic stress-displacement curves (After

10
Hirayama 1990)
10
One-dimensional finite element model constructed by

elastic bar elements used in the method proposed by
D ep th (m )
20 20
Hirayama is shown in Figure 11. The detailed
discussion of the background principle and procedure to
30 30 determine the required parameters in using this method
can be seen in the published paper of Hirayama (1990).
The major advantage of the method is that input
40 40 parameters required can be determined from the results
of routine in-situ tests such as SPT and CPT. The
predicted load-displacement curves agreed well with
50 50 those of the actual measurement according to the
authors’ experience in Bangkok subsoil.
Figure 9 Load distribution along pile shaft of test pile
152
In crem en tal ap p lie d lo ad : Q Table 4 Input shaft and end resistance values used in
analyses
z 1 Input data in analyses
1 k f1 Preliminary analysis Back
1
Tangential spring constant

Depth Soil SPT Case 1 Case 2 analysis
Increm ental nodal displacem ent
z2 2 (m) Type N fult qult fult qult fult qult

k f2 (KPa) (MPa) (Kpa) (MPa) (KPa) (MPa)
2
0-4 Sand 7 35 35 30
Elem ent No.

z 3 4-12 Clay 7 70 70 30
Node No.
3 12-14 Sand 18 90 90 60
14-16 Gravel 20 100 100 60
16-18 Sand 30 150 150 60
18-22 Clay 24 150 150 135
zn n 22-24 Sand 60 200 200 135
24-30 Clay 50 150 150 135
n k fn 30-32 Sand 44 150 150 135
32-36 Clay 40 150 150 135
36-42 Clay 40 150 150 135
z n+1 42-44 Sand- 85 200 8.5 200 25.5 110 40
n+1 ke gravel
Assumed
displacement
Figure 11 One-dimensional non-linear finite element 0.2 % of pile 0.2 % of pile 0.2 % of pile
required to fully
diameter diameter diameter
model of a pile used in analyses (after Hirayama 1990) mobilize shaft
resistance
Preliminary Analyses Assumed
displacement
5 % of pile 5 % of pile 5 % of pile
Two preliminary analyses were carried out. The input required to fully
diameter diameter diameter
mobilize end
soil parameters required in the analyses were resistance
established based on the recommendation of Hirayama
Pile toe condition Normal toe Stiff toe Stiffer toe
(1990) as shown in Table 3 with the following key
assumptions. Note : SPT-N indicate in the table are average values of
relevant soil layers
• Same shaft resistance values (derived from SPT-N
values of original site investigation) for both cases
The load-displacement curves of preliminary and back
• Preliminary Case 1 : No pile toe improvement so that
analyses along with that of actual measurement from
normal toe condition exists – lower factor (100) is
static pile load test are depicted in Figure 12.
used in ultimate end resistance derivation.
Load (KN)
• Preliminary Case 2 : Toe-grouting improves pile toe
0 10000 20000 30000
so that stiffer toe condition than case 1 – higher 0
factor (300) is used in ultimate end resistance Preliminary Case 1
derivation. 4 Preliminary Case 2

Pile Head Displacement (mm)
Back Analysis
Table 3 Factors used in estimation of shaft and end 8 Measured
resistance from SPT-N values for preliminary analyses

12
Case
Parameter Soil Type 16 Case 1 : turns to plunging
Case 1 Case 2 failure over 23 ,000 KN
Ultimate shaft Sand and 20

5N (≤ 200 KPa)
Resistance, fult (KPa) gravel
24
clay 10N (≤150 KPa)
Ultimate end 28
gravel 100N 300N
Resistance, qult (KPa)
Figure 12 Load-displacement curves of preliminary and
Assigned pile toe condition Normal toe Stiff toe back analyses in comparison with measured data
Note : Normal toe condition in above table represents non-
grouted or plain pile toe condition Assessment of Load-displacement Behavior
Back Analysis Based on Pile Load Test Results Reviewing the analyses results shown in Figure 12 in
conjunction with Table 4, load-displacement
Fixing the ultimate shaft resistance derived from strain characteristics of the test pile may be evaluated as
gauges data obtained from static pile load test, a series follows. Same pattern and magnitude of pile head
of trail runs were made to back calculate the ultimate displacement as the actual measurement can be
end resistance of the test pile. Table 4 shows the input observed for preliminary Case 1 before reaching to
data of shaft and end resistance used in the preliminary 18,000 KN applied load. However, displacement
and back analyses. increases abruptly over 18,000 KN and turns to the
153
plunging pattern at 23,000 KN, which suggests that the CONCLUSIONS
pile would experience plunging failure before reaching to
the maximum test load of 25,000 KN, if the normal toe The practical aspects of toe-grouting in sand-gravel
condition for non-grouted or plain bored pile assumed in formation of Chiang Mai City has been presented. Direct
Case 1 existed. Load-displacement pattern of Case 2 is observation from SPT conducted at pile base before and
found to be very similar to that of the actual after toe-grouting clearly confirms the improvement of
measurement. The factor (300N) adopted in the pile base by tube-a-manchette toe-grouting technique
derivation of end resistance shown in Table 3 is employed in the project. Target pressure, 20 bar proved
therefore considered reasonable. to be sufficient for toe-grouting in sand-gravel formation.
With the use of maximum shaft friction values obtained A practical method for examining the load against pile
from the load test results and end resistance value of displacement behavior is illustrated by a series of load-
40 MPa as input parameters, the load-displacement displacement curves with various pile end resistance
curve of back analysis best fits the actual measurement values. Load-displacement analyses with the application
which means that the selected value of end resistance of hyperbolic load-transfer method provided an effective
in back analysis fairly represents that of actual tool in assessment and evaluation of pile toe condition.
magnitude. In other words, the measured load- In addition to SPT conducted at pile base, improvement
displacement curve strongly suggests that end of pile end resistance by toe-grouting is demonstrated by
resistance of the pile is markedly high; detrimental soft back analyses results. Without applying the toe-grouting,
toe condition does not exist at pile base. test pile is likely to experience plunging failure before
Load (KN) Measured displacement reaching to the maximum test load. Toe-grouting in this
at 25 ,000 KN case could therefore be considered as extra safety
0 10000 20000 30000 40000 measure to ensure the sufficient capacity of the pile.
0
The back analyzed load-displacement curves also
20
demonstrate that the pile might be able to resist the load
up to 37,000 KN without reaching to the plunging failure.
Pile Head Displacement (mm)
40
Therefore, the specified design load 12,500 KN might
60
Input End Resistance 10 MPa have been underestimated for the toe-grouted bored pile
80 Input End Resistance 15 MPa in sand-gravel formation. Instead of conducting a proof-
loading test on the working pile, static pile load test
100 Input End Resistance 30 MPa
should have been performed on the pilot or trial pile with
120 Input End Resistance 40 MPa the higher load up to failure to determine the ultimate
Input End Resistance 60 Mpa
capacity of the pile and to optimize the pile design which
140
might provide some cost-saving for the project.
Measured
160
180 REFERENCES
Figure 13 Load-displacement curves of back analyses BAUM, F., BRAUN, E. V., HAHN L., HESS, H., KOCH,
with various input end resistance values K. E., KRUSE, G., QUARCH H. and SEIBENHUNER M.
1970. On the geology of northern Thailand. Geologishe
To examine the influence of end resistance on the load-
Jahrbuch, Hannover, Beiheft 102.
displacement mechanism, a series of additional back
analyses were carried out using different values of end HIRAYAMA, H. 1990. Load settlement analysis for bored
resistance with the fixed shaft resistance values piles using hyperbolic transfer functions. Soil and
(obtained from the load test results) as presented in Foundations, Vol. 30, No. 1, Japanese Society of Soil
Figure 13. It is evident from Figure 13 that load- Mechanics and Foundation Engineering, pp. 55-64.
displacement behavior is not excessively sensitive to
variation for end resistance values in the range of 30 to INTRASUTA, T., 1983. Ground water potential in
60 MPa. For end resistance less than 15 MPa, st
unconsolidated sediments in Chiang Mai Basin. 1 Syp.
displacement increases significantly and load- on geomorphology and quaternary geology of Thailand,
displacement curve turns to a plunging failure pattern, Bangkok, pp. 179-195.
which is more significant for the case of 10 MPa as can
be seen in the figure. In summary, it should be pointed RHODES, B. P., BLUM, J., DEVINE, T. and
out that test pile would have experienced plunging type RUANGVATANASIRIKUL, K., 1997. Geology of the Doi
of failure if it had seated on the normal pile toe condition Suthep metamorphic complex and adjacent Chiang Mai
(non-grouted or plain pile). It is therefore considered Basin. Proc. of the International Conference on
that pile end resistance was effectively improved by the Stratigraphy and Tectonic Evolution of Southeast Asia
toe-grouting method employed. and the South Pacific, Bangkok, Thailand, pp. 305-313.
From Figure 13, it is reasonable to conclude that the pile TEPARAKSA, W., THASNANIPAN N., and ANWAR M.
end resistance is in the range of 30 to 40 MPa. The load- A., 1999. Base grouting of wet process bored piles in
th
displacement curves of end resistance 30 and 40 MPa Bangkok subsoils. The 11 Asian Regional Confe. on
suggest that the higher load up to 37,000 KN can be Soil Mechanics and Geotechnical Engineering, Seoul,
applied on pile without reaching to the plunging failure. Korea, p.p. 269-272.
154
Under the Royal Patronage of His Majesty the King
Organized by :
Department of Civil Engineering, Khon Kaen University
with support from Petroleum Authority of Thailand and
Department of Civil Engineering from Universities in Thailand
In Corporation with :
Department of Agricultural Engineering and Engineering Alumni Association, Khon Kaen University
Advanced Technologies for the

Development of Civil
Engineering, Volume 2
Proceedings of the Eighth National Convention on
Civil Engineering
APPLICATION OF POLYMER-BASED SLURRY FOR WET-PROCESS BORED

PILES CONSTRUCTION IN MULTI-LAYERED SOIL OF BANGKOK
Zaw Zaw Aye1, Kamol Singtogaw2, Chanchai Submaneewong3

Project Manager1, Project Director2, Geotechnical Engineer3
SEAFCO Co., Ltd., Thailand
155
การประยุกตใชสารละลายโพลีเมอรในการกอสรางเสาเข็มระบบเจาะเปยกในชั้นดินกรุงเทพฯ
APPLICATION OF POLYMER-BASED SLURRY FOR WET-PROCESS BORED PILES
CONSTRUCTION IN MULTI-LAYERED SOIL OF BANGKOK
Zaw Zaw Aye1, Kamol Singtogaw2, Chanchai Submaneewong3

Project Manager1, Project Director2, Geotechnical Engineer3
SEAFCO Co., Ltd., Thailand
บทคัดยอ: อีกทางเลือกหนึ่งแทนการใชสารละลายเบ็นโทไนท คือการนําสารละลายโพลีเมอรมาใชในการกอสรางเสาเข็มเจาะใน

กรุงเทพฯ สารละลายโพลีเมอรมีคุณสมบัติที่สงผลกระทบตอสิ่งแวดลอมนอยและสามารถลดปญหาที่เกี่ยวกับการกําจัดสารเจือปน
ในดินซึ่งเปนมลภาวะในหนวยงานกอสรางใหนอยลง บทความนี้นําเสนอการประยุกตใชและสมรรถนะของเสาเข็มเจาะขนาดหนาตัด
ปานกลางถึงขนาดใหญที่กอสรางโดยใชสารละลายโพลีเมอรในกรุงเทพฯ รวมทั้งใหขอเสนอแนะและปจจัยที่สําคัญในการใชโพลีเมอร
ในการกอสรางเสาเข็มเจาะในชั้นดินกรุงเทพฯ
ABSTRACT: As an alternative to the classic bentonite slurry, polymer-based slurry has been increasingly
used for wet-process bored piles in Bangkok. Polymer-based slurry having environmentally friendly
properties, is found to be a good solution in minimizing the problems associated with the disposal of
contaminated soils as well as the site pollution. This paper presents the practical application of polymer-based
slurry for construction of medium to large diameter deep-seated bored piles in Bangkok. Important factors in
using polymer-based slurry in Bangkok subsoil are also discussed.
KEYWORDS : Polymer-based slurry, bentonite slurry, filtercake, fluid loss, wet-process bored pile
1. Introduction 2. Wet-Process Bored Pile Construction Method
Drilling slurry or supporting fluid plays a major role Wet process method as its name implies makes the
in construction of wet-processed medium to large pile under wet condition by using drilling slurry as
diameter deep-seated bored piles in multi-layered supporting fluid for borehole stability. Depending on
soil of Bangkok. Rotary drilling technique coupled the subsoil condition, temporary casing of 14-18m in
with wet-process method using bentonite slurry as length is also used as a support in soft clay layer.
support fluid was commonly applied to construct Drilling is commenced by dry process with auger
these piles for over two decades. As bentonite slurry applying rotary drilling action in the soft clay layer
had proven acceptable to make qualified bored piles and stiff clay layer. Before reaching to the first sand
in the past, most of them were constructed only by layer, slurry is fed to the borehole and drilling is
using this mineral based slurry. Recently, polymer- continued with the bucket to the final depth. A
based slurry (PBS) has been increasingly popular in cleaning bucket or airlift is used to clean the bottom
Bangkok as an alternative supporting fluid in wet- of the pile prior to reinforcement cage installation.
process bored pile drilling. The primary reason, in Concreting is carried out by tremie method under the
the initial stage, for using this alternative slurry was slurry. Slurry from borehole is normally pumped
to minimize the problems associated with using back to the storage tank during concreting and
classic bentonite such as site pollution and the reused for next piles after treating to get specified
disposal of excavated soils which are usually properties. Depending on the level of design pile
contaminated. This paper presents the application of cut-off, bored piles are normally overcast about 1 to
polymer-based slurry for construction of wet-process 3m above cut-off level to avoid presence of
bored piles in multi-layered soil of Bangkok. contaminated concrete within the design pile length.
157
3. Basic Principles of Different Drilling Slurries 4. Practical Application of Polymer-Based
Drilling Slurry In Bangkok
It is important that practicing engineers involved in
bored pile construction understand the basic In Bangkok, the primary reason for switching from
principles on reaction of soil layers with different classic pure bentonite slurry to polymer-based slurry
drilling slurries so that the correct procedures and in most cases was to minimize the problems
methods can be applied in operation to avoid the associated with environmental issues caused by the
detrimental effect of wet-process bored piles. use of bentonite slurry. In the initial stage of using
Bentonite and polymer slurry react differently with polymer-based slurry, the major concerns were the
the soil layers. Pure bentonite slurry stabilized the potential negative impact on stability of the borehole
borehole wall by its density and formation of a and the shaft friction capacity as this alternative
membrane or filtercake particularly in permeable slurry was not yet accustomed to the bored piling
layers. Bentonite slurry is capable of maintaining the industry in Bangkok. However, no adverse effects
soil particles in suspension. Therefore, soil particles from the use of polymer-based slurry have been
in bentonite slurry column stay in suspension for experienced to date and numbers of projects have
long hours and slowly settled toward the borehole been successfully completed with this environmental
bottom. Formation of filtercake reduces the slurry friendly drilling slurry.
filtration into the permeable layers, thick first and
second sand layers typically encountered in 4.1 Mixed-Ratio and Procedure of Mixing
Bangkok.
Dry anionic partially hydrolyzed polyacrylamide
Unlike pure bentonite slurry, polymer-based slurry (PHPA) polymer powder premixed with fresh water
does not suspend sand especially coarse grain and a small percentage of bentonite is mainly used
particles. Hence, in polymer slurry, sand particles for long bored piles in Bangkok. First, a small
(mainly from cuttings during drilling) tend to fall out dosage of bentonite powder is premixed with fresh
of suspension and settle on the base of borehole water (bentonite 5-15kg per cubic meter of water)
significantly sooner than that of pure bentonite as and leave in the storage tank for minimum 12 hours.
evident from field observation. In Bangkok, The predetermined dosage of dry polymer (typically
depending on subsoil profile and composition of polymer powder 0.5kg per cubic meter of premixed
polymer-based slurry, 0.1m to 1m of sediment is bentonite slurry) is then added and mixed to get so
commonly observed on base of borehole in 2 to 8 called polymer-based slurry. O’Neill and Reese
hours. The sedimentation of sand particles is found (1999 [4]) cited that vigorous mixing of polymers
to be stable with time. supplied in the dry form should be avoided, since
polymer chains can be broken down and the polymer
Polymer molecules tend to wrap around clay and silt slurry rendered ineffective. Polymer-based slurry is
particles and produce small agglomerated structure. stored in the steel tanks before supplying to the
By its own weight, the resulting agglomerated bored pile drilling. The reason for adding bentonite
structures tend to settle out of suspension slowly and for long bored piles is to minimize the fluid loss
accumulate mushy sediments on the firstly settled (high filtration of slurry into formation soils), which
sand sediments at the bottom of the borehole. Some was observed from the trial boreholes drilled only
of the agglomerated particles also tend to float with pure polymer slurry. Depending on the
temporarily on the surface of polymer slurry stay in thickness and permeability of sand layers present at
the suspension. The polymer molecules then proceed site and the embedded length of the pile in sand
to attach themselves to larger and appear as a bulky layers, addition of bentonite may not be necessary.
material that some observers have termed “oatmeal” No specific method or formula is available to
(O’Neill and Reese, 1999 [4]). In Bangkok as small determine the required best composition of polymer-
percentage of bentonite is added in polymer-based bentonite ratios. Therefore, at each site, different
slurry, similar to polymer-clay system (Majano and mixed-ratios should be made on trial in the initial
O’Neill, 1993 [3]), it can be assumed that a thin stage to determine the most appropriate ratio. Table
polymer-bentonite layer is formed in the soil-slurry 1a and 1b present the mixed composition of
interface of permeable layers. This thin membrane polymer-based slurry of bored piles constructed at
of polymer-based slurry is capable of reducing the 10 different locations. According to the load-
high filtration in sand layers of Bangkok. However, settlement curves obtained from static pile load tests,
overall volume of fluid loss for polymer-based bored observed settlements of all piles were less than those
piles is found to be higher than those of bentonite of predicted values using bentonite pile parameters.
bored piles.
158
Table 1a. Technical data of polymer-based slurry based slurry is commonly introduced only after
bored piles at different locations in Bangkok drilling reaches to stiff clay layer. However, the top
Description
Pile ID level of slurry is required to maintain about 1m
T1 T2 T3 T4 T5 below the existing ground at all times once borehole
Pile Diameter (m) 0.80 0.80 1.20 1.80 1.50
Pile Length (m)
excavation reaches to the first sand layer. The
49 50 51 62 54
Mixed
results of the sonic caliper measurement conducted
Polymer 0.50 0.50 0.50 0.50 0.50
Ratio per on hundreds of piles proved that borehole can be
3
1m (kg) Bentonite 12 12 10 10 12 well stabilized with the use of polymer-based slurry
Total standing time prior at total construction time prior to concreting even
to concreting 17:20 12:20 14:00 23:00 15:30
(Hr : Mins)
over 72 hours. Figure 1 shows the borehole profile
Density test conducted after 24 hours leaving the hole filled
1.06 1.03 1.01 1.04 1.04
Properties of slurry
with polymer-based slurry without any agitation.

prior to concreting
(g/ml)
Viscosity
40 45 44 51 44 Y'
(sec/qt)
pH 8 8 8 8 8 X X'
Sand
TOP OF
EXISTING X X' Y Y Y' CASING
content 0 0.5 0 0 0.5 GROUND 0
(%)
TOP OF
Approx. SOFT SLURRY
First sand 12 8 7 20 5 5
Pile
CLAY
embedded
in sand Second 10
7 9 13 18 13
layers (m) sand
MEDIUM
CLAY 15
Table 1b. Technical data of polymer-based slurry
bored piles at different locations in Bangkok STIFF 20
Pile ID CLAY
Description
T6 T7 T8 T9 T10 25
Pile Diameter (m) 1.00 1.00 0.80 0.60 0.60
Pile Length (m) 51 30 41 26 24
DENSE 30
Mixed Polymer 0.50 0.50 0.50 0.50 0.50 SAND
Ratio per
3
1m (kg) Bentonite 5 5 10 5 0 35
Total standing time (after
drilling completion)
7:00 3:45 6:55 2:30 0:45
prior to concreting Figure 1 Profile of polymer-based pile: test
(Hr : Mins) conducted after 24 hours leaving the hole (filled
Density
1.02 1.03 1.01 1.02 1.02 with PBS) without agitation
Properties of slurry
prior to concreting
(g/ml)
Viscosity*
36 45 44 43 42
(sec/qt) 4.3 Properties of Polymer-Based Slurry
pH 8 8 8 8 8
Sand
content 0 0 0.1 0 0
It is important that properties of slurry used are
(%) within the applicable ranges to avoid the detrimental
Approx. effect on the performance of bored piles. O’Neill &
Pile First sand 9 12 12 8 1
Reese (1999, [5]) cited that there is no perfect set of
embedded
in sand Second slurry specifications that can be used on every job,
15 - 11 - -
layers (m) sand but specifications should be tailored to fit the
* Marsh Funnel Viscosity requirements of a particular job at a particular
location, where possible. Today, though use of
4.2 Borehole Stability polymer slurry is widely acceptable in bored piling
industry, information on the desirable properties is
Ata & O’Neill (1998, [2]) reported that polymer still extremely limited. Table 2 shows the ranges of
slurry when introduced with the right dosage and properties of polymer slurries specified by the
prior to the borehole reaching the piezometric level, International Association of Foundation Drilling
maintained the stability of the boreholes for more Standards and Specification Committee (ADSC,
than 18 hours without significant change of the 1999 [1]), those successfully used on several
borehole diameter at any depth. In Bangkok subsoil, projects as reported by O’Neill and Reese (1999,
as temporary casing of about 15m length is installed [4]) from the work of Majano et al. (1994) and the
to protect the collapse of soft clay layer, polymer- properties commonly applied in Bangkok subsoil.
159
Table 2 Properties of polymer slurries specified by Table 3 Summarized comparison of bentonite and
ADSC and Polymer-based slurry used in Bangkok polymer-based slurry in application
Majano et al., 1994 Common Description Bentonite Slurry Polymer-based slurry
ADSC Emulsified Dry Practice in Mixing Add bentonite • First add bentonite
Property
(at 20 C) PHPA PHPA Bangkok method powder in water powder in water
(Vinyl) (PBS**) and mixed by and mixed similar
Density 1.02 0.995 to 0.995 to 1.05 • Electric mixer to normal bentonite
(g/cc) Maximum 1.01 1.01 Maximum (or) mixing
Viscosity 40 – 90* 33 - 45 50 - 120 40 to 90 • Water jet mixer • Adding polymer
(sec/qt) (+/- 5) (using powder in
pH 7 - 12 8 – 11.7 8 – 11.7 8 – 11 submersible premixed bentonite
pumps in tank slurry and mix with
API Sand 1.0 0 to 1 0 to 1 Less than and mix with electric mixer
Content Maximum 1% jetting action) • Vigorous mixing of
*or as recommended by manufacturer and approved by polymer powder
geotechnical engineer, ** Polymer-based slurry must be avoided
Main function Unit weight • Unit weight
in borehole (density) and (density) and
4.4 Integrity of Polymer-Based Bored Piles stability filtercake effect binding of soil
particle by polymer
According to sonic integrity and sonic logging or strand
sonic coring tests conducted on large numbers of • Possible forming of
polymer-based bored piles in Bangkok, there is no very thin filter cake
evidence that integrity of the pile is affected by the due to addition of
use of polymer-based slurry. bentonite
Sand Ability to suspend Sand particles tend to
suspension in sand particles full out of suspension
4.5 Effective Base Cleaning for Polymer-Based slurry column and settle on base of
Bored Piles in Bangkok borehole.
Desanding Use desanding Desanding equipment
O’Neill & Reese (1999 [4]) pointed out that process equipment is not required. Use
vigorous mechanical agitation such as pumping by sedimentation tank to
allow fine sand and
centrifugal pumps can shear the polymer chains
silt settle before
severely. As soil-polymer reaction system is largely reuse.
depending on the polymer chemical chain structure, Pile base Cleaning bucket Mainly cleaning
excessive breaking of it would affect the cleaning or air lift or bucket. Occasionally
performance of the slurry. Therefore, high turbulent recycling by airlift
means of base cleaning methods such as recycling Effect on long Possible Shaft friction
exposure time degradation of degradation has not
by centrifugal pumps should not be performed for shaft friction been encountered
the polymer-based bored piles. Field experience capacity if (from available
shows that base cleaning by air lift method cannot continuous literatures and
effectively clean thick sediments at the base of the agitation is not Bangkok experience)
polymer-based bored pile. In Bangkok, the effective applied
Environmental • Particulate • Non-particulate
base cleaning for long bored piles may be achieved impact nature nature
by the following procedure. • Potentially • Non-hazardous
(1) Carry out drilling to 1 to 2m above the final hazardous to • Can be disposed on
design depth under polymer-based slurry for the aquatic life site in drainage
first bored pile • Not be disposed system or sewers
(2) Leave the borehole for 2 to 8 hours (depending in drainage
system or
on the rate and stabilization trend of sewers
sedimentation observed from trail borehole at
each site)
(3) Move the drilling rig to commence the drilling 5. Review of Shaft Friction Capacities Affected
of second bored pile by Drilling Slurry
(4) Drilling rig come back to the first bored pile and
carry out final drilling for (remaining 1 to 2m) Properties of drilling slurries influence on the shaft
with the cleaning bucket capacity has been studied by many researchers. A
(5) Immediate installation of reinforcement cage number of researchers reported the reduction of shaft
and concrete pouring capacity caused by excessive formation of the filter
cake on borehole wall of bored piles constructed
160
with bentonite slurry. It is mainly contributed by Load (kN) Unit Skin Friction (kN/m 2)
long exposure time without slurry agitation as cited 1

0 4000 8000 12000 0 50 100 150
by them. 0 0 .0 0
So ft to
M edium Clay
10 1 0.0 10
Few researches are available on the comparison of Stiff to
shaft friction capacity between bored piles 20
Very Stiff 2 0.0 20
constructed with bentonite and polymer. Ata and

Clay
Depth (m.)
M edium Dense
O’Neill (1998 [2]) reported that developed shaft 30 to V.Dense

Sand
3 0.0 30
friction factors β (for sand) and α (for clay) of bored Hard Clay
40 4 0.0
piles constructed with polymer were higher than
40
M edium to
those recommended by USA-FHWA. 50 5 0.0 50

Mobilized at 9 9 0 0 kN
6. Performance of Polymer-Based Slurry Bored

Qult (calculated)
60 6 0.0 60
Piles in Bangkok
Figure 2. Load distribution curves and average unit
Instrumented static pile load test results of three test shaft friction of T1 (Dia. 0.80x49 m)
piles constructed by wet-process with polymer-based Lo ad (k N) Unit Skin Friction (kN/m 2)
slurry in Bangkok are presented in this paper. 1
0 50 100 150
0 4000 8000 12000
0 0 0
6.1 Slurry Properties and Construction Records of

So ft clay
Test Piles 10
Lo o se Sand
10 10
Table 1a shows the properties of polymer-based

20
Depth (m.)
Stiff to Very 20 20
Stiff Clay
slurry and construction records of test piles (T1, T2, 30 30 30

and T3). It is to be noted that all these test piles were
Dense to V ery
Dense Sand
constructed by the same contractor. Hence it is 40

H ard Clay
40 40
assumed that influence of other variables such as Dense to V ery
equipment as well as construction procedure used in 50 50 50
drilling, slurry mixing and quality control measures

Mobilized at 9 9 0 0 kN
Qult (calculated)
applied for all piles would be minimized in 60 60 60
evaluation of the load test results. Figure 3. Load distribution curves and average unit
shaft friction of T2 (Dia. 0.80x50 m)
6.2 Instrumented Pile Load Test Results
Load (kN) Unit Skin Friction (kN/m 2)
Shaft frictional resistance of polymer-based bored
1
0 6000 12000 18000 0 50 100 150 200 250
0
piles in sand layers were of major interest to the
0 0
Soft Clay
authors as it is considered that drilling slurry has less 10 10 10
influence on clay layers according to the previous
Stiff Clay
studies on bentonite bored piles. 20 20 20
V. Stiff to
Depth (m.)
Hard clay
Load distribution curves along the test pile shafts at 30

De nse Sand
30 30
various applied loads of 3 test piles are illustrated in Hard Clay
Figure 2, 3 and 4. The values of average unit skin 40

V
De nse to
D
40 40
friction developed by the maximum test load at the

different soil layers along the shaft of test piles in 50 50 50
comparison with those of ultimate unit skin friction Mobilized at 1 8 0 0 0 kN

Qult (calculated)
calculated using the empirical formula are also 60 60 60
illustrated in these figures. As can be seen in the Figure 4. Load distribution curves and average unit
figures, developed average unit shaft friction in sand shaft friction of T3 (Dia. 1.20x51 m)
layers are significantly higher than those of
theoretically calculated values using normally used A summarized comparison between measured total
formula and parameters applied for bentonite bored shaft friction resistances at maximum test load and
piles in Bangkok. This finding suggested that those of theoretically calculated values using
polymer-based slurry bored piles produced better bentonite bored piles parameters of 3 test piles is
shaft friction resistance in sand layers than that of presented in Table 4.
bentonite bored piles.
161
Table 4 Calculated and Measured shaft resistance of higher shaft friction resistance could be obtained
three instrumented test piles if pile is constructed with polymer-based slurry.
Description
Test Pile ID (5) Design parameters currently used for bored
T1 T2 T3 piles in Bangkok were developed from bentonite
Diameter (m) 0.80 0.80 1.20 bored piles. These parameters are found to be
Length (m) 49 50 51
conservative for polymer-based bored piles.
Design Load (kN) 3300 3300 7200
Max. Test Load (kN) 9900 9900 18000 Further research is needed to confidently
establish the design parameters for bored piles
Total shaft Calculated 6900 6200 10820
resistance of
constructed with polymer-based slurry in
Measured at Bangkok subsoil.
pile (kN) 9710 9670 17060
max. test load
Measured
Calculated
1.41 1.56 1.58 Acknowledgement
The authors wish to express their appreciation to Mr.

Thasnanipan et. al. (2002, [5]) reported that load-
Narong Thasnanipan, Managing Director, SEAFCO
settlement behavior of eleven polymer-based bored
Co., Ltd. for his encouragement to write this paper.
piles constructed in Bangkok subsoil are better than
that of prediction using bentonite bored piles
For Further Details
parameters.
Please contact Mr. Zaw Zaw Aye, SEAFCO Co.,
7. Main Factors on Shaft Friction Capacity Ltd. at 26/10 Rarm Intra 109 Road, Bangkok 10510
Improvement of Polymer-based Bored Piles Tel.02-9190090, E-mail : seafco@seafco.co.th
Shaft friction capacity improvement of bored piles
References
constructed under polymer-based bored piles in
comparison with those of bentonite bored piles is [1] ADSC Standards and Specifications for the
considered to be contributed by the following Foundation Drilling Industry, ADSC, Revised 1999,
factors. Incorporating ACI 336.1R-98, 1999.
• No thick mudcake or filtercake is formed by
polymer-based slurry [2] Ata A. A. and O’Neill M. Side-wall stability and
• Polymer strands tend to bind the soil particle in side-shear resistance in bored piles constructed with
the formation which generates the drag forces and high-molecular-weight polymer slurry, Third
cohesion (Unlike plate-shaped bentonite, polymer International Geotechnical Seminar on Deep
molecules are hair-shaped strands) Foundation on Bored and Auger Piles, Ghent,
Belgium, 19-21 October 1998, Balkema, Rotterdam,
8. Conclusion 111-117, 1998.
(1) The practical application of polymer-based [3] Majano R. E. and O’Neill M. W. Effect of
slurry in construction of wet-process bored piles mineral and polymer slurries on perimeter load
in Bangkok subsoil has been presented. transfer in drilled shafts, A Report to ADSC,
(2) The measurement of borehole profile by sonic University of Houston, Dept. of Civil Engineering,
caliper and construction records proved that UHCE, 1993.
there is no adverse effect of using polymer-
based slurry in Bangkok subsoil as far as the [4] O’Neill M. W. and Reese L. C. Drilled Shafts :
stability of borehole is concerned. Construction Procedures and Design Methods, U.S
(3) Overall shaft resistances of bored piles Department of Transportation Federal Highway
constructed with polymer-based slurry are Administration, Publication No. FHWA-IF-99-025,
higher than that of theoretically calculated ADSC Publication No. ADSC-TL 4, 1999.
values using bentonite bored pile parameters in
the order of 1.5. [5] Thasnanipan N., Z. Z. Aye, Submaneewong C.
(4) The developed average unit shaft frictions in and Teparaksa W. Performance of Wet-Process
sand layers of 3 instrumented bored piles Bored Piles Constructed with Polymer-Based Slurry
constructed under polymer-based slurry are in Bangkok Subsoil, Proc. of the International Deep
found to be higher than those of ultimate values Foundations Congress 2002, Geotechnical Special
calculated by empirical formula using bentonite Publication No. 116, Volume One, ASCE, February
piles parameters. This finding suggested that 14-16 2002, Orlando, Florida, USA, 143-157, 2002.
162
Fourth Regional Symposium on Infrastructure
Development in Civil Engineering, RSID4
Bangkok, Thailand
April 3rd – 5th 2003
Organized by :
Department of Civil Engineering, Kasetsart University

Tokyo Institute of Technology
University of Philippines - Diliman
Development in construction of bored pile and barrette for

infrastructure projects in Thailand, A Country Report
Narong Thasnanipan, Kamol Singtogaw, Zaw Zaw Aye and Sumate Pravesvararat
163
DEVELOPMENT IN CONSTRUCTION OF BORED PILE AND BARRETTE FOR
INFRASTRUCTURE PROJECTS IN THAILAND, A COUNTRY REPORT
Narong Thasnanipan, Kamol Singtogaw, Zaw Zaw Aye and Sumate Pravesvararat
SEAFCO Co., Ltd. , 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah
Bangkok 10510, Tel. 0-2919 0090, Fax. 0-2919 0098
E-mail : seafco@seafco.co.th, Homepage : www.seafco.co.th
ABSTRACT
This paper presents the authors’ experiences in development of bored piles and
barrettes for infrastructure projects in Thailand with particular focused on construction issues.
Some literatures related to wet-processed bored piles and barrettes in Thailand published
through out the past 30 years are also summarized together with recent research works.
KEYWORDS : bored pile, barrette, deep foundations, bentonite, polymer-based slurry
INTRODUCTION
Cast-in-place deep foundations, particularly wet-processed deep-seated large diameter

bored piles and barrettes have been used for foundations of various infrastructure projects in
Thailand for over two decades. The versatility of the construction method and the high-load
capacity which in turn offered the constructibility and cost-saving are the main factors
contributed to the increasing use of deep-seated large diameter bored piles and barrettes. In
the initial stage there were a number of questions on design and construction aspects of these
foundation systems particularly in Bangkok subsoil. With the passage of time, the
construction equipment, installation techniques and testing of deep foundation elements have
been developed. Large numbers of instrumented full-scale static pile load tests were
conducted through out 1990’s which provided better understanding on behavior of these deep-
seated foundations. Researches focused on the design parameters and methods were produced
based on these test results. The design parameters were so well established that wet-process
cast-in-place foundations became regarded as reliable foundations for not only in the property
sectors but also for practicing engineers involved in the infrastructure projects. This paper
presents the development of wet-processed bored piles and barrettes in Thailand. The
information contained in the paper are mainly from the infrastructure developments built in
Bangkok and adjacent areas.
CONSTRUCTION METHOD
Large diameter deep-seated wet-processed bored piles can be constructed by 2

methods, reverse-circulation and rotary drilling method. As construction procedure of historic
reverse-circulation method has been well documented in many literatures and it is less
frequently utilized nowadays in Thailand, only a rotary-drilling method is briefly presented in
In rotary drilling method, a temporary casing of appropriate length (12 to 18m in
Bangkok depending on the thickness of soft clay) with required diameter (internal diameter
not less than that of design bored pile diameter) is first installed to ensure the stability of the
borehole in the top soft or loose soil layers. In some projects where the vibration is strictly
165
limited, standard length of casing with oscillator or short temporary casing of 5 to 8 m is
pushed down in combination with pre-boring process. Drilling is commenced by auger to drill
out the soil within the temporary casing and up to the top of the first water-bearing sand layer
or bottom of the casing in case of using short casing method. Drilling slurry or supporting
fluid is then supplied to the borehole and drilling is continued with a bucket down to the
design final depth of the pile. Before lowering the reinforcement cage, a special cleaning
bucket is used to clean the base of borehole. If bentonite slurry is used, recycling method by
air-lift or pump is applied as base cleaning process. Reinforcement cages are then lowered
into the borehole and concreting is carried out by tremie method.
A mechanical or hydraulic cable-suspended grab is commonly used for barrette
construction. Excavation of the trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the clamshell operated by cables.
Different from bored pile construction, a guide wall of depth 1 to 1.5 m with inside clear
dimensions slightly larger than the nominal size of the barrette is used to guide the grab
during initial bites. Bentonite slurry is introduced to the trench as soon as initial excavation
commenced. The excavation is continued under the bentonite slurry to the final depth. After
recycling of slurry and lowering the rebar cage, concreting is done by tremie method.
OVERVIEW OF APPLICATION
Bored pile
The first wet-process large diameter bored pile was constructed for Pinklao Bridge in
Bangkok 30 years ago. Using reverse circulation method, diameter 1.50m bored pile was
installed up to 45m in the second sand layer. Three major bridges were constructed by bored
pile of reverse circulation method in Bangkok from early 1970 to 1980. The summarized
information of these bridges are tabulated below.
Table 1 First 3 major bridges constructed by large diameter wet-processed

bored piles in Bangkok
Project Name Year of Construction Method Diameter Depth

Construction (m) (m)
Pinklao Bridge 1971 Reverse circulation 1.50 45
Sathorn Bridge 1979 Reverse circulation 1.50 46
New Memorial 1982 Reverse circulation 1.50 49
Bridge
The first wet-processed bored pile of rotary-drilling method down to first sand layer
was constructed in Bangkok in late 1970 for high-rise building project, Royal Orchid Hotel.
Since then bored piles constructed by rotary-drilling method have been extensively used for
foundations of various heavy structures such as high-rise buildings, elevated expressways,
overpass-bridges, underground car park buildings, waste-water treatment plants and most
recently underground train stations of Bangkok first subway project. By mid 1980, bored pile
became the foundation of choice for heavy structures particularly in urban area of Bangkok.
The versatility of the construction method and the high-load capacity which in turn offered
the constructibility and cost-saving are the main factors contributed to the increasing use of
deep-seated large diameter bored piles and barrettes. Most of the early wet-process bored
piles (1980’s) in Bangkok were constructed up to 50 m depth. Diameter of bored piles
166
commonly constructed in early days were ranging from 0.6 to 1.5 m. Table 2 below
summarizes the information of early major high-rise building projects in Bangkok.
Table 2 First high-rise building projects constructed by large diameter wet-processed bored
piles in Thailand (from 1979 to 1983)
Project Name Construction Method Diameter (m) Depth (m)

Royal Orchid Hotel Rotary-drilling with 0.80 - 1.00 33.0
(1979) auger and bucket
Taiping Tower Rotary-drilling with 0.80 - 1.50 32.0
River City Hotel Rotary-drilling with 0.80 – 1.00 27.5
Asoke Tower Rotary-drilling with 0.80 - 1.50 50.0
Time Square Rotary-drilling with 0.80 - 1.50 50.0
Building (1983) auger and bucket
Barrette
In some projects where bored piles were not feasible due to the site constraints,
applicable construction method and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimension ranging from
0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000 kN
have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in late 1970 for the
foundation of Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x 2.5m
and toe depth of 33m in late 1970. Barrettes can be constructed with flexible layout plan for
both vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous
row or column, radial, alternating long and short axis of barrette and a combination of two or
more of such patterns. One type of equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the mobilization cost for additional
equipment. In addition to the large bearing capacity requirements, on site difficulties such as
limited head room where piling rigs cannot be utilized, under such situations like presence of
overhead high voltage power cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the barrettes. A summary of
observation on barrettes of completed 26 projects completed with respect to selection criteria
is presented in Table 3.
Table 3 Summary of Barrette Selection
Criterion No. of Project Remarks

High Load Capacity 4 Incorporated with bored piles
Minimize Construction 6 Alternative for bored piles
Equipment
Limited Head Room for 3 Under existing structures such as bridges,
Excavation elevated expressway and power lines
Diaphragm Wall
Foundation as well as 3 Provision for the future requirement
portion of column
167
Static load test up to 5290 ton conducted on barrette set the record as the highest load
ever tested for a single cast-in-situ deep foundation in Thailand (Thasnanipan et. al. 2002a)
PROBLEMS OF WET-PROCESSED BORED PILES AND BARRETTES

IN EARLY DAYS
Basic but extensive problems were experienced in early stages of bored pile construction in
Thailand as summarized below.
• Limited availability and capacity of equipment
• Lack of skills in operation of equipment (particularly drilling rig)
• Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
using bentonite slurry)
• Limited knowledge and less advance technique in control of bentonite slurry
• Limited experience in construction method and related negative impact
• Quality of concrete for tremie concreting method
• Lack of experienced engineers and foremen
• Improper construction and quality control specification and guidelines for deep-seated
piles in local soil condition
• Improper design for constructibility
DEVELOPMENT IN CONSTRUCTION
Over the past 3 decades, along with the development of wet-processed piling
technology world-wide, equipment, construction method, design as well as better
understanding and realization of construction impact on the performance have been
significantly improved in Thailand. Table 4 summarizes the areas of improvement in bored
pile and barrette construction and main factors contributed to these.
Research works presenting some of these developments were published both locally
and internationally. The design, construction and behavior of bored cast in-situ concrete piles
in Bangkok Subsoil was presented by Thasnanipan et. al., (1998a). The construction and
performance of barrettes in Bangkok Subsoil was also reported by Thasnanipan et. al.,
(1998b). Effect of construction time and bentonite viscosity on shaft friction of bored piles
was also pointed out by Thasnanipan et al., (1998c). Teparaksa et. al., (1999) published the
base grouting of wet process bored piles in Bangkok subsoil which demonstrated the method
of base grouting and reported the grout spreading mechanism for long bored piles and
performance of base grouted bored piles.
(a) (b)
Fig. 1 (a) Guide wall for cruciform barrette (b) View of cruciform barrette after installation of
reinforcement
168
Construction of cruciform barrette (shown in Figure 1) for monopole type high-
voltage power transmission line was described in the work of Thasnanipan et. al., 2000a.
Above-mentioned research works reflect the development history of bored pile and barrettes
in Thailand and offered useful information to the local construction industry and perhaps to
the international deep foundation engineering society.
Table 4 Summary of development in bored pile and barrette construction
Area of In the past At present Main factor

development / (Early 1980’s) contributed to
improvement development
Speed of Minimum 3 days Less than 1 day to Better equipment,
construction required to complete complete diameter operating skills as
diameter 1.5m tip 50m 1.5m tip 50m bored well as improved-
bored pile pile knowledge in
construction method,
management and
local soil condition
Pile size and Bored pile Bored Pile Better equipment,
depth Maximum diameter 2.0m Maximum Diameter operating skills as
and common depth 25- 2.0m and common well as improved-
50m for bored pile. depth 25-60m knowledge in
Barrette Barrette construction method,
Limited in size and depth Various sizes & management and
depth over 60m local soil condition
Base grouting Not available Available Equipment
availability and
advance technology
Application of Not available Extensively use Material availability
polymer-based and research works
slurry (for bored
pile only)
Construction Not well understood Improving Experience from past
impact on quality projects and research
and performance works
Quality control in Not well established and Well established Experience from past
construction systematic and systematic projects and research
process works
Quality control Less methods available Better equipment Advance equipment,
test method and and limited knowledge in available and better experience from past
interpretation interpretation knowledge in projects and research
interpretation works
Experience from past projects and extensive research works provided better
understanding on construction impact on quality and performance of these wet-processed
deep foundations. Marked difference between outcome quality of bored piles constructed in
1980s and 1990s can be observed by significant less defects found in the latter. It can also be
observed from load test results that bored piles constructed in late 1990 have higher capacity
than those of 1980s as shown in Table 5.
169
Table 5 Load test results of bored piles constructed in 1970 to 1980 and 1990s
Year of Project Name Pile Dimension Design Test Total settlement

construction (Dia. & Depth) Load Load at max. test load
(ton) (ton) (mm)
1971 Pinklao bridge 1.5m x 45m 410 820 5
1995 Central Plaza 1.2m x 45m 700 1400 27
Pinklao
1980 Taiping tower 1.0m x 32m 300 1125 118
1990 High-rise building 1.0m x 32m 390 975 23
at Ekamai Rd.
Fig. 2 Progress of piling work for the Rama VIII main bridge
Bored piles have been extensively used for foundations of the majority of the elevated
expressways since 1991. Thousands of base-grouted bored piles were constructed for these
infrastructure projects including Second Stage Expressway (1991), Don Muang Tollway
Extension (1997), Bangna-Bang Pli-Bangprakong Expressaway (1998), and Wat Nakorn-In
project (2001). Over 700 bored piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed (1999-2000) by polymer-based slurry for the foundation of
Rama VIII Bridge, one of the initiatives of his majesty the King Bhumibol Adulyadej.
Progress of piling works at the bank of Chao Phraya river for the main bridge of Rama VIII
Bridge is depicted in Figure 2.
The availability of more reliable and powerful equipment and tools for drilling makes
possible to construct the rock-socket bored piles. Figure 3 shows the drilling rig equipped
with rock-auger used for construction of highway bridge across Mool River in 2001, in the
north-eastern part of Thailand, where drilling was carried out through weathered-sandstone by
powerful rotary drilling rig with core barrel and rock-auger.
Fig. 3 Rock-socket bored pile construction for Mool River Project - Ubon Ratchathani Province
170
DEVELOPMENT IN DESIGN CONCEPT AND PARAMETERS
In the initial stage of introducing bored piles in Thailand, the design concepts and
parameters were mainly based on the available literatures from the research works carried out
in other parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966),
Bowles (1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the
first research data available for design parameters of bored pile in Bangkok soft clay. The
author conducted the research based on 6 dry-processed small diameter bored piles with load
cells installed at the pile toe in AIT campus in Phathumthani. Though the piles were simply
constructed by dry-processed with casing method, this early-stage research study provided
some important information such as load-settlement behavior, adhesion factor (α) and bearing
capacity factor (N) of bored pile in Bangkok soft clay.
With the passage of time, design method and parameters for local subsoil were
improved as a result of research works carried out in 1980s. Differences between the behavior
of driven piles and bored piles were well realized from these researches. (Ng, 1983) presented
the load distribution characteristics of wet-processed bored piles founded in both first and
second sand layers of Bangkok based on the instrumented (strain gauges) pile load test
results. (Chiewcharnsilp, 1988) reported the shaft friction factor β values of sand layers in
Bangkok based on the instrumented load test results. In addition to the researches previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon, 1981; Ng, 1983; Chiewcharnsilp,
1988) Suchada (1989) determined the shaft friction factor (adhesion factor, α) of Bangkok
clay layers based on 11 bored piles of diameter ranging from 0.50 to 0.80m with the
embedded length varying between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were mainly based on the estimation of
shaft friction loads from plain static pile load test results with numbers of assumption since
instrumented pile load test results were limited.
With the peak of construction-boom, large numbers of instrumented full-scale static
load tests on bored piles were conducted throughout 1990’s which provided better
understanding on behavior of these deep-seated foundations. Researches focused on the
design parameters and methods were produced based on these test results. The design
parameters were more accurate and well established that wet-process cast-in-place
foundations became regarded as reliable foundations in construction industry of Thailand.
With more confidence on soil parameters and behavior of these deep foundations, the
designers designed to carry significantly higher loads for bored piles and barrettes constructed
in late 1990 than those of 1980s. (Thasnanipan et. al., 1999) reported the failure mechanism
of long bored piles in layered soil of Bangkok. The authors cited that for the bored piles
embedded in the multi layered soils of Bangkok, estimation of ultimate shaft friction capacity
needs to consider the brittle type of failure mechanism of stiff to hard clay layers and the
α values selected need to be adjusted accordingly. Peak and residual α values mobilized in
the stiff clay layers analyzed by the authors were plotted as shown in Figure 4 along with the
suggested curves by different researchers. It can be seen from the figure that the residual
α value of 2nd stiff clay layer (undrained shear strength values of 25 ton/m2) at the maximum
test load drops below the curve suggested by (Suchada, 1989). So the α values proposed in
Figure 4 for the stiff to hard clay layers gives overestimate of the ultimate shaft friction under
theses conditions.
Introduction of polymer-based slurry for wet-process bored piles marked a major
breakthrough for both construction and design engineers. Thasnanipan et al.,(2002b), reported
that bored piles constructed with polymer-based slurry have higher capacity than those
constructed with bentonite slurry. Figure 5 shows the shaft friction factors β of sand layers
for polymer-based bored piles in comparison with the design line of bentonite bored piles.
171
1.20
1.00
Tomlinson,1957
0.80
α
Adhesion Factor ,
Suchada,1989
0.40
Peak
0.20
Residual
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
2
Fig. 4 Comparison of adhesion factor α , suggested by different researchers with the actual
mobilized in the stiff clay layers (after Thansnanipan et. al 1999)
1 .0
0 .8
β (Ks tan δ)

( 1999 )
0 .4
0 .2
0 .0
28.0 30.0 32.0 34.0 36.0 38.0
Fig. 5 Back-calculated β values of polymer bored piles at maximum test load plotted on
design line of bentonite bored piles constructed in Bangkok subsoil
(after Thansnanipan et. al 2002b)
.
DEVELOPMENT IN QUALITY CONTROL TESTING
Improvement of testing equipment and powerful computer facilities are the key factors
contributed to the development in quality control testing. Interpretation skills relevant to local
soil condition and construction method of these tests were significantly improved in local
industry. For instance, in early 1980, interpretation of the sonic integrity (seismic test) test
results were needed to send to the specialists abroad which made the testing cost more
expensive. Significant cost-saving were achieved in some major projects of late 1990, as more
practical and precise interpretation were made to verify and establish the acceptance criteria in
proving the quality of suspected piles with anomalies.
172
Koden drilling monitoring system
In Thailand, prior to the availability of Koden testing equipment, borehole verticality
was checked by mechanical type equipment such as plumb float and adjustable globe.
However, the reliability of these methods was doubtful. Koden drilling monitor system was
believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can be
monitored rapidly from the electronic plot of continuous profile by Koden equipment. This
system is very useful for the verticality-control of long barrettes - unlike bored piles, guiding
by temporary casing is not available for barrette drilling.
Sonic integrity test

Sonic integrity test, also known as seismic test is the most common method of
integrity testing for both driven and bored piles in Thailand. Sonic integrity test is usually
selected for both quality check (control test) and retrospective investigation. It is the cheapest
in terms of cost and the simplest in terms of testing process. The main advantage of this test
is that since no particular preparation is necessary during the pile construction phase it is more
flexible to select which pile is to be tested. However, interpretation of sonic integrity testing
needs considerable experience and knowledge in testing, subsoil condition and construction
method. In many projects it is a part of the contractual requirement to conduct sonic integrity
test. Minimum 10 % to maximum 100 % of production piles are commonly tested. It is also
a reasonably acceptable method for applying as a retrospective investigation in determining
integrity of the pile. The signal characteristics and their interpretations of sonic integrity test
on piles founded in Bangkok subsoil were reported in details by Thasnasipan et al. (1998d).
Cross-hole sonic logging test

Sonic logging test is relatively expensive. It is mainly employed as a pre-planed site
quality control testing. The major advantage of this method is that test can be carried out
shortly after the pile construction. Hence, rectification measures can be implemented if the
pile is detected with defect while the foundation contractor is on site. However, this method is
generally not applicable if pile integrity is in question due to post-construction activities as
access tubes are usually grouted after completion of the test. The first sonic logging test was
believed to be conducted in 1982 for the wet-processed bored piles of Memorial Bridge
Project where bad concrete zones were detected at depth about 20m and 1m (Ng, 1983). Use
of Sonic logging test for checking pile integrity has increased in Thailand in recent years.
The results from sonic logging test conducted on model piles in Bangkok helped to extend the
knowledge of the signal characteristics and interpretation (Thasnanipan et. al., 2000b).
High strain dynamic load test

High strain dynamic integrity test has become a well-accepted method especially for
evaluating the pile capacity in today’s foundation industry of Thailand and it is applied for
both driven and bored piles.. A large number of related technical papers and case histories of
the test have been published and it is a part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dynamic Testing of Piles). Thasnanipan et. al.
2000b, reported the application of dynamic load testing in Thailand.
CONCLUDING REMARKS
Development of bored pile and barrettes in Thailand has been presented. According to
the authors’ experience as a contractor, development in both construction and design aspects
of these deep foundations in past decades are significant. In authors’ opinion, however, there
are many works to be done to improve further with particular focus on constructibility issues,
concrete technology for wet-processed bored piles and barrette, reliable but cost-effective
173
quality control testing and value-engineering. Starting from planning stage, site investigation,
design, construction and inspection should be integrated such that designers, contractors and
construction inspectors participating as a team with common goal. Appropriate and practical
specification should be established jointly by these parties for local soil condition and
construction method. Practical acceptance criteria should be developed to verify bored piles or
barrettes with suspected anomalies. Continuing education should be promoted for both
designers, inspection engineers and contractors. The Geotechnical Chapter of the Engineering
Institute of Thailand under the royal patronage of his majesty the King Bhumibol Adulyadej,
has started to establish the standard code of practice and guidelines for wet-processed bored
piles which will serve as a yardstick for the deep foundation industry upon completion in near
future.
REFERENCES
Bowles, J. E. (1968) Foundation analysis and design, McGRaw-Hill, New York

Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles – a summary, Soils-Soils, No.
18-19, pp. 21-32
Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored piles (in Thai), M.
Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213, Asian Institute of
Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil mechanic and
foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second sand layer, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay, Proceedings of 10th
conference on soil mechanic and foundation engineering, Vol. 2, pp. 815-818
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian Institute of Technology,
Bangkok, Thailand
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999), Base grouting of wet process bored piles in Bangkok
subsoils, 11th Asian Regional Confer. on soil mechanics and geotech. Engg., Seoul, Korea, pp. 269-272.
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and behavior of bored
cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louise, Missouri, U.S.A., p.p.281-287
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils, construction and
performance, Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and Bentonite Viscosity on
Shaft Capacity of Bored Piles, Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded in Bangkok
Subsoil-Signal Characteristics and Their Interpretations, Proceedings: Fourth International Conference
on Case Histories in Geotechnical Engineering, St. Louis, Missouri, U.S.A., p.p.1086-1092
Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles in Layered Soils
of Bangkok, Civil and Environmental Engineering Conference, Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference on Geotechnical & Geological Engineering,
Melbourne, Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive Integrity Testing on
Piles Founded in Bangkok Subsoil, 6th International Conference on Application of Stress-Wave Theory
to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume Two, ASCE, Febuary 14-16 2002,
Orlando, Florida, USA, p.p. 1073-1087
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b) Performance of
Wet-Process Bored Piles Constructed with Polymer-Based Slurry in Bangkok Subsoil, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication No. 116, Volume
One, ASCE, February 14-16 2002, Orlando, Florida, USA, p.p. 143-157
174
Proceeding of the 4th International Geotechnical Seminar on Deep
Foundations on Bored and Auger Piles
2-4 June 2003, Ghent, Belgium

and Auger Piles
Edited by:
Van Impe, W. F.
Behaviour of Polymer-based Slurry for Deep-

seated Bored Piles in Multi-layered Soil of
Bangkok
Narong Thasnanipan, Zaw Zaw Aye and Thayanan

Boonyarak
SEAFCO Co., Ltd. Bangkok, Thailand
175
Behaviour of Polymer-based Slurry for Deep-seated Bored Piles in
Multi-layered Soil of Bangkok
Narong Thasnanipan, Zaw Zaw Aye and Thayanan Boonyarak
ABSTRACT: This paper focuses the behaviour of polymer-based slurry both in the practical construction of
deep-seated bored piles and in the laboratory. The behaviour of polymer slurry with different mixed-ratios is
studied in the laboratory by means of model tests. Test models were constructed with the sand samples ob-
tained from 25 to 40m depth, First and Second Sand layers of Bangkok. The model tests showed that signifi-
cant strength improvement can be achieved in slurry-sand interface if polymer slurry (polymer 0.05% by
weight) is mixed with a small percentage of bentonite (1% by weight). The model tests also proved that filtra-
tion of slurry in permeable layers can be reduced by adding a small percentage of bentonite in polymer slurry.
1 INTRODUCTION
In Bangkok, deep-seated large diameter bored piles 0 Weathered Crust
0
of depths ranging from 35 to 60m and diameter of Bangkok Soft Su= 10-25 KPa
γ t = 14-16kN/m 3
800mm to 1800mm are commonly used for founda- 10
Clay
-
10
tion of heavy structures such as high rise buildings,

elevated expressways, flyovers or overpass bridges Medium to
Depth (m)
20 Stiff Clay Su= 50-14020- KPa

γ t = 17-21kN/m 3
and recently for underground subway stations. Con-
Act
ventionally, these deep-seated bored piles in the

ual
Medium -
30 SPT-N= 20-40
multi-layered soil of Bangkok were constructed by Dense Sand
Pie
30
Hyd
(First Sand)
γ t = 20kN/m 3
zom
ros
rotary drilling under wet process or slurry replace-

etric
tatic
40 -
ment method, using pure bentonite slurry as support Hard Clay Su> 200 KPa
γ = 21KN/m 3
40
Dra
Line
t
fluid in boring process. As an alternative to ben-
wdo
tonite slurry, polymer-based slurry has been increas- Dense to Very

wn
-
50 50
Lin
ingly used in Bangkok recently. Polymer-based (Second Sand) γ t = 21KN/m 3

e
slurry with environmentally friendly properties, is 60 -

60
found to be a good solution in minimizing the prob-

lems associated with the disposal of contaminated Figure 1. Typical subsoil profile and piezometric profile of
Bangkok (Thasnanipan et. al. 2002)
soil as well as the site pollution. This paper focuses
the behaviour of polymer-based slurry both in the
practical construction of deep-seated bored piles and 3 WET-PROCESS BORED PILE
in the laboratory. CONSTRUCTION METHOD
Wet process method as its name implies makes the

2 GENERAL DESCRIPTION OF BANGKOK pile under wet condition by using drilling slurry as
SUBSOIL supporting fluid for borehole stability. Depending on
the subsoil condition, temporary casing of 14-18m in
Subsoil profile is relatively consistent at different lo-
length is also used as a support in soft clay layer.
calities in Bangkok. A typical subsoil profile is
Drilling is commenced by dry process with auger
characterized by the alternating layers of clay and
applying rotary drilling action in the soft clay layer
sand deposits as shown in Figure 1.
and stiff clay layer. Before reaching to the first sand
177
layer, slurry is fed to the borehole and drilling is layer is formed in the soil-slurry interface of perme-
continued with the bucket to the final depth. A able layers. This thin membrane of polymer-based
cleaning bucket or airlift is used to clean the bottom slurry is capable of reducing the high filtration in
of the pile prior to reinforcement cage installation. sand layers of Bangkok. However, overall volume of
Concreting is carried out by tremie method under the fluid loss for polymer-based bored piles is found to
slurry. Slurry from borehole is normally pumped be higher than that of bentonite bored piles accord-
back to the storage tank during concreting and re- ing to the field observation of deep-seated bored pile
used for next piles after treating to get specified construction in Bangkok. Figure 2 (a) and (b) dem-
standard properties. Depending on the level of de- onstrates the generalized behaviour of bentonite and
sign pile cut-off, bored piles are normally overcast polymer slurry in borehole respectively.
about 1 to 3m above cut-off level to avoid presence
of contaminated concrete within the design pile Thick filter
length. Suspended
soil particle
cake is
formed
in bentonite
slurry
4 BEHAVIOUR OF SLURRIES IN OVERVIEW

Bentonite
slurry
Bentonite and polymer slurry react differently with
the soil layers. Pure bentonite slurry stabilized the
borehole wall by its density and formation of a
membrane or filtercake particularly in permeable Soft sediment
at base of
layers. Bentonite slurry is capable of maintaining the bored hole
soil particles in suspension. Therefore, soil particles (a)

in bentonite slurry column stay in suspension for
long hours and slowly settled toward the borehole Thin filter
bottom. Formation of filtercake reduces the slurry cake is
formed
filtration into the permeable layers, thick First and Soil particle
is settling
Second sand layers typically encountered in Bang- (b) Sand
kok as shown in figure 1. Particle
Unlike pure bentonite slurry, polymer-based Polymer slurry

mix with very
slurry does not suspend sand especially coarse grain little amount of
Bentonite
particles. Hence, in polymer slurry, sand particles
(mainly from cuttings during drilling) tend to fall out Dense sediment
of suspension and settle on the base of borehole sig- at base of
bored hole
Polymer Chain
Structure
nificantly sooner than that of pure bentonite as evi-

dent from field observation. In Bangkok, depending
on subsoil profile and composition of polymer-based (b)
slurry, 0.1m to 1m of sediment is commonly ob- Figure 2. Generalized behaviour of slurries in borehole (a) ben-
served on base of borehole in 2 to 8 hours. The sand tonite slurry (b) polymer-based slurry
sedimentation process under polymer-based slurry is
found to be stable with time. Polymer molecules
tend to wrap around clay and silt particles and pro- 5 BEHAVIOUR OF SLURRIES IN
duce small agglomerated structure. By its own LABORATORY
weight, the resulting agglomerated structures tend to
settle out of suspension slowly and accumulate Laboratory study was conducted to investigate the
mushy sediments on the firstly settled sand sedi- followings properties of Bangkok sand layers in dif-
ments at the bottom of the borehole. Some of the ag- ferent types of slurries (Boonyarak, 2002).
glomerated particles also tend to float temporarily on • Filtration through the sand layers
the surface of polymer slurry stay in the suspension. • Shear strength of slurry-sand interface (by pocket
The polymer molecules then proceed to attach them- penetrometer)
selves to larger and appear as a bulky material that • Friction test for slurry-sand interface
some observers have termed “oatmeal” (O’Neill and
Reese, 1999). In Bangkok as small percentage of
5.1 Sand specimen preparation
bentonite is added in polymer-based slurry, similar
to polymer-clay system Majano and O’Neill, (1993) The procedure involved in preparation of sand
, it can be assumed that a thin polymer-bentonite specimen is outlined below.
178
• Sand samples were collected from the actual 5.3 Filtration tests
bored piles drilling – excavated soil of Bangkok In order to study the laboratory behaviour with re-
First and Second sand layers from depth about gard to the filtration rate of different slurries in sand
25m and 40m below existing ground level respec- formation, series of filtration tests were performed.
tively. The procedure involved in the test is outlined in the
• Basic properties (minimum and maximum den- following section. The test apparatus and arrange-
sity, grain size distribution, hydraulic conductiv- ment used were similar to that of standard API filter
ity) of collected sand were tested and recorded. press as shown in Figure 3. The applied pressure
Determined the natural density of sand. was predefined for the laboratory filtration test in
• Sand samples were then mixed with water to ob- simulation of actual in-situ bored pile drilling condi-
tain water content of 10%. tion – the differential fluid pressure between the
• Poured sand into the testing mold. slurry and pore pressure of formation soil at speci-
• Compacted the sand in the mold to the desired fied level, taking piezometric draw-down condition
natural density determined above of Bangkok sand layers (as shown in Figure 1) into
consideration.
5.2 Slurry preparation and physical Tests • Poured the slurry into the test chamber on pre-
prepared sand specimen as described above
Slurries were mixed by propeller-type mixer follow- • Applied predefined pressure through the pressure
ing the procedure outlined below. inlet.
• Mixed bentonite powder with clean water until • The slurry was flowed through the sand medium
blended. Leave the slurry for few hours and re- by applied pressure depositing solids in suspen-
mixed. sion at the slurry-sand interface (filter cake for-
• Mixed polymer powder with clean water until all mation)
powder dissolved. • The filtrate was collected in a graduated cylinder
• Mixed premixed bentonite and polymer slurry to- • The cumulative filtrate volume was recorded at
gether (except for pure bentonite and polymer regular interval until the target test time (60 min-
slurry) utes)
• Tested basic properties of the slurry • Released air pressure and pumped out the slurry
The results of the physical tests of different from the chamber (sample was then ready for
slurry types from the works of Boonyarak (2002) are penetrometer test)
presented in Table 1. PRESSURE
INLET
Table 1. Mixed ratios of different slurry types and results of

physical tests (Boonyarak, 2002)
Mixed ratios of
slurry (percentage to Physical test results
water by weight) SLURRY
Slurry
Polymer Bentonite Slurry Marsh pH
Type FILTER CAKE
(%) (%) Density Funnel (SLURRY-SAND INTERFACE)
(g/cm3) Viscos- SAND
ity (sec)
1 0.05 1 1.007 41.8 9.16
2 0.05 1.5 1.012 64.3 9.20
3 0.05 2 1.019 71.3 9.38 FILTRATE
4 0.08 0 1.000 43.3 7.78

5 0.08 1 1.008 49.8 8.87
6 0.08 2 1.020 98.3 9.13 Figure 3. Picture and schematic of filtration test
7 0.10 0 0.999 45.5 7.69
8 0.10 1 1.008 58.6 9.14 The filtration test results of Bangkok First Sand
9 0.10 2 1.020 110 9.35 layer is presented in Figure 4. An influence of filter
10 0.15 0 0.999 50.0 7.75 cake is evident in Figure 4 for pure bentonite slurry
11 0.15 1 1.009 71.8 9.28
(Bentonite 5%) - the filtrate volume was very low
and became constant within a relatively shorter pe-
12 0.15 2 1.019 120.4 9.30
riod. For the pure polymer slurry, filtration volume
13 0.20 0 1.000 60.3 7.81
was high and no constant rate of filtration was ob-
14 0.20 1 1.008 83.2 9.18
served until the target testing time. This is appar-
15 0.20 2 1.020 160.3 9.32
ently due to the absence of filter cake formation. In
16 0 5 1.050 70 9.12 case of polymer-based slurry (polymer 0.05% with
179
bentonite 1%), filtration volume was significantly The relationship between shear strength of inter-
reduced in comparison with pure polymer slurry as face and dosage of polymer in slurry is depicted in
can be observed in the Figure. This finding is consis- Figure 6. Based on the regions in which the shear
tent with the field application in actual bored pile strength values fall, the test results can be catego-
construction for deep-seated bored piles in Bangkok. rized in 3 groups – interface of pure bentonite slurry,
4000
pure polymer slurry and polymer-bentonite slurry.
Bentonite 5%
The results indicate that interface of polymer-
3500
Polymer 0.05% Bentonite 1%
bentonite slurry is significantly stronger than that of
pure bentonite and polymer. The pictures showing
Filtrate Volume (cm )
3000
3
Polymer 0.15%
2500 the condition of interface produced by pure polymer
2000
slurry and polymer-bentonite slurry can be seen in
Figure 7 (a) and (b) respectively. Stronger surface
1500
condition of polymer-bentonite interface is evident
1000 from these pictures.
500
0
0 10 20 30 40 50 60
Time (min)
Figure 4. Filtration of different slurry types through Bangkok

First Sand sample.
5.4 Shear strength of slurry-soil interface

Shear strength of slurry-sand interface formed by (a) Interface of slurry-sand (pure polymer)
different types of slurries was tested by means of
pocket penetrometer. These tests were conducted
immediately after completion of filtration test de-
scribed above. Picture and schematic of pocket
penetrometer test are shown in Figure 5.
Pocket
Penetrometer
slurry-sand (b) Interface of slurry-sand (polymer-bentonite)

interface
Sand Figure 7. Condition of slurry-soil interface after the filtration

test with (a) pure polymer slurry (b) polymer-bentonite slurry
Figure 5. Picture and schematic of pocket penetrometer test ap- 5.5 Friction tests
plied on slurry-soil interface
Though it is almost impossible to model the field
250 condition of shaft friction resistance influence by
slurry, attempt was made to study the frictional re-
Interface Shear Strength (kN/m )
2
200 sistance of slurry-sand interface in the lab. A series

Polymer + Bentonite Slurry of friction tests were conducted using the apparatus
Pure Bentonite
150 Slurry Pure Polymer shown in Figure 8 following the procedure outlined
Slurry below.
100 • After completion of pocket penetrometer test the
sample was cut in such a way that less disturbed
50 Bentonite 0% Bentonite 1% slurry-sand interface was obtained.
Bentonite 1.5% Bentonite 2%
• Cement-mortar plate was prepared few days prior
Bentonite 5%
0 to the test
0 0.05 0.1 0.15 0.2 0.25
• Set up the sample in such a way that slurry-sand
Polymer Dosage (%) interface was placed (upside-down position) di-
Figure 6. Test results of penetrometer test applied on slurry-soil rectly on the top of cement-mortar as depicted in
interface Figure 9 (a).
180
• Applied a range of normal and shear loads as • Frictional resistance of slurry-sand interface of
shown in schematic of Figure 9. polymer-bentonite (polymer 0.1% + bentonite
• Recorded loads and displacement at regular 1.5%) is slightly higher than that of pure ben-
interval until friction load reached maximum tonite (5% bentonite).
value or displacement reached 25% strain • The difference of frictional resistance of slurry-
• The test was also conducted on pure sand sand interface between polymer-bentonite (poly-
sample directly placed on the cement-mortar mer 0.1% + bentonite 1.5%) and pure bentonite
following above procedure to make a rough (5% bentonite) was far less than expected. Likely
comparison of frictional resistance between factor contributed to this result is that the contact
slurry-sand interface and pure sand against time between the bentonite slurry and sand sam-
mortar plate. ple was not long enough to develop an exces-
sively thick filter cake.
100
2nd sand-mortar
polymer 0.1%+bentonit 1.5%
80
Max. Friction (kN/m2)

bentonite 5%
60
40
20
0
0 50 100 150 200
Normal Stress (kN/m2)
Figure 8. Laboratory friction test apparatus

Figure 10. Friction test results of 3 different types of samples
N o rm a l
L oad
S and
s lu rr y -s a n d 6 FIELD APPLICATION AND BEHAVIOUR OF
in te r fa c e F ric tio n POLYMER-BASED SLURRY IN BANGKOK
L oad
C e m e n t-m o rta r
In Bangkok, the primary reason for switching from
(a) classic pure bentonite slurry to polymer-based slurry
in most cases was to minimize the problems associ-
ated with environmental issues caused by the use of
bentonite slurry. In the initial stage of using poly-
mer-based slurry, the major concerns were the po-
tential negative impact on stability of the borehole
and the shaft friction capacity as this alternative
slurry was not yet accustomed to the bored piling in-
dustry in Bangkok. However, no adverse effects
(b) from the use of polymer-based slurry have been ex-
Figure 9. (a) Schematic demonstration of friction test (b) pho- perienced to date and numbers of projects have been
tograph of friction test successfully completed with this environmental
friendly drilling slurry.
The relationship between normal stress and maxi-
mum friction stress of different samples are plotted 6.1 Properties of polymer-based slurry
and presented in Figure 10. Following points can be
drawn from this Figure.
• Frictional resistance of slurry-soil interface of It is important that properties of slurry used are
both pure bentonite (5% bentonite) and polymer- within the applicable ranges to avoid the detrimental
bentonite (polymer 0.1% + bentonite 1.5%) were effect on the performance of bored piles. O’Neill &
less than that of sand-mortar – suggested that fric- Reese (1999) cited that there is no perfect set of
tional resistance was reduced by application of slurry specifications that can be used on every job,
slurry. but specifications should be tailored to fit the re-
181
quirements of a particular job at a particular loca- No specific method or formula is available to deter-
tion, where possible. Today, though use of polymer mine the required best composition of polymer-
slurry is widely acceptable in bored piling industry, bentonite ratios. Therefore, at each site, different
information on the desirable properties is still ex- mixed-ratios should be made on trial in the initial
tremely limited. Table 2 shows the ranges of proper- stage to determine the most appropriate ratio. Table
ties of polymer slurries specified by the International 3 presents the mixed composition of polymer-based
Association of Foundation Drilling Standards and slurry of bored piles constructed at 5 different loca-
Specification Committee (ADSC, 1999), those suc- tions in Bangkok. According to the load-settlement
cessfully used on several projects as reported by curves obtained from static pile load tests, observed
O’Neill and Reese (1999) from the work of Majano settlements of all piles were less than those of pre-
et al. (1994) and the properties commonly applied dicted values using bentonite pile parameters Thas-
for bored piles in Bangkok subsoil. nanipan et al. (2002).
Table 2. Properties of polymer slurries specified by ADSC and

Polymer-based slurry used in Bangkok Table 3. Technical data of polymer-based slurry constructed at
Majano et al., 1994 Common different localities in Bangkok
ADSC Emulsified Dry Practice in Pile ID
Property Description
(at 20 C) PHPA PHPA Bangkok T1 T2 T3 T4 T5
(Vinyl) (PBS**) Pile Diameter (m) 0.80 0.80 1.20 1.80 1.50
Density 1.02 0.995 to 0.995 to 1.05 Pile Length (m) 49 50 51 62 54
(g/cc) Maximum 1.01 1.01 Maximum Mixed Ratio per P* 0.05 0.05 0.05 0.05 0.05
1m3 slurry (% in
Viscosity 40 – 90* 33 - 45 50 - 120 40 to 90 weight) B** 1.2 1.2 1.0 1.0 1.2
(sec/qt) (+/- 5) Total standing time
pH 7 - 12 8 – 11.7 8 – 11.7 8 – 11 prior to concreting 17:20 12:20 14:00 23:00 15:30
API Sand 1.0 0 to 1 0 to 1 Less than (Hr : Mins)
Content Maximum 1% Density
1.06 1.03 1.01 1.04 1.04
Proper- (g/ml)
*or as recommended by manufacturer and approved by geo-
ties of Viscosity
technical engineer, ** Polymer-based slurry 40 45 44 51 44
slurry (sec/qt)
prior to pH 8 8 8 8 8
concret- Sand
ing content 0 0.5 0 0 0.5
6.2 Mixed-ratio and procedure of mixing (%)
Approx.
Pile em- First sand 12 8 7 20 5
Dry anionic partially hydrolyzed polyacrylamide
bedded
(PHPA) polymer powder premixed with fresh water in sand Second
and a small percentage of bentonite is mainly used layers sand
7 9 13 18 13
for long bored piles in Bangkok. First, a small dos- (m)
age of bentonite powder is premixed with fresh wa- * polymer, **Bentonite
ter (bentonite 5-15kg per cubic meter of water) and
leave in the storage tank for minimum 12 hours. The 6.3 Effective base cleaning for polymer-based
predetermined dosage of dry polymer (typically bored piles in Bangkok
polymer powder 0.5kg per cubic meter of premixed
bentonite slurry) is then added and mixed to get so Since bentonite and polymer slurry react differently
called polymer-based slurry. with the soil layers as described in section 4 above,
O’Neill and Reese (1999) cited that vigorous mix- base cleaning method for two different slurries
ing of polymers supplied in the dry form should be should be selected based on their behaviour. O’Neill
avoided, since polymer chains can be broken down & Reese (1999) pointed out that vigorous mechani-
and the polymer slurry rendered ineffective. Poly- cal agitation such as pumping by centrifugal pumps
mer-based slurry is stored in the steel tanks before can shear the polymer chains severely. As soil-
supplying to the bored pile drilling. The reason for polymer reaction system is largely depending on the
adding bentonite for long bored piles is to minimize polymer chemical chain structure, excessive break-
the fluid loss (high filtration of slurry into formation ing of it would affect the performance of the slurry.
soils), which was observed from the trial boreholes Therefore, high turbulent means of base cleaning
drilled only with pure polymer slurry. Depending on methods such as recycling by centrifugal pumps
the thickness and permeability of sand layers present should not be performed for the polymer-based
at site and the embedded length of the pile in sand bored piles. Field experience shows that base clean-
layers, addition of bentonite may not be necessary. ing by air lift method cannot effectively clean thick
182
sediments at the base of the polymer-based bored slurry in the USA are higher than those of bentonite
pile since the sediments deposited at the base of the piles. Similar finding was reported by Thasnanipan
pile are comparatively dense for bored piles con- et. al. (2002) from the results of instrumented test
structed with polymer-based slurry. According to the pile in Bangkok.
field observation on polymer-based bored piles, rate Load distribution curves along the pile shaft at
of sedimentation at pile base was generally found to various applied loads of instrumented test pile (T3 of
be stabilized few hours after stopping the drilling Table 3) constructed by polymer-based slurry is il-
operation. Figure 11 presents the thickness of the lustrated in Figure 12. The values of average unit
sediments at base of polymer-based bored piles re- skin friction developed by the maximum test load at
corded at regular interval. The records shown in the different soil layers along the shaft of test piles
Figure 11 were made at the same site in Bangkok for in comparison with those of ultimate unit skin fric-
diameter 1000mm bored piles of about 55m in tion calculated using the empirical formula are also
length. The soil profile at the site is similar to that of illustrated in these figures. As can be seen in the fig-
Figure 1, a typical subsoil profile of Bangkok. This ure, developed average unit shaft friction in sand
sedimentation behaviour of polymer-based bored layers are significantly higher than those of theoreti-
pile should be taken account in base cleaning cally calculated values using normally used formula
method and process. and parameters applied for bentonite bored piles in
Bangkok. This finding suggested that polymer-based
45 slurry bored piles produced better shaft friction re-
Thickness of sediment (cm)
40
sistance in sand layers than that of bentonite bored
35
piles.
30
25 1 Load (kN) Unit Skin Friction (kN/m2)
0 50 100 150 200 250
20 0 0 6000 12000 18000
Pile No.1 0 0
15 Pile No.2 Soft Clay
10 Pile No.3 10
10 10
Pile No.4
5
Pile No.5 Stiff Clay
0 20 20
V. Stiff to 20
Depth (m.)
0 50 100 150 200 Hard clay
30 Dense Sand
Time after stopping drilling operation (min) 30 30
Hard Clay
Figure 11. Sediments thickness with time at pile base recorded 40
Dense to
40 40
from polymer-based bored piles Very Dense

50 Sand 50 50
Qult (calculated)
The effective base cleaning for long bored piles in 60 60 60
Bangkok, may be achieved by the following proce- Figure 12. Load distribution curves and unit skin friction of in-
strumented test bored pile diameter 1.20m, 51m deep con-
dure. structed by polymer-based slurry in Bangkok
• Carry out drilling to 1 to 2m above the final de-
sign depth under polymer-based slurry for the
first bored pile 7 DISCUSSION ON LAB TEST RESULTS AND
• Leave the borehole for 2 to 8 hours (depending on FILED PERFORMANCE
the rate and stabilization trend of sedimentation
observed from trail borehole at each site) Polymer-based slurry used in construction of bored
• Move the drilling rig to commence the drilling of piles in Bangkok is commonly composed of 0.05%
second bored pile and 1% (by weight per 1m3 slurry) of polymer and
• Drilling rig come back to the first bored pile and bentonite respectively. The performance of bored
piles constructed with this composition was proved
carry out final drilling for (remaining 1 to 2m)
to be significantly better than that of bored piles
with the cleaning bucket
constructed with pure bentonite slurry. Using pure
• Immediate installation of reinforcement cage and polymer in the field, construction of deep-seated
concrete pouring bored piles in multi-layered soil of Bangkok is found
to be impractical particularly at the site where thick
6.4 Shaft friction resistance of polymer-based bored sand layer presence due to the excessive loss of
piles slurry into the formation. Table 4 presents the find-
ings of laboratory tests in comparison with field ob-
Ata and O’Neill (1998) concluded that unit shaft re- servation. Though perfect comparison of laboratory
sistances of test piles constructed with polymer findings and field observation is not possible, Table
183
4 provides some useful summary on the behaviour which is proved to be performing better than classic
of slurry in the laboratory and of practical applica- pure bentonite slurry in both construction and load-
tion in the field. ing capacity aspects.
Table 4. Summarized comparison of parameters between lab

findings and field observation REFERENCES
Parameters Laboratory Field Discussion
test observation notes ADSC Standards and Specifications for the Foundation Drill-
Filtration Polymer Polymer Laboratory ing Industry. 1999, ADSC, Revised 1999, Incorporating
(0.05%) + (0.05%) + results agree ACI 336.1R-98
bentonite bentonite well with Ata A. A. and O’Neill M. 1998. Side-wall stability and side-
(1%) sig- (1%) signifi- field observa- shear resistance in bored piles constructed with high-
nificantly cantly re- tion molecular-weight polymer slurry, Third International Geo-
reduces fil- duces filtrate technical Seminar on Deep Foundation on Bored and Au-
trate vol- volume ger Piles, Ghent, Belgium, 19-21 October 1998 : 111-117.
ume than (slurry loss) Balkema : Rotterdam,.
pure poly- Boonyarak T. , Behaviour of Polymer Slurry for Wet-process
mer bored Piles Construction in Bangkok Subsoil, Master The-
Shear Slurry pro- Slurry pro- Indirect com- sis (in Thai), Chulalongkorn University, 2002
strength of duced by duced by parison of lab Majano R. E. and O’Neill M. W. 1993. Effect of mineral and
slurry- polymer polymer shear strength polymer slurries on perimeter load transfer in drilled
sand inter- mixed with mixed with vs. actual shafts, A Report to ADSC, University of Houston, Dept. of
face small per- small per- shaft friction Civil Engineering, UHCE.
centage of centage of measured O’Neill M. W. and Reese L. C. 1999. Drilled Shafts : Con-
bentonite has bentonite has from instru- struction Procedures and Design Methods, U.S Department
significant significant mented pile of Transportation Federal Highway Administration, Publi-
higher higher shaft load test cation No. FHWA-IF-99-025, ADSC Publication No.
strength than friction ca- agrees well ADSC-TL 4.
pure polymer pacity than Thasnanipan, N., Aye, Z. Z., Submaneewong, C. and Te-
or bentonite that of pure paraksa, W. 2002. Performance of Wet-Process Bored Piles
bentonite Constructed with Polymer-Based Slurry in Bangkok Sub-
Frictional Slurry pro- Slurry pro- Reduction of soil, Proc. of the International Deep Foundations Congress
resistance duced by duced by frictional re- 2002, Geotechnical Special Publication No. 116, Volume
of slurry- polymer polymer sistance is not One, American Society of Civil Engineers, February 14-16
sand inter- mixed with mixed with significant in 2002 : 143-157. Orlando, Florida, USA : ASCE
face small per- small per- laboratory. In
centage of centage of lab, contact
bentonite has bentonite has time between
higher significant bentonite
strength higher shaft slurry and
(slightly) than friction ca- sand sample
pure polymer pacity than was not long
or bentonite that of pure enough to de-
bentonite velop an ex-
cessively
thick filter
cake
8 CONCLUSION
Behaviour of polymer-based slurries in both labora-

tory and field have been presented. The laboratory
findings of this research work provided backup and
useful information for the actual construction of
bored piles in the field. The laboratory model tests
reveal that significant strength improvement can be
achieved in slurry-sand interface if polymer slurry
(polymer 0.05% by weight) is mixed with a small
percentage of bentonite (1% by weight). This mix
composition significantly reduced the filtrate volume
in the laboratory test in comparison with that of pure
polymer. Similar dosage of bentonite is commonly
added in preparing polymer-based slurry for con-
struction of deep-seated bored piles in Bangkok
184
G EO T ECHN ICA L SPEC IAL PUBL IC AT ION NO . 124
GEOSUPPORT 2004
DRILLED SHAFTS, MICROPILING, DEEP
MIXING, REMEDIAL METHODS, AND
SPECIALTY FOUNDATION SYSTEMS
PROCEEDINGS OF SESSIONS OF THE GEOSUPPORT CONFERENCE :

INNOVATION AND COOPERATION IN THE GEO-INDUSTRY
January 29-31, 2004

Orlando, Florida
SPONSORED BY
International Associations of Foundation Drilling (ADSC)
EDITED BY
John P. Turner
Paul W. Mayne
Effectiveness of Toe-Grouting for Deep-Seated Bored Piles in Bangkok Subsoil

Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong
185
Effectiveness of Toe-Grouting for Deep-Seated Bored Piles in Bangkok Subsoil
Narong Thasnanipan1, Zaw Zaw Aye2, Chanchai Submaneewong3
ABSTRACT: A comprehensive study on the effectiveness of two different toe-

grouting methods, known as tube-â-manchette and drill-and-grout, commonly
applied for deep-seated bored piles in Bangkok is presented in this paper. Mobilized
unit end bearing of bored piles constructed by two toe-grouting methods are
compared. The problems encountered in practical application of different methods
for large diameter deep seated bored piles are highlighted. Despite considerable extra
cost, various construction problems encountered and risks involved in successful
execution of grouting, drill-and-grout method does not offer particular advantage
over tube-â-manchette method as far as pile capacity improvement is concerned.
INTRODUCTION
In Bangkok, due to the prevailing subsoil and groundwater conditions, all the
deep-seated bored piles (depth ranging from 25 to over 60 m) are constructed by wet-
processed or slurry displacement method. Although the wet-processed method is
most suitable and its application has been well established in Bangkok subsoil, the
method itself cannot avoid undesired effects especially loosening and upheaval of
soil at pile base as well as accumulated suspension of loose materials or sediments
from drilling operation. These effects naturally may create a loose or soft pile toe
causing negative impact on the pile capacity or load vs. pile head movement
behavior. One of the special measures to improve the soft pile toe is to apply the
post-grouting at pile toe. Two different methods, known as tube-â-manchette and
drill-and-grout are commonly applied in Bangkok. A comprehensive study on the
effectiveness of two toe-grouting techniques applied for deep-seated bored piles in
Bangkok is presented in this paper.
1
Managing Director, 2Project Manager, 3Geotechnical Engineer
SEAFCO Company Limited, Bangkok, Thailand, seafco@seafco.co.th
187
SUBSOIL PROFILE AND EXISTING PIEZOMETRIC PROFILE
Subsoil profile is relatively consistent at different localities in Bangkok. A

typical subsoil profile is characterized by the alternating layers of clay and sand
deposits as shown in Figure 1. Pore water pressure profile illustrating the current
piezometric drawdown condition of Bangkok is also shown in Figure 1.

0 Weathered Crust
0

Clay γ t = 14-16kN/m 3
10 -
10
Medium to
Depth (m)
20 Stiff Clay Su= 50-14020-KPa

γ t = 17-21kN/m 3
Act
ual
Medium
-
Pie
30
Hyd
(First Sand)
γ t = 20kN/m 3
zom
ros
etri
tatic
-
cD
40
γ = 21KN/m 3
Line
t
raw
dow
50 Dense to Very - nL
50
ine
(Second Sand) γ t = 21KN/m 3
60 -
60
Figure 1. Typical soil profile of Bangkok with piezometric drawdown

condition (Thasnanipan, 2002)
WET-PROCESS BORED PILE CONSTRUCTION METHOD
In Bangkok, medium to large diameter (0.6 m to 1.8 m diameter) bored piles

with depth ranging from 24 m to over 60 m are constructed by wet process due to the
prevailing subsoil and ground water condition. Temporary casing of 14-18 m in
length is typically used as a support in soft clay layer. Drilling is commenced by dry
process with auger applying rotary drilling action in the soft clay layer and stiff clay
layer. Before reaching to the first sand layer, supporting fluid or slurry is fed to the
borehole and drilling is continued with the bucket to the final depth. Depending on
the type of slurry used, either a cleaning bucket or airlift method is commonly
applied d to clean the bottom of the borehole prior to reinforcement cage installation.
Concreting is done by tremie method for wet-processed bored piles.
TOE GROUTING METHODS IN BANGKOK
Two grouting methods known as tube-â-manchette and drill-and-grout

method used in some major projects where the authors were involved are presented
in this paper. In Bangkok, both toe-grouting methods are mainly applied for the piles
with the tip embedded in the sand layer though there were few cases where it was
applied in clay layer.
188
Tube-â-manchette Toe-grouting Method (TAM Method)
This grouting method has been widely used in other parts of the world
according to the available published papers. A review of past and present toe-
grouting methods used throughout the world has been reported by Mullins et. al.
(2000). In Bangkok, the first toe-grouting for large diameter bored piles diameter
2.0m seated at depth 35m in the first sand layer was carried out by tube-a-manchette
method in 1985 for the construction of Rama IX cable stayed bridge (Morrison,
1987). Since then, tube-a-manchette method has been commonly used for toe-
grouting of wet-processed bored piles in Bangkok. It is estimated that over 5,000
bored piles have been constructed using this toe-grouting method in Bangkok.
Grout pipes configuration for tube-â-manchette method. In normal practice of toe-

grouting by this method in Bangkok, a system of grouting circuit consists of two U-
shape loops formed by two pairs of PE pipes with manchette placed at 5 to 10 cm
above pile tip. The manchette is formed by the perforated steel pipe wrapped with
rubber sleeves as shown in Figure 2. The manchette allows the grouting material to
flow out of the pipe from the perforations but prevents return of grouting material
when the grout injection is stopped. PE tubes are fixed along the rebar cages to the
top. In some projects where the specification required cross-hole sonic logging tests,
same access tubes for the sonic test apparatus are used in lieu of PE pipes.
Vertical Steel Tubes or

Pile Head Flexible PE Pipes Vertical Steel Tubes or
Flexible PE Pipes
2 No. Horizontal
Bored Pile Steel Tubes with
Wall perforation
Rubber Sleeve
(Manchette)
Reinforcement
Cage
Perimeter of Bored
Pile (Bored Pile
Wall)
Pile Base
Rubber
Horizontal
Sleeve
Steel Tube
Figure 2. Position and layout configuration of grouting circuit in

Tube-â-manchette method
189
Grout Mixture Composition and Mixing Process. The typical cement grout consists
of cement and water with the ratio of 0.5-0.6 by weight. Bentonite or other reagent
can also be added but trial mix should be made to find out the characteristics of the
grout. Cement and water are added at right proportion in a high turbulence mixer and
mixed about 2 to 4 minutes. Cement mixture is then conveyed to the agitator. The
grout must be agitated continuously or at regular interval during the period between
the time of grout mixing and the time it enters the pump. Usually, first grout injection
should be started within 10 to 15 minutes after mixing the first batch to minimize the
stiffening of the cement grout during grouting operation. It is essential that grouting
team understands the required properties such as viscosity, density, setting time of
grout are met to effectively grout the toe of the pile. Grouting operation should be
completed within the initial setting time of the grout. Mixing rate, storage capacity
and grouting rate should be well planned to complete the grouting work within the
time limit.
Grouting procedure. The process involved in grouting by tube-a-manchettes method

is illustrated in Figure 3. First, the grout pipes are flushed with water to ensure that
the grouting loops are clear from any blockage. The pile toe is then cracked with
relatively high pressure water flow, 12 to 24 hours after concreting, to open the
manchettes and to make way for forth coming grout. The commencement of cement
grouting depending on the preset pressure criteria. If specified maximum pressure is
low (less than 20 bar / 2000 kN/m2), grouting is commenced immediately after the
pile toe cracking. Grouting is normally commenced 7 to 10 days after concreting if
higher maximum control pressure is specified to avoid damage to pile by high
pressure grouting. Grouting is stopped when the specified volume or pressure is
achieved. In current practice, pressures ranging from 20 to 60 bar (2000 to 6000
kN/m2) and grout volumes 500 to 1000 liters are normally applied.
WATER GROUT UNDER

IN OUT INJECTION PRESSURE
(1) (2) (3) (4)

Flushing grout Cracking with Pressurized Grouted pile base
tubes with water water grouting
Figure 3. Grouting process in Tube-â-manchette method

190
Drill-and-grout Method (DAG Method)
Toe-grouting by this method is in fact not frequently used in Bangkok. It is to

the author’s knowledge, not commonly used in the current practice of toe-grouting of
bored piles world wide. To date, less than 500 bored piles have been completed in
Bangkok by this method.
Grout pipes configuration for drill-and-grout method. Depending on the size of

bored piles, three to four black steel pipes used for sonic test access tubes are mostly
utilized as grouting pipes in this method. Steel pipes are equally spaced and fixed
around inside perimeter of reinforcement cage as shown in Figure 4. The bottom of
steel pipes are securely closed with steel cap to prevent intrusion of foreign material
during installation and concreting.
Vertical Steel
Tubes 4 No. Vertical Steel
Pile Head Tubes closed at
bottom-end
Pile
Wall
Reinforcement
Cage Perimeter of Bored Pile
(Bored Pile Wall)
Pile Base Pile Toe Coring
(After Concreting
and Sonic Testing)
Figure 4. Position and layout configuration of grouting circuit in

Drill-and-grout method
Grout Mixture Composition and Mixing Process. Cement to water ratio of 0.6-0.7
by weight is normally used in drill-and-grout method. Cement and water are filled at
right proportion and mixed about 2 to 4 minutes in a high turbulence mixer.
Bentonite is then added at 0.5 % of cement in weight and mixed for another 2
minutes and the readily mixed grout is conveyed to the storage tank equipped with
agitator. The control factors such as viscosity, density and setting time are also
equally important as mentioned in the tube-â-manchette method.
Grouting procedure. The procedure for drill-and-grout method presented in this

paper is for diameter 1.50m bored pile with 4 sonic access tubes. The sonic logging
test is performed minimum 7 days after casting of the pile. Coring is then carried out
through the pile base to the depth 20 cm below the pile toe. First, drilling rod
attached with core-barrel is lowered into the water-filled sonic access tubes (steel
pipes) and drilled through the bottom steel cap and the concrete at pile base. Drill rod
is then withdrawn to change core-barrel with the soil sampler and lower again to drill
the soil beneath the pile base about 20 cm. It is important that sonic tube is
completely filled with water during the drilling process to avoid excessive blow-in of
soil into the tubes. After completion of toe coring and sampling, the grout pipes are
191
flushed in a sequence to ensure that flow path exists between the tubes. To do this,
after initial cleaning, two pipes (inlet pipes) are connected to the pressure pump and
low pressure up to 10 bar (10 kN/2) is applied using clean water. The other two pipes
(outlet pipes) are filled with water and are left open. As the pressure from the pump
is gradually increased, water will start to flow from one or more of the other two
open pipes. Once the flow is observed from the first outlet pipe, it is closed off and
the water flushing continued until water flow is observed from the second pipe. The
process may be repeated in like sequence from each of the grout pipes until
satisfactory flow is observed from each pipe. Once the flushing of the preinstalled
grout pipes are completed, water-supply hoses are disconnected from 2 inlet grout
pipes and connected with the grout-supply hoses. Two outlet pipes are then opened
and cement grout is filled from two inlet pipes with low pressure until the grout
discharge from other 2 outlet pipes.
GROUT GROUT UNDER
IN OUT FILLING PRESSURE
PILE
TOE CORING
(1) (2) (3) (4)

Pile toe coring Flushing grout Filling cement Pressurized
pipes with water grout in pipes grouting
Figure 5. Grouting process in Drill-and-grout method
Initial discharged cement grout is normally contaminated with the in-situ

sand so that it is required to check the purity of outcoming grout. Once clean cement
grout is flowed out, 2 inlet pipes are immediately connected with the grout-supply
hoses and pressurized grouting commences. Grouting is started with the grout
injection rate (IR) between 25 to 30 liter per minute. Grout pressure development is
closely monitored from the pressure gauges installed at the top of each pipe. If the
grouting pressure does not reach to 20 bar (2000 kN/m2) after injecting the 300 to
500 liter of grout, the injection rate is reduced to 12-15 liter per minute. Otherwise
grouting is continued until the control criteria is achieved. Figure 5 illustrates the
working procedure of pile toe grouting by drill-and-grout method used in Bangkok
subsoil for 3 major projects where the authors involved.
Grout Spreading Mechanism
Tube-â-manchette method. Teparaksa et al. (1999) investigated the penetration of

grout into the sand layer from the toe-grouting of bored piles in Bangkok by tube-â-
manchette method. Toe-grouting was performed 24 hours after concreting with
192
maximum pressure 40 bar (4000 kN/m2) for grouting the base of large diameter
bored piles constructed by wet-processed method under bentonite slurry seated in
sand layer. From the series of SPT and soil sampling performed below the pile toe
and immediate vicinity of the shaft, the authors concluded that grout did not
permeate into the sand layer at pile base. Upward migration of grout along the pile
shaft is most likely in tube-â-manchette method, as grout injection was applied under
relatively high pressure within 24 hours after concreting of the piles, plane of
weakness was formed at pile-soil interface which is most likely to be contributed by
the following conditions.
• Stress relaxation from the drilling process was not yet fully restored along the
shaft as time gap between pile drilling and grouting was relatively short
• Presence of bentonite filter-cake along the shaft
Drill-and-grout method. Grout spreading mechanism of drill-and-grout method are

discussed in relation to the steps involved in grouting operation (illustrated in Figure
5) as summarized below.
• Step 1 : drilling 4 boreholes of diameter 48 mm up to 20 cm below pile toe
(using water-filled sonic tubes as access pipes to the bottom of the pile)
• Step 2 : water flushing the sonic tubes to remove the sand inside the tubes and to
form flow-paths between the tubes
• Step 3 : filling the sonic tubes with the grout
• Step 4 : pressurized grouting until pressure reaches 6,000 kN/m2 and maintain
for 5 minutes
Step 1 creates 4 boreholes in the soil underneath the pile base. These
boreholes become larger as pocket of sand between them is washed out during Step 2
and created a cavity or relatively large opening beneath the pile base prior to the
commencement of Step 3. This cavity is filled by grout in Step 3 which eventually
formed a mass grout bulb underneath the pile base as illustrated in Figure 5.
Pressurized grout injection in Step 4 thus compacted the grouted mass which in turn
densified the grout-soil interface. Basically grout did not enter soil pores but
remained in a compacted homogeneous mass. Upward migration of grout along the
pile shaft is considered unlikely since toe-grouting was carried out minimum 14 days
after pile concreting so that time gap between concreting and grouting would have
allowed some positive improvement of pile shaft condition (influence of stress
relaxation and bentonite filter cake).
PERFORMANCE OF TOE-GROUTED BORED PILES
Typical Load Transfer Mechanism of Toe-Grouted Bored Piles in Bangkok
Table 1 shows the measured mobilized shaft resistance of bored piles with
different toe grouting method at specific load in comparison with design working
load and calculated ultimate values using empirical formula. The details of the
empirical formula commonly used in Bangkok subsoil has been presented by
193
Thasnanipan et. al. (2002). As can be seen in the table under design working load,
shaft friction carried the large portion of the load (over 90% of applied load). Only
the small percentage of the total load was transferred to the pile tip. Therefore, even
with the toe-grouting application, the end bearing component of bored piles is still
under-utilized and thus remaining unmobilized end bearing capacity serves only as a
reserve that forms the additional factor of safety.
Table 1. Summary of mobilized shaft resistance at specific load in comparison with

design working load and calculated ultimate value
Design Total shaft resistance (KN) Ratio of
Test shaft
Pile Dia. Grouting Working Calculated Measured Measured Q fm /
Pile resistance
xL method Load, DL Ultimate at DL, at 2 x DL, DL
No.
(m) (kN) (Qfm2 / Qfu)
Value, Qfu Qfm Qfm2
TT-1 1.2 x 44.6 5000 6900 4480 8560 0.90 1.24
TT-2 1.2 x 53.8 TAM 6500 9500 6120 10830 0.94 1.14
TT-3 1.2 x 46.0 5000 7900 4730 9250 0.95 1.17
TD-1 1.50 x 50.0 13000 14600 12200 22160 0.94 1.52
TD-2 1.50 x 56.5 DAG 13000 16500 12590 24270 0.97 1.47
TD-3 1.50 x 52.0 13000 16400 12570 24410 0.97 1.49
Mobilized Unit End Bearing
Figure 6 shows the mobilized unit end bearing vs. pile head movement of
TAM and DAG grouted bored piles. In general, as can be seen in the figure,
mobilized unit end bearing at specific pile head movement of piles constructed with
tube-â-manchette grouting method (TAM) is higher than those of piles constructed
with drill-and-grout method (DAG). It is also evident from Figure 6 that large pile
head movement is required to mobilize the end bearing despite toe-grouting was
applied for all piles. It should however to be noted that DAG grouted bored piles of
Figure 6 (TD1, TD2 and TD3) were not tested to failure.
3000
Unit end bearing (kN/m2)
2500
TT1 (TAM)
2000
TT2 (TAM)
1500 TT3 (TAM)
1000 TD1 (DAG)

TD2 (DAG)
500
TD3 (DAG)
0
0 5 10 15 20 25
Pile head movement (mm)
Figure 6. Mobilized unit end bearing of TAM and DAG grouted bored piles
194
Though it is not advisable to simply draw the conclusion on particular toe-
grouting method effects on performance of piles solely by mobilized unit end bearing
criteria from the available limited test results, the data provided in this research study
however could be taken as preliminary guidance to assess the effectiveness of the
different toe-grouting methods.
According to the recent full-scale instrumented static pile load tests carried
out in late 2002 and early 2003 on 13 drill-and-grout bored piles of diameter 1.50m
seated at depth over 50m (a complete set of data is not available while preparing this
paper), developed end bearing values of these piles were found to be scattered or
inconsistent despite the fact that grouting pressure of all piles were the same -
reached 60 bars (6,000 KN/m2). Therefore in authors’ opinion a careful justification
is to be made to design the deep-seated bored piles in Bangkok subsoil with high end
bearing values regardless of high pressure achievement in toe-grouting application.
Load vs. Pile Head Movement Comparison
Figure 7 shows the Load vs. Pile Head Movement of ungrouted, grouted
piles of tube-â-manchette and drill-and-grout methods. Table 2 summarized and
compared the load test results of the test piles of Figure 7.
Load (KN)
0 10000 20000 30000 40000
0
10
20
A1 - Dia. 1.5x60.0m (TAM)

30
A2 - Dia. 1.5x60.0m (Ungrout)
B1 - Dia. 1.5x63.7m (TAM)
40
B2 - Dia. 1.5x63.7m (Ungrout)
50 C1 - Dia. 1.5x56.5m (DAG)

D1 - Dia. 1.5x52.0m (DAG)
60
Figure 7. Load vs. pile head movement of ungrouted and toe-grouted piles
Following conclusions can be drawn from Figure 7 and Table 2.
• Limit loads at 20mm pile head movement are improved (14% and 21% for Project
A and B respectively) by tube-a-manchette toe grouting method (TAM) in
comparison with ungrouted piles.
• No significant difference of load-settlement behavior is found for ungrouted piles
in comparison with TAM grouted and DAG grouted piles at design load and 2
times of design load.
195
Table 2. Summarized comparison of load test results
Project A Project B Project C Project D
Pile No. A1 Pile No. A2 Pile No. B1 Pile No. B2 Pile No. C1 Pile No. D1
(TAM Grout) (Ungrout) (TAM Grout) (Ungrout) (DAG Grout) (DAG Grout)
Diameter (m) 1500 1500 1500 1500 1500 1500
Length (m) 60.00 60.50 63.70 63.70 56.50 52.00
Qd (KN) 10,000 10,000 10,000 10,000 13,000 13,000
QL
(at 20mm total 26,500 23,200 34,280 21,750 25,500 26,000*
settlement)
% increase in QL 14.20 21.57 N.A N.A
Total pile at Qd 3.50 3.00 4.20 3.75 6.10 5.80
head 1.5 Qd 6.50 6.50 7.20 6.00 11.90 11.00
movement 2 Qd 10.50 12.00 10.00 9.70 21.50 17.50
at
Note : * At 17.50mm pile head movement
Therefore, despite the various construction problems encountered and risks

involved in successful execution of grouting, drill-and-grout method does not offer
particular advantages over tube-â-manchette method.
EFFECTIVENESS OF THE DIFFERENT METHOD IN PRACTICAL

APPLICATION
Problems Encountered in Practical Application of Different Method
Table 3 summarizes the most critical problems encountered in practical

application of different toe-grouting methods used in Bangkok. In summary, it is
obvious that higher risks are involved in the drill-and-grout method. The most critical
risk of this method for the long piles is to maintain a good verticality of the grout
access tubes.
Table 3. Summaries of the problems encountered in practical application of different

methods
Problems Encountered and Risks Involved to Complete
Toe-grouting
Toe Grouting Unsuccessful or Verticality of sonic / Unsuccessful
Method ineffective grout tube coring of pile toe
cracking of
manchette
Tube-â- Possible but Not required Not required
manchette rarely occur
Drill-and-grout Not required to High verticality is High risk. If coring
crack manchette extremely important. If is unsuccessful, toe
access tube is inclined grouting can’t be
drilling rod can not be performed.
lowered to the bottom of
the pipe for toe coring
196
Impact Caused by Different Toe-grouting Method on Bored Pile Construction
The procedure involved in preparation and execution of two methods of toe-

grouting poses different impact on bored pile construction. Since verticality is the
main governing criteria in the successful application of grouting by drill-and-grout
method as explained in above section the major impact caused in this method is due
to the need of significant extra time in the following activities to produce the good
verticality of the pile itself and the grout access tubes.
• Drilling
• Extensive borehole monitoring for verticality by sonic caliper method
• Grout pipe installation
Due to the excessive time consumption in the above mentioned operations in

construction of the piles to further carry out the toe-grouting by drill-and-grout
method, to produce 1 bored pile of diameter 1.50 m seated over 50m below ground,
could take 2 to 3 days excluding time consumes for grouting work. For normal bored
pile of same size and length, it normally takes less than 24 hours to produce 1 pile.
Therefore considerable additional cost would incur for applying drill-and-grout
method due to the requirement of significant extra construction time.
Assessment on the Effective Toe-grouting Method in Practical Construction
Assessment on Constructibility. Constructibility is one of the most important issues

for most construction projects. If the construction method is impractical or if it is
involved various problems and high risk in actual execution, it should not be
employed in the first place. Careful justification on assessment of the constructibility
is essential in selection of the toe-grouting method and in establishing the technical
criteria. The party who is responsible for drawing the specification such as project
consultant or designer should take the following points into consideration.
• The potential risks involve in the practical execution and consequential impact in
minimizing the risk imposed by the method itself (e.g. delay in construction, extra
cost)
• Additional cost involved for toe-grouting against saving from higher bearing
capacity achivable by toe-grouting
• Clear and practical acceptance and rejection criteria in case the problems are
encountered in actual execution
Assessment on the Performance of Pile. As presented in the earlier section, deep-

seated large diameter bored piles in Bangkok are mainly supported by shaft friction
and very small percentage of the load is transferred to pile toe under working load
and double of working load. Therefore, even with the toe-grouting application, the
end bearing component of bored piles is still under-utilized. Thus remaining
unmobilized end bearing capacity serves only as a reserve that forms the additional
factor of safety.
197
CONCLUSION
Based on the field application and measured data of instrumented pile load tests,
following conclusions can be drawn.
(1) Under design working load for large diameter deep-seated bored piles with toe
grouting in Bangkok subsoil, the large portion of load (over 90% of applied
load) is carried by the shaft friction. Only the small percentage of the total load
is transferred to the pile tip. No significant difference of load-settlement
behavior is found for non-grouted piles against TAM grouted and DAG grouted
piles at design load and 2 times of design load. Therefore, for deep-seated bored
piles in Bangkok subsoil, even with the toe grouting application, the end bearing
component of bored piles could be still under-utilized. Thus the remaining
unmobilized end bearing capacity serves only as a reserve which forms the
additional factor of safety.
(2) The drill-and-grout method has higher risks than the tube-â-manchette method in
successfully constructing the toe grouted bored pile. Considerable additional
cost would incur in applying drill-and-grout method due to the requirement of
significant extra construction time to complete a qualified pile.
(3) Despite the considerable extra cost, various construction problems encountered
and risks involved in successful execution of grouting, drill-and-grout method
does not offer particular advantage over tube-â-manchette method as far as pile
capacity improvement is concerned.
(4) The potential risks involve in the practical execution and consequential impact in
minimizing the risk imposed by the method itself should be taken into account in
selecting the effective toe-grouting method.
REFERENCES
Morrison, I.M. (1987) “Bored piled foundation for Chao Phya river crossing at Wat
Sai, Bangkok”, Proceedings of the 9th South East Asian Geotechnical Society
Conference, Bangkok, pp. 6-207 to 6-218.
Mullins A. G., Dapp S. D. and Lai P. (2000) “Pressure-Grouting drilled shaft tips in
sand” New Technological and Design Developments in Deep Foundations,
Proceedings of the Sessions of Geo-Denver 2000, Geotechnical Special Pub. No.
100, ASCE, August 5-8, 2000, Denver, Colorado, USA, p.p. 1-17
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999). “Base grouting of wet
process bored piles in Bangkok subsoils” 11th ARC. on soil mechanics and
geotechnical Engineering., Seoul, Korea, pp. 269-272.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa
(2002) “Performance of wet-process bored piles constructed with polymer-based
slurry in Bangkok subsoil”, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume One, ASCE,
February 14-16 2002, Orlando, Florida, USA, p.p. 143-157
198
ASCE Geo-Institute Geotechnical Special Publication in Honor of
Professor Gorge Gobble
27-31 July 2004, Los Angeles, California, USA
GeoTrans 2004
Current Practice and Future Trends of Deep

Foundations in Bangkok, Thailand
Narong Thasnanipan, Zaw Zaw Aye Kamol Singtogaw and

Sumate Pravesvararat
199
CURRENT PRACTICE AND FUTURE TRENDS
OF DEEP FOUNDATIONS IN BANGKOK, THAILAND
Narong Thasnanipan, Zaw Zaw Aye, Kamol Singtogaw & Sumate Pravesvararat
SEAFCO Co., Ltd. , 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah
Bangkok 10510, Tel. 02-919 0090, Fax. 02-9190098
E-mail : seafco@seafco.co.th, Homepage : www.seafco.co.th
ABSTRACT
This paper presents the authors’ experiences in development of deep foundations,

bored piles and barrettes in particular in Thailand. Some literatures related to wet-processed
bored piles and barrettes in Thailand published through out the past 30 years are also
summarized together with recent research works.
KEYWORDS : bored pile, barrette, deep foundations, bentonite, polymer-based slurry
INTRODUCTION
Cast-in-place deep foundations, particularly wet-processed deep-seated large diameter

bored piles and barrettes have been used for foundations of various infrastructure projects in
Thailand for over two decades. The versatility of the construction method and the high-load
capacity which in turn offered the constructibility and cost-saving are the main factors
contributed to the increasing use of deep-seated large diameter bored piles and barrettes. In
the initial stage there were a number of questions on design and construction aspects of these
foundation systems particularly in Bangkok subsoil. With the passage of time, the
construction equipment, installation techniques and testing of deep foundation elements have
been developed. Large numbers of instrumented full-scale static pile load tests were
conducted through out 1990’s which provided better understanding on behavior of these deep-
seated foundations. Researches focused on the design parameters and methods were produced
based on these test results. The design parameters were so well established that wet-process
cast-in-place foundations became regarded as reliable foundations for not only in the property
sectors but also for practicing engineers involved in the infrastructure projects. This paper
presents the development of wet-processed bored piles and barrettes in Thailand. The
information contained in the paper are mainly from the infrastructure developments built in
Bangkok and adjacent areas.
CONSTRUCTION METHOD
Large diameter deep-seated wet-processed bored piles can be constructed by 2

methods, reverse-circulation and rotary drilling method. As construction procedure of historic
reverse-circulation method has been well documented in many literatures and it is less
frequently utilized nowadays in Thailand, only a rotary-drilling method is briefly presented in
In rotary drilling method, a temporary casing of appropriate length (12 to 18m in
Bangkok depending on the thickness of soft clay) with required diameter (internal diameter
not less than that of design bored pile diameter) is first installed to ensure the stability of the
borehole in the top soft or loose soil layers. In some projects where the vibration is strictly
limited, standard length of casing with oscillator or short temporary casing of 5 to 8 m is
201
pushed down in combination with pre-boring process. Drilling is commenced by auger to drill
out the soil within the temporary casing and up to the top of the first water-bearing sand layer
or bottom of the casing in case of using short casing method. Drilling slurry or supporting
fluid is then supplied to the borehole and drilling is continued with a bucket down to the
design final depth of the pile. Before lowering the reinforcement cage, a special cleaning
bucket is used to clean the base of borehole. If bentonite slurry is used, recycling method by
air-lift or pump is applied as base cleaning process. Reinforcement cages are then lowered
into the borehole and concreting is carried out by tremie method.
A mechanical or hydraulic cable-suspended grab is commonly used for barrette
construction. Excavation of the trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the clamshell operated by cables.
Different from bored pile construction, a guide wall of depth 1 to 1.5 m with inside clear
dimensions slightly larger than the nominal size of the barrette is used to guide the grab
during initial bites. Bentonite slurry is introduced to the trench as soon as initial excavation
commenced. The excavation is continued under the bentonite slurry to the final depth. After
recycling of slurry and lowering the rebar cage, concreting is done by tremie method.
OVERVIEW OF APPLICATION
Bored pile
The first wet-process large diameter bored pile was constructed for Pinklao Bridge in
Bangkok 30 years ago. Using reverse circulation method, diameter 1.50m bored pile was
installed up to 45m in the second sand layer. Three major bridges were constructed by bored
pile of reverse circulation method in Bangkok from early 1970 to 1980. The summarized
information of these bridges are tabulated below.
Table 1 First 3 major bridges constructed by large diameter wet-processed

bored piles in Bangkok
Project Name Year of Construction Method Diameter Depth

Construction (m) (m)
Pinklao Bridge 1971 Reverse circulation 1.50 45
Sathorn Bridge 1979 Reverse circulation 1.50 46
New Memorial 1982 Reverse circulation 1.50 49
Bridge
The first wet-processed bored pile of rotary-drilling method down to first sand layer
was constructed in Bangkok in late 1970 for high-rise building project, Royal Orchid Hotel.
Since then bored piles constructed by rotary-drilling method have been extensively used for
foundations of various heavy structures such as high-rise buildings, elevated expressways,
overpass-bridges, underground car park buildings, waste-water treatment plants and most
recently underground train stations of Bangkok first subway project. By mid 1980, bored pile
became the foundation of choice for heavy structures particularly in urban area of Bangkok.
The versatility of the construction method and the high-load capacity which in turn offered
the constructibility and cost-saving are the main factors contributed to the increasing use of
deep-seated large diameter bored piles and barrettes. Most of the early wet-process bored
piles (1980’s) in Bangkok were constructed up to 50 m depth. Diameter of bored piles
commonly constructed in early days were ranging from 0.6 to 1.5 m. Table 2 below
summarizes the information of early major high-rise building projects in Bangkok.
202
Table 2 First high-rise building projects constructed by large diameter wet-processed bored
piles in Thailand (from 1979 to 1983)
Project Name Construction Method Diameter (m) Depth (m)

Royal Orchid Hotel Rotary-drilling with 0.80 - 1.00 33.0
Taiping Tower Rotary-drilling with 0.80 - 1.50 32.0
River City Hotel Rotary-drilling with 0.80 – 1.00 27.5
Asoke Tower Rotary-drilling with 0.80 - 1.50 50.0
Time Square Rotary-drilling with 0.80 - 1.50 50.0
Building (1983) auger and bucket
Barrette
applicable construction method and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimension ranging from
0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000 kN
have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in late 1970 for the
foundation of Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x 2.5m
and toe depth of 33m in late 1970. Barrettes can be constructed with flexible layout plan for
both vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous
row or column, radial, alternating long and short axis of barrette and a combination of two or
more of such patterns. One type of equipment can be used for constructing both barrettes and
diaphragm walls in particular projects thus reduces the mobilization cost for additional
equipment. In addition to the large bearing capacity requirements, on site difficulties such as
limited head room where piling rigs cannot be utilized, under such situations like presence of
overhead high voltage power cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the barrettes. A summary of
observation on barrettes of completed 26 projects completed with respect to selection criteria
is presented in Table 3.
Table 3 Summary of Barrette Selection
Criterion No. of Project Remarks

High Load Capacity 4 Incorporated with bored piles
Minimize Construction 6 Alternative for bored piles
Equipment
Limited Head Room for 3 Under existing structures such as bridges,
Excavation elevated expressway and power lines
Diaphragm Wall
Foundation as well as 3 Provision for the future requirement
portion of column
Static load test up to 5290 ton conducted on barrette set the record as the highest load
ever tested for a single cast-in-situ deep foundation in Thailand (Thasnanipan et. al. 2002a)
203
PROBLEMS OF WET-PROCESSED BORED PILES AND BARRETTES
IN EARLY DAYS
Basic but extensive problems were experienced in early stages of bored pile construction in
• Lack of skills in operation of equipment (particularly drilling rig)
• Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
• Limited knowledge and less advance technique in control of bentonite slurry
• Limited experience in construction method and related negative impact
• Quality of concrete for tremie concreting method
• Improper construction and quality control specification and guidelines for deep-seated
piles in local soil condition
• Improper design for constructibility
DEVELOPMENT IN CONSTRUCTION
Over the past 3 decades, along with the development of wet-processed piling
technology world-wide, equipment, construction method, design as well as better
understanding and realization of construction impact on the performance have been
significantly improved in Thailand. Table 4 summarizes the areas of improvement in bored
pile and barrette construction and main factors contributed to these.
Research works presenting some of these developments were published both locally
and internationally. The design, construction and behavior of bored cast in-situ concrete piles
in Bangkok Subsoil was presented by Thasnanipan et. al., (1998a). The construction and
performance of barrettes in Bangkok Subsoil was also reported by Thasnanipan et. al.,
(1998b). Effect of construction time and bentonite viscosity on shaft friction of bored piles
was also pointed out by Thasnanipan et al., (1998c). Teparaksa et. al., (1999) published the
base grouting of wet process bored piles in Bangkok subsoil which demonstrated the method
of base grouting and reported the grout spreading mechanism for long bored piles and
performance of base grouted bored piles.
(a) (b)
Fig. 1 (a) Guide wall for cruciform barrette (b) View of cruciform barrette after installation of
reinforcement
voltage power transmission line was described in the work of Thasnanipan et. al., 2000a.
204
Above-mentioned research works reflect the development history of bored pile and barrettes
in Thailand and offered useful information to the local construction industry and perhaps to
the international deep foundation engineering society.
Table 4 Summary of development in bored pile and barrette construction

construction required to complete complete diameter operating skills as
diameter 1.5m tip 50m 1.5m tip 50m bored well as improved-
bored pile pile knowledge in
management and
Pile size and Bored pile Bored Pile Better equipment,
depth Maximum diameter 2.0m Maximum Diameter operating skills as
and common depth 25- 2.0m and common well as improved-
50m for bored pile. depth 25-60m knowledge in
Barrette Barrette construction method,
Limited in size and depth Various sizes & management and
depth over 60m local soil condition
availability and
advance technology
Application of Not available Extensively use Material availability
polymer-based and research works
slurry (for bored
pile only)
and performance works
process works
Quality control Less methods available Better equipment Advance equipment,
interpretation works
understanding on construction impact on quality and performance of these wet-processed
deep foundations. Marked difference between outcome quality of bored piles constructed in
1980s and 1990s can be observed by significant less defects found in the latter. It can also be
observed from load test results that bored piles constructed in late 1990 have higher capacity
than those of 1980s as shown in Table 5.
Table 5 Load test results of bored piles constructed in 1970 to 1980 and 1990s
205
Year of Project Name Pile Dimension Design Test Total settlement
construction (Dia. & Depth) Load Load at max. test load
(ton) (ton) (mm)
1971 Pinklao bridge 1.5m x 45m 410 820 5
1995 Central Plaza 1.2m x 45m 700 1400 27
Pinklao
1980 Taiping tower 1.0m x 32m 300 1125 118
1990 High-rise building 1.0m x 32m 390 975 23
at Ekamai Rd.
Fig. 2 Progress of piling work for the Rama VIII main bridge
infrastructure projects including Second Stage Expressway (1991), Don Muang Tollway
Extension (1997), Bangna-Bang Pli-Bangprakong Expressaway (1998), and Wat Nakorn-In
project (2001). Over 700 bored piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed (1999-2000) by polymer-based slurry for the foundation of
Rama VIII Bridge, one of the initiatives of his majesty the King Bhumibol Adulyadej.
Progress of piling works at the bank of Chao Phraya river for the main bridge of Rama VIII
Bridge is depicted in Figure 2.
The availability of more reliable and powerful equipment and tools for drilling makes
possible to construct the rock-socket bored piles. Figure 3 shows the drilling rig equipped
with rock-auger used for construction of highway bridge across Mool River in 2001, in the
north-eastern part of Thailand, where drilling was carried out through weathered-sandstone by
powerful rotary drilling rig with core barrel and rock-auger.
Fig. 3 Rock-socket bored pile construction for Mool River Project - Ubon Ratchathani Province
DEVELOPMENT IN DESIGN CONCEPT AND PARAMETERS
206
parameters were mainly based on the available literatures from the research works carried out
in other parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966),
Bowles (1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the
first research data available for design parameters of bored pile in Bangkok soft clay. The
author conducted the research based on 6 dry-processed small diameter bored piles with load
cells installed at the pile toe in AIT campus in Phathumthani. Though the piles were simply
constructed by dry-processed with casing method, this early-stage research study provided
some important information such as load-settlement behavior, adhesion factor (α) and bearing
capacity factor (N) of bored pile in Bangkok soft clay.
With the passage of time, design method and parameters for local subsoil were
improved as a result of research works carried out in 1980s. Differences between the behavior
of driven piles and bored piles were well realized from these researches. (Ng, 1983) presented
the load distribution characteristics of wet-processed bored piles founded in both first and
second sand layers of Bangkok based on the instrumented (strain gauges) pile load test
results. (Chiewcharnsilp, 1988) reported the shaft friction factor β values of sand layers in
Bangkok based on the instrumented load test results. In addition to the researches previously
1988) Suchada (1989) determined the shaft friction factor (adhesion factor, α) of Bangkok
clay layers based on 11 bored piles of diameter ranging from 0.50 to 0.80m with the
shaft friction loads from plain static pile load test results with numbers of assumption since
load tests on bored piles were conducted throughout 1990’s which provided better
understanding on behavior of these deep-seated foundations. Researches focused on the
design parameters and methods were produced based on these test results. The design
parameters were more accurate and well established that wet-process cast-in-place
foundations became regarded as reliable foundations in construction industry of Thailand.
With more confidence on soil parameters and behavior of these deep foundations, the
designers designed to carry significantly higher loads for bored piles and barrettes constructed
in late 1990 than those of 1980s. (Thasnanipan et. al., 1999) reported the failure mechanism
of long bored piles in layered soil of Bangkok. The authors cited that for the bored piles
embedded in the multi layered soils of Bangkok, estimation of ultimate shaft friction capacity
needs to consider the brittle type of failure mechanism of stiff to hard clay layers and the
α values selected need to be adjusted accordingly. Peak and residual α values mobilized in
the stiff clay layers analyzed by the authors were plotted as shown in Figure 4 along with the
suggested curves by different researchers. It can be seen from the figure that the residual
α value of 2nd stiff clay layer (undrained shear strength values of 25 ton/m2) at the maximum
test load drops below the curve suggested by (Suchada, 1989). So the α values proposed in
Figure 4 for the stiff to hard clay layers gives overestimate of the ultimate shaft friction under
theses conditions.
Introduction of polymer-based slurry for wet-process bored piles marked a major
breakthrough for both construction and design engineers. Thasnanipan et al.,(2002b), reported
that bored piles constructed with polymer-based slurry have higher capacity than those
constructed with bentonite slurry. Figure 5 shows the shaft friction factors β of sand layers
for polymer-based bored piles in comparison with the design line of bentonite bored piles.
207
1.20
1.00
Tomlinson,1957
0.80
α
Adhesion Factor ,
Suchada,1989
0.40
Peak
0.20
Residual
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
2
Fig. 4 Comparison of adhesion factor α , suggested by different researchers with the actual
1 .0
0 .8
β (Ks tan δ)

( 1999 )
0 .4
0 .2
0 .0
28.0 30.0 32 .0 34.0 36 .0 38.0
Fig. 5 Back-calculated β values of polymer bored piles at maximum test load plotted on
design line of bentonite bored piles constructed in Bangkok subsoil
.
DEVELOPMENT IN QUALITY CONTROL TESTING
Improvement of testing equipment and powerful computer facilities are the key factors
contributed to the development in quality control testing. Interpretation skills relevant to local
soil condition and construction method of these tests were significantly improved in local
industry. For instance, in early 1980, interpretation of the sonic integrity (seismic test) test
results were needed to send to the specialists abroad which made the testing cost more
expensive. Significant cost-saving were achieved in some major projects of late 1990, as more
practical and precise interpretation were made to verify and establish the acceptance criteria in
proving the quality of suspected piles with anomalies.
208
However, the reliability of these methods was doubtful. Koden drilling monitor system was
believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can be
monitored rapidly from the electronic plot of continuous profile by Koden equipment. This
system is very useful for the verticality-control of long barrettes - unlike bored piles, guiding
by temporary casing is not available for barrette drilling.

is that since no particular preparation is necessary during the pile construction phase it is more
flexible to select which pile is to be tested. However, interpretation of sonic integrity testing
needs considerable experience and knowledge in testing, subsoil condition and construction
method. In many projects it is a part of the contractual requirement to conduct sonic integrity
test. Minimum 10 % to maximum 100 % of production piles are commonly tested. It is also
a reasonably acceptable method for applying as a retrospective investigation in determining
integrity of the pile. The signal characteristics and their interpretations of sonic integrity test
on piles founded in Bangkok subsoil were reported in details by Thasnasipan et al. (1998d).

Sonic logging test is relatively expensive. It is mainly employed as a pre-planed site
quality control testing. The major advantage of this method is that test can be carried out
shortly after the pile construction. Hence, rectification measures can be implemented if the
pile is detected with defect while the foundation contractor is on site. However, this method is
generally not applicable if pile integrity is in question due to post-construction activities as
access tubes are usually grouted after completion of the test. The first sonic logging test was
believed to be conducted in 1982 for the wet-processed bored piles of Memorial Bridge
Project where bad concrete zones were detected at depth about 20m and 1m (Ng, 1983). Use
of Sonic logging test for checking pile integrity has increased in Thailand in recent years.
The results from sonic logging test conducted on model piles in Bangkok helped to extend the
knowledge of the signal characteristics and interpretation (Thasnanipan et. al., 2000b).

both driven and bored piles.. A large number of related technical papers and case histories of
2000b, reported the application of dynamic load testing in Thailand.
CONCLUDING REMARKS
Development of bored pile and barrettes in Thailand has been presented. According to
the authors’ experience as a contractor, development in both construction and design aspects
of these deep foundations in past decades are significant. In authors’ opinion, however, there
are many works to be done to improve further with particular focus on constructibility issues,
concrete technology for wet-processed bored piles and barrette, reliable but cost-effective
209
quality control testing and value-engineering. Starting from planning stage, site investigation,
design, construction and inspection should be integrated such that designers, contractors and
construction inspectors participating as a team with common goal. Appropriate and practical
specification should be established jointly by these parties for local soil condition and
construction method. Practical acceptance criteria should be developed to verify bored piles or
barrettes with suspected anomalies. Continuing education should be promoted for both
designers, inspection engineers and contractors. The Geotechnical Chapter of the Engineering
Institute of Thailand under the royal patronage of his majesty the King Bhumibol Adulyadej,
has started to establish the standard code of practice and guidelines for wet-processed bored
piles which will serve as a yardstick for the deep foundation industry upon completion in near
future.
REFERENCES

Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles – a summary, Soils-Soils, No.
18-19, pp. 21-32
Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored piles (in Thai), M.
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213, Asian Institute of
Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil mechanic and
foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second sand layer, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay, Proceedings of 10th
conference on soil mechanic and foundation engineering, Vol. 2, pp. 815-818
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian Institute of Technology,
Bangkok, Thailand
Teparaksa W., Thasnanipan N., and Anwar M. A. (1999), Base grouting of wet process bored piles in Bangkok
subsoils, 11th Asian Regional Confer. on soil mechanics and geotech. Engg., Seoul, Korea, pp. 269-272.
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and behavior of bored
cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louise, Missouri, U.S.A., p.p.281-287
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils, construction and
performance, Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and Bentonite Viscosity on
Shaft Capacity of Bored Piles, Third International Geotechnical Seminar on Deep Foundation on
Bored and Auger Piles, Ghent, Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded in Bangkok
Subsoil-Signal Characteristics and Their Interpretations, Proceedings: Fourth International Conference
on Case Histories in Geotechnical Engineering, St. Louis, Missouri, U.S.A., p.p.1086-1092
Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles in Layered Soils
of Bangkok, Civil and Environmental Engineering Conference, Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference on Geotechnical & Geological Engineering,
Melbourne, Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive Integrity Testing on
Piles Founded in Bangkok Subsoil, 6th International Conference on Application of Stress-Wave Theory
to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok, Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No. 116, Volume Two, ASCE, Febuary 14-16 2002,
Orlando, Florida, USA, p.p. 1073-1087
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b) Performance of
Wet-Process Bored Piles Constructed with Polymer-Based Slurry in Bangkok Subsoil, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication No. 116, Volume
One, ASCE, February 14-16 2002, Orlando, Florida, USA, p.p. 143-157
210
Proceedings of the 16th Southeast Asian Geotechnical Conference
8-11 May 2007, Kuala Lumpur, Malaysia
16th Southeast Asian

Geotechnical Conference
Edited by:
Ooi Teik Aun, Yee Yee Weng and Ting Wen Hui
Development and Achievements of Deep-seated

Bored Piles and Barrettes Construction in
Thailand for the Past Forty Years, A Country
Report
Narong Thasnanipan, Zaw Zaw Aye, Chanchai

Submaneewong and Thayanan Boonyarak
SEAFCO Public Co., Ltd. Bangkok, Thailand
211
Development and Achievements of Deep-seated Bored Piles and Barrettes
Construction in Thailand for the Past Forty Years, A Country Report
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong and Thayanan Boonyarak
Seafco Public Company Limited, Bangkok, Thailand
zaw@seafco.co.th
Abstract: This paper presents the development and achievements of deep-seated bored piles and barrettes construction in Thailand
based on the authors’ experiences, in the past four decades. Literature related to bored piles and barrettes in Thailand published
through out the past forty years is summarized together with recent research works. Problems and difficulties in construction of wet-
processed bored piles and barrettes in early days are reported. The areas of improvement, advancement and achievement in bored pile
and barrette construction and main factors contributed to these developments are also presented. An attempt is also made to briefly
discuss the future trend of bored pile and barrette foundations.
1 INTRODUCTION
During the past four decades, owing to the acceleration of Geotechnical Pore Pressure (kN/m 2)
Parameters
development in mega infrastructure and numerous tall 0 Weathered Crust
0
0 200 400 600
building projects, techniques and practice of cast-in-place Bangkok Soft Su= 10-25 KPa
 t = 14-16kN/m 3
deep foundations, particularly wet-process deep-seated large 10
Clay
-
10
diameter bored piles and barrettes have experienced enormous

Depth (m)
development and achievements in Thailand. The versatility of 20

Medium to
Stiff Clay Su= 50-14020- KPa
 t = 17-21kN/m 3
the construction method and the high-load capacity which in
Ac
tua
turn offered the constructibility and cost-saving are the main 30
Medium
-
lP
30
Hy
 t = 20kN/m 3
factors contributed to the increasing use of those cast-in-place (First Sand)
iez
dro
om
deep foundations. In the initial stage, there were a number of
sta
-
etr
40
tic
 = 21KN/m 3
t
questions on design and construction aspects of these
ic D
Lin
raw
e
foundation systems particularly in Bangkok subsoil. With the 50 Dense to Very
Dense Sand
-
50
do
SPT-N= 50-100
passage of time, the construction equipment, installation (Second Sand)  t = 21KN/m 3
wn
Lin
techniques and testing of deep foundation elements have been 60 -
60
e
developed. Large numbers of instrumented full-scale static
pile load tests were conducted through out the 1990s mainly in
Fig. 1. Typical soil profile of Bangkok with piezometric
mega projects, which provided better understanding on
drawdown condition (Thasnanipan, et al., 2002)
behavior of these deep-seated foundations.
Research focused on the design parameters and methods
were produced based on these test results. The design 3. CONSTRUCTION METHOD IN GENERAL
parameters became so well established that wet-process cast-
in-place foundations became regarded as reliable foundations Due to the prevailing subsoil and groundwater conditions,
for practitioners involved in construction industry in Thailand. deep-seated bored piles of toe depth over 24m are constructed
This paper presents the past and current practice of by wet-process or slurry displacement method. Wet-process
wet-processed bored piles and barrettes in Thailand together bored piles can be constructed by 2 methods, reverse-
with a brief discussion on future trend of these cast-in-place circulation and rotary drilling. As construction procedure of
deep foundations. The information contained in the paper is historic reverse-circulation method has been well documented
mainly from the mega projects constructed in Bangkok and in various studies and since it is less frequently utilized
adjacent areas. nowadays in Thailand, only a rotary-drilling method is briefly
presented in this article as follows.
In the rotary drilling method, a temporary casing of
2. SUBSOIL AND EXISTING PIEZOMETRIC PROFILE
appropriate length (12 to 18m in Bangkok depending on the
thickness of soft clay) with required diameter (internal
Subsoil profile and the present piezometric drawdown
diameter not less than that of design bored pile diameter) is
condition of Bangkok are presented in Fig. 1 below. A typical
first installed to ensure the stability of the borehole in the top
subsoil profile is relatively consistent in different localities in
soft or loose soil layers. In some projects where the vibration
Bangkok. It is characterized by alternating layers of clay and
is strictly restricted, a standard length of casing with oscillator
sand deposits as shown in Fig. 1.
or short temporary casing of 5 to 8 m is pushed down in
combination with the pre-boring process. Drilling is
commenced by auger to drill out the soil inside the temporary
casing. Auger-drilling is commonly continued up to the top of
213
the first water-bearing sand layer or bottom of the casing when the construction method and the high-load capacity which in
using the short casing method. Drilling slurry or supporting turn offered the constructability and cost-saving, are the main
fluid is then supplied to the borehole and drilling is proceeded factors contributing to the increasing use of deep-seated large
with a bucket down to the design final depth of the pile. diameter bored piles and barrettes. Most of the early wet-
Before lowering the reinforcement cage, a special cleaning process bored piles (1980s) in Bangkok were constructed up to
bucket is used to clean the base of borehole. If bentonite slurry 50 m depth. Sizes of bored piles constructed in early days
is used, recycling method by air-lift or pump is applied as the ranged from diameter 0.6 to 1.5 m. Summarized information
base cleaning process. Reinforcement cages are then lowered of early major high-rise building projects in Bangkok is
into the borehole and concreting is carried out by tremie presented in Table 2.
method.
A mechanical or hydraulic cable-suspended grab is
commonly used for barrette construction. Excavation of the 4.2 Barrette piles
trench is carried out by the cyclic-process of lifting and
lowering of the grab under gravity and tangential force of the In some projects where bored piles were not feasible due to
clamshell operated by cables (mechanical grab) or hydraulic the site constraints, applicable construction methods, and/or
action (hydraulic grab). Different from bored pile extensive bearing capacity requirements, the use of barrette
construction, a guide wall of depth 1 to 1.5 m with inside clear foundations would make a suitable alternative. Barrettes with
dimensions slightly larger than the nominal size of the barrette dimensions ranging from 0.80 m x 2.7 m to 1.5 m x 3.0 m for
is used to guide the grab during initial bites. Bentonite slurry safe working load capacity from 11 000 to 25 000 kN have
is introduced into the trench as soon as the initial excavation been used in some major projects.
commenced. The excavation is continued under the bentonite The first barrette in Bangkok was believed to have been
slurry to the final depth. After recycling of slurry and lowering constructed in the late 1970s for the foundation of the
the rebar cage, concreting is done by tremie method. Bangkok Bank Head Office Building at Silom road with the
size of 0.6-0.8 x 2.5m and the toe depth of 33m. Barrettes can
be constructed with flexible layout plans for both vertical and
4. OVERVIEW OF APPLICATION lateral loads. The layout pattern of barrettes can be arranged
in a continuous row or column, radial, alternating long and
short axis of barrette and a combination of two or more of
4.1 Bored piles
such patterns. To minimize the need of additional equipment,
Using tripod rigs, small diameter dry-processed bored pile a mechanical grab mounted on crawler crane can be used for
construction was started in early 1970. The first wet-process constructing both barrettes (as piled-foundation) and
large diameter bored pile was constructed for Pinklao Bridge diaphragm walls (as earth retaining structure) in some suitable
in Bangkok 36 years ago. Using reverse circulation method, a projects without employing a bored piling rig, which could
1.50m diameter bored pile was installed up to 45m in the save the extra mobilization cost. In addition to the large
second sand layer. Three major bridges were constructed by bearing capacity requirements, on site difficulties such as
bored pile using the reverse circulation method in Bangkok limited head room where piling rigs cannot be utilized, under
from early 1970 to 1980. The summarized information of such situations like presence of overhead high voltage power
these bridges is tabulated in Table 1 below. cables, existing overpasses or structures for elevated
expressways and planned subway stations, also demands the
Table 1. First 3 major bridges constructed by large diameter use of barrettes. Static load test up to 52 900 kN conducted on
wet-processed bored piles in Bangkok. barrette set the record as the highest load ever tested for a
single barrette foundation in Thailand (Thasnanipan et. al.
Project Year of Construction Diameter Depth 2002a). A summary of completed 35 projects constructed with
Name Construction Method (m) (m) barrettes with respect to selection criteria is presented in
Pinklao 1971 Reverse 1.50 45 Table 3.
Bridge circulation
Sathorn 1979 Reverse 1.50 46
Bridge circulation 5. PROBLEMS AND DIFFICULTIES IN
New 1982 Reverse 1.50 49 CONSTRUCTION OF WET-PROCESS BORED PILES
Memorial circulation AND BARRETTES IN EARLY DAYS
Bridge
Basic but extensive problems were experienced in early stages
The first wet-processed bored pile utilizing the rotary- of bored pile construction in Thailand as summarized below.
drilling method down to first sand layer was constructed in  Limited availability and capacity of equipment
Bangkok in the late 1970s for high-rise building project, the  Lack of skills in operation of equipment (particularly the
Royal Orchid Hotel located at the bank of Chao Phraya river. drilling rig)
Since then bored piles constructed by rotary-drilling method  Adverse effect due to a slow rate of drilling (e.g. excessive
have been extensively used for foundations of various heavy formation of filter-cake by using bentonite slurry)
structures such as high-rise buildings, elevated expressways,  Limited knowledge and less advanced techniques in the
overpass-bridges, underground car park buildings, waste-water control of bentonite slurry
treatment plants and most recently underground train stations  Limited experience in construction method and related
of Bangkok’s first subway project. By mid 1980s, bored piles negative impact
became the foundation of choice for heavy structures  Quality of concrete used for tremie concreting
particularly in the urban area of Bangkok. The versatility of  Lack of experienced engineers and foremen
214
 Improper construction and quality control specifications of barrettes in Bangkok Subsoil was also reported by
and guidelines for deep-seated piles in local soil Thasnanipan et al., (1998b). Effect of construction time and
 Improper design for constructability bentonite viscosity on shaft friction of bored piles was also
pointed out by Thasnanipan et al., (1998c). Thasnanipan et
Table 2. First high-rise building projects constructed with al., (2004a) also reported a comprehensive study on the
large diameter wet-processed bored piles in Thailand (from effectiveness of two different toe-grouting methods, known as
1979 to 1983). tube-â-manchette and drill-and-grout, applied in Bangkok.
Project Construction Diameter (m) Depth Table 4. Summary of development in bored pile and barrette
Name Method (m) construction.
Royal Rotary-drilling 0.80 - 1.00 33.0
Orchid with auger and Area of In the past At present Main factor
Hotel bucket development (Early contributed
(1979) / 1980’s) to
Taiping Rotary-drilling 0.80 - 1.50 32.0 improvement development
Tower with auger and Speed of Minimum 3 Less than 1 Better
(1980) bucket construction days day to equipment,
River City Rotary-drilling 0.80 – 1.00 27.5 required to complete operating
Hotel with auger and complete diameter 2.0 skills as well
(1982) bucket diameter m tip 60m as
Asoke Rotary-drilling 0.80 - 1.50 50.0 1.5m tip 50m bored pile improved-
Tower with auger and bored pile knowledge
(1983) bucket in
Time Square Rotary-drilling 0.80 - 1.50 50.0 construction
Building with auger and method,
(1983) bucket management
and local
soil
Table 3. Summary of Barrette Selection in major projects. condition
Pile size and Bored pile Bored Pile Better
Selection Criterion No. of Remarks depth Maximum Maximum equipment,
Project diameter Diameter operating
1.5m and 2.0m and skills as well
High Load Capacity 7 Incorporated with bored common common as
piles depth 25- depth 25- improved-
50m for 60m knowledge
Minimize 7 Alternative for bored bored pile. Barrette in
Construction piles Barrette Various sizes construction
Equipment Limited in & depth over method,
Limited Head Room 3 Under existing structures size and 60m management
for Excavation such as bridges, elevated depth and local
expressway and power soil
lines condition
Combination with 16 As diaphragm wall legs Base Not available Available Equipment
Diaphragm Wall grouting availability
and advance
Foundation as well 3 Provision for for the technology
as portion of futures requirement Application Not available Extensive Material
column of polymer- use availability
based slurry and research
(for bored
pile only)
6. IMPROVEMENT IN CONSTRUCTION Construction Not well Improving Experience
impact on understood from past
Over the past three decades, along with the development of quality and projects and
wet-processed piling technology in other parts of the world, performance research
equipment, construction technique and design methods as well Quality Not well Well Experience
as better understanding of construction impact on the control in established established from past
performance of this type of foundation have significantly construction and and projects and
improved in Thailand. Table 4 summarizes the areas of process systematic systematic research
improvement in bored pile and barrette construction and main Quality Fewer Better Advance
factors contributed to these developments. Some of these control test methods equipment equipment,
improvements were published both locally and internationally. method and available and available and experience
interpretation limited better from past
The design, construction and behaviour of bored cast in-situ
knowledge knowledge projects and
concrete piles in Bangkok Subsoil were presented by
in in research
Thasnanipan et al., (1998a). The construction and performance interpretation interpretation
215
Experience from past projects and extensive research works
provided better understanding of construction impact of
quality and performance of these wet-processed deep
foundations. Marked difference between outcome quality of
bored piles constructed in 1980s and 1990s can be observed
by significant fewer defects found in the latter. It can also be
observed from load test results that bored piles constructed in
late 1990 have higher capacities than those of 1980s as shown
in Table 5.
Table 5. Load test results of bored piles constructed in 1970

to 1980 and 1990s.
Year Project Pile size Design Test max

Name (Dia. & Load Load (mm)
Depth) (kN) (kN)
1971 Pinklao 1.5 m x 4100 8200 5
bridge 45 m Fig. 2b. Rama VIII cable-stayed bridge and connecting
1995 Central 1.2 m x 7000 14000 27 elevated flyovers.
Plaza 45 m
Pinklao The availability of more reliable and powerful equipment
1980 Taiping 1.0 m x 3000 11250 118 and tools for drilling makes it possible to construct the rock-
Tower 32 m socket bored piles by rotary drilling method. Fig. 3(a) and 3(b)
1990 High-rise 1.0 m x 3900 9750 23 show the drilling rig equipped with a rock auger used for
building at 32 m construction of highway bridge across Mool River in 2001, in
Ekamai the north-eastern part of Thailand. Drilling was carried out
Rd. through weathered-sandstone by powerful rotary drilling rig
with core barrel and rock auger.
Bored piles have been extensively used for foundations of
the majority of the elevated expressways since 1991.
Thousands of base-grouted bored piles were constructed for
these infrastructure projects including Second Stage
Expressway, constructed in 1991; Don Muang Tollway
Extension, constructed in 1997; Bangna-Bangpli-
Bangprakong Expressaway constructed in 1998; and
Wat Nakorn-In project, constructed in 2001. Over 700 bored
piles of diameter ranging from 0.8m to 1.5m with depth from
35m to 54m were constructed between 1999 and 2000 using
polymer-based slurry for the foundation of Rama VIII Bridge,
one of the initiatives of his majesty the King Bhumibol
Adulyadej. Progress of piling and superstructure works of the
Rama VIII Bridge at the bank of Chao Phraya river is depicted
in Fig. 2(a) and 2(b).
Fig. 3 a. Rock-socket bored pile construction for Mool River

Project Ubon Ratchathani Province
Fig. 2a. Rama VIII cable-stayed bridge construction – more

than 700 deep-seated bored piles were constructed using
polymer-based slurry. Fig. 3 b. Rock-auger for Mool River bored pile construction
216
Bored pile of diameter 2m founded at 62m depth for the
foundation of Rama 8 bridge, a cable-stayed bridge across
Chaophaya river constructed in early 2005 was the largest
deep-seated bored pile in Bangkok installed by rotary drilling
method at that time. In the same project, both conventional
static pile load tests and bi-directional load tests were carried
out on 2m diameter bored piles. Simplified version of load vs.
settlement of these pile tests are presented in Fig.4.
Apparently, from the test results, the test piles could carry a
significantly higher load than the original pre-defined safe
design load of 1600 ton. Thasnanipan et. al. (2006) reported
the experiences gained in practical construction of these
largest diameters bored piles at the river bank. Pictures of
some activities during construction of bored piles in the said
project are presented in Figs. 5(a), 5(b), 6(a), 6(b), 7(a) and
7(b).
Fig. 5b View of progress of bored pile diameter 2m drilling at

LOAD (kN) the bank of Chaophaya River
0 10000 20000 30000 40000 50000
0
5
Settlement (mm)
10
15
20
Conventional Static Pile Load Test
25 Bi-directional Load Test
30
Fig.4. Load-settlement relationship of diameter 2.0m bored

piles (conventional static pile load test and bi-directional load
test)
Fig. 6a View of progress in rebar cage preparation for bi-

directional load test
Fig. 5a View of progress in temporary casing installation for

bored pile diameter 2m at the bank of Chaophaya River Fig. 6b View of progress in installation of rebar cage
217
Fig. 8b View of cruciform barrette after installation of
reinforcement
Fig. 7a View of conventional static pile load test Due to the extensive bearing capacity requirements, T-
shape barrettes (Thickness=1m, Width =3m) of 55m deep
were used in one of the recent major projects. View of the
guide wall and reinforcement installation of T-shape barrette
are shown in Fig. 10(a) and 10(b). Fig. 11 shows the T-shape
barrettes after exposing to the cut-off level.
The above-mentioned works reflect the development and
achievement history of bored piles and barrettes in Thailand
and offered useful information to the local construction
industry as well as to that of the international deep foundation
construction.
Fig. 7b View of bi-directional pile load test
Deep-seated T-shape and L-shape barrettes were

constructed as leg-piles for diaphragm wall in recently
completed projects in Bangkok. Construction of cruciform
barrette (shown in Fig. 8(a) and 8(b)) for monopole type high-
voltage power transmission line was described by
Thasnanipan et al., (2000a).
The availability of more powerful and efficient hydraulic
grabs offers faster construction of deep-seated barrettes in
comparison with mechanical cable-hung grabs. Fig. 9 shows Fig. 9 View of barrette construction by hydraulic grab
the construction of barrettes using hydraulic grab.
Fig. 8a Guide wall for cruciform barrette Fig. 10a View of guide-wall construction for T-shape barrette
218
instrumented (strain gauges) pile load test results.
Chiewcharnsilp (1988) reported the shaft friction factor 
values of sand layers in Bangkok based on the instrumented
load test results. In addition to the literature previously
available (Chiruppapa, 1968; Suwanakul, 1969; Promboon,
1981; Ng, 1983; Chiewcharnsilp, 1988), Pimpasugdi (1989)
determined the shaft friction factor (adhesion factor, ) of
Bangkok clay layers based on 11 bored piles of diameters
ranging from 0.50 to 0.80m with the embedded length varying
between 21.50m and 46m. It should be noted that the design
parameters obtained from the research works of 1980s were
mainly based on the estimation of shaft friction loads from
plain static pile load test results with numbers of assumptions
since instrumented pile load test results were limited.
With the peak of construction-boom in Thailand,
particularly in Bangkok, large numbers of instrumented full-
scale static load tests on bored piles were conducted
Fig. 10b View of reinforcement installation for T-shape throughout 1990’s which provided better understanding on
barrette behavior of these deep-seated foundations. Researches focused
on the design parameters and methods were published based
on these test data. With improved design methods the wet-
process cast-in-place foundations came to be regarded as
reliable foundations in the construction industry of Thailand.
With more confidence on soil parameters selection and better
understanding on behaviour of these deep foundations, the
designer designed higher load capacity bored piles and
barrettes in late 1990 than those in 1980s. Thasnanipan et. al.,
(1999) reported the failure mechanism of long bored piles in
layered soil of Bangkok. The authors cited that for the bored
piles embedded in the multi layered soils of Bangkok,
estimation of ultimate shaft friction capacity needs to consider
the brittle type of failure mechanism of stiff to hard clay layers
and the  ��values selected need to be adjusted accordingly.
Peak and residual  �values mobilized in the stiff clay layers
analyzed by the authors were plotted as shown in Fig. 12
Fig. 11 View of T-shape barrette after excavation to the final along with the suggested curves by different researchers. It
depth can be seen from the figure that the residual � value of 2nd
stiff clay layer (undrained shear strength values of 25 ton/m2),
at the maximum test load, drops below the curve suggested by
7. IMPROVEMENT IN DESIGN CONCEPT AND
Pimpasugdi (1989).��So the � values proposed by the
PARAMETERRS SELECTION
author in Fig.12 for the stiff to hard clay layers overestimate
the ultimate shaft friction under these conditions.
In the initial stage of introducing bored piles in Thailand, the
Introduction of polymer-based slurry for wet-process bored
design concepts and parameters were mainly based on the
piles marked a major breakthrough for both construction and
available literature from research carried out in other parts of
design engineers. Thasnanipan et al.,(2002b), reported that
the world such as Tomlinson (1957), Skempton (1959), Broms
bored piles constructed with polymer-based slurry have higher
(1966), Bowles (1968), Meyerhof (1976) etc. The work of
capacity than those constructed with bentonite slurry. Fig. 13
Chiruppapa (1968) was believed to be the first research data
shows the shaft friction factors  of sand layers for polymer-
available for design parameters of bored piles in Bangkok soft
based bored piles in comparison with the design line of
clay. The author conducted the study based on 6 dry-processed
bentonite bored piles. The higher load capacity of bored piles
small diameter bored piles with load cells installed at the pile
constructed with polymer-based slurry allows the use of a
toes on AIT campus in Phathumthani. Though the piles were
single, large diameter deep-seated bored pile in place of a
simply constructed using the dry-process with casing method,
group of smaller size shallow-seated bored piles or driven
this early-stage study provided some important information
piles. Fig. 14 depicted the static pile load test result of 1.80m
such as load-settlement behaviour, adhesion factor () and
diameter bored pile of 60m depth constructed with polymer-
bearing capacity factor (N) of bored piles in Bangkok soft
based slurry in urban area of Bangkok. The test was conducted
clay.
in late 2003. The maximum applied load 48 000 kN in this test
With the passage of time, design method and selection of
is believed to be the highest static pile load test performed on
parameters for local subsoil were improved as a result of
a single bored pile in Thailand. Apparently, from the test
research works carried out in 1980s. Differences between the
result, the pile (diameter 1.80m x 60m deep) could carry a
behaviour of driven piles and bored piles were well realized
significantly higher load than the original pre-defined safe
from these studies. Ng (1983) presented the load distribution
design load of 16 000 kN.
characteristics of wet-processed bored piles founded in both
the first and second sand layers of Bangkok based on
219
1.20
Bored Cast Insitu Piles performance monitoring and quality control testing.
Interpretation skills relevant to local soil conditions and
1.00
construction methods of these tests were significantly
Tomlinson,1957 improved in local industry. For instance, in early 1980, the
0.80 sonic integrity (seismic test) test results were needed to send

to the specialists abroad for interpretation which in turn made

Adhesion Factor ,
Suchada,1989
the testing cost more expensive. Significant cost-saving were
achieved in some major projects of late 1990s, as more
practical and precise interpretation were locally made to verify
0.40
Peak
and establish acceptance criteria in proving the quality of
suspected piles with anomalies.
0.20
Residual
9. KODEN DRILLING MONITORING SYSTEM

0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
2 In Thailand, prior to the availability of Koden testing
equipment, borehole verticality was checked by mechanical
type equipment such as plumb float and adjustable globe.
Fig. 12 Comparison of adhesion factor�,  suggested by However, the reliability of these methods was doubtful. The
different researchers with the actual mobilized in the stiff clay Koden drilling monitor system was believed to be first used in
layers (after Thansnanipan et. al 1999) Thailand in 1980. The verticality of slurry-filled borehole can
be monitored rapidly from the electronic plot of the
1.0 continuous profile by Koden equipment. This system is very
Submaneewong (1999) : Bentonite bored Piles (2nd Sand)
Submaneewong (1999) : Bentonite Bored Piles (2 nd Hard Clay) useful for the verticality-control of long barrettes - unlike
0.8 Polymer Bored Pile TP-1 (1st Sand) bored piles, guiding by temporary casing is not available for
Polymer Bored Pile TP-2 (1st Sand) barrette drilling.
0.6 Polymer Bored Pile TP-1 (2nd Sand)
 (Ks tan 
Design Line for

bentonite bored piles
0.4
Polymer Bored Pile TP-3 (2nd Sand) Submaneewong (1999) 10. SONIC INTEGRITY TEST
0.2 Sonic integrity test, also known as seismic test is the most
common method of integrity testing for both driven and bored
piles in Thailand. Sonic integrity test is usually selected for
0.0
both quality check (control test) and retrospective
28 30 32 34 36 38 investigation. It is the cheapest in terms of cost and the
Angle of Internal Friction, '
simplest in terms of testing process. The main advantage of
this test is that since no particular preparation is necessary
Fig. 13 Back-calculated  values of polymer bored piles at during the pile construction phase, it is more flexible to select
maximum test load plotted on design line of bentonite bored which pile is to be tested. However, interpretation of sonic
piles constructed in Bangkok subsoil integrity testing needs considerable experience and knowledge
(after Thansnanipan et. al 2002b) in testing, subsoil condition and construction method. In many
Applied Load (KN) projects it is a part of the contractual requirement to conduct
0 5000 10000 15000 20000 25000 30000 35000 40000 45000 50000 sonic integrity test. Minimum 10 % to maximum 100 % of
0 production piles are commonly tested. It is also a reasonably
Cycle1 acceptable method as a retrospective investigation in
Pile head displacement (mm)
Cycle2
10 Cycle3
determining integrity of the pile. The signal characteristics and
their interpretations of sonic integrity test on piles founded in
20 Bangkok subsoil were reported in details by Thasnasipan et al.
(1998d).
30
11. CROSS- HOLE SONIC LOGGING TEST

40
The first sonic logging test was believed to be conducted in

50
1982 for the wet-processed bored piles of the Memorial
Bridge Project where bad concrete zones were detected at
Fig. 14 Static pile load test result of 1.80m diameter bored depth about 20m and 1m Ng, 1983. Use of the Sonic logging
pile of depth 60m in Bangkok test for checking pile integrity has increased in Thailand
particularly in mega projects. Cross-hole sonic logging test is
relatively expensive. It is mainly employed as a pre-planned
8. DEVELOPMENT IN PERFORMANCE MOINTORING
site quality control testing. The major advantage of this
AND QUALITY CONTROL TESTING
method is that test can be carried out shortly after pile
construction. Hence, rectification measures can be
Improvement of testing equipment and powerful computer
implemented if the pile is defective while the foundation
facilities are key factors that contributed to the development in
contractor is on site. However, this method is generally not
220
applicable if pile integrity is in question due to post- Thailand. However, in the authors’ opinion, there is much
construction activities, as access tubes are usually grouted work to be done with particular focus on constructability
after completion of the test. The results from sonic logging test issues, concrete technology for wet-processed bored piles and
conducted on model piles in Bangkok helped to extend the barrette, reliable but cost-effective quality control testing and
knowledge of the signal characteristics and interpretation value-engineering.
(Thasnanipan et. al., 2000b). Thasnanipan et al., (2004b) also Starting from the planning stage, site investigation, design,
demonstrated the application of cross-hole sonic logging test construction and inspection should be integrated so that
in identifying over-cast length of bored piles prior to exposing designers, contractors and construction inspectors can
the pile heads. participate as a team with a common goal. Appropriate and
practical specifications should be established jointly by these
parties for local soil conditions and construction methods.
12. HIGH STRAIN DYNAMIC LOAD TEST Practical acceptance criteria should be developed to verify
bored piles and barrettes with suspected anomalies.
High strain dynamic integrity test has become a well-accepted Continuing education should be promoted for designers,
method especially for evaluating the pile capacity in today’s inspection engineers, and contractors. The Geotechnical
foundation industry of Thailand and it is applied for both Chapter of the Engineering Institute of Thailand under the
driven and bored piles. A large number of related technical royal patronage of his majesty King Bhumibol Adulyadej, has
papers and case histories of the test have been published and it started to establish the standard code of practice and
is a part of standards and specifications such as ASTM guidelines for wet-processed bored piles which will serve as a
D4945-89 (Standard Method for High-strain Dynamic Testing yardstick for the deep foundation industry upon its completion
of Piles). Thasnanipan et al. (2000b), reported the application in the near future.
of dynamic load testing on piles in Thailand.
REFERENCES
13. FURURE TRENDS OF CAST-IN-PLACE DEEP
FOUNDATIONS IN THAILAND Bowles, J. E. 1968. Foundation analysis and design,
McGRaw-Hill, New York
As the practitioners in construction industry have gained more
Broms B. B. 1966. Methods of calculating the ultimate
experience and confidence in using higher capacity bored piles
bearing capacity of piles – a summary, Soils-Soils, No. 18-
and barrettes, these types of deep foundations are expected to
be of more in demand in the future. The higher load capacity 19 : 21-32
achievable by deep-seated large bored piles and barrettes will Chiewcharnsilp, S. 1988. Static load transfer and predictions
allow the use of a single foundation element in place of a of carrying capacity of bored piles (in Thai), M. Eng.
group of smaller size shallow-seated bored piles and driven Thesis, Chulalongkorn University, Bangkok, Thailand
piles. It is expected that improvement in installation Chiruppapa, P. 1968. Cast in-situ bored piles in Bangkok
techniques will be of key advance in the future of deep Clay, M. Eng. Thesis No. 213, Asian Institute of
foundation industry in Thailand. Technology, Bangkok, Thailand
Meyerhof, G. G. 1976. Bearing capacity and settlement of pile
It is anticipated that barrettes will be more popular in the foundations, Journal of soil mechanic and foundation
future for the following reasons.
division, Proceeding ASCE Vol. 102, GT3 : 197-223
 The availability of more powerful and efficient hydraulic
Ng, K. C. 1983. The construction problem and performance of
grabs which will offer faster construction of deep-seated
barrettes large bored piles in second sand layer, M. Eng. Thesis No.
 Less noise which will offer main advantages in congested GT 82-26, Asian Institute of Technology, Bangkok,
and sensitive neighbourhood Thailand
 Less equipment requirement that will offer advantageous Pimpasugdi Suchada 1989. Performance of Bored, Driven and
construction in limited space Auger Press Piles in Bangkok Subsoils, M. Eng. Thesis
 More versatility (e.g. able to construct in the area with No. GT 88-12, Asian Institute of Technology, Bangkok,
limited headroom) Thailand
 Better in safety aspects as less equipment are required Promboon, S . and Brenner, R.P. 1980. Large diameter bored
(especially for the construction along the public roads, piles in Bangkok clay, Proceedings of 10th conference on
subway stations, elevated expressway etc.)
soil mechanic and foundation engineering, Vol. 2 : 815-
818
14. CONCLUDING REMARKS Sakaret Naka 2004. Bidirectional pile test on bored piles in
Bangkok subsoils (in Thai), M. Eng. Thesis,
In Thailand, according to the authors’ experience as a deep- Chulalongkorn University, Bangkok, Thailand
foundation contractor, development and achievement in both Suwanakul, V. 1969. Cast in-situ piles in Bangkok clay, M.
construction and design aspects of wet-process deep bored Eng. Thesis No. 260, Asian Institute of Technology,
piles and barrette foundations in the past four decades were Bangkok, Thailand
significant. With recognition of technical and economic
advantages of using these high capacity cast-in-place
foundations by local practitioners, it is expected that they will
be more popular in the future construction industry of
221
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., Thasnanipan Narong, Zaw Zaw Aye, Thayanan Boonyarak
1998a. Design, construction and behavior of bored cast in- and Natamon Kampananon 2004b. Quantifying Pile Head
Situ concrete piles in Bangkok Subsoil, Proceedings: Condition Before Basement Excavation by Cross-hole
Fourth International Conference on Case Histories in Sonic Logging Tests, The Seventh International
Geotechnical Engineering, St. Louise, Missouri, U.S.A. : Conference on Application of Stress-Wave Theory to
281-287. Piles, Kuala Lumpur, Malaysia, August 9 – 11, 2004 :
Thasnanipan N., Maung A.W., Tanseng P. 1998b. Barrettes 233-238.
founded in Bangkok subsoils, construction and Thasnanipan N., Manosuthisan C., Boonyarak T., Chanchad
performance, Thirteenth Southeast Asian Geotechnical C., 2006. Experience in Construction of Largest Diameter
Conference, Taipei, Taiwan, ROC :.573-578. Deep-seated Bored Piles at the Bank of Chaopraya River,
Thasnanipan N., Ganeshan B., Anwar M.A. 1998c. Effect of Proceedings of the Eleventh National Convention on Civil
Construction Time and Bentonite Viscosity on Shaft Engineering, Phuket, April 20 – 22, 2006 (in Thai)
Capacity of Bored Piles, Third International Geotechnical
Seminar on Deep Foundation on Bored and Auger Piles,
Ghent, Belgium : 171-177.
Thasnanipan N., Maung A.W., Ganeshan B. 1998d. Sonic
Integrity Test on Piles Founded in Bangkok Subsoil-
Signal Characteristics and Their Interpretations,
Proceedings: Fourth International Conference on Case
Histories in Geotechnical Engineering, St. Louis,
Missouri, U.S.A. : 1086-1092
Thasananipan N., M. A. Anwar Maung A. W 1999. Failure
Mechanism of Long Bored Piles in Layered Soils of
Bangkok, Civil and Environmental Engineering
Conference, Bangkok, Thailand : V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A.
2000a. Barrettes : A Versatile Foundation for
Transmission Line Towers, An International Conference
on Geotechnical & Geological Engineering, Melbourne,
Australia, 19-24 November 2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye
2000b. Non-Destructive Integrity Testing on Piles
Founded in Bangkok Subsoil, 6th International Conference
on Application of Stress-Wave Theory to Piles, September
2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa.
2002a. Barrette of Over 50,000 kN Ultimate Capacity
Constructed in the Multi-Layered Soil of Bangkok,
Proceedings of the International Deep Foundations
Congress 2002, Geotechnical Special Publication No.
116, Volume Two, ASCE, Febuary 14-16 2002, Orlando,
Florida, USA : 1073-1087.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong
and Wanchai Teparaksa 2002b. Performance of Wet-
Process Bored Piles Constructed with Polymer-Based
Slurry in Bangkok Subsoil, Proceedings of the
International Deep Foundations Congress 2002,
Geotechnical Special Publication No. 116, Volume One,
ASCE, February 14-16 2002, Orlando, Florida, USA :143-
157.
Thasnanipan, N., Zaw Zaw Aye and Chanchai Submaneewong
2004a. Effectiveness of Toe-grouting for Deep-seated
Bored Piles in Bangkok Subsoil, Proceedings of sessions
of the Geosupport Conference: Innovation and
Cooperation in the Geo-Industry, Geotechnical Special
Publication No. 124, ASCE, January 29-31, 2004,
Orlando, Florida, USA : 561-572.
222
Proceedings of Geotechnical and Foundation Conference 2012
26-27 September 2012, Bangkok, Thailand
Geotechnical and Foundation

Conference 2012
Barrette for Highrise Buildings and Heavy

Structures in Bangkok Subsoil
Narong Thasnanipan, Zaw Zaw Aye, Phongpat Kitpayuck and

Nutthapon Thasnanipan
Thayanan Boonyarak
Hong Kong University of Science and Technology
223
Barrette for Highrise Buildings and Heavy Structures
in Bangkok Subsoil
Narong Thasnanipan1, Zaw Zaw Aye1, Thayanan Boonyarak2, Phongpat Kitpayuck1
Nutthapon Thasnanipan1
1
Seafco Public Company Limited: zaw@seafco.co.th
2
Hong Kong University of Science and Technology: thayanan@hotmail.com
Based on the authors’ experiences, this paper is intended to summarize the development of
deep-seated barrettes construction in Thailand, in comparison with that of the past. Literature
related to wet-processed bored piles and barrettes in Thailand published through out the past
30 years is summarized together with recent research works. Results of the static load test
performed on barrettes in three projects are summarized. Increasingly use of barrettes for the
highrise building projects is described. Recent application of heavy structures barrettes for
elevated mass rail transport project, MRT Purple Line is also presented. An attempt is also
made to briefly discuss the future trend of deep foundations in Thailand.
INTRODUCTION
During the past three decades, owing to the acceleration of development in mega
infrastructure projects, techniques and practice of cast-in-place deep foundations, particularly
wet-processed deep-seated large diameter bored piles and barrettes have experienced
enormous progress in Thailand. The versatility of the construction method and the high-load
capacity which in turn offered the constructability and cost-saving are the main factors
contributed to the increasing use of those cast-in-place deep foundations. In the initial stage,
there were a number of questions on design and construction aspects of these foundation
systems particularly in Bangkok subsoil. With the passage of time, the construction
equipment, installation techniques and testing of deep foundation elements have been
developed. Large numbers of instrumented full-scale static pile load tests were conducted
through out the 1990s mainly in mega projects, which provided better understanding on
behavior of these deep-seated foundations.
Research focused on the design parameters and methods were produced based on
these test results. The design parameters became so well established that wet-process cast-in-
place foundations became regarded as reliable foundations for practitioners involved in
construction industry in Thailand. This paper focus the past and current practice of barrettes
in Thailand together with a brief discussion on future trend deep foundations in Bangkok
subsoil. The information contained in the paper is mainly from the highrise buildings and
infrastructure projects constructed in Bangkok and adjacent areas.
225
SUBSOIL AND EXISTING PIEZOMETRIC PROFILE
Subsoil profile and the present piezometric drawdown condition of Bangkok are
presented in Figure 1 below. A typical subsoil profile is relatively consistent in different
localities in Bangkok. It is characterized by alternating layers of clay and sand deposits as
shown in Figure 1. 2
Geotechnical Pore Pressure (kN/m )
0 Weathered Crust
0

Clay  t = 14-16kN/m 3
10 -
10
Medium to
Su= 50-14020- KPa
Depth (m)
20 Stiff Clay  t = 17-21kN/m 3
Act
u
Medium
al P
-
30
H yd
 t = 20kN/m 3
iezo
(First Sand)
ros
m
etri
ta
40 -
tic L
Hard Clay Su> 200 KPa
c
40
 = 21KN/m 3
D ra
t
ine
Dense to Very wdo
50 wn
-
50
Lin
(Second Sand)  t = 21KN/m 3
e
60 -
60
Figure 1. Typical soil profile of Bangkok with piezometric drawdown condition

(Thasnanipan, et. al., 2002)
BARRETTE CONSTRUCTION METHOD IN BANGKOK SUBSOIL
Due to the prevailing subsoil and groundwater conditions, barrettes are constructed by
wet-process or slurry displacement method.
Both mechanical and hydraulic cable-suspended grab are used for barrette
construction. In recent years, use hydraulic grabs are more common. Excavation of the trench
is carried out by the cyclic-process of lifting and lowering of the grab under gravity and
tangential force of the clamshell operated by cables (mechanical grab) or hydraulic action
(hydraulic grab). Different from bored pile construction, a guide wall of depth 1 to 1.5 m with
inside clear dimensions slightly larger than the nominal size of the barrette is used to guide
the grab during initial bites. Bentonite slurry is introduced to the trench as soon as the initial
excavation commenced. The excavation is continued under the bentonite slurry to the final
depth. After recycling of slurry and lowering the rebar cage, concreting is done by tremie
method.
226
GUIDE WALL
TREMIE PIPE
REINFORCEMENT
CAGE
BENTONITE
SLURRY
EXCAVATION INSTALLATION OF CONCRETING

REINFORCEMENT
Figure 2. Schematic of barrette construction sequences
OVERVIEW OF BARRETTE APPLICATION IN CONSTRUCTION PROJECTS
applicable construction methods, and/or extensive bearing capacity requirements, the use of
barrette foundations would make a suitable alternative. Barrettes with dimensions ranging
from 0.80 m x 2.7 m to 1.5 m x 3.0 m for safe working load capacity from 11,000 to 23,000
kN have been used in some major projects.
The first barrette in Bangkok was believed to be constructed in the late 1970s for the
foundation of the Bangkok Bank Head Office Building at Silom road with size of 0.6-0.8 x
2.5m and toe depth of 33m. Barrettes can be constructed with flexible layout plans for both
vertical and lateral loads. The layout pattern of barrettes can be arranged in a continuous row
or column, radial, alternating long and short axis of barrette and a combination of two or more
of such patterns. To minimize the need of additional equipment, a mechanical or hydraulic
grab mounted on crawler crane can be used for constructing both barrettes (as piled-
foundation) and diaphragm walls (as earth retaining structure) in some suitable projects
without employing a bored piling rig, which could save the extra mobilization cost. In
addition to the large bearing capacity requirements, on site difficulties such as limited head
room where piling rigs cannot be utilized, under such situations like presence of overhead
high voltage power cables, existing overpasses or structures for elevated expressways and
planned subway stations, also demands the barrettes. Static load test up to 52,900 KN
conducted on barrette set the record as the highest load ever tested for a single barrette
foundation in Thailand (Thasnanipan et. al. 2002a). A summary of completed projects
constructed with barrettes with respect to selection criteria in 2003 in comparison with 2012 is
presented in Table 1. As can be seen in the table, significant increase of barrette foundation
can be realized particularly in the project with high load capacity requirement. It is also
evident that barrette in combination with diaphragm wall is increasingly popular in the
construction industry of Thailand.
227
Table 1 Summary of Barrette Selection in major projects
Selection Criterion Accumulated number of projects Remarks
as of 2003 as of 2012
August
High Load Capacity 4 21 Incorporated with bored
piles
Minimize Construction 6 14 Alternative for bored piles
Equipment
Limited Head Room for 3 8 Under existing structures
Excavation such as bridges, elevated
expressway and power lines
Combination with 10 25 As diaphragm wall legs
Diaphragm Wall
Foundation as well as 3 4 Provision for the future
portion of column requirement
General configuration of barrettes as leg piles integrated with diaphragm wall and bored piles
of building with deep basement is shown in Figure 3.
Figure 3 Schematic of barrettes as leg piles integrated with diaphragm walls and bored piles
PROBLEMS AND DIFFICULTIES IN CONSTRUCTION OF BARRETTES IN

EARLY DAYS
Basic but extensive problems were experienced in early stages of barrette construction in
 Limited availability and capacity of equipment
 Lack of skills in operation of equipment (particularly the drilling rig)
228
 Adverse effect due to a slow rate of drilling (e.g. excessive formation of filter-cake by
 Limited knowledge and less advanced techniques in control of bentonite slurry
 Limited experience in construction method and related negative impact
 Quality of concrete for tremie concreting method
 Lack of experienced engineers and foremen
 Improper construction and quality control specifications and guidelines for deep-
seated piles in local soil
 Improper design for constructability
IMPROVEMENT IN CONSTRUCTION OF BARRETTE
Over the past three decades, along with the development of wet-processed piling
technology in other parts of the world, equipment, construction technique and design methods
as well as better understanding of construction impact on the performance of this type of
foundation have significantly improved in Thailand. Table 2 summarizes the areas of
improvement in bored pile and barrette construction and main factors contributed to these
developments. Some of these improvements were published both locally and internationally.
The design, construction and behavior of bored cast in-situ concrete piles in Bangkok Subsoil
was presented by Thasnanipan et. al., (1998a). The construction and performance of barrettes
in Bangkok Subsoil was also reported by Thasnanipan et. al., (1998b). Effect of construction
time and bentonite viscosity on shaft friction of bored piles was also pointed out by
Thasnanipan et al., (1998c). Thasnanipan et. al., (2004a) also reported a comprehensive study
on the effectiveness of two different toe-grouting methods, known as tube-â-manchette and
drill-and-grout, applied in Bangkok.
(a) (b)
Figure 4. (a) Guide wall for cruciform barrette (b) View of cruciform barrette after
installation of reinforcement

voltage power transmission line was described in the work of Thasnanipan et. al., (2000a).
Construction of deep-seated T-shape and L-shape barrettes as leg-piles for diaphragm wall are
now more common in Bangkok.
229
Above-mentioned research works reflect the development history of bored pile and
barrettes in Thailand and offered useful information to the local construction industry and
perhaps to the international deep foundation engineering society.
Table 2 Summary of development in barrette construction

construction required to complete complete barrette operating skills as
barrette 1.5mx3m of toe 1.2mx3m toe depth well as improved-
depth 57m 60m knowledge in
management and
Pile size and Barrette Barrette Better equipment,
depth Limited in size and depth Various sizes & operating skills as
depth over 60m well as improved-
knowledge in
management and
availability and
advance technology
and performance
process
Quality control Fewer methods available Better equipment Advance equipment,
interpretation
understanding of construction impact of quality and performance of these wet-processed deep
foundations.
The availability of more powerful and efficient hydraulic grabs offers faster
construction of deep-seated barrettes in comparison with mechanical cable-hung grabs. Figure
5 shows the construction of barrettes using hydraulic grab.
230
Figure 5. View of barrette construction by hydraulic grab
IMPROVEMENT IN DESIGN CONCEPT AND PARAMETERS SELECTION
Usually the designers use the same design concept and parameters for barrettes and
bored piles constructed under bentonite slurry. Following paragraphs describe the design
practice of both elements commonly applied in the past and improvement with time.
parameters were mainly based on the available literature from research carried out in other
parts of the world such as Tomlinson (1957), Skempton (1959), Broms (1966), Bowles
(1968), Meyerhof (1976) etc. The work of Chiruppapa (1968) was believed to be the first
research data available for design parameters of bored piles in Bangkok soft clay. The author
conducted the study based on 6 dry-processed small diameter bored piles with load cells
installed at the pile toes on AIT campus in Phathumthani. Though the piles were simply
constructed using the dry-process with casing method, this early-stage study provided some
important information such as load-settlement behavior, adhesion factor () and bearing
capacity factor (N) of bored piles in Bangkok soft clay.
With the passage of time, design method and selection of parameters for local subsoil
were improved as a result of research works carried out in 1980s. Differences between the
behavior of driven piles and bored piles were well realized from these studies. Ng (1983)
presented the load distribution characteristics of wet-processed bored piles founded in both
the first and second sand layers of Bangkok based on instrumented (strain gauges) pile load
test results. Chiewcharnsilp (1988) reported the shaft friction factor  values of sand layers in
Bangkok based on the instrumented load test results. In addition to the literature previously
1988), Pimpasugdi (1989) determined the shaft friction factor (adhesion factor, ) of Bangkok
clay layers based on 11 bored piles of diameters ranging from 0.50 to 0.80m with the
shaft friction loads from plain static pile load test results with numbers of assumptions since
231
load tests on bored piles and a few barrettes were conducted throughout 1990’s which
provided better understanding on behavior of these deep-seated foundations. Researches
focused on the design parameters and methods were published based on these test data. With
improved design methods the wet-process cast-in-place foundations came to be regarded as
reliable foundations in the construction industry of Thailand. With more confidence on soil
parameters selection and better understanding on behavior of these deep foundations, the
designer designed higher load capacity bored piles and barrettes in late 1990 than those in
1980s. Thasnanipan et. al., (1999) reported the failure mechanism of long bored piles in
layered soil of Bangkok. The authors cited that for the bored piles embedded in the multi
layered soils of Bangkok, estimation of ultimate shaft friction capacity needs to consider the
brittle type of failure mechanism of stiff to hard clay layers and the values selected need to
be adjusted accordingly. Peak and residual values mobilized in the stiff clay layers analyzed
by the authors were plotted as shown in Figure 10 along with the suggested curves by
different researchers. It can be seen from the figure that the residual value of 2nd stiff clay
layer (undrained shear strength values of 25 ton/m2), at the maximum test load, drops below
the curve suggested by Pimpasugdi (1989).So the values proposed by the author in Figure
6 for the stiff to hard clay layers overestimate the ultimate shaft friction under these
conditions.
1.20
1.00
Tomlinson,1957
0.80

Adhesion Factor ,
Suchada,1989
0.40
Peak
0.20
Residual
0.00
0.00 5.00 10.00 15.00 20.00 25.00 30.00
2
Figure 6 Comparison of adhesion factor , suggested by different researchers with the actual
232
1.0
Submaneewong (1999) : Bentonite bored Piles (2nd Sand)
Submaneewong (1999) : Bentonite Bored Piles (2 nd Hard Clay)
0.8 Polymer Bored Pile TP-1 (1st Sand)
0.6 Polymer Bored Pile TP-1 (2nd Sand)
 (Ks tan 
Design Line for

bentonite bored piles
Polymer Bored Pile TP-3 (2nd Sand) Submaneewong (1999)
0.4
0.2
0.0
28 30 32 34 36 38
Angle of Internal Friction, '
Figure 7 Back-calculated  values of polymer bored piles at maximum test load plotted on
design line ( values in sand layer) of bentonite bored piles based on back-calculated
DEVELOPMENT IN PERFORMANCE MOINTORING

AND QUALITY CONTROL TESTING OF BARRETTES
Improvement of testing equipment and powerful computer facilities are key factors
that contributed to the development in performance monitoring and quality control testing.
Interpretation skills relevant to local soil conditions and construction methods of these tests
were significantly improved in local industry. For instance, in early 1980, the sonic integrity
(seismic test) test results were needed to send to the specialists abroad for interpretation which
in turn made the testing cost more expensive. Significant cost-saving were achieved in some
major projects of late 1990, as more practical and precise interpretation were locally made to
verify and establish acceptance criteria in proving the quality of suspected piles or barrettes
with anomalies.

However, the reliability of these methods was doubtful. The Koden drilling monitor system
was believed to be first used in Thailand in 1980. The verticality of slurry-filled borehole can
be monitored rapidly from the electronic plot of the continuous profile by Koden equipment.
This system is very useful for the verticality-control of long barrettes - unlike bored piles,
guiding by temporary casing is not available for barrette drilling.
233
is that since no particular preparation is necessary during the pile construction, phase, it is
more flexible to select which pile is to be tested. However, interpretation of sonic integrity
testing needs considerable experience and knowledge in testing, subsoil condition and
construction method. In many projects it is a part of the contractual requirement to conduct
sonic integrity test. Minimum 10 % to maximum 100 % of production piles are commonly
tested. It is also a reasonably acceptable method as a retrospective investigation in
determining integrity of the pile. The signal characteristics and their interpretations of sonic
integrity test on piles founded in Bangkok subsoil were reported in details by Thasnasipan et.
al. (1998d).
The first sonic logging test was believed to be conducted in 1982 for the wet-
processed bored piles of the Memorial Bridge Project where bad concrete zones were detected
at depth about 20m and 1m Ng, 1983. Use of the Sonic logging test for checking pile
integrity has increased in Thailand particularly in mega projects. Cross-hole sonic logging
test is relatively expensive. It is mainly employed as a pre-planned site quality control testing.
The major advantage of this method is that test can be carried out shortly after pile
construction. Hence, rectification measures can be implemented if the pile is defective while
the foundation contractor is on site. However, this method is generally not applicable if pile
integrity is in question due to post-construction activities, as access tubes are usually grouted
after completion of the test. The results from sonic logging test conducted on model piles in
Bangkok helped to extend the knowledge of the signal characteristics and interpretation
(Thasnanipan et. al., 2000b). Thasnanipan et. al., (2004b) also demonstrated the application of
cross-hole sonic logging test in identifying over-cast length of bored piles prior to exposing
the pile top.
both driven and bored piles. A large number of related technical papers and case histories of
(2000b), reported the application of dynamic load testing on piles and barrettes in Thailand.
STATIC PILE LOAD TESTS ON BARRETTES
Figure 8, 9 and 10 show the load vs. pile head movement and load-transfer mechanism
of two standard rectangular-shape barrettes and T-Shape barrette respectively.
234
Applied Load (Tons)
0 1,000 2,000 3,000 4,000 5,000 6,000
0
Pile Head Settlement (mm)
20
40
60
Barrette 1.5x3.0 m, depth 57.5 m
80
(a)
Su (UC) t/m2 Load (Ton)

0 5 10
0 1000 2000 3000 4000 5000
0 Weathered
Crust
Soft
Clay
10
First Stiff
20
Clay
First
30
Sand
40
Hard
Clay 3.00 m
50
Dense Sand Barrette
Hard Clay
Second Tip Level
60 Sand -57.5 m
VWSG
0 50 100 60 Mechanical
SPT-N (blows/30cm) Extensometer
(b)
Figure 8 (a) Load vs. pile head movement and (b) load-transfer mechanism of base-grouted
barrette 1.50x3.00m founded at 57.5 m in Bangkok
235
Applied Load (Tons)
0 1,000 2,000 3,000 4,000 5,000 6,000 7,000
0
10
20
30
Barrette 1.2x3.0 m, depth 65 m
40
Figure 9 Load vs. pile head movement of barrette 1.20x3.00m founded at 65 m in Bangkok
Applied Load (Tons)

0 1,000 2,000 3,000 4,000 5,000 6,000
0
10
20
30
Barrette 1.0x2.0x3.0 m, depth 55 m
40
Figure 10 Load vs. pile head movement of T-shape barrette (1.00x2.00x3.00m) founded at
55m in Bangkok (constructed in tested in 2008)
236
Table 3 Summary of load-settlement curve from three case studies
Test No. 1 Test No. 2 Test No. 3
(1.5m x 3.00m) Barrette Test 2 T-shape barrette
Description
Base-grouted (1.2m x 3.0m) (1.0m x 2.0m x
3.0m)
Dimension 1.50m x 3.00m 1.2m x 3.0m 1.0m x 2.0m x 3.0
m (T)
Tip Level (m) -57.5 -65.0 -55.0
Shaft friction area (m2) 518 546 770
End Bearing area (m2) 4.50 3.60 5.00
Slurry Used Bentonite Bentonite Bentonite
Total sand layer thickness 18 33 6
along the shaft (m)
Design load (ton) 2,300 2,500 2,600
Maximum pile head movement 62 25 28
at maximum applied load (mm)
Load at 5mm pile head 2,515 2,470 2,600
movement (ton)
Load at 20mm pile head
5,290 6,100 5,270
movement (ton)
Load at fully mobilized shaft
5,060 6,470 4,600
friction (ton)a
Pile head movement at fully
20 24 17
mobilized shaft friction (mm)
Unit end bearing at maximum
70b 65b 130b
test load (ton/m2)
Note : a. Assumed shaft friction fully mobilized at plunging point of load-settlement curve
b. Unit end bearing not fully mobilized
Though it is not advisable to remark major conclusions from 3 barrette tests, following
summary can be drawn.
 Barrette head displacement (settlement) at design load is less or equal to 5mm
 Barrette head displacement (settlement) at assumed fully mobilized shaft friction is in
the order of 1.3 to 2% of shorter width of barrette (thickness)
 End bearing does not contribute to the ultimate capacity of barrette
 Thickness of sand layer along the pile shaft may have influence on the barrette shaft
friction capacity (as can be seen in Test No. 2 where sand layer thickness along the
shaft is significantly thicker than other 2 test barrettes)
BARRETTES FOR HIGH-RISE BUILDING PROJECTS
More use of barrettes for highrise buildings can be observed in recent years. Points
below are considered key factors contributed to increasing use of barrettes for highrise
buildings particularly in congestive urban area.
 Buildings foot-prints are relatively small in comparison with the heights in majority of
the highrise buildings. Hence higher load capacity foundation is required within
available area
 Equipment utilized for barrette construction are relatively quiet and less vibration so
that disturbance to the neighbouring area is less than that of piling equipment
237
 Less equipment require for barrette construction – one set of barrette excavation
equipment can be used for construction of barrette and diaphragm wall for the project
with limited space
 Removal of excavated soil from the site can be easier as barrette grab can dispose
directly into the dump trucks without needing to store on site provided that dumping
trucks are able to move in and out from the site during excavation time (important
issue in most of the congestive area where traffic restriction is stringent)
Bored pile
T-shape barrette
Figure 11 Mat foundation of a high-rise building supported by barrette and bored pile
Figure 11 shows an example of mat foundation of a high-rise building supported by

combination of barrette and bored pile. As explained earlier, one of major benefits of using
barrette to support high-rise building is larger load bearing capacity of barrette compared with
that of bored pile. Barrette can be designed for the high load concentration area while bored
pile can be used in less concentrated load area as shown in the figure. In addition, barrette
provides a versatile option to be constructed in various shapes such as T-shape (Fig. 10), X-
shape or cruciform (Fig. 4) and L-shape to meet requirement of load bearing and bending
moment capacity.
BARRETTES FOR HEAVY STRUCTURES - ELEVATED TRAIN PROJECT
infrastructure projects including Second Stage Expressway, constructed in 1991; Don Muang
Tollway Extension, constructed in 1997; Bangna-Bang Pli-Bangprakong Expressaway
constructed in 1998; and Wat Nakorn-In project, constructed in 2001. Over 700 bored piles of
diameter ranging from 0.8m to 1.5m with depth from 35m to 54m were constructed between
1999 and 2000 using polymer-based slurry for the foundation of Rama VIII Bridge, one of the
initiatives of his majesty the King Bhumibol Adulyadej.
However, in 2010, large numbers of barrettes were constructed for new MRT project
called Purple Line. Originally, the majority of foundations to support heavy elevated-
238
structures were designed for large bored piles of diameter 1m, 1.2m and 1.5m depending on
the loading requirement. After awarding the contracts to the main contractors, the value
engineering option using barrettes to replace bored piles proposed by the piling contractor
was accepted (as shown in Figure 12 and 13). Apart from the safety aspects, the key reasons
of changing from bored piles to barrettes are that (i) construction of barrettes requires less
equipment on the middle of the congestive road than bored piles (ii) pile cap (footing) for
barrette is much smaller than that of bored pile. Construction of over 1000 barrettes of
different sizes, 1x3m, 1.2x3m and 1.5x3m with toe depth ranging from 50m to 65m were
completed in the beginning of 2012.
Figure 12. Schematic of bored piles (original design) and proposed option of barrettes with
equivalent capacity– size of footing for barrette (20m2 in plan area) is significantly less than
that of bored pile (60m2 in plan area)
Figure 13. Construction of barrette for Elevated Train Project in Bangkok
Comparison of equipment and construction aspects bored pile and barrette is shown in Table
4 below.
239
Table 4. Comparison of equipment and construction aspects of barrette and bored pile for
construction on the public roads
Item Description Barrette Bored Pile
1 Total high-tech equipment requirement High Medium
2 Experience and skill of operation crew High Medium
3 Risk involved (safety aspect) Medium High
4 Cost of equipment High Medium
5 Area requirement on road Less More
6 Site and road cleanness Excellent Average
7 Traffic problem due to method of soil disposal Less More
LATER LOAD TEST ON T-SHAPE BARRETTE AND DIAMETER 2M BORED PILE
The comparison of field test results and 3D finite element analysis on full-scale
instrumented static lateral load tests on T-shape barrette (Figure 14) and bored piles
constructed in Bangkok were presented by Submaneewong C (2009). Figure 15 presents the
Load-movement graphs of barrettes and bored piles under lateral loading. The findings
reported by the authors provide good reference for future design of similar elements subjected
to lateral loading, particularly result of stiffness ratio vs shear strain level of barrette and
bored pile as can be seen in Figure 16.
Soil Description
1
1
0 0 -1.50
5 Very Soft to 5
P
-7.00
Soft Clay
10 10
-12.00
15 Medium Stiff Clay 15

Stiff to Very Stiff Clay
20 20
Medium Dense Sand
-24.00
25 25
Depth below Ground Surface (m)
30 Very Stiff Clay 30
35 35
40 40 -40.00
45 Hard Clay 45
50 50
Hard Clay
Pile Tip at
55 55
- 55.00m
(a) Section and plan configuration of 60 Very Dense Sand

1.00m 2.00m
lateral load test on T-shape barrettes 65
70 S-3 S-4
S-5
Hard Clay
75
3.00m 1.00m
80
S-1 S-6
Inclinometer Tube
S-2
Strain Guage
(b) Soil profile and instrumentation

details of tested T-shape barrettes
Figure 14. Lateral load test configuration for T-shape barrettes and instrumentation details
240

5000
4000
3000
Load (kN)
T-shape barrette (T4BP13)

2000 T-shape barrette (T4BP6)
Bored pile (T4P13)
1000 Bored pile (T4P14)
0
0 10 20 30 40 50 60 70 80 90 100
Movement at Pile Head (mm)
Figure. 15. Movement at pile head under lateral load of barrettes and bored piles

700
600
500
400
G/Su
300
200
100
0
0.01 0.1 1
Shear Strain (%)
Figure 16. Strain level of T-barrette and bored pile (reported by Submaneewong 2009)
plotted on stiffness ratio vs shear strain for Bangkok Soft Clay reported by Prinzl and Davies
(2006)
241
FURURE TRENDS OF CAST-IN-PLACE DEEP FOUNDATIONS IN THAILAND
As the practitioners in construction industry have gained more experience and

confidence in using higher capacity bored piles and barrettes, these types of deep foundations
are expected to be of more demanding in the future. The higher load capacity achievable by
deep-seated large bored piles and barrettes will allow the use of a single foundation element
in place of a group of smaller size shallow-seated bored piles and driven piles. It is expected
that improvement in installation techniques will be of key advance in the future of deep
foundation industry in Thailand.
It is anticipated that barrettes will be more popular in the future for the following
reasons.
 The availability of more powerful and efficient hydraulic grabs which will offer faster
construction of deep-seated barrettes
 Less noise which will offer main advantages in congestive and sensitive
neighbourhood
 Less equipment requirement which will offer advantages construction in limited space
 More versatility (e.g. able to construct in the area with limited headroom)
 Better in safety aspects as less equipment are required (especially for the construction
along the public roads, subway stations, elevated expressway etc.)
CONCLUDING REMARKS
In Thailand, according to the authors’ experience as a deep-foundation contractor,

development in both construction and design aspects of wet-process barrette foundations in
past decades were significant. With recognition of technical and economic advantages of
using these high capacity cast-in-place foundations by local practitioners, it is expected that
they will be more popular in the future construction industry of Thailand. However, in the
authors’ opinion, there is much work to be done with particular focus on constructability
issues, concrete technology for wet-processed bored piles and barrette, reliable but cost-
effective quality control testing and value-engineering.
Starting from the planning stage, site investigation, design, construction and
inspection should be integrated so that designers, contractors and construction inspectors can
participate as a team with a common goal. Appropriate and practical specifications should be
established jointly by these parties for local soil conditions and construction methods.
Practical acceptance criteria should be developed to verify bored piles and barrettes with
suspected anomalies. Continuing education should be promoted for designers, inspection
engineers, and contractors.
REFERENCES
Broms B. B. (1966) Methods of calculating the ultimate bearing capacity of piles – a

summary, Soils-Soils, No. 18-19, pp. 21-32
242
Chiewcharnsilp, S. (1988) Static load transfer and predictions of carrying capacity of bored
piles (in Thai), M. Eng. Thesis, Chulalongkorn University, Bangkok, Thailand
Chiruppapa, P. (1968) Cast in-situ bored piles in Bangkok Clay, M. Eng. Thesis No. 213,
Asian Institute of Technology, Bangkok, Thailand
Meyerhof, G. G. (1976) Bearing capacity and settlement of pile foundations, Journal of soil
mechanic and foundation division, Proceeding ASCE Vol. 102, GT3, pp. 197-223
Ng, K. C. (1983) The construction problem and performance of large bored piles in second
sand layer, M. Eng. Thesis No. GT 82-26, Asian Institute of Technology, Bangkok,
Thailand
Pimpasugdi Suchada (1989) Performance of Bored, Driven and Auger Press Piles in Bangkok
Subsoils, M. Eng. Thesis No. GT 88-12, Asian Institute of Technology, Bangkok,
Thailand
Promboon, S . and Brenner, R.P. (1980), Large diameter bored piles in Bangkok clay,
Proceedings of 10th conference on soil mechanic and foundation engineering, Vol. 2,
pp. 815-818
Sakaret Naka (2004) Bidirectional pile test on bored piles in Bangkok subsoils (in Thai), M.
Submaneewong C. (2009) Behavior of T-shape barrette and bored pile subject to Vertical and
Lateral Loading, Ph.D Dissertation, Chulalongkorn University, Bangkok, Thailand
Suwanakul, V. (1969) Cast in-situ piles in Bangkok clay, M. Eng. Thesis No. 260, Asian
Institute of Technology, Bangkok, Thailand
Thasnanipan N., Teparaksa W., Maung A.W., Ganeshan B., (1998a) Design, construction and
behavior of bored cast in-Situ concrete piles in Bangkok Subsoil, Proceedings: Fourth
International Conference on Case Histories in Geotechnical Engineering, St. Louise,
Missouri, U.S.A., p.p.281-287.
Thasnanipan N., Maung A.W., Tanseng P., (1998b) Barrettes founded in Bangkok subsoils,
construction and performance, Thirteenth Southeast Asian Geotechnical Conference,
Taipei, Taiwan, ROC, p.p.573-578.
Thasnanipan N., Ganeshan B., Anwar M.A., (1998c), Effect of Construction Time and
Bentonite Viscosity on Shaft Capacity of Bored Piles, Third International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent,
Belgium. p.p. 171-177.
Thasnanipan N., Maung A.W., Ganeshan B., (1998d), Sonic Integrity Test on Piles Founded
in Bangkok Subsoil-Signal Characteristics and Their Interpretations, Proceedings:
Fourth International Conference on Case Histories in Geotechnical Engineering, St.
Louis, Missouri, U.S.A., p.p.1086-1092
243
Thasananipan N., M. A. Anwar Maung A. W (1999) Failure Mechanism of Long Bored Piles
in Layered Soils of Bangkok, Civil and Environmental Engineering Conference,
Bangkok, Thailand, p.p. V-69 to V-73.
Thasnanipan N., Tanseng P., Maung A.W. & Anwar M. A. (2000a), Barrettes : A Versatile
Foundation for Transmission Line Towers, An International Conference on
Geotechnical & Geological Engineering, Melbourne, Australia, 19-24 November
2000
Thasnanipan N., Maung A. W. & Navaneethan T.., Z. Z. Aye (2000b), Non-Destructive

Integrity Testing on Piles Founded in Bangkok Subsoil, 6th International Conference
on Application of Stress-Wave Theory to Piles, September 2000, Sao Paulo, Brazil
Thasnanipan N., Zaw Zaw Aye and Wanchai Teparaksa (2002a) Barrette of Over 50,000 kN
Ultimate Capacity Constructed in the Multi-Layered Soil of Bangkok, Proceedings of
the International Deep Foundations Congress 2002, Geotechnical Special Publication
No. 116, Volume Two, ASCE, Febuary 14-16 2002, Orlando, Florida, USA, p.p. 1073-
1087.
Thasnanipan, N., Zaw Zaw Aye, Chanchai Submaneewong and Wanchai Teparaksa (2002b)
Performance of Wet-Process Bored Piles Constructed with Polymer-Based Slurry in
Bangkok Subsoil, Proceedings of the International Deep Foundations Congress
2002, Geotechnical Special Publication No. 116, Volume One, ASCE, February 14-16
2002, Orlando, Florida, USA, p.p. 143-157.
Thasnanipan, N., Zaw Zaw Aye and Chanchai Submaneewong (2004a) Effectiveness of Toe-
grouting for Deep-seated Bored Piles in Bangkok Subsoil, Proceedings of sessions of
the Geosupport Conference: Innovation and Cooperation in the Geo-Industry,
Geotechnical Special Publication No. 124, ASCE, January 29-31, 2004, Orlando,
Florida, USA, p.p. 561-572.
Thasnanipan Narong, Zaw Zaw Aye, Thayanan Boonyarak and Natamon Kampananon
(2004b), Quantifying Pile Head Condition Before Basement Excavation by Cross-hole
Sonic Logging Tests, The Seventh International Conference on Application of Stress-
Wave Theory to Piles, Kuala Lumpur, Malaysia, August 9 – 11, 2004. p.p. 233-238.
Thasnanipan N., Manosuthisan C., Boonyarak T., Chanchad C., (2006) Experience in
Construction of Largest Diameter Deep-seated Bored Piles at the Bank of Chaopraya
River, Proceedings of the Eleventh National Convention on Civil Engineering, Phuket,
April 20 – 22, 2006
244
PROCEEDINGS OF THE THIRTEENTH
SOUTHEAST ASIAN GEOTECHNICAL
CONFERENCE
November 16-20 1998, Taipei, ROC
Performance of a Braced Excavation in Bangkok Clay,

Diaphragm Wall Subject to Unbalanced Loading Conditions
N. Thasnanipan, A. W. Maung, P. Tanseng and S.H. Wei
Sponsored by;
Chinese Institute of Civil and Hydraulic Engineering

Southeast Asian Geotechnical Society
245
Thirteenth Southeast Asian Geotechnical Conference, 16-20 November, 1998, Taipei, Taiwan, ROC
Performance of A Braced Excavation in Bangkok Clay,

Diaphragm Wall Subject to Unbalanced Loading Conditions
N. THASNANIPAN SEAFCO Co., Ltd., Bangkok, Thailand
A. W. MAUNG SEAFCO Co., Ltd., Bangkok, Thailand
P. TANSENG SEAFCO Co., Ltd., Bangkok, Thailand
S.H. WEI SEAFCO Co., Ltd., Bangkok, Thailand
SYNOPSIS 0.8m thick cast-in-situ diaphragm wall having toe depth of 28m with two level temporary bracing was used
to construct the basements of a structure which is located near the river and surrounded by buildings, including a historical one
in Bangkok, Thailand. Due to the site condition, unbalanced lateral loading on the wall was expected and an excavation down
to -12.7m below the existing ground level was carried out with instrumentation, consisting of (8) inclinometer tubes installed
in the wall panels, settlement plates around excavation zone and tiltmeters and beam sensors on the existing structures. This
paper presents computer model analysis and performance of the wall including results of instrumentation. Behaviour and
performance of the wall is compared with those of other projects in Bangkok area.
INTRODUCTION
In Bangkok, the growing land price and need for space has
necessitated deeper and larger basement excavations, even in
some unfavorable subsoil and site conditions and in limited
spaces. Subsoil conditions in Bangkok is generally a very
soft clay of 12m to 18m thick layer underlain by stiff to hard
clay and series of sand layers. Excavation in such soft soil
requires efficient retaining structures and cast-in-situ
diaphragm walls have therefore come in use frequently.
This paper presents performance of a bracing excavation
with diaphragm wall adjacent to the river and surrounding
structures, including a historical building. This historical
building, having archeological and cultural values, not only
limited the height of the building but also influenced the
construction time. Since the location of planned building is
in a current limited height zone in which up to 4 storey
height is permitted, 3 level basement was included to
increase the usable floor area. Fig. 1 Layout of project site and instrumentation
3.0m in depth, sloping towards the mid-stream to a depth of

SITE CONDITION AND SUBSOIL about 10m to 12m. River water level is about 1.6m below
the ground level during dry season and sometimes in the
The building site is located nearby the Chao Phaya River rainy season is above the ground level, causing inundation.
and surrounded by a historical building and other existing
structures supported by pile foundation (Fig. 1). Accurate A primary site investigation was carried out by drilling two
information on foundation of these existing buildings was boreholes and two field vane shear tests. Prior to design for
not available. The diaphragm wall along the river (D1) was bracing and basement excavation, drilling of additional two
constructed about only 4.0m away from the existing old boreholes and two field vane shear tests were performed to
river wall. The river bed near the river wall is about 2.2m- check variability of subsoil conditions. The subsoil
655
247
0m 6m 10m 20m
SPT-N Value (blows/ft) Su (t/m2)

0 20 40 60 80 100 0.0 2.0 4.0
0
Weathered
RIVER Clay
5
Soft to
RIVER WALL
Very Soft
10 Clay
Med. Clay
15
FOUNDATION 0.80 THK.
PILES
D-WALL Stiff to
0.80 THK.
20 D-WALL Very Stiff
Clay
25
Dense to
Very
30
Dense
Sand
Fig. 2 Basement section and soil profile
properties obtained from the boreholes and test data were For the wall sections adjacent to floor slab openings,
summarized in Table 1. especially in the water storage area, additional
reinforcements were provided for bending stress in
Table 1. Summary of subsoil properties horizontal direction.
Layer W ?s cu SPT
Soil Type Top in PILE FOUNDATIONS
Depth % kN/m 3
kPa N
m
Soft Clay 0.-3.0 35-78 16-19 30 A total of 165 bored piles consisting of 0.8m, 1.0m, 1.2m
Medium Clay 12.7 30 19 71 and 1.5m in diameter being founded in depth of 48m was
constructed to support the structure. Out of 165 piles, 5
Stiff Clay 14 22-34 19-21 43-300 14-52
numbers of 0.8m diameter piles and 8 numbers of 1.0m
Dense Sand 25 14-25 20-23 - >39
diameter piles were incorporated with diaphragm wall panels
as a leg to carry the load transferred through the wall.
DIAPHRAGM WALL
INSTRUMENTATION
A 800mm thick diaphragm wall was designed for excavation
down to 12.7m below the ground with two levels of Due to the locality and particular conditions of the site,
temporary bracing. The diaphragm wall toe was embedded more instrumentation than those used by other projects in
down to 28m to achieve the overall stability of the Bangkok were installed and monitored. Three types of
excavation which is right on the river bank (Fig. 2). instrumentation - Eight inclinometer tubes were installed in
Originally excavation stages were modeled by using a one- diaphragm wall panels to monitor displacement of wall, ten
dimensional finite element computer program (Nonlinear settlement plates of 1.0m and 5.0m in depth around
Beam Column Analysis). Soil elements were modeled as excavation zone, and ten tiltmeters and four vertical beam
spring and wall elements were modeled as beam. Four types sensors on surrounding buildings were installed to monitor
of wall were designed to suit the temporary construction tilting. Layout of instrumentation is presented in Fig. 1 .
stages and permanent conditions. The walls were reinforced
to withstand bending stresses up to 1000 kN.m/m in vertical
direction. The maximum movement of the wall was
expected to be 24.2mm.
656
248
BASEMENT EXCAVATION
The principal steps of basement excavation and construction

sequence are described below;
1. Construct capping beam and excavation to 2.5m below
ground level.
2. Install first level bracing at -1.5m and pre-load the struts
3. Excavate down to -7.0m
4. Install second level bracing at -6.5m and pre-load the
struts Fig. 3 Deformed mesh of two-dimensional model
5. Continue excavation down to the final depth at -12.7m
6. Construct mat foundation (Basement 3), Basement 2 and excavation level, the river water level rose to a maximum
remove second level bracing level about 0.5m above the ground level.
7. Construct Basement 1 and remove the first level bracing
Since the excavation work is located adjacent to the river TEMPORARY BRACING SYSTEM
and surrounded by old buildings, the diaphragm walls are
subject to three different lateral load conditions resulting A simple cross-lot bracing system with continuous wale
from (1) full depth of the earth, (2) steep downward slope of beams was used to support the wall during basement
riverbed, and (3) full depth of earth with possible surcharges construction stage (Fig. 1). 20m long steel king posts of
from the adjacent buildings. In particular, the walls H300x300 sections were driven into stiff clay to support the
alongside the river (D1) and the opposite walls (D2), were bracing system and temporary deck for excavation and
expected to undergo an unbalanced loading condition. construction equipment. A summary of steel sections used
During temporary bracing design, two dimensional in bracing system is presented in Table 2.
computer modeling were carried out to study the effects of
unbalanced lateral loading on the wall and bracing. An Table 2. Summary of temporary bracing system
additional two dimensional model analysis was carried out
prior to designing a temporary bracing. The model analysis
Bracing Wale Beam Strut Expected
indicated relatively less movement of Wall D1 towards
Level Force
excavation (Fig. 3). However, the following measures were
(kN/m)
adopted in excavation work to prevent potentially adverse
First 1xWF400x400 WF350x350 484.0
behavior of wall D1 and to keep the lateral wall movements
within tolerable limits; Second 2xWf400x400 WF400x400 789.0
1. Using a simple, but efficient temporary bracing system

2. Pre-loading of the struts (200kN/m and 400kN/m for first Between wale beams and wall, lean mixed concrete were
and second levels respectively) on the one end of the poured into the gaps between wale beams and wall to
strut on Wall D2 only (Fig. 1). achieve a good load transfer. Pre-loading was carried out
3. Excavating the soil in front of Wall D1 side first at any using hydraulic jacks. During pre-loading, the movements
excavation stage of the strut were measured. Horizontal movements of
4. Frequent monitoring of wall movements 0.82mm to 9.06mm and 0.15mm to 29.83mm in the
5. Minimizing construction time direction of jacking for first level struts and second level
struts respectively were recorded. For second level struts,
During the initial excavation to 2.5m for installation of first the vertical movement of up to 1.68mm was measured at the
level bracing, a historic foundation was unexpectively jacking position.
discovered. Excavation was suspended for about 3 months
and resumed after further excavation was permitted by the
archaeological department. INSTRUMENTATION RESULTS
Excavation for final depth was made in the rainy season
and a berm made of sand bags for flood protection was Inclinometer Monitoring - Wall movements measured from
constructed around the perimeter of the wall. After lean inclinometers and predicted displacements by one-
mixed blinding concrete had been cast at the final dimensional finite element analysis are shown in Fig. 4.
Movements of Wall D1 and a section of Wall D3 close to
the river, near the garden are found to be cantilever shape
indicated by I-6, I-7 and I-8 respectively while movements
657
249
of Walls D2 and D3 have inward bulging shape indicated by within the range (about 0.1% to 1.2% of excavation depth)
I-1 to I-5. of projects completed in Bangkok area.
(Max. Wall Movement.)/(Excavation Depth), %

Displacement (mm) Displacement (mm)
0 20 40 40 20 0
3.0
0.00
-1.50
Excavate to 7.0m
RIVER
2.5
Excavate to 12.7m
5.00
-6.50
2.0
10.00
-12.70
0.9
Depth (m)
1.5
15.00 1.0
1.1
PREDICTED
OBSERVATION 1.0
20.00
1.4
0.64
0.5 2.0
25.00 0.42
3.0 0.24
0.14
Inclinometer Inclinometer
30.00
No. I-7 No. I-2 0.0
10 100 1000 10000
System Stiffness, (EI)/(γw/h4)
0 10 20 30 40 40 30 20 10 0
Fig. 5 Wall movement and excavation depth ratio

0.00
-1.50
5.00
-6.50
Settlement Monitoring - Ground settlement of up to 16mm
10.00 was observed after completion of excavation works. The
-12.70
maximum settlement was found to be near the location of
Depth (m)
15.00 the inclinometer I-6 which indicates the largest lateral

PREDICTED movement at the top of the wall.
20.00 OBSERVATION
Tiltmeter and Beam Sensor Monitoring - Readings of

tiltmeter and beam sensors installed in the surrounding
25.00
30.00
Inclinometer
No. I-3
Inclinometer
No. I-5
structures are shown with the progress of excavation in Fig.
6. A comparison between results from tiltmeter and beam
Fig. 4 Predicted and measured lateral wall movements sensor and typical values for maximum building slope or
settlement for damage risk assessment is presented in Table
Generally the predicted and measured lateral wall 3.
movements are in a good agreement, except for the top
portion of wall in which the measured wall movements
exceed the predicted movements. The differences in wall Table 3. Comparison between observed values and
movement were found to be caused by the following; maximum damage risk assessment (Lake et. al.))
Max. Slope Max.

1. Advancement of excavation in front of Wall D1 further of Settlement Description
than other walls, allowing Wall D1 to stand longer prior Building of building
to strut installation than the others (mm)
2. Wall D1 was in cantilever condition for about 3 months Risk <1/500 <10 Negligible : superficial
after initial excavation for first level bracing. Category 1 damage unlikely
3. No direct nor immediate pre-loading on Wall D1. Tiltmeter 1/2082* -
4. Slight over excavation to install first level strut for all
Walls. Vertical 1/3165* -
Beam Sensor
Risk 1/500 10 - 50 Slight: possible
The walls are in fix end conditions as fixity of walls Category 2 superficial damage
indicated by all inclinometer reading are found to be at depth which is unlikely to
of about 20.0m which is 8m above the walls’ toe depth. have structural
Maximum lateral wall movements after installation of significant
struts were in the range of 0.14% to 0.64% of the * monitoring results for the project presented in this paper
corresponding excavation depths (Fig. 5) and were also
658
250
No building settlements nor cracks on the buildings were DISCUSSION
observed till completion of the basement construction. The
buildings were evaluated to be in risk category 1. The risk Pre-loading is an effective way to reduce further wall
assessment and monitoring results confirm a good movement after strut installation by providing a good intact
performance of the braced excavation. between wall and supporting system. This is clearly seen in
measured wall movements compared between I-7 and I-2,
0 50 100
Time (days)
150 200 250 and I-3 and I-5 which are located in the walls generally
0
facing each other (Fig. 4 and 1)
Depth (m)
5 Install 1st strut
10
Install 2nd strut Delay in strut installation and construction time affect the
wall movements, especially at initial excavation stage before
Excav. near the river
Excav. away from the river
15
installation of first level bracing.

0.10
0.05
∆T T-1 (B3)
T-2 (B3)
Wall D1 which is alongside the river was found to be
∆T (mm)
0.00
T-3 (B3)
T-4 (B3)
T-5 (B3)
TILTMETER
minimally affected by the unbalanced load conditions due to
deep embedment (about 8m below the fixity) of the wall and
T-6 (B2)
T-7 (B1)
T-8 (B1)
-0.05 T-9 (B1)
T-10 (B1)
presence of the old river wall. This was also shown by the
-0.10
0 50 100 150 200 250 two dimensional analysis (Fig. 3).
0.50
EL-1 (5th Floor of B3)
∆EL
The predicted movements of the wall in cantilever condition

0.25 EL-4 (2nd Floor of B1)
∆EL (mm)
EL-5 (2nd Floor of B1)
0.00
VERTICAL
BEAM
at initial excavation stage in soft clay layer is considerably
-0.25
SENSOR
small due to the high modulus of soft clay adopted. In this
case, construction time and sequence of excavation needed
-0.50
0 50 100 150 200 250 to be strictly controlled to keep the wall within the
70 movement limit predicted by the one dimensional analysis.
Max. Horiz. Movement (mm)
I-1
I-2
60 I-3
I-4
I-5
50 I-6
40
I-7
I-8 Two dimensional analysis with facilities to model the
30
20
INCLINOMETER
geometry of site is required for the diaphragm wall subject
10 to unbalanced lateral loading conditions.
0
0 50 100 150 200 250
Fig. 6 Tilting of adjacent buildings and horizontal CONCLUSION

movement of diaphragm wall with construction time
Braced excavation using diaphragm wall subject to
unbalanced loading due to adjacent river and surrounding
PLANE STRAIN BACK ANALYSIS buildings, was successfully achieved with proper
instrumentation and monitoring.
A two-dimensional analysis was carried out by using finite
element program named PLAXIS (Veermeer and Performance of the wall based on the instrumentation results
Brinkgreve, 1995). A trial and error procedure in which the are presented and discussed.
values of undrained Young’s modulus (Eu) were adjusted
iteratively was adopted to obtain the best fit between Back analysis was carried out with two dimensional
calculated and measured lateral wall movements. The modeling to determine the soil modulus. The soil modulus
excavation stages were modeled by using plane strain obtained from back analysis and the modulus adopted in
analysis. Mohr-Coulomb constitutive model was used for wall designed were compared.
clay and sand. Walls and strut were modeled as plate
element and elastic-perfectly plastic spring element
respectively. Pre-loading in struts are also included in the ACKNOWLEDGEMENT
analysis.
Results of analysis show that modulus value of soft clay is The authors express their appreciation to the colleagues,
about 500Su and for stiff clay modulus value is about especially to Mr. Ganeshan Baskaran and Mr. Muhammad
2,000Su. For one-dimensional analysis, the spring stiffness Ashfaq Anwar for their invaluable suggestion and assistance
used is equivalent to Eu=800Su for both soft clay and stiff in the preparation of this paper. Initial analysis works
clay. carried out by Dr. Vichai Vitayasupakorn and Mr.Young
Zou are acknowledged. 659
251
REFERENCES
Lake L. M., Rankin W. J. and Hawley J. (1996) “Prediction

and Effects of Ground Movements Caused by Tunnelling
in Soft Ground beneath Urban Area” CIRIA Report No.
30, Construction Industry Research and Information
Association, UK.
Vermeer, P. A. and Brinkgreve, R. B. .J. (1995). “Finite

Element Code for Soil and Rock Analyses, Balkema,
Rotterdam.
Mana I. A. and Clough G. W. (1981) “Prediction of

Movements for Braced Cuts in Clay”, Journal of
Geotechnical Engineering Division, ASCE, Vol. 107,
No. GT6, pp. 36-42.
Gan C. H. (1997). “Review and Analysis of Ground

Movements of Braced Excavation in Bangkok Subsoil
using Diaphragm Walls”, M. Eng Thesis, Asian Institute
of Technology , Bangkok, Thailand.
Tanseng, P. (1997). “Instrumented Deep Excavations in

Bangkok Subsoil ”, M. Eng Thesis, Asian Institute of
Technology, Bangkok, Thailand.
660
252

Engineering
Prediction and performances of short embedded cast

in-situ diaphragm wall for deep excavation in Bangkok
subsoil
Thasnanipan N. , Maung A. W. and Wei S. H.
Teparaksa W.
Rolla, Missouri
253
686
Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri,
March 9-12, 1998.
PREDICTION AND PERFORMANCES OF SHORT EMBEDDED CAST IN-SITU

DIAPHRAGM WALL FOR
DEEP EXCAVATION IN BANGKOK SUBSOIL
Dr. Wanchai Teparaksa Narong Thasnanipan Aung Win Maung Wei Shixin Paper No. 5.04
Civil Engineering Dept. SEAFCO CO. Ltd. SEAFCO Co., Ltd. SEAFCO Co., Ltd.
Chulalongkorn University Bangkok 10320 Bangkok 10320 Bangkok 10320
Bangkok Thailand Thailand Thailand Thailand
ABSTRACT
This paper presents a case history of performances and predictions of a diaphragm wall which is subject to serve differently from its
initial purpose for which the wall was constructed. The updated requirements associated with the additional excavation created a
situation whereby the wall would become short embedded and behave like free-end supported wall. As the wall had been already
constructed during the time of modification it was only possible to modify the excavation method and bracing system. Accordingly,
the wall stability and possible toe movement were analyzed, to accommodate the updated requirements, by using finite element
methods. Continuous inclinometer monitoring has been carried out during the construction and results are being compared and
analyzed for predicted values. Performance of the wall based on comparison between the inclinometer monitoring results during
different excavation stages and predicted results, are discussed. It has been found that excavation depth for first bracing layer and
construction period are very important for diaphragm wall performance. Large initial movements in the wall strongly influence the
wall movements in the successive excavation works. Construction practice plays a major role in deep excavation work.
KEY WORDS
Diaphragm wall, Lateral movements, Inclinometer, Settlements, Free-end, Bracing, Excavation, Critical height, Cantilever
255
687
INTRODUCTION
Economic growth in Thailand during the past decade has been

forcing the construction industry to look for all feasible
methods to enhance the infrastructure of the country
especially for the Bangkok city. Huge number of high-rise
buildings with multilevel basements are being erected within a
short time possible to accommodate the booming economy.
Value of urban land is increasing in many folds beyond the
limits. Unlike in developed countries, the building rules and
regulations of this city are only in documented form and
hardly established all in practice in a short time. As a result,
developers and owners enjoy relatively more freedom to
modify the structural layout, even during and after
construction stages, to facilitate their growing demands which
are being conceptualized within a short time duration in the
fast developing city, aiming to maximize the area utilization. Fig. 1 Soil Profile at the Site
The project reported in this paper is an example for such case.
The soil properties used for Diaphragm wall design are
The project site is located in the vicinity of central business summarised in the table below
area of Bangkok. The structure is a multi-storied building for
a business complex. The load of the structure would be Table 1 Unit weight and Undrained shear strength of soil
carried by bored cast-in-situ bored piles and barrette piles with layers.
toe depth of 60m below the ground level, embedded into the
second sand layer of the Bangkok subsoil. Unit Weight Undrained Shear
Soil Type γ Strength
The constructed cast in-situ diaphragm wall is 0.8m thick and (t/m3) Su (t/m2)
the toe is embedded into stiff clay at a depth 18m below the Soft Clay CH 1.6 1-2.9
ground level. Initially the wall was designed for construction
Medium Clay CH 1.93 3.5
of four basements including mat foundation of 3.5m thick.
The maximum excavation depth was set to be 14m and a three Stiff Clay CH 2.0 13.7-17.95
level of temporary bracing system was considered to construct
the basements with conventional bottom up construction
method.
SITE CONDITION
Embedded depth below the final excavation for the mat
foundation was designed to be 4m. After construction of the The project site is located in the corner of a main road and a
wall it was decided to increase the number of basements to be street of the central business district in Bangkok. The planned
five for part of the building area (Tower area) and excavation building is facing both the main road and street. The building
depth needed to be increased by 1.8m from the originally plan was set out in accordance with building code having a
designed depth, leaving the wall embedment depth of 2.2m setback of 6m from an adjacent property boundaries. Behind
only. it are two story buildings located more than 10m away. Deep
excavation is ideal in such site condition. Figure 2 shows the
layout plan of the building.
SUBSOIL CONDITION
The sub soil conditions at the project site is not uncommon to DIAPHRAGM WALL DESIGN
usual conditions observed in Bangkok comprising a 15m thick
soft to medium clay layer on the top underlain by stiff clay to Originally diaphragm wall was designed with a 3 level
a depth of about 22m. Below the stiff clay is a series of temporary bracing to allow a maximum excavation depth of
alternating layers of hard silty clay, sandy clay and dense 14m below the existing ground for construction of mat
sand. foundation (see Fig. 3). Computer programs of finite element
methods (WALLAP and CRISP) and finite difference method
(FLAC) were used to simulate the staged excavation and
predict the wall behaviour and performance. In the computer
modeling, undrained strength of soil with total stress
parameters derived from laboratory and field test results were
256
688
used. 0.8m thick Diaphragm wall with toe depth of 18m was
found to be adequate enough for the planned basement
construction in such soil condition.
Fig. 3 The Original Construction Sequence with three

temporary bracing levels
MODIFIED CONSTRUCTION SEQUENCE

Fig. 2 Layout Plan of Diaphragm Wall & Piles
Finite element computer modeling was again carried out for
Estimated maximum bending moment of 90 t-m/m on the the desired excavation depth. The model analysis showed that
excavation side and 56 t-m/m on soil side of the wall. The 4 level temporary bracing was necessary with strut pre-
maximum deflection of the wall predicted was about 35- loading. It also indicated the diaphragm wall in free-end
37mm with some degrees of toe movement up to 11mm. condition with some degrees of toe movement in a range of
23mm. The maximum predicted wall movement at the top was
Bored piles of 1.20m in diameter were in-cooperated within 77mm.
and below the toe of the wall to support the column load and
dead load of the wall. Foundation piles (dia. 1.20m), To prevent progressive toe movement with time, followings
including bored piles and barrette piles (2.70mx0.80m) were were recommended for excavation works.
founded at the depth of 60m in dense sand layer. Four
inclinometer tubes I-1 to I-4 (20m to 25m in length) were (i) frequent monitoring of diaphragm wall movements when
installed in the wall panels with bored pile legs to observe the excavation close to the final excavation depth.
behaviour and performance of the wall.
(ii) casting a reinforced lean concrete slab with a thickness of
Diaphragm wall and foundation piles were constructed in 20cm and 6m wide in front of the wall just after reaching the
parallel. After completion of diaphragm wall construction, excavation depth of -13.8m.
but construction of bored piles were still in progress, overall
design of the building was revised to suit the requirements of (iii) excavation to -15.8m must be carried out leaving a 12m
the property market. Number of basement floors was wide soil berm with lean mixed concrete slab on top of the
increased from 4 to 5 in the main tower area. As a result, the berm all around the wall.
diaphragm wall constructed was deemed to support for the
maximum excavation depth of 15.8m. Apart from four inclinometer tubes installed in the wall, a
total of 16 sets of settlement observation stations were
Firstly diaphragm wall design and construction stages were installed around the wall perimeter. Each station comprised
then reviewed. Since the wall has been constructed, two to three settlement points located at 3m intervals and 3m
excavation sequence and bracing system can only be modified away from the wall.
so that the modified deeper excavation can be done without
impairing the structure and stability of the wall.
257
689
in the soil of active side about 2-3m away from the wall.
After installation of struts and pre-loading them, the lateral
movements of the wall ceased off. An increment of 2mm-
4mm was recorded with further excavation stages (See Fig. 5).
Struts from first to fourth levels were pre-loaded 40 t/m, 75
t/m, 75 t/m and 40 t/m respectively, 20%-25% of design load
of bracing system, prior to further excavation. The pre-
loading of the strut was very effective to reduce lateral
movement of the wall and ensuring good intact between wall
and bracing system.
Fig. 4 The Modified Construction Sequence
The risk of over excavation was carefully assessed not only

for diaphragm wall but also for the existing buildings near by
the site. Review of the site condition indicated that the
existing buildings surrounding the project site were more than
10m away from the excavation zone and they were found
subject to no risk.
Fig. 5 Maximum lateral movements at the top of wall vs.
Construction Time.
BRACING SYSTEM
The bracing system included continuous waling beams along

the diaphragm wall and longitudinal and transverse struts.
The spacing of strut was generally 6.4m to 6.8m. Table no. 2
shows the type of wide flange structural steels used for the
bracing.
Table 2 Structural steel sections used for temporary bracing

Bracing Strut Wale
Layer
1st 2W300x300x94kg/m 2-W300x300x94kg/m
2nd 2W350x350x137kg/m 2W350x350x137kg/m
rd
3 2W350x350x137kg/m 2W350x350x137kg/m
4th W350x350x137kg/m 2W350x350x137kg/m
LATERAL MOVEMENTS OF THE WALL
Monitoring of wall movements was carried out by reading the

inclinometer tubes profiles installed in the wall at every major
construction stages. The wall constructed was designed for Fig. 6A. Lateral Movements of the wall shown by
initial excavation of -2.5m to install the first bracing layer. Inclinometer Tubes I-1 and I-2.
However the excavation has been done to about -3.0m to -
3.2m at once and a movement of 40mm was observed. Time The pattern of movements of the wall shown by the
delay during construction of capping beam and over inclinometer readings is similar to that of cantilever condition.
excavation deeper than the critical height (Hcr = 2C/γ) was Because of large movements were induced Initially, braced
unsupported around three months resulted an increase in wall mode was not developed. After reaching the final excavation
movements to 55mm -70mm. Tension cracks were developed depth the observed movements of the wall at the toe level
258
690
were in a range of 1.2mm to 16mm while the predicted toe
movement was about 23mm. Predicted and observed wall
movements is tabulated as below;
Table 3 Predicted and observed wall movements
Location Predicted Predicted Observed

on the for 14.0m for 15.8m on Site
Wall Excavation Excavation (mm)
(mm) (mm)
Top 35-37 77 55-98
Toe 11 23 1.2-16
The maximum lateral movements were found in I-4. At the

final excavation stage, movements of the wall increased
considerably below the first bracing level down to toe of the
wall. It was noted that in I-4 area the longitudinal and Fig. 7 Relationship between maximum ground settlements
transverse struts of the bracing system become diagonal to the and maximum lateral wall movements
wall sections due to the geometry of wall layout. The stiffness
of the bracing in this particular zone is considered to be much ANALYSIS ON PERFORMANCE OF THE WALL
less than that of others.
Based on the deformation of the wall measured from the
The wall’s fixity was found to be at depth 20m-22.5m from inclinometer readings, the bending moments were computed
inclinometer readings when the final excavation reached. from the well known equation to check structural performance
of the wall
M=Ec.Ief.[dS/dy] (1)
where Ec = Modulus of Elasticity of concrete

Ief = Effective moment of inertia of the wall section
S = slope given by inclinometer reading
y = elevation of the slope given
Generally the bending moments computed were within the

design moments and the performance of the wall is found to
be satisfactory. However a direct comparison between
bending moment envelopes from computer simulation and
bending moments derived from inclinometer readings shows
some differences, especially at the strut levels and below the
depth of 12m. These differences can be due to the following;
1. Stiffness of medium to stiff clay layer used based on the

field and laboratory tests for simulation is much lower than
the actual in-situ soil stiffness. In simulation for the wall
design, influence on soil stiffness below the final excavation
level by bored piles was not considered.
2. Efficiency and stiffness of bracing system, including pre-

Fig. 6B Lateral movements of the wall shown by Inclinometer loading forces. In the original design calculation, much lower
Tubes I-3 and I-4. pre-loading forces were used.
From the settlement observation data, it was found that the 3. large initial movements during the wall in cantilever
maximum ground settlements behind the wall were condition.
approximately 34% to 50% of the lateral wall movements,
except for the area near the I-4 where the ground settlements 4. Inclinometer tubes (20-25m in length) were installed in
equal the lateral wall movements. diaphragm wall panels with bored pile legs. The bored pile
259
691
legs provide the fixity of the wall and increase the bending The relationship between wall movement and system stiffness
moments at a location near the bottom of the wall. (Fig. 9) shows that because of large movements during the
wall is in cantilever condition, the excavation to 5.5m with
5. Moment redistribution in the wall is considered not fully one bracing deviates from the normal trend of braced
activated. excavation.
Fig. 10 Relationship between Normalized Maximum lateral

movement versus factor of safety against basal heave.
The normalized lateral movements of the wall shown in Fig.

10 is within the range of 0.5-1.5% which is slightly higher
Fig 8 Design envelope of bending moments from computer than those of similar projects in Bangkok subsoil. The higher
modeling and bending moments computed from inclinometer movement ratio occurred during excavation in the soft clay
readings at final excavation layer with slight over excavation and delay in first strut
installation. It is observed that the first excavation depth and
duration in bracing layer installation are very critical for the
performance of the wall at further excavation stage.
CONCLUSION
1. Movements in first excavation stage is governing the

performance of Diaphragm wall for further excavation.
2. First excavation deeper than the critical height of the soil

needs to be supported immediately.
3. Apart from support system and soil properties,

construction practice influenced the movement of the wall in
this particular project.
4. A successful deep excavation with short embedded

diaphragm wall has been achieved using a combination of
prediction by computer simulations and instrumented field
Fig. 9 Relationship amongst system stiffness, non- observations.
dimensionalized wall movement and factor of safety against
basal heave .
260
ACKNOWLEDGEMENT
692
especially to Mr. Ganeshan Baskaran, Mr. Young Zou, Mr.
Muhammad Ashfaq Anwar, Mr. Pornpot Tanseng and Mr.
Chetsadar Plongkrathok for their assistance in the preparation
of this paper.
REFERENCES
Balasubramaniam A.S., Bergado D.T., Chai J.C. and Sutabutr

T. [1994]. “Deformation Analysis of Deep Excavations in
Bangkok Subsoils” XIII Internation Conference on Soil
Mechanic and Foundation Engineering , New Delhi, India. pp.
909-914.
Day P. W. [1994]. “Factors Influencing the Movement of

Retaining Structures” XIII Internation Conference on Soil
Mechanic and Foundation Engineering , New Delhi, India. pp.
109-114.
Gan C. H. [1997]. “Review and Analysis of Ground

Movements of Braced Excavation in Bangkok Subsoil using
Diaphragm Walls.” M. Eng Thesis, Asian Institute of
Technology. Bangkok, Thailand.
Mana I. A. and Clough G. W. [1981] “Prediction of

Movements for Braced Cuts in Clay”, Journal of Geotechnical
Engineering Division, ASCE, Vol. 107, No. GT6, pp. 36-42.
Peck R. B. [1969] “Deep Excavation and Tunneling in Soft

Ground. Proceeding of 7th International Conference on Soil
Mechanics and Foundation Engineering, Mexico. State of Art
Volume, pp. 225-290.
Saxena S. K. [1975]. “Measured performance of a rigid

concrete wall at the World Trade Centre.” Diaphragm Walls
and Anchorages. Institution of Civil Engineers, London, UK.
pp. 107-112.
261
PROCEEDINGS OF THE 5TH INTERNATIONAL SYMPOSIUM ON FILED MEASUREMENTS
IN GEOMECHANICS – FMGM99/SINGAPORE/ 1 – 3 DECEMBER/ 1999
Filed Measurements
In Geomechanics
Edited by
C.F.Leung, S.A.Tan & K.K.Phoon
Department of Civil Engineering, National University of Singapore
Lessons from the collapse during construction of an inlet

pumping station : Geotechnical instrumentation aspects
Teparaksa W.,
Thasnanipan N., Maung A.W. and Tangseng P.

263
Field Measurements in Geomechanics, Leung, Tan & Phoon (eds) © 1999 Balkema, Rotterdam, ISBN 90 5809 066 3
Lessons from the collapse during construction of an inlet pumping station :

Geotechnical instrumentation aspect
W. Teparaksa
Department of Civil engineering, Chulalongkorn University,Bangkok, Thailand

ABSTRACT: In the wastewater treatment plant, an Inlet Pumping Station (IPS) is one of the most important
structures to collect and pump all wastewater to the treatment plant. The IPS presented in this paper consists
of 20m deep underground chambers. A reinforced concrete diaphragm wall system 1.0m thick and 25m. deep
was used in construction of the IPS. Four Inclinometer tubes were installed in the diaphragm wall to monitor
the lateral wall movements during excavation work for basement construction. When excavation approached
the final depth, excessive lateral wall movements were observed, but they were not taken seriously. Later,
some cracks on the capping beam and noise from the bracing system were noticed. Then the IPS collapsed
and was buried under sliding soil mass. At the time IPS collapsed five main bracing layers and only two of
the five intermediate bracing layers had been installed. This paper emphasizes the importance of instrumenta-
tion. Additionally it presents the results of investigation into the causes of the collapse, as well as the recon-
struction work.
1. INTRODUCTION Canal Wall

nal
Ca
ng
isti
Bangkok metropolis is sectioned into four parts for Ex Pump Sta
tion Reconstructed
Canal
the implimentation of four mega projects, or Inlet Pumping

Station (IPS)
"phases" that comprise a wastewater treatment sys-
Chao Phraya River

Storm Pumping
tem to serve the rapidly growing industrial and do-

Station (SPS)
mestic needs. Three of four phases are under con-

struction. All projects are a turnkey system where
the main contractors are responsible for both design
Treatment Plant
and construction. The engineers will be the consult- (STP)
ants as representatives of the owner to review the
conceptual and detailed design and to supervise the
construction. The aim of this paper is to highlight
the importance of instrumentation in deep under-
ground construction, as proved by the collapse of Figure 1. Layout of Inlet Pumping Station
the Inlet Pumping Station (IPS) at the second phase.
This paper also briefly describes its reconstruction.
The IPS has two distinct chambers, a pumping
chamber and an inlet/bypass chamber, all built un-
2. PROJECT DESCRIPTIONS derground. The pumping chamber is a circular shaft
while the inlet/bypass chamber forms a box on the
The second phase Bangkok wastewater project pre- north side of the pumping chamber. The pumping
sented herein was started in 1995 to serve the cen- chamber is 20.3m in diameter and 20.2m deep. It
tral area of Bangkok City. The project was planned consists of a central shaft 8.75m. in diameter and an
to be completed by 1999 (ie. within 4 years). In the annulus 4.875m. wide. The inlet chamber is essen-
sewage treatment plant (STP), the IPS is one of the tially a large box receiving the inlet sewers and also
most important structures to collect and pump all housing screens for removal of debris. The inlet
wastewater to the sewage treatment plant. Figure 1 chamber is 11.5m x 9.0m approximately in plan.
presents the location of IPS on the sewage treatment The bypass chamber is adjacent to an inlet chamber
plant (STP) site. and divided by a concrete wall from base level at
247
265
19.5m below ground to roof level. The inlet/bypass Legend
-2.20
chamber has three levels. The base level is at Main strut +Diagonal Strut
Intermediate Strut -3.95
19.50m below the ground. Levels 1 and 2 are at -5.70
10.5m and 5.5m depth respectively. The roof slab is I-2 -7.95
at 1.0m depth. A
-10.20
The IPS was located at the northwest corner of -11.95

I-1
the STP site, approximately four to ten meters from -13.70
I-3 B
a canal. The canal is protected by concrete walls Diagonal
-15.45
supported by battered concrete piles. Private resi-

Struts -17.20
dences consisting of wooden house were located

-19.10
Main Strut +
Intermediate Strut
close to the property line in the west. B+ -21.0 -21.0
I-4 I-3
Space limitation necessitated reinforced concrete A+ SECTION A-A SECTION B-B
diaphragm walls 1.0m thick and about 25.0m deep

as permanent structures of the IPS. The diaphragm Figure 2. Temporary bracing system (Initial design)
walls were also used as temporary structures during
excavations up to 22.0m deep. The one-meter-thick
wall was cast in panels with lengths of around 2.5 -
3.0m using the bentonite slurry trench construction
method. The connection between adjoining panels
was constructed by using U-shaped stop-ends with
water bars.
For excavation work, five levels of temporary
steel bracing were designed to support the straight
wall portion and an additional five intermediate
bracing layers was planned to frame the edges
where the straight wall meets the circular enclosure
(Fig. 2). Four inclinometer tubes were installed at
critical points of the diaphragm wall to observe lat-
eral movements during excavation for underground
construction. Figure 3. Final Temporary bracing system (after Kanok-
The bracing system was classified as temporary Nukulchai et al., 1998).
and thus no submission was made for review. The
final temporary bracing arrangement (Fig. 3) was
Table 1. Summary of soil properties
changed from the initial design in Figure 2. In the
initial design, two lacing beams shown as a dashe Depth m Soil γt Su
line in Figure 2 were planned to strengthen the strut- from to Description kN/m3 kN/m2
ting system. However these were cancelled in the 0.0 3.0 Fill 18.0 -
final temporary bracing system (Fig. 3). 3.0 16.0 Soft Clay 16.0 12-30
16.0 20.0 Stiff Silty Clay 19.0 100-150
20.0 30.0 Very Stiff Silty Clay 20.0 175-250
3. SOIL CONDITIONS >30 Dense Silty Sand 20.0 -
At the start of the project, the site consisted of very

soft clay (mud) about 2 - 3m deep. The soft clay 4. COLLAPSE OF INLET PUMPING STATION
apparently extended westward, into the residential
area of the wooden houses. When work on the pro- The reinforced concrete diaphragm wall and instal-
ject commenced, mud was removed from the site up lation of four inclinometers were completed on 15th
to the property boundaries. Approximately a 3.0m January 1997 by the diaphragm wall sub-contractor.
thick of sand backfill was placed on the site. Then After completion of the diaphragm wall, the same
the soil conditions consisted of 3.0m thick sand sub-contractor pointed out to the main contractor the
backfill, followed by soft to medium dark grey clay importance of an adequate bracing system and
approximately 16.0m deep. Then stiff to very stiff staged construction sequences. The excavation and
silty clay was encountered at about 30.0m depth, strutting was carried out by another subcontractor,
overlaying the first dense silty sand layer. The sen- while the other permanent structures such as cap-
sitivity of the soft to medium clay was in the order ping ring beam were constructed by the main con-
of 2.5 to 6.0. A piezometric level was found at tractor. The capping ring beam had been casted to
about 21.0m below ground level, as a result of deep two thirds of its height by the main contractor on 8
well pumping. A summary of general soil proper- June 1997. Two inclinometer tubes (I-1 and I-2)
ties and soil stiffness is presented in Table 1. were found damaged and this was reported. But the
248
266
damaged tubes were left unrepaired and could not ments indicated that the edges of the wall between
be used for monitoring. the straight wall and the circular enclosure had been
The excavation began in mid-June, 1997. Back- under torsional stress. It was reported that tension
hoes were used for excavating inside the shaft and cracks on the capping beam were observed in late
all materials were collected with a bucket. The July 1997. I-4 recorded wall movements with a
bucket was hung from a service crane located be- maximum of 13mm (Fig. 4b). Movements of the
hind the diaphragm wall perimeter near I-1 location wall toe could not be measured due to lack of the
(collapsed zone) on the ground. The excavation and inclinometer tube penetration into the soil below the
installation of struts was carried out step by step. wall toe level.
However, only 5 diagonal bracing layers, 5 main After observation of tension cracks on the cap-
struts and two out of five intermediate struts were ping beam, immediate installation of the remaining
installed without instrumentation. intermediate struts to support the edges of the circu-
Lateral wall movements were recorded by I-3, lar enclosure and straight wall area were recom-
located at one edge of circular enclosure and I-4, lo- mended. Occurence of more cracks on the capping
cated in the intact zone (Fig.2). Wall movements in ring beam and noise from the bracing system was
A-axis direction of the inclinometer access tube I-3 noticed. At this stage excavation depth inside the
were indicative of the wall in cantilever/unsupported circular enclosure had been reached the final depth.
condition while those in B-axis direction were in- However, only two of the five intermediate bracing
dicative of the wall being supported (Fig. 4a). layers had been installed at the time the IPS col-
lapsed. Figure. 5 shows the photo of the bracing
system taken one day before collapse. The IPS col-
Displacement (mm)
-100
Displacement (mm)
-50 0 50
lapsed in the morning on 17 August 1997.
-50 0 50 100
0.00 0.00 After the collapse, soil mass caved in. Several adja-
cent wooden houses were also damaged and slipped
5.00 5.00
down (Fig. 6a). The soil mass buried the collapsed
portions of the wall, the steel bracing, and the exca-
10.00 10.00
vation equipment (Fig. 6b). Some sections of the
wall on the western side collapsed while the remain-
Depth (m)
Depth (m)
15.00 15.00
ing wall sections stood in place. The collapse also
20.00 20.00
sheared off the capping beam at the collapse loca-
20-Jun-97
tion, causing severe cracking in the capping beam
section, and opening panel joints on remaining sec-
20-Jun-97
4-Jul-97 4-Jul-97
12-Jul-97 12-Jul-97
25.00 25.00
18-Jul-97
25-Jul-97
18-Jul-97
25-Jul-97 tions of the wall. No human casualties were re-
ported in the incident as the loud noise prior to
5-Aug-97 5-Aug-97
13-Aug-97 13-Aug-97
30.00 30.00
A-Axis B-Axis the
Figure 4a. Inclinometer reading of I-3

-20 -10 0 10 -20 -10 0 10
0.00 0.00
5.00 5.00
10.00 10.00
Depth (m)
Depth (m)
15.00 15.00
20.00 20.00
27-Jun-97 27-Jun-97
25.00 4-Jul-97 25.00 4-Jul-97
12-Jul-97 12-Jul-97
25-Jul-97 25-Jul-97
5-Aug-97 5-Aug-97
30.00 30.00
A-Axis B-Axis
Figure 4b. Inclinometer reading of I-4
However, the wall movements in A-axis direc-

tion were considerably large and thus the main
beams supporting the edges of the circular enclosure Figure 5. Bracing system seen on 16, Aug. 1997 (after Thas-
seemed not effective. The two differential move- nanipan, 1997)
249
267
Figure 6a. After collapse soil mass moved in (after Thas-
nanipan, 1997) Figure 7. Underground construction of similar shape to IPS
(after Katzenbach et al. 1998)
posed walls. Phien-Wej & Sriruanthong (1999) re-

ported that at the final excavation depth in IPS re-
construction, the toe of the diaphragm wall in the
Soft Clay
collapsed zone was found in original position and
there was no evidence of basal heave.
The importance of temporary bracing has been
Stiff Clay
emphasized in various design guides, manuals and
reports by Padfied & Mair (1984), William & Waite
Figure 6b. Soil mass buried everything inside the pit. (after (1993) and Feld & Carper (1996). In the case of
Phien-Wej et al. 1999) improper temporary bracing, this issue was identi-
fied as the main cause of collapse. Results of the
analysis by means of FEM (Kanok-Nukulchai, et al.
collapse provided warning to the people working in- 1998) concluded that the main strut supporting the
side. edges of the circular enclosure was grossly over-
stressed. The report also stated that these struts
5. CAUSE OF COLLAPSE would have progressively reached their capacity at
the initiation of the collapse and may have become
The cause of collapse has been investigated by vari- totally ineffective. It was clearly indicated by the
ous parties involved. The preliminary causes of the photographs taken just one day before the collapse
collapse were considered to be: of the IPS (Fig. 8).
• Improper shape of IPS permanent structure Figure 8 shows separations of the built-up wall-
• Failure of diaphragm wall ing beams and struts, which would reduce the ca-
• Failure of internal temporary bracing during pacity of the monolithic section by 30%. The three
excavation lower intermediate struts required at the edges of the
• Failure due to improper sequence of excavation circular enclosure were not present eventhough the
• Failure due to poor workmanship in temporary excavation works reached the final depth (Fig. 5).
works This improper bracing also contributed to the cracks
Three main causes of collapse are identified; the observed on the capping beam. Additionally, lateral
improper shape of the IPS, failure of the diaphragm movements of the walls, shown by inclinometer
wall, and improper temporary bracing.
In the case of improper shape of the IPS, similar
underground structure had been completed in Frank-
furt, Germany (Katzenbach et al. 1998) as shown in
Figure 7. Pit excavation was about 20.0m deep and
a similar temporary strutting system was also used.
Evidence also indicated that collapse of the IPS
was not caused by the diaphragm wall. It was found
that the results of analysis of the wall designed by
both the designer and diaphragm wall sub-contractor
were in conformity with completed projects in
Bangkok. The completed projects have been re-
ported by Teparaksa et al. (1998) and Thasnanipan Figure 8. Separating of the built-up walling beams and struts
(1998). No defect or leakage was found on the ex- in layer 3 (after Thasnanipan, 1997)
250
268
251
-50 0 50 100 -100 -50 0 50
0.00 0.00
Legend
Grouting
Zone
5.00 5.00
Inclinometer I-3
Tube
10.00 10.00
I-6
Depth (m)
Depth (m)
15.00 15.00
I-2 Existing Diaphragm Wall
1.00m thk.
20.00 20.00 I-4

Outer Diaphragm Wall
1.50m thk.
25.00 25.00
I-1
FEM Analysis
Excavate 16.5 m Excavate 16.5 m
Excavate 19.0 m Excavate 19.0 m
30.00 30.00
A-Axis B-Axis I-5

Figure 9. Prediction vs. actual wall behavior (I-3)
monitoring from I-3 indicated that the walls were in Figure 10. Location of inclinometers and treated area by Jet
cantilever mode. The wall behavior did not reflect grouting (after EIT 1999)
predicted behavior indicating inefficiency of the Legend
bracing system (Fig. 9). The cracks occurred in the Inclinometer Tube
capping beams confirmed cantilever wall move- VW Strain Gauge
ments. I-3
6. REMEDIAL ACTION / RECONSTRUCTION I-6 Strain Gauge

Location
Reconstruction of the IPS was started at the end of I-2

1997. The process involved construction of an outer Strain Gauge
diaphragm wall perimeter to enclose the collapsed

Location
I-4
IPS location. The IPS could not be relocated due to
presence of affiliated structures. Before starting the
outer loop diaphragm wall, Jet Grouting techniques I-1
were used to stabilize the soil in the surrounding
area for purposes of diaphragm wall construction as
well as for excavation works (Fig. 10). Jet grouting I-5
and diaphragm wall construction was undertaken by

another subcontractor. Jet grouting stabilized only
Figure 11. Arrangement of strutting system with stain gauges
soft clay up to about 16.0m deep. The outer loop (after EIT 1999)
diaphragm wall which was designed as a temporary
structure was 1.5m thick and about 32.0 m. deep.
During reconstruction, all parties realized the im-
portance of instrumentation. Six inclinometers were
installed, five numbers in the wall panels, and one in
the jet grouted area in the collapsed zone (Fig. 10).
During excavation, two bracing layers with heav-
ily built-up steel sections were used as diagonal
struts at depths of 2.0m and 11.5m (Fig. 11). In
both layers, strain gauges were also installed on two
main struts and readings were recorded each time a
depth of every 2.0m. had been excavated or every 2
days.
Installation and monitoring of (VWSG) on struts
provided information on their performance during
construction so that necessary action against insta- Figure 12. Excavation in progress for IPS reconstruction
bility could be taken.
269
Displacement (mm) Displacement (mm) Feld, J. and Carper K.L (1996) Construction failure (2nd edi-
-50 -25 0 25 50 75 0 25 50 75 100 tion) John Wiley & Sons. Inc.
0.00 Kanok-Nukulchai, W. and Phien-Wej, N. (1998). Investigation
-2.00
on Collapse of an Inlet Pumping Station- Looking into
modeling, Design and Construction Considerations, ACE-
5.00 COMS News and Views, July - September, AIT, Thailand
Katzenbach, R., Moormann, C., and Quick H (1998), A new
10.00
concept for the excavation of deep building pits in inner
-11.50
urban areas combining top/down method and piled-raft
foundation, 7th Int. Conf. and Exhibition on Pile and Deep
15.00 Foundation, Vienna, Austria.
Depth (m)
Padfied, C. J. and Mair, R. J. (1984) Design of Retaining walls

Embedded in stiff clay CIRIA Report 104.
20.00 -21.20 Phien-Wej, N. (1991) Types of Design and limitations of vari-
ous retaining cofferdam system, Technical seminar in
25.00 Foundation and underground construction works, organ-
ized by EIT, Feb. (in Thai) Bangkok, Thailand
29-Mar-98
13-Apr-98 13-Apr-98
17-May-98 17-May-98 Phien-Wej, N. and Sriruanthong, M. (1999) Remedial Con-
30.00 30-May-98
21-Jun-98
30-May-98
21-Jun-98 -32.00
struction of Inlet Pump Station, wastewater Project, Yan-
2-Aug-98 2-Aug-98 nawa, Civil Engineering Magazine, EIT. Jan- March. (in
35.00
13-Sep-98 13-Sep-98
Thai.)
Inclinometer No. I-6 Inclinometer No. I-2 STS (1995) Preliminary factual report of Geotechnical Inves-
tigation for Bangkok Wastewater Treatment on Sewage
Figure 13. Deflections recorded by I-2 & I-6
Treatment Plant, Bangkok, Thailand
Teparaksa, W., Thasnanipan, N., Muang, A. W. and Wei, S.
Figure 12 shows the status of excavation and strut- H. (1998) Prediction and Performances of Short embedded
ting before casting the new IPS structures inside the cast in-situ diaphragm wall for deep excavation in Bang-
temporary diaphragm wall. The lateral wall move- kok subsoil. 4th Intl. Conf. On Case Histories in Geotechni-
ments were recorded by I-2 and I-6 with maximum cal Engineering St. Louis, Missouri, USA.
Thasnanipan, N. (1997) The Cause of failure of the cofferdam
wall deflections of about 50.0mm 70.0 mm respec- under soil excavation and bracing installation stage at an
tively (Fig. 13). The reconstruction work was thus Inlet Pumping Station, Sewage treatment plant, SEAFCO,
successfully completed with proper instrumentation. Internal Report (in Thai).
Thasnanipan, N., Muang, A.W, Tanseng, P. and Wei, S.H.
(1998) Performance of A Brace Excavation in Bangkok
clay, Diaphragm wall subject to unbalanced loading
7. CONCLUSIONS & RECOMMENDATIONS conditions. 13th Southeast Asian Geotechnical Conference.
Taipei, Taiwan, ROC.
It was learnt that ignorance of instrumentation for William, B. P. and Waite, D. (1993) The Design and
deep underground construction can cause a serious Construction of Sheet Piled Cofferdams CIRIA special
disaster. publication 95, Thomas Telford.
Adequate instrumentation should be provided for
deep excavation to avoid remedial work.
Monitoring and maintenance of instrumentation
should be carried out throughout the work.
Parties involved in deep underground construc-
tion should have enough knowledge in geotechical
works and instrumentation.
Proper instrumentation enabled the successful
reconstruction of IPS.
8. ACKNOWLEDGEMENT
The authors wish to express their appreciation to the

colleagues, especially to Mr. Thiruchelvam
Navaneethan for his invaluable advice and assis-
tance in preparation of this paper.
9. REFERENCES
EIT (1999), Case study of problems related to settlement and
failures Geotechnical Engineering in the Past 2 years,
Subcommittee of Geotechnical Engineering, Engineering
Institute of Thailand (EIT)
252
270
PROCEEDINGS OF THE 5TH INTERNATIONAL SYMPOSIUM ON FILED MEASUREMENTS
IN GEOMECHANICS – FMGM99/SINGAPORE/ 1 – 3 DECEMBER/ 1999
Filed Measurements
In Geomechanics
Edited by
C.F.Leung, S.A.Tan & K.K.Phoon
Department of Civil Engineering, National University of Singapore
Behavior and performance of diaphragm walls under

unbalanced lateral loading along the Chao Phraya River
Thasnanipan N., Maung A.W. and Tangseng P.
Teparaksa W.,
271
Field Measurements in Geomechanics, Leung, Tan & Phoon (eds) © 1999 Balkema, Rotterdam, ISBN 90 5809 066 3
Behavior and Performance of Diaphragm Walls under Unbalanced

Lateral Loading along the Chao Phraya River
W. Teparaksa
Department of Civil Engineering, Chulalongkorn University, Bangkok, Thailand
ABSTRACT: This paper presents the behavior and performance of diaphragm walls constructed for base-
ments of two buildings along the bank of the Chao Phraya River bank, Bangkok, Thailand. These buildings
were constructed at different periods under one contract. A case history on basement excavation of the first
building has been reported previously by Thasnanipan et al. (1998). For the second building, more numbers
of instrumentation including VWSGs in a diaphragm wall panel and pressure gauges in struts to monitor
stress in the wall and the performance of bracing system were used respectively. The construction sequence
adopted for excavation, performance of the walls and bracing based on instrumentation results are discussed.
1 INTRODUCTION buildings respectively. They are also situated in the

heart of an old established culturally-significant
Retaining structures and supported systems for deep zone. Space and building height limitations necessi-
excavation in urban areas, particularly in the vicinity tate multi-level basements to increase usable floor
of a river, are often subjected to unbalanced lateral areas although the planned buildings are right on the
load due to asymmetric loading and surcharges. The river bank.
various publications and design guides by GCO no.
1/90 (1996), Williams & Waite (1993) and Padfield
& Mair (1991) give special guidance on overall sta- Chao Phraya River
bility of retaining structures and strutting systems in

such conditions. Case histories on braced excava- 5 Storey
tions under unbalanced/non-symmetrical lateral
5 Sto
Building
First Building Second Building
loading also have been reported by Thasnanipan et
rey B
Excav. Depth = 12.70m Excav. Depth = 9.70m

uildin
al. (1998), De Rezende Lopes (1985) and Kotoda et

g
al. (1990).
This paper presents two basement excavation 5 Storey 5 Storey
Building Building
works up to 12.7m below the ground level for two

separate buildings using braced diaphragm walls
along the bank of the Chao Phraya River, Bangkok.
The first one had previously been reported by Thas- Figure 1. Layout of building sites
nanipan et al. (1998). Construction of the second
building commenced about one year after comple-
tion of the first one. At the building sites, the river is about 205m wide
and 10-12m deep mid-stream. The riverbed near the
river wall is about 2.2-3.0m in depth with a gentle
2 SITE AND SUBSOIL CONDITIONS slope. The river water level in dry season is about
1.6m below ground level and in the rainy season
The building sites are located 40.0m apart, and sepa- sometimes rises above ground level, causing a flood.
rated by an existing 5 storey building between them. Primary site investigations included one borehole
They are located along the river and surrounded by 60m deep and one field vane shear test for each site.
existing structures including a historical building For the first building, prior to designing temporary
(Fig. 1). The existing buildings were supported by bracing and basement excavation work, drilling of
pile foundation. The distance between diaphragm additional three boreholes 20m deep and two field
walls along the river and the existing old river wall vane shear tests were carried out to check subsoil
267
273
variations. For the second building, two additional nite element program predicted maximum wall
60.0m deep boreholes were dug. The shear strength movements up to 24.2mm for the first building.
of soft clay in the planned second building site is Based on the performance of walls for the first
higher than the soft clay in the first site. However, building, a maximum wall movement of 51.0mm
there is no significant variation in subsoil conditions was allowed during excavation for the second build-
within each site, particularly along the depth of ing. The diaphragm walls for its basements were
planned diaphragm walls. Subsequently the profile analyzed with elasto-plastic soil model using
of the sites and surcharges from surrounding build- WALLAP program.
ings are of major concern for unbalanced lateral
loading conditions in excavation works. Table 1
shows a summary of subsoil properties obtained 4 PILE FOUNDATIONS
from the boreholes and test data.
Foundation piles having pile toe at 48m depth were
Table 1. Summary of soil properties constructed using bored piling under bentonite
Layer W γs Cu SPT slurry. Quantities of piles are tabulated in Table 2.
Soil Top in
3
Type Depth m % kN/m kPa N
Soft Clay 0-3.0 35-78 16-19 30
Med. Clay 12.7 30 19 71 Table 2. Quantity of foundation piles
Stiff Clay 14 22-34 19-21 43-300 14-52 Pile Dia.
Dense Sand 25 14-25 20-23 - 35-50
Building 800mm 1000mm 1200mm 1500mm
Silty Clay 36.5 17-21 20-23 175-240 30-45
Dense Sand* 42-45 20-26 21 - >58 First 32* 18* 31 6
* Some 4-5m thick hard clay seams present at depth 48-52m Second 17* 33 5 23
* Some piles were incorporated in diaphragm wall panels.
3 DIAPHRAGM WALLS
5 INSTRUMENTATION
For both buildings, 800mm thick cast in-situ con-
crete diaphragm walls of 28m toe depth with two As construction sites were located in a very sensitive
level temporary bracings were designed for base- urban area and subjected to unbalanced lateral load-
ment excavation. The maximum excavation depths ing conditions, various types of instrumentation
for the first and second buildings were 12.7m and were installed and systematically monitored. Lay-
9.7m respectively. 28m deep walls were necessary outs of instrumentation for the first and second
for overall stability of excavation as they were lo- buildings are shown in Figures 3a and 3b respec-
cated on the riverbank. tively.
For the second building, 5 levels of vibrating wire
strain gauges (VWSG) in pairs were installed in one
+14.80
+11.30
0 3 5 10
diaphragm wall panel and 2 sets of earth pressure
+7.80
gauges were installed in struts to observe stress in
the diaphragm wall and strut forces respectively.
+4.30 Vane Shear Strength (kPa) SPT-N (Blows/ft)
0 20 40 60 80 0 50 100
+0.80 +0.70
0.00
Types of instrumentation are presented in Table 3.

Weathered Clay
-3.20
-5.80 5.00
Soft to
-8.40 Medium
CLAY 10.00
0.80m Thick
D-Wall
Bored
Pile Stiff to Very
15.00
Table 3. Quantity of Instrumentation
Stiff CLAY 20.00
-28.00 (D-Wall tip)

25.00 Type of First Second Measurements
Medium to
Very Dense 30.00
First Bldg.
Instrumentation Building Building made
SAND
Inclinometer 8 6
Second Bldg.
-48.00 (Pile Tip)

35.00
Wall deflections
Figure 2. Basement section and soil profile (Second Building) Settlement Plate 10 20 Ground settle-
ments
Tiltmeter 10 10 Tilting of Build-
ings
Different wall types were classified by design ac- Vertical Beam 5 5 Tilting of build-
cording to load conditions, slab supporting locations Sensor ings
and geometry of ground profiles. The walls were
designed generally as vertical beams to withstand VWSG - 1 location* Strain in rebars
bending moments up to 1100 kN-m/m. In the slab
opening areas, additional horizontal bars were pro- Earth Pressure - 2x2 Sturt forces
vided for secondary bending moments. A non-linear Gauge
Beam-Column Analysis using a one-dimensional fi- * A pair of VWSG in 5 layers along the depth of wall panel
274
268
Tiltmeter
Chao Phray
1. Using a simple and efficient temporary bracing
a River
system
Inclinometer
Vertical Beam Sensor I-7
Surface & Deep Settlement Plates D1 I-6
Bored Piles
Preload Position
2. Pre-loading on one end of struts on the opposite
I-8 walls
3. Excavating first soil in front of the wall closest to
D3 B3
the river at any excavation stage
D3
4. Frequently monitoring of wall movements
5 Storey Building
I-5
5. Minimizing construction time
I-1 I-2
D2
Historical
Building B1
I-3
(2 Storey) D3
D2 I-4
0 5 10 15 B2
Scale 7 Storey Building
Figure 3a. Layout of temporary bracing and instrumentation

(First Building)
Chao Phraya River
D10 S10
D4 S4 S3
I-1 D9 S9
D3
I-2
T1 D1
T10
EL5
Figure 4. Behavior of walls shown by defromed mesh of two-
dimensional model
S1
I-3
5 Stor
I-6 T9 5 Storey
EL1 S8 Building
ey Bui
S2 D8
T2 D2
lding
EP-1 D7
T8
EL4
Moreover, an observational construction approach
T3
I-5 D5
(Ikuta et al. 1994), using the "most probable" condi-
tions and parameters in the design with a contin-
S6 D6 S7
S5
I-4 VWSG-1
T7 EL3
gency plan for "most unfavorable" conditions was

T4 T5 EL2 T6 5 Storey Building
5 Storey Building
Legend
0 6 10 20
employed. Firstly a monitoring plan was established
before excavation. Secondly the instrumentation
I = Inclinometer in D-Wall D = Deep settlement plate (5.0m)
T = Tilt plate EP = Earth pressure gauge
EL = Vertical beam sensor VWSG = Vibrating wire strain gauge
S = Surface settlement plate (1.0m) Pre-load position
monitoring data were used to trigger the contigency
Figure 3b. Layout of temporary bracing and instrumentation
plan. Generally predicted wall movements were set
(Second Building) as primary trigger values for the contingency plan.
After successful completion of the first building, the
same approach was also adopted in excavation work
6 EXCAVATION WORK for the second building. For the first level bracing
of the second building, single strut arrangement (less
Conventional bottom-up method with two levels of rigid but economical compared to that of the first
temporary bracing was adopted for excavation to building) was adopted with provision of strut-force
construct foundation and basement floors of both monitoring.
buildings. Prior to excavation work for the first A simple cross-lot bracing system with continuous
building, a two-dimensional computer analysis was wale beams was used for both buildings. 20m and
carried out using PLAXIS finite element computer 19m-long, steel king posts of H300 x 300 in section
program to study possible behavior of the wall sys- were used to support the working platform and brac-
tem under unbalanced lateral loading conditions. ing system for the first and second buildings respec-
Such conditions were considered to be resulting tively.
from (1) full depth of the earth, (2) a step sloping In both cases, the initial excavation to 2.5m for in-
river bed and (3) full depth of earth with possible stalling the first level bracing revealed historic foun-
surcharges from the adjacent buildings. dations. Temporary bracing and excavation works
A major concern was that the diaphragm walls
alongside the river would be thrust from the opposite Table 4. Summary of steel sections for the temporary bracing
walls. These bore a higher lateral load through axial Building Bracing Bracing Strut Design Strut
force of struts as excavation progressed in stages. Level Elevation Sections Force kN/m
This wall behavior was indicated by computer mod- First I -2.0 2 x WF350 x 350 484.0
II -6.5 2 x WF400 x 400 789.0
eling (Fig. 4), and previous reports by De Rezende
Second I -2.0 1 x WF400 x 400 279.4
Lopes (1985) and Kotoda et al. (1990). To prevent II -7.0 2 x WF350 x 350 332.1
any adverse wall behavior alongside the river, the
following measures were taken;
275
269
were delayed for about 3 and 2 months for the first Time (Days)
0 100 200 300 400

and second buildings pending permission by the ar- 0.0
chaeological department for further excavation. 2.0 B1
Depth (m)
Excavation Depth
4.0
B2
6.0
8.0 B3
7 INSTRUMENTATION RESULTS 10.0

80
I-4
Max. horiz. movement

Construction activities and the corresponding results 60
of instrumentation monitoring with construction
(mm)
40 I-2
time are presented in Figures 5a and 5b. These fig- 20
ures, particularly for the second building indicate a 0
good relationship between construction activities on 3000

Design Line
one hand and responses of the walls, existing build- 2500
Strut Load (kN)

ings and ground on the other hand. It can be ob- 2000
First Strut
served that significant changes in these responses 1500
generally occured during the initial excavation stage 1000
in which walls were unsupported. 500

Second Strut
0
1200
Time (days)
Strain (micro strain)

0 50 100 150 200 250 50% of ultimate tensile strain Soil Face
800 SF (4.5m)
Tension = positive
0 SF (7.0m)
Excav. near the river SF (9.5m)
Depth(m)
Excav. aw ay f rom the river SF (13.0m)

5
400 SF (17.0m)
10
15 0
200
Differential in Tilting
Tilt Meter
-400
100 T-1 (B3)
T-2 (B3)
(second
1200
)
T-3 (B3)
0 T-4 (B3)
T-5 (B3)
T-6 (B2) Excavation Face
Strain (micro strain)
T-7 (B1) 800 50% of ultimate tensile strain EF (4.5m)

Tension = positive
-100 T-8 (B1)

T-9 (B1) EF (7.0m)
T-10 (B1) EF (9.5m)
-200 EF (13.0m)
400 EF (17.0m)
-300
0.50 Vertical Beam Sensor

EL-1 (5th Floor of B3) 0
0.25 EL-3 (7th Floor of B2)
Displacement
EL-4 (2nd Floor of B1)

EL-5 (2nd Floor of B1) -400
20
Settlement (mm)
(mm)
0.00
Surface settlement plates
0 S6 (Land side)
-0.25 S7 (Land side)
-20 S9 (River side)

-0.50 S10 (River side)
100 -40
Inclinometer
I-1
. Movement
80 I-2
I-3 -60
I-4
60
I-5
I-6
20
I-7
Deep settlement plates
(mm)
I-8 D5 (Land side)

Max. Horiz
40 0
Settlement (mm)
D6 (Land side)
D9 (River side)
20
-20 D10 (River side)
0
-40
Figure 5a. Instrumentation results with construction time (First -60

Building) 200 Tilt plates
Differential in Tilting
T5
100 T6
(Second)
During the period of delay, monitoring of 0
instrumentation was frequently carried out to -100
compare the monitoring data and the trigger values -200
for planning a contingency plan against possible 10

Vertical beam sensor
risks to adjacent structures from unsupported

EL2
Displacement (mm)
5 EL3
excavations to 2.5m depth. 0
-5
7.1 Inclinometer monitoring -10
For the first building, generally the predicted and Figure 5b. Instrumentation results with construction time
measured lateral wall movements were in good (Second Building)
agreement, except for those at the top portion of
walls (Fig. 6a). The wall-top movements exceeded For the second building, wall movements were
predicted movements in the case of both buildings monitored weekly, sometimes every 2-3 days when
due to delay in installing the first level bracing, leav- necessary. At I-6 location, wall movements were
ing the walls in cantilever condition for about 2 found to reach the trigger/predicted value, especially
months or more.
270 276
at the top of wall when further excavation reached I-4 suggested that the wall alongside the river had
7.5m depth after installation of the first level brac- been pushed against the retaining soil by the oppo-
ing. site wall (Fig. 6b).
All inclinometer readings for both buildings indi-
cated that the walls were in fixed-end condition with
Displacement (mm) Displacement (mm) fixity at depths of about 20.0m. Maximum wall
0.00
0 20 40 40 20 0
movements after installation of bracing ranged from
RIVER -1.50
0.14% to 0.64% of the corresponding excavation
5.00
depth for the first building, and 0.27% to 0.83% for
-6.50
the second building. These were also within the
10.00
range (about 0.1% to 1.2% of excavation depth) of
-12.70 wall movements in other projects completed in
Bangkok area.
Depth (m)
15.00
PREDICTED
20.00
OBSERVATION
7.2 Pressure gauges
Readings from the pressure gauges installed on
25.00
struts of the second building were recorded daily.
Inclinometer Inclinometer After removal of second level bracing upon comple-
30.00 No. I-7 No. I-2
tion of base slab and during construction of base-
Figure 6a. Lateral wall movements (First Building) ment 2 floor, measured strut force of the first level
bracing reached the trigger value. However, in gen-
eral the readings indicated that the bracing system
0 20 40 60 80 80 60 40 20 0
used was adequate.
0.00
-2.00
River
Table 5. Predicted, allowed and measured strut forces (Second
5.00
Day 67
Building).
Day 133 -7.00 Day 67 Item First level Second level
Day 155
Day 221
Day 133
kN/m kN/m
-9.70 Day 155
10.00 Day 221 Predicted by WALLAP 354.5 307.3
Allowed 279.4 332.1
Measured 331.3 206.2
Depth (m)
15.00 Prediction
20.00
7.3 Vibrating Wire Strain Gauges
Strains developed in the wall of the second building
were measured by using VWSGs attached to main
Day Description
25.00 67 Excavate to -2.5m

133
155
Excavate to -7.5m
Excavate to -9.70m reinforcement bars of the wall panel where I-4 was
221 Remove strut at -7.00m
also located. A comparison between the measured
30.00
Inclinometer Inclinometer strains and the allowed strains of the wall was made
No. I-2 No. I-4
as shown in Figure 5. The comparison indicated that
Figure 6b. Lateral wall movements (Second Building)
strains induced in the wall were well below the al-
lowed values.
Lateral wall movement reached a maximum rate of 7.4 Settlement Plates

6.5mm/day. Then a tension crack on the ground sur-
For the first building, a maximum settlement of
face appeared about 8.0m away from the wall. A
16mm was recorded. For the second building, in
close examination of the bracing system indicated
most cases, ground settlements (surface and deep
that one strut had swayed slightly about 130mm at a
settlements at 1.0m and 5.0m depth) had been found
distance of 15.0m from the wall. Moreover, the sec-
to stop after installation of the first bracing. In some
tions of the strut had been connected at an angle. An
cases, the surface settlements stopped only after in-
additional strut was then immediately stacked on the
stallation of the second level bracing, resulting in
defective one and then wall movements were found
maximum settlements up to 30-50mm. The large
to cease. Installation of the second level bracing
settlements were found to be associated with sur-
was immediately carried out for this area.
charges from stockpiles of construction materials
Regarding unbalanced lateral loading condition,
and in one case, from unsupported concrete steps,
the wall movements of the first building were found
under demolition, which had become detached from
unaffected by this condition (Fig. 6a). However, for
the adjacent buildings.
the second building, monitoring results from I-2 and
277
271
7.5 Tiltmeter and Beam sensor 9 CONCLUSION
Readings are also shown in Figure 5. Results from
With proper instrumentation, two basement excava-
monitoring and typical values for maximum building
tions using braced diaphragm walls subjected to un-
slope or settlement for damage risk assessment are
balanced lateral loading from the river and adjacent
compared in Table 6.
buildings were completed.
Performance of the walls based on the instrumen-
tation results are presented and discussed.
Table 6. Comparison between observed values and maximum For underground construction work, monitoring
damage risk assessment (Lake et. al.) with instrumentation plays a major role, especially
when delay occurs during excavation. Moreover,
Max. Slope Max.
of Settlement Description overall construcion cost, time and risk can be mini-
Building of building mised with the use of instrumentation.
mm
Risk <1/500 <10 Negligible: superficial dam-
Category 1 age unlikely
Tiltmeter 1/2082* - 10 ACKNOWLEDGEMENT
1/2008** -
-
Vertical 1/3165* - The authors express their appreciation to Mr. Thi-
Beam Sensor 1/985** - ruchelvam Navaneethan for his assistance in the
Risk 1/500 10 - 50 Slight: possible superficial
Category 2 damage which is unlikely to
preparation of this paper. We would also like to ac-
have structural significant knowledge initial analysis work carried out by Dr.
* monitoring results for the first building, Vichai Vitayasupakorn and Mr. Young Zou.
** for the second building presented in this paper
REFERENCES
8 DISCUSSION
De Rezende Lopes, F. (1985). The problem of asymmetricsur-
The planned excavation sequence and deep embed- charges in strutted excavations. Ground Engineering, vol.
ment (about 18m and 8m below the excavation and 18, no.2, pp 31-35. GCO Publication no.1/90 (1996). “ Re-
view of design methods for excavations” Geotechnical En-
the fixity respectively) of the walls are considered to gineering Office, Civil Engineering Department, Hong
have contributed to minimizing the effect of unbal- Kong.
anced load conditions, particularly for the walls GCO Publication no.1/90 (1996). Review of design methods
alongside the river. for excavations. Geotechnical Engineering Office, Civil
Instrumentation results, particularly from the Engineering Department, Hong Kong.
monitoring of wall movements have given effective Ikuta, Y., Maruoka M., Aoki M. and Sato E. (1994). Applica-
tion of the observational method to a deep basement exca-
warning against inproper installation of bracing sys- vated using the top-down method. Geotechnique 44, No. 4,
tem and excavation sequences. This allowed pp 655-664.
improvement or modification of the bracing system Kotoda, K., Kazama, S., Ogawa, H. and Kato, K. (1990). Be-
where necessary. havior of earth retaining structure for excavation subjected
No damage was found in adjacent buildings, as to non-symmetrical lateral loading. Tenth Southeast Asian
Geotechnical Conference, 16-20 April 1990, Taipei, Tai-
confirmed by instrumentation results, particularly wan. pp. 539-544.
those from vertical beam sensors and tiltmeters. Lake, L. M., Rankin, W. J. and Hawley, J. (1996). Prediction
In comparing bracing systems for the first and sec- and effects of ground movements caused by tunnelling in
ond building works, stiffness of the first level brac- soft ground beneath urban area. CIRIA Report No. 30,
ing used for the second building is estimated to be Construction Industry Research and Information Associa-
tion, UK.
only 50% of that for the first building while ex- Padfield, C. J. and Mair, R. J. (1991). Design of retaining
pected strut force is 25% less. Moreover single strut walls embedded in stiff clay. CIRIA Report 106. Construc-
arrangement in the first level bracing for the second tion Industry research and Information Association. Lon-
building is less rigid compared to dual strut ar- don, UK.
rangement which can act as a composite beam for Peck, R. B (1969). Advantages and limitations of the observa-
the first building. Instrumentation on struts con- tional method in applied soil mechanics. Geotechnique 19,
No. 2, pp. 171-187.
firmed that bracing system used was adequate. Thasnanipan, N., Maung A. W., Tanseng P. and Wei S. H.
In both cases, early installation of bracing, par- (1998). Performance of a braced excavation in Bangkok
ticularly for the first level is very important for clay, diaphragm wall subject to unbalanced loading condi-
minimizing wall top movements, subsequently re- tions. Thirteenth Southeast Asian Geotechnical Confer-
ducing accumulated wall movements for further ex- ence, 16-20 November, 1998, Taipei, Taiwan. Pp. 655-660.
Williams, B. P and Waite D. (1993). The design and construc-
cavation. tion of sheet-piled cofferdams. CIRIA Special Publication
98, Thomas Telford Publications.
272278

th

8-12 November 1999
Bangkok, Thailand
Conference Proceedings
Volume 2 (Part I)
Editors:
D. G. Lin
D. T. Bergado
N. Phien-wej
P. Nutalaya
Analysis of lateral wall movement for deep excavation in

Bangkok subsoils
Teparaksa W.
N. Thasnanipan & Pornpot Tanseng

279
ANALYSIS OF LATERAL WALL MOVEMENT FOR DEEP BRACED

EXCAVATION IN BANGKOK SUBSOILS
Dr. Wanchai Teparaksa

Department of Civil Engineering, Chulalongkorn University
Bangkok 10330, Thailand
Narong Thasnanipan & Pornpot Tanseng
SEAFCO Co.Ltd., Bangkok, Thailand
ABSTRACT
This paper presents the case histories and back analysis of diaphragm walls for deep-
based excavations in Bangkok subsoils. Three case studies with different boundary conditions
are detailed as under unbalanced loading, underpass project under very narrow clearance under
the first stage expressway, and short embedded wall for modified deeper excavation. The
back-figured values for the soil stiffness in terms of Young’s modulus are calculated as Eu/Su
= 500 and 2000 for the soft Bangkok clay and stiff clay, respectively. The results have later on
been confirmed with the first self-boring pressuremeter tests in Bangkok subsoils.
INTRODUCTION
Deep excavations in Bangkok City are still necessary even the country has faced sever
economic crisis. There are still required the necessary civil works to serve the demands and
also to solve the environmental problems such as the first underground subway project, and
the wastewater treatment schemes. Subsoils condition in Bangkok is generally consists of a
very soft dark gray of about 12-14 m thick layer underlain by stiff and very stiff clay to about
23-25 m depth then followed by the first dense silty sand layer. Excavation in such soft dark
gray clay requires efficient retaining structures such as temporary sheet pile for shallow
excavation, and cast in-situ reinforced concrete (RC) diaphragm wall as permanent structures
for deep excavation. There are a lot of diaphragm wall activities under construction and under
excavation stages for the first blue line subway project, Metropolitan Rapid Transit (MRT).
This subway project consists of 18 underground stations, about 20 km long twin tunnels
running from North of Bangkok at Bangsue station and turn around the business area, bus
terminal and ended at the main state railway station namely Hua Lumpong station (Teparaksa,
1999). Excavation depth for the underground stations is between 22-33 m.
This paper aims to evaluate the soil stiffness parameters of the soft and stiff Bangkok
clay for RC diaphragm wall based on back analysis of three case studies with different
construction boundaries.
SOIL PARAMETERS FOR DIAPHRAGM WALL DESIGN
The RC diaphragm wall can be designed by various methods such as Finite Difference
Method, Beam on Elastic Foundation Method, and also the Finite Element Method (FEM).
The most well known and commonly used method is the two dimensional (2D) FEM method
or plain strain method. The 3D FEM analysis is rarely used. The FEM analytical method can
simulate the soil-structure interaction, and can predict the lateral wall movement as well as the
II ~ 67
281
ground surface settlement. The appropriate soil modelings as well as the input soil parameters
are the key issues to simulate the system. The most important soil parameter is the stiffness of
each soil layer, which is normally specified in terms of Young’s modulus (E). The E value is
very important especially for the soft clay layer that lead to induce the lateral wall movement
and ground settlement during excavation works. Normally, E is assumed to be linear and
determined from the conventional triaxial compression test and is very low as compared to the
values obtained from the empirical correlation. This is because of the high strain levels at
which the E is measured in the conventional triaxial test. The E value is recently known to be
dependent on the order of shear strain level. Mair (1993) compared the soil stiffness at various
shear strain levels for different structural systems as well as various kinds of laboratory tests as
shown in Figure 1. In case of concrete diaphragm walls, the shear strain is of the order of
0.01% to 0.10% only.
The research works on the soil behaviors at the low strain level in soft Bangkok clay
was presented by Shibuya et al (1997) based on laboratory tests as well the in-situ field vane
shear tests. Authors presented the soil stiffness in terms of Maximum Shear Modulus (Gmax).
Gmax is about 300*Su (300 times of Undrained Shear Strength) to 500*Su and if converted to
the Undrained Young’s modulus (Eu) will be equal to 900*Su to 1500*Su for Poisson ratio (ν)
of 0.5. Teramast (1998) also presented the soil stiffness at low strain levels based on the
Bender Element Test as Gmax = 400*Su to 570*Su or Eu = 1200*Su to 1710*Su. As shown
in Figure 1, Mair (1993) mentioned that the Bender Element Test is suitable for very low
strain levels such as applied to the dynamics or earthquake problems. Most practical value of
Eu for Bangkok subsoils is based on the empirical correlation proposed by Duncan &
Buchigani (1976) as shown in Figure 2. The Eu values were presented relative to the
undrained shear strength, which is dependent on the order of Over Consolidation Ratio (OCR).
For Bangkok soft clay where OCR is about 1.6, Eu/Su = 300-600.
Typical strain Ranges:
Retaing Walls
Stiffness G
Foundations
Tunnels
0.0001 0.001 0.01 0.1 1 10

Bender
Shear Strain εs: %
Resonant Coulomn
Local
Special Triaxial
Conventional
Figure 1 : Typical Strain Range (Mair, 1993) Figure 2: Empirical Young’s modulus
Values, Duncan & Buchigani, 1976.
BACK ANALYSIS CASE STUDIES
As mentioned above that the soil stiffness varies and depends on the shear modulus.
However, the values from laboratory tests give a high variation with the empirical correlation.
Therefore, the back analysis from case histories of diaphragm wall construction is really
282
II ~ 68
necessary to verify the appropriate soil stiffness. The back analysis was carried out for three
case studies at the Thammasat University project, Dindang underpass project and Sathorn
Complex project. Using the 2D-FEM program namely “PLAXIS” (Brinkgreve & Brand,
1996) back analysis for the said projects have been carried out. The analytical plain strain
concept and the Mohr-Coulomb soil model or bi-linear modeling was used.
Thamasart University Project
The building site is located nearby the Chao Phaya River and surrounded by a
historical building seated on shallow raft foundation as shown in Figure 3. The diaphragm
wall along the river was constructed about 3-4 m away from the existing old and shallow river
wall as shown in the building section in Figure 4. The river bed near the river wall is about 3
m in depth and sloping towards the mid-stream to a depth of about 10-12 m (Thasananipan et
al, 1998). This riverbank lead to induce the unbalanced lateral earth pressure on the
diaphragm wall system between land side and riverside. A 800 mm thick diaphragm wall was
designed for excavation down to 12.7 m below the ground with two levels of temporary
bracing at –1.5 m and –6.5 m depth. The diaphragm wall toe was embedded down to 28 m to
achieve the overall stability of the excavation which is right on the river bank (Fig. 4). A site
investigation was carried out by two boreholes and four field vane shear tests, the subsoil
properties are summarized in Table 1.
Table 1. Summary of subsoil properties
Soil Type Depth Wn Unit Weight, γ s Cu SPT-N

(m) (%) (kN/m3) (kPa) (Blows/ft)
Crust 0.0-2.5 50 19 30 -
Soft Clay 2.5-12.7 70-85 26 15-30 -
Medium Clay 12.7-14.0 50-55 17 30-45 -
Siff Clay 14.0-25.0 20-25 20 130-250 14-40
Dense Sand >25.0 - 20 - >39
Instrumentation
Due to the locality and particular conditions of the site, more instrumentation than those used
by other projects in Bangkok were installed and monitored. Three types of instrumentation -
Eight inclinometer tubes are installed in diaphragm wall panels to monitor displacement of
wall, ten settlement plates of 1.0m and 5.0m in depth around excavation zone, and ten
tiltmeters and four vertical beam sensors on surrounding buildings were installed to monitor
tilting. Layout of instrumentation is presented in Fig. 3. Since the excavation work is located
adjacent to the river and surrounded by old buildings, the diaphragm walls are subject to the
unbalanced lateral load The model analysis indicated relatively less movement of Wall D1
towards excavation in the land side and move less in the river side as well as induce forward
movement of river wall to the river (Fig. 5). This analysis led the engineer to solve the
damages of the river wall due to unbalance loading by using the following construction
methods.
283
II ~ 69
Chao Praya Riv
er
I-7
I-6
D1
I-8
Tiltmeter
B3
Inclinometer
Vertical Beam Sensor
D3
Surface & Deep Settlement Plates
Bored Piles
D3
Preload Position
5 Storey Building
I-2 I-5
I-1 D2
B1
D3
Historical Building
(2 Storey) I-3
D2 I-4
0 5 10 15 B2
Scale 7 Storey Building
Figure 3: Project Layout
0m 6m 10m 20m
0
RIVER
RIVER
WALL
10
15 FOUNDATION
PILES
0.80 THK.
0.80 THK. D-WALL
20 D-WALL
25
30
35
Figure 4: Section perpendicular to the river
1. Arrange the strut as a simple straight grid line

2. Pre-loading of the struts (200kN/m and 400kN/m for first and second levels
respectively) on the one end of the strut on the land side wall.
3. Excavating the soil in front of Wall river side first at any excavation stage
Figure 5 : Deformed Mesh of Two-Dimensional model
284
II ~ 70
Analytical Results
The analysis was trial by changing the soil stiffness and compares the results with the
field performance. Figure 6 shows the results of FEM analysis compared with inclinometer
measurements based on soil stiffness of Eu/Su = 500 and 2000 for soft Bangkok clay and stiff
clay, respectively. . Movements of Wall D1 close to the river are found to be cantilever shape
indicated by I-7 while movements of Walls D2 and D3 have inward bulging shape indicated
by I-1 to I-5. Generally the predicted and measured lateral wall movements are in a good
agreement, except for the top portion of wall in which the measured wall movements exceed
the predicted movements. The differences in wall movement were found to be caused by the
following;
• Advancement of excavation in front of Wall D1 further than other walls, allowing Wall D1
to stand longer prior to strut installation than the others.
• Wall D1 was in cantilever condition for about 3 months after initial excavation for first
level bracing.
SPT (blow/ft) Su FV (kN/m 2) Excavate
Excavate Excavate
0 20 40 60 80 0 10 20 30 40 2.00 m 7.00 m 12.70 m
0.00 0.00 0.00
Crust
5.00 5.00 5.00

Soft to
Very
Soft Clay
10.00 10.00 10.00
Med.
Depth (m)
Clay
15.00 15.00 15.00
Stiff to
Very
Stiff Clay 20.00 20.00 20.00
25.00 Field 25.00 25.00
Dense to PLAXIS
Very Dense
Sand
30.00 30.00 30.00
0 10 20 30 40 0 10 20 30 40 0 10 20 30 40
Displacement (mm) Displacement (mm) Displacement (mm)
Figure 6 : Comparison between calculations and observation (Thammasat University Project)
Dindang Underpass Project
The Dindang underpass was the first Bangkok underpass project at Dindang intersection.
Underpass the main road beneath the main foundations of the expressway. This project is also
very interesting, specially designed equipment for D-wall construction was employed under
very narrow clearance under the first stage expressway (Thasananipan et al. 1996). There was
one temporary bracing layer at 2 m. depth and excavation to 4.9 m. depth. Figure 7 shows the
results of FEM analysis compared with inclinometer measurements based on soil stiffness of
Eu/Su = 500 and 2000 for soft Bangkok clay and stiff clay, respectively. The predicted and
measured lateral wall movements are in a good agreement.
285
II ~ 71
SPT (blow/ft) Su FV (kPa) Excavate Excavate
0 20 40 60 80 0 20 40 60 80 2.7 m 4.90 m
0 0 0.00 0.00
Crust
5 5 5.00 5.00
Soft to Very Soft

Clay
10 10 10.00 10.00
Medium Stiff Clay
15 15 15.00 15.00
Stiff
20
to Very Stiff 20 20.00 20.00
Clay
Field
PLAXIS
25 25 25.00 25.00
-10 0 10 20 -10 0 10 20
Figure 7 : Comparison between prediction and observation (Dindang Underpass Project)
Sathorn Complex project
The project site is located in the vicinity of central business area of Bangkok. The
structure is a multi-storied building for a business complex. The load of the structure would
be carried by bored cast-in-situ bored piles and barrette piles with toe depth of 60m below the
ground level, embedded into the second sand layer of the Bangkok subsoil. The constructed
cast in-situ diaphragm wall is 0.8m thick and the toe is embedded into stiff clay at a depth 18m
below the ground level. Initially the wall was designed for construction of four basements
including mat foundation of 3.5m thick. The maximum excavation depth was set to be 14m
and a three level of temporary bracing system was considered to construct the basements with
conventional bottom up construction method. Embedded depth below the final excavation for
the mat foundation was designed to be 4m. Figure 8 shows the layout plan of the building
After construction of the wall it was decided to increase the number of basements to be
five for part of the building area (tower area) and excavation depth needed to be increased by
1.8m from the originally designed depth, leaving the wall embedment depth of 2.2m only.
The first author was invited to solve such a short embedded diaphragm wall (Teparaksa et al,
1998). The temporary bracing was changed from three layers to be four layers with using the
soil berm of 12 m. wide to prevent the excessive wall movement as shown in Figure 9.
286
II ~ 72
I-3
Excavation -15.5m
Bored Pile
Barrette
I-4
Inclinometer Tube
I-2
89.35
Excavation -14.0m
I-1
120.25
Figure 8: Layout of Project
+1.70
16
-1.50
15
-2.40
2 14
1 -4.20 1
13
-5.10
3 12
-6.90
11
-9.00
4 10 -9.60
7
-11.80
-12.30 5 9
8 -13.30
-14.00
6
-15.50
-18.00 SCHEMATIC SECTION OF BASEMENT CONSTRUCTION -18.00
LEGEND
12 CONSTRUCTION STAGE
TEMPORARY BRACING
Figure 9: Schematic Section
Subsoil Conditions
The sub soil conditions at the project site is not uncommon to usual conditions observed in
Bangkok comprising a 15m thick soft to medium clay layer on the top underlain by stiff clay
to a depth of about 22m. Below the stiff clay is a series of alternating layers of hard silty clay,
sandy clay and dense sand. The soil properties used for Diaphragm wall design are
summarized in the table 2 belows:
Table 2 Soil Properties used for D-Wall Design

Unit Weight Undrained Shear
Soil Type Strength, Su
γ s (kN/m3)
(kPa)
Soft Clay, CH 16 10-29
Medium Clay, CH 18 35
Stiff Clay, CH 20 137-180
287
II ~ 73
Analytical Results
. The wall constructed was designed for initial excavation of -2.5m to install the first
bracing layer. However the excavation has been done to about -3.0m to -3.2m at once and a
movement of 40mm was observed. Time delay during construction of capping beam and over
excavation deeper than the critical height (Hcr = 2C/γ) was unsupported around three months
resulted an increase in wall movements to 55mm -70mm. Tension cracks were developed in
the soil of active side about 2-3m away from the wall. After installation of struts and pre-
loading them, the lateral movements of the wall ceased off. Struts from first to fourth levels
were pre-loaded 40 t/m, 75 t/m, 75 t/m and 40 t/m respectively, 20%-25% of design load of
bracing system, prior to further excavation. The pre-loading of the strut was very effective to
reduce lateral movement of the wall and ensuring good intact between wall and bracing
system. Figure 10 shows the comparison of the results of FEM analysis compared with
inclinometer measurements based on soil stiffness of Eu/Su = 500 and 2000 for soft Bangkok
clay and stiff clay, respectively. The measuring data was neglect the first movement due to the
delay of the first cantilever diaphragm wall around three months. The predicted and measured
lateral wall movements are in a good agreement.
Horiz. Displacement, mm
-10 0 10 20 30 40
0.00
-2.40
5.00 -5.10
-9.00
10.00
Depth, m
-11.80
-14.00
15.00 -15.50
-18.00
20.00
Inclinometer Tube
Penetrate into Pile Leg
PLAXIS
I-1
25.00
I-2
Figure 10 : Comparison of wall movement from FEM analysis and inclinometer results
COMPARISONS OF ANALYSIS WITH PRESSUREMETER TEST
The trial soil stiffness based on the back analysis shows that the Young modulus of soft
Bangkok clay and stiff clay are Eu/Su = 500 and 2000, respectively. Recently, six self-boring
pressuremeter tests were performed along the route of Metropolitan Rapid Transit (MRT)
northern contract by the Cambridge In-situ of Little Eversden (1997). The self-boring
pressuremeter tests were carried out by reload/loading cycles in order to determine the soil
stiffness at various shear strain levels. Figure 11 shows the results of pressuremeter test and
288
II ~ 74
shows the variation of shear modulus with shear strain, Figure 11.1 is for soft clay while
Figure 11.2 is for first stiff clay. For normal practice of diaphragm wall construction for deep
basements in Bangkok soft clay, the shear strain is in the order of 0.1-0.2%, where the G/Su is
about 160 or Eu/Su =480. In case of the first stiff silty clay layer where shear strain is in the
order of 0.05-0.1%, the G/Su is about 340 or Eu/Su = 1020. The soil stiffness of the soft clay
from pressuremeter test results is very close to the results of back analysis, however, the
values of stiff clay from pressuremeter test results is lower than the results of back analysis,
this is might due to the effect of bored pile stiffness inside of the excavation area. The
correlation from the self-boring pressuremeter was the non-linear soil model that will be used
to carry out the future back analysis.
1500 1 5 00
1250 1 2 50
1000 1 0 00
G /Su
G /Su
750 750
500 500
250 250
0 0
0 .0 1 0 .1 0 1 .0 0 0 .01 0 .10 1 .00
Sh ear S train (%) Sh ear Strain (%)
Figure 11.1: Self-boring Pressuremeter Figure 11.2: Self-boring Pressuremeter

Test Result of Bangkok Soft Clay Test Result of Bangkok Stiff Clay
.
CONCLUSIONS
The analysis and the design of the diaphragm wall are normally based on the
assumption of soil stiffness. Back analysis to verified the soil stiffness based on three
different cases as under unbalanced loading, underpass project under very close to the
foundations of the first stage expressway, and short embedded wall. The results shows that the
soil stiffness in terms of Young modulus is Eu/Su = 500 and 2000 for soft Bangkok clay and
stiff clay, respectively. These values were also corresponding to the results of self-boring
pressuremeter tests of the first blue-line subway project in Bangkok.
REFERENCES
Brinkgreve, R. and Brand, P.A.(1996): Application of PLAXIS for soil and rock plasticity,
Short Course on Numerical Analysis in Geotechnical Engineering, AIT.
Cambridge In-situ (1997): Self-boring pressuremeter MRTA initial system project, North
Contract.
Duncan, J.M. and Buchigani, A.L.(1976): An engineering manual for settlement studies,
Geotechnical Report of Civil engineering Department, university of California at
Berkeley.
289
II ~ 75
Mair,R.J.(1993): Development in geotechnical engineering research, application to tunnels and
deep excavations, Proceedings of the Institution of Civil Engineers and Civil
engineering.
Shibuya, S., Hanh, L.T., Wilailak K., Lohani T.N., and Tanaka H. (1997): Characterizing
stiffness and strength of soft Bangkok clay from in-situ and laboratory tests, First Int.
Conf. on Site Characteristics.
Thasananipan N., Muang A. Singtokaew K. (1996): Experiences in construction of diaphragm

wall for underpass project, Seminar on Underground Construction’96, Engineering
Institute of Thailand.
Thasananipan N., Teparaksa W, Muang A. and Baskaran G., (1998): Design, construction and
behavior of bored cast in-situ concrete pile in Bangkok subsoils, Proc. 4th Int. Conf. on
Case Histories in Geotechnical Engineering, St. Louis, Missouri, USA.
Thasananipan N., Muang A., Tanseng P. and Wei S. H., (1998): Performance of a braced
excavation in Bangkok Clay, diaphragm wall subject to unbalanced loading conditions,
13th Southeast Asian Geotechnical Conference, Taipei, Taiwan ROC.
Teparaksa W., Thasananipan N., Muang A., and Wei S. (1998): Prediction and performances
of short embeded cast in-situ diaphragm wall for deep excavation in Bangkok subsoils,
Proc. 4th Int. Conf. on Case Histories in Geotechnical Engineering, St. Louis,
Missouri, USA.
Teparaksa W. (1999): Geotechnical aspects of the Bangkok subway design and construction,
guest speaker for the Society of Professional Engineers Thailand (S.P.E.T.) member
meeting at British Club, 18 February. .
Theramast N. (1998): Characterization of pseudo-elastic shear modulus and shear strength of

Bangkok clay, M. Eng. Thesis, AIT.
290
II ~ 76
GeoEng2000
An International Conference on
Geotechnical & Geological Engineering
19-24 November 2000

Melbourne Exhibition and Convention Centre
Melbourne, Australia
Diaphragm wall and barrette construction for Thiam Ruam Mit

Station Box, MRT Chaloem Ratchamongkhon Line, Bangkok
Thasnanipan N., Maung A.W. and Ganeshan B.
ISBN No. 1-58716-068-4
291
GeoEng2000 An International Conference on Geotechnical and Geological Engineering 19-24 November 2000, Melbourne, Australia
DIAPHRAGM WALL AND BARRETTE CONSTRUCTION FOR

THIAM RUAM MIT STATION BOX, MRT CHALOEM
RATCHAMONGKHON LINE, BANGKOK
Narong Thasnanipan1, Aung W. Maung2 and Ganeshan Baskaran3
ABSTRACT
Construction experience of cast-in-situ diaphragm walls and foundation barrettes for the Thiam Ruam Mit
Station, Chaloem Ratchamongkhon line of Metropolitan Rapid Transit Authority (MRTA) is briefly reported
in this paper. The construction activities in the midst of heavily congested traffic and in the vicinity of un-
derground utilities to meet the specified quality and safety controls are also presented.
INTRODUCTION
The Chaloem Ratchamongkhon line is the first underground mass rapid transit project in Bangkok, Thai-
land being constructed under the supervision of Metropolitan Rapid Transit Authority (MRTA). The project
involves the construction of about 22km-long, twin bored, single track tunnels, 18 station boxes and cut-
and-cover tunnels. The project was divided into two separate contracts, South and North. The south contract
was awarded to BCKT Joint Venture consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch. Karnchang
Public Co., Ltd., Kumagi Gumi Co., Ltd. and Tokyu Construction Co., Ltd. The north contract was awarded
to ION Joint Venture, comprising Italian Development Public Co., Ltd., Obayashi Corporation and Nishi-
matsu Construction Co., Ltd. This paper presents experience on construction of diaphragm walls and foun-
dation barrettes for Thiam Ruam Mit Station (station no. 12) of the North contract where heavy traffic condi-
tions prevail.
LEGEND :
MRT (Underground Structures-North)
The station is one of the biggest stations
MRT (Underground Structures-South)
MRT Interchange Station
MRT Station
and located on the Ratchadaphisek Road. The
station box is enclosed by principally 1.0m
SRT System (Hopewell)
BMA System (BTS) BANG SU STATION
thick side walls and 1.20m thick end walls

ER
RIV
32.0-42.0m deep. A Tunnel Boring Machine

YA
RA
PH
(TBM) launching shaft was included in the

AO
CH
THIAM RUAM MIT STATION

DEPOT
north side of the station. To facilitate top-
down construction, 63 Barrettes (1.2mx3.0m)
HUA LAMPHONG STATION
embedded 44.5-55.0m deep in conjunction
with pre-placed stanchions at the top were
used.
Figure 1. Layout of MRT Chaloem Ratchamongkhon line
Three basement levels were planned for the station. The excavation depth inside the station box was up
to 24.5m for tunneling and basement slab construction. As deep excavation required stringent tolerances for
the diaphragm walls and barrettes, a high level of quality control was necessary.
1
Managing Director, Seafco Co., Ltd. Bangkok, Thailand
2
Project Manager, Seafco Co., Ltd. Bangkok, Thailand
3
Project Engineer, AVOPILING Pty., Ltd. NSW 2087, Australia
293
Construction of diaphragm walls and barrettes of the Thiam Ruam Mit station was carried out by an ex-
perienced local foundation contractor in which the authors worked while international contractors undertook
similar work for the remaining 17 stations.
SUBSOIL CONDITIONS
0m
SOUTH NORTH The subsoil profile, as shown in Fig. 2 consists of
made-ground up to 2.5m thick for pavement underlain
STATION BOX
10m SOFT CLAY by a series of clay and sand layers. Below the made-
TUNNEL
ground is 12.0m to 14.8m-thick Bangkok Soft Clay.
The first stiff clay layer occurs below 14.0m depth and
20m
TUNNEL
30m
extends to 21.0m, overlying the first sand layer. The
SILTY SAND sand layer is about 20.0m thick at this location and has
40m two thin (about 2-4m thick) very stiff clay layers at the
top. Below the sand layer is stiff to hard clay layer
50m STIFF CLAY which extends beyond 60.0m in depth. Within the clay
60m
SILTY SAND layer, from 48.0m depth, a 14m thick sand layer occurs
in the north and lenses out to the south. The properties
of soil layers are summarized in Table 1.
Figure 2. Profile of soil layers and station box.
Table 1. Summary of soil properties

Soil Type Depth to Top w γs Cu SPT- N
of Layer (m) % kN/m3 kPa Blow/ft
Made-Ground 0.0 40 17.0-19.0 30-45 5-9
Bangkok Soft Clay 0.2-2.6 64-100 14.6-16.8 15-31 -
Stiff Clay 14.2-17.3 16-51 18.3-21.8 107-250 13-17
Sand 24-28.6 - 20.8-21.1 - 27-59
Stiff to Hard Clay 40.8 16-20 21.0-21.1 61-410 11-61
The depth of ground water table at the site generally varies from 0.9m to 2.2m. Piezometers installed in
the project area indicate the pore pressure is hydrostatic below 23m depth due to pumping of ground water.
SITE CONDITION AND PREPARATION
2-SC
THAI
N
R.S.TOWER INSURANCE
2-SC 2-SC
2-SC
2-SC
Mr. P FRESH
OPEN AREA 4-CF
MILK
4-CF
ENTRANCE 12.1 ENTRANCE 12.2
C
L
C
L C
L
TBM
TO RAMA IX ROAD DRIVE
C
L
RATCHADAPHISEK SHAFT 27.000 TO LATPHRAO ROAD
E
C
L
RD. C
L
GE LIN
ENTRANCE
12.4 ENTRANCE 12.3
ORAN
VENT.
VENT.
SHAFT 1
SHAFT 2
359.000
Figure 3. Layout of station box
Traffic Management
The construction site was on a 35.0m-wide main road with 4 inbound and 4 outbound traffic lanes. Dur-
ing the construction period a maximum of only two traffic lanes could be occupied at a time, allowing 3 or 4
lanes for each traffic direction. The width of the working area available was about 11.0m. Sometimes, the
294
width of the work area needed to be widened locally to position the grab crane for diaphragm wall panel ex-
cavation.
Utility Diversion
The site is bounded by some underground utilities, including a 1.0m diameter water pipe and telecommu-
nication cables in the close vicinity of the work area, particularly on the east side. Above ground, 69kV elec-
tricity lines, telephone cables, drainage and sewage pipes (1.5m dia.) lay on both sides. The water pipe ran
along the planned diaphragm wall at about 3.0m depth and thus it was diverted by the other subcontractor to
the west, outside the station box prior to construction. The face-to-face clearance distance between the di-
verted water pipe and the outer face of the station wall was only 35cm due to space constraints. A reinforced
concrete wall to protect the pipe was constructed and it formed the bottom part of the guide wall for dia-
phragm wall panel excavation.
Two duct banks of the telecommunication cables were located 5-10cm away from the wall face at a depth
of 3.04m. 15mm thick steel plates were used in these locations as both guide walls and protection systems
for the duct banks.
CONSTRUCTION SPECIFICATION
Generally construction specifications were in line with the commonly practised specification for founda-
tion works. Regarding construction tolerances, verticality of the walls and barrettes is specified within 1:200
while for pre-founded steel stanchions, the verticality allowed is within 1:400.
Table 2. Test and compliance values for bentonite

slurry Properties of supporting fluid for wall and
Test Prior to barrette excavation were maintained within the
Property Specification Freshly Placing ranges as shown in Table 2. In addition, design
API RP13B- Mixed Rebar & assumptions required that total construction
1 Section Concrete time below the soft clay layer was limited to
Density 3 < 1.05g/ml < 1.10g/ml minimize the effects of bentonite on shaft fric-
Viscosity 4 29-50 29-50
Fluid loss 5 < 30ml < 30ml tion and in order to achieve optimum shaft fric-
Sand content 7 < 1% < 2% tion capacity of foundation elements.
pH 9 8-12 8-12
CONSTRUCTION PHASES
Throughout the project, the main construction schedule for the station comprised several interfacing and
parallel work activities such as for utilities and traffic diversions, temporary decking, substructure work for
station box and the TBM driving shaft which other subcontractors undertook. This meant that guide wall
construction in particular needed to follow utility diversion as the same area was to be excavated for reloca-
tion of the utilities, removal of abandoned water pipes and construction of guide walls. The temporary deck-
ing and excavation for the TBM launching shaft followed the diaphragm wall construction.
DIAPHRAGM WALL AND BARRETTE CONSTRUCTION
Equipment and Plant

Mechanical rope-suspended grabs with crawler cranes (80tons) and service cranes (50-80tons) were used
as main construction equipment. Grabs, 1.0mx2.0-3.0m and 1.2mx3.0m were used to cut the required di-
mensions of panel excavations. Bentonite slurry was used as supporting fluid for panel and barrette excava-
tion. Silos having a total of 300 cu.m storage capacity and de-sanding and de-silting units (80cu.m/hr capac-
ity) were used to supply the slurry required for up to two-panel excavation at a time.
295
29.200
6.300 6.300
3.000
OF TNL.
OF TNL.
4.000
L
C
TBM
L
C
15.000
LAUNCHING
SHAFT
4.500 4.850
Figure 4. General view of the site Figure 5. Panel layout of TBM launching shaft
Guide Walls
Generally 1.5m deep guide walls were used for panel excavation. In the location of underground water
pipes and duct banks, up to 3.0m-deep guide walls were used. Individual panels were sized with the follow-
ing considerations: (1) type of reinforcement, (2) size of grabs used, (3) thickness and depth of panels, (4)
location of slab and wall openings for entrances and TBM, (5) stability of trench in connection with location
of underground utility and (6) location of public accesses and streets.
Panel Layout
The walls were divided into 184 panels with horizontal lengths varying from 2.5m to 4.85m. L-shaped
and T-shaped panels were used to form box structures of the station and TBM launching shaft (see Fig. 5).
Sequence of Construction
Construction activities took place day and night. To maintain the traffic flow, comply with local author-
ity regulations, construction sequence and activities were planned to allow concrete pouring and spoil re-
moval to be done before and after rush hours, in most cases at night. Generally panel excavation was carried
out at least 6.0m away from the recently cast panel or 16hrs after casting of the adjacent panels to avoid any
damage induced by excavation.
Supporting Fluid
Locally available bentonite powder was used for preparation of slurry. This product had been used for
more than a dozen completed diaphragm wall projects,. Approximately 496 tons of bentonite powder was
consumed for excavation work. The last portion of slurry displaced by concrete or contaminated with ce-
ment was usually discarded. Soil volumes of about 26,274cu.m and 10,740cu.m were excavated for dia-
phragm walls and barrettes respectively.
Reinforcement
Steel bars of SD50 were used for the main reinforcement bars. The reinforcement cages were 30.0m to
42.2m long and up to 4.5m wide, with 4 levels of box-out for slab connections for diaphragm wall panels.
For barrettes, cages were 48.1 to 54.6m long. 4 to 5 cage sections with lengths of 6-12m were joined to form
one continuous cage for individual panels and barrettes. The cage sections were fabricated off site and no
more than two were joined for transporting to the site. Cage sections were connected prior to and during
lowering into the trench. In order to place the box-out in position and to achieve an exact match of the cage
sections, markings were made on the main bars of each cage. U-shaped bolts were used for cage connec-
tions. Couplers were generally used for slab connection and both couplers and bent-out bars were used for
beam connections at the wall openings for entrances and slab openings. The large panel cages weighed up to
34.8 tons.
For panels with glass fibre reinforcement polymer (GFRP) for openings of TBM break-through, a tempo-
rary steel frame was necessary to hold the cage section due to different stiffness between GFRP and steel.
The frame was then cut away section by section while lowering the cage. Diaphragm wall panels were rein-
296
forced with 136-218 kg/m2 of steel bars while barrettes were reinforced with about 50kg/m3 of steel bars.
The wastage of steel was about 7% in which most quantity was consumed for cage hanging bars and frames
for soft eye openings. A total of 4,756.15 tons of reinforcement bars was used for diaphragm wall and bar-
rettes. Steel pipes 150mm in diameter were installed inside the diaphragm wall panels at 5 locations as a
void former for inclinometer access tubes. 28 sets of vibrating wire strain gauges and total jack-out pressure
cells were also installed in the wall panels of TBM launching shaft.
Concrete Casting
Ready mixed concrete grades 40 (cube strength 40MPa at 28 days) and 35 (35MPa at 28days) to BS5328
were used for diaphragm walls and barrettes respectively. Two sets of tremie pipes were used in pouring
concrete for both diaphragm walls and barrettes. The largest single concrete pour was about 232cu.m for L-
shaped panels. Total concrete volumes of 27,480cu.m and 6,611cu.m were poured for diaphragm walls and
barrettes respectively. Concrete over-consumption in diaphragm wall caused by old, sand back-fill in the
area of abandoned water pipes was estimated to be an average of 7.8% with a concrete wastage of up to
38.0% occurring in one panel. However, the average concrete wastage was 5.0%. For barrettes, the average
concrete wastage was up to 14%. The high wastage compared to building and elevated highway projects (up
to 7%) was caused by overcasting of concrete well above the cutoff level of 22m depth to ensure that sound
concrete reached above the cutoff. Aggregates were used to backfill the open trench of barrettes above the
concrete after casting.
Installation of Stanchion in Barrettes

Conventional plunging method was not adopted in view of the site conditions, low cutoff level, the stan-
chion design and horizontal bar arrangements in the reinforcement cage. Two specially designed installation
frames with adjustable screws were used for installation of stanchions to ensure high accuracy for verticality
and plan position. A removable 10m-long, steel column was fixed on top of the stanchion to check the verti-
cality and alignment.
PRODUCTION RATES
The construction time required for a typical panel and barrette is detailed
in Table 3.
Table 3. Average construction time for typical diaphragm wall panels

(1.0x4.5x35.0m, cutoff 0.0m) and barrettes (1.2x3.0x44.5m, cutoff -
22.5m).
Activities Diaphragm wall Barrette

Hours Hours
Excavation 18 16.5
Checking verticality with Koden 1 2
Desanding 4 5
Cage installation 2.5 4.5
Stanchion installation - 3
Tremie preparation 2.5 3.5
Concrete pouring 4 3.5
Figure 6. Guide frame for steel

stanchion installation in barrettes
Throughout the construction period, a total of about 867hrs, approximately 15% of construction time (ex-
cluding mobilization and site preparation period) was consumed by equipment maintenance and repair. Grab
teeth were broken and changed often during chipping off the concrete overbreak behind the stop-end plate
caused by loose material below the guide walls. Additionally, steel wire ropes for the grab also needed to be
changed frequently as they were worn out by both the chiseling action and broken concrete pieces. 17% of
construction period was rainy, but this usually had no significant effect on the production rate. A flash flood
297
occurred at one time by a heavy downpour. Standby time (14% of construction period) took place and was
contributed by (1) obstructions by overbreak concrete during panel excavation and stop-end plate removal
and (2) standing time after panel excavation reached below the soft clay to allow subsequent construction ac-
tivities such as spoil removal and concrete delivery taking place outside rush hours. Traffic diversions were
carried out in multi-phases and took at least 5% of construction period.
QUALITY, SAFETY AND ENVIROMENTAL CONTROLS
As the alignment of diaphragm wall and barrettes is critical, every excavated panel and trench was
checked with Koden monitoring equipment. Monitoring was carried out after excavation reached about
22.0m (ie. at the levels of base slab and barrette cutoff) and toe levels. Necessary corrections to trench verti-
cality were made during excavation if required. Panel ends, especially for those panels with soft-eyes for
TBM break-through were also checked to achieve high levels of accuracy for positioning. During excava-
tion, reference posts were used for checking the grab position and observing rope position relative to the
trench sides at all times. Prior to lowering the reinforcement cage into the trench, any mudcake built-up on
the trench faces and panel joints was scraped off by a grab to which brushes were attached. Finally, any
sediment or loose materials deposited at the bottom of the trench were also removed by using the grab.
Reference bars attached to the top of reinforcement cage were extended above ground level to check the
position of the cage with survey equipment during installation of the cage into the trench.
Delivered concrete was checked for slump and cohesiveness prior to casting. As per specifications, ade-
quate embedding length of tremie pipes in the concrete and slurry properties were maintained to achieve
good-quality concrete casting. Plugging materials and shutters were introduced in the tremie pipes to sepa-
rate first concrete pour and slurry in the trench. Samples of reinforcement bars and concrete were taken as
specified for testing in the lab. 37 barrettes were provided with 6 steel tubes each for sonic logging to test
the barrette concrete quality. No significant anomalies were detected by sonic logging test, and integrity and
quality of barrette were found adequate.
As the site was in a busy public area, site and public accesses were kept clean all the time. Flagmen were
also provided for traffic control at each construction zone. All workers were inducted in safety procedure
prior to assuming duty. Construction activities were also supervised by a full-time safety officer. The safety
officer ensured that all labor used mandatory personal protection equipment and observed the safety regula-
tions. All heavy equipment and cranes were checked for safety and certified for operation prior to use and
for regular maintenance. Additionally, regular safety patrols and inspection were also conducted by the main
contractor and supervising engineers.
CONCLUSION
Construction work on the public road, involving various activities parallel with different
teams/subcontractors demanded comprehensive construction sequences and planning to meet the targeted
milestones. Proper coordination, planning, supervision and strict safety and environmental control by the
parties involved thus brought about the successful completion of diaphragm wall and barrette construction on
one of the busiest roads in Bangkok without any serious accident. When the paper was finalized, excavation
inside the station box had been completed and it revealed good quality of workmanship without requiring
any significant remedial work.
ACKNOWLEDGEMENT
The authors wish to express their appreciation to the MRTA and Nishimatsu Construction Co. Ltd. for
their permission to publish this paper.
REFERENCES
Thasnanipan, N. Maung, A. W. and Tanseng, P. (1998), "Barrettes Founded in Bangkok Subsoils, Construc-
tion and Performance", Thirteenth Southeast Asian Geotechnical Conference, Taipei, Taiwan, ROC, No-
vember 16-20, 1998. pp. 573-578.
298
Under the Royal Patronage of His Majesty the King
Organized by :
Chulalongkhon University
• Department of Civil Engineering
• Department of Environmental Engineering
• Department of Survey Engineering
• Department of Water Resources Engineering
In Corporation with :
Department of Civil Engineering from Universities in Thailand
Engineering for Sustainable

Development
Proceedings of the Seventh National Convention on
Civil Engineering
Monitoring of diaphragm wall displacement and associated ground

movement, brace excavation adjacent to historical building at the bank of
Chao Phraya River
Pornpot Tanseng
Zaw Zaw Aye and Chanchai Submaneewong

ISBN 974-13-0107-3
299
การประชุมวิชาการวิศวกรรมโยธาแหงชาติ ครั้งที่ ๗ จุฬาลงกรณมหาวิทยาลัย ๑๗ - ๑๘ พฤษภาคม ๒๕๔๔
การตรวจวัดการเคลื่อนตัวของกําแพงชนิดขุดเจาะและการเคลื่อนตัวของดินที่
เกิดจากงานขุดดินใกลอาคารประวัติศาสตรริมฝงแมน้ําเจาพระยา
MONITORING OF DIAPHRAGM WALL DISPLACEMENT AND ASSOCIATED GROUND
MOVEMENT, BRACED EXCAVATION ADJACENT TO HISTORICAL BUILDING
AT THE BANK OF CHAO PHRAYA RIVER
Pornpot Tanseng1, Zaw Zaw Aye2 and Chanchai Submaneewong3
1
Lecturer
2
Project Manager, 3Geotechnical Engineer
SEAFCO Company Limited, Bangkok, Thailand
บทคัดยอ: อาคารสมัยใหมที่กอสรางในพื้นที่กรุงเทพมหานครจํานวนมากมักจะออกแบบใหมีชั้นใตดินอยูใกลกับอาคารหรือสิ่งกอสรางเดิม
การขุดดินฐานรากของอาคารตางๆ ดังกลาวโดยใชกําแพงกันดินชนิดขุดและหลอคอนกรีตในดิน (Diaphragm wall) ไดรับความนิยมอยางแพร
หลายเนื่องจากเปนระบบปองกันดินที่สามารถปรับเปลี่ยนเปนโครงสรางถาวรได ภายหลัง การกอสรางจะตองมีการตรวจวัดการเคลื่อน
ตัวระหวางการขุดอยางเปนระบบโดยเฉพาะอยางยิ่งในกรณีที่ตองขุดใกลกับโครงสรางเดิมที่สําคัญและโครงสรางที่เกิดการเคลื่อนตัวงาย บท
ความนี้จะนําเสนอระบบการตรวจวัดการเคลื่อนตัวของกําแพงชนิดขุดและหลอคอนกรีตในดินและการทรุดตัวของดินบริเวณดานหลังกําแพง
และบริเวณอาคารเกาขางเคียงรวมถึงอาคารประวัติศาสตรโดยใชอุปกรณเครื่องมือวัดหลายชนิดในงานขุดเพื่อกอสรางหองใตดินริมฝงแมน้ํา
เจาพระยาซึ่งเปนงานขุดดินในสภาวะที่มีแรงดันดินไมสมดุลย (Unbalanced loading condition) และไดทําการประเมินความเสี่ยงตอความเสียหาย
ตออาคารขางเคียงจากขอมูลการวัดการเคลื่อนตัวที่วัดไดจริงเปรียบเทียบกับกฎเกณฑที่เสนอไวในงานวิจัยกอนหนา รวมทั้งนําเสนอการปรับ
เปลี่ยนขั้นตอนการกอสรางใหสามารถทําการตรวจวัดการเคลื่อนตัววิกฤตไดตลอดเวลา
ABSTRACT: Modern buildings in Bangkok frequently require the excavation works for the basement facility
adjacent to the existing structures. Diaphragm walls have been commonly used as permanent retaining walls for
excavation works in various projects. Comprehensive monitoring system is of essential for any excavation works
particularly for those adjacent to the sensitive buildings. This paper presents the monitoring of diaphragm wall
displacement and associated movement of the ground and adjacent structures including historical building using
a number of instruments for basement construction works located at the bank of Chao Phraya River. The bracing
system and construction sequence adopted for excavation with the consideration of the presence of an
unbalanced loading condition are also briefly discussed. Building damage risk assessment using a simple
approach was carried out based on the instrumentation results and compared with the damage criteria established
by published research work. A course of action undertook in response to the observed critical movement is also
presented.
KEYWORDS: Diaphragm wall, Instrumentation, Unbalanced Loading

For further details, contact Mr. Zaw Zaw Aye, SEAFCO Co., Ltd. 26/10 Prayasuren Road, Bangchan,
Khlong Sam Wah, Bangkok 10510, Thailand
301
1. INTRODUCTION
Excavation works in soft ground inevitably induced the ground movements which in turn may
damage the adjacent existing structures. Diaphragm walls are commonly used as permanent
retaining walls for basement excavation of many projects in Bangkok. It is necessary to use this
type of rigid retaining wall in excavation work to minimize the ground movement particularly
where the work is to be carried out in close proximity to the existing sensitive buildings. This
paper presents the monitoring of diaphragm walls displacement and associated movement of the
ground and the adjacent structures including historical building using a number of instruments
for completed construction works located at the bank of Chao Phraya River, Bangkok. Design
and construction aspects of the project have been reported by Thasnanipan et al. (1999).
2. PROJECT LOCATION AND CONSTRAINTS IMPOSED ON SITE

The project is situated in the heart of an old established culturally-significant area. The project
consisted of basement excavation works for two separate buildings with the support of braced
diaphragm walls.
Chao Phraya River

Chao Phraya River D1-6 D2-4 S2-4
D2-10 S2-10
S2-9 S2-3
I1-6 I2-1 D2-9
D1-1 D2-3
I1-7 T1-1
E1-2
D2-1 T2-10
B3 T2-1 E2-5
S2-1
T1-2 I2-3
ing I2-6 T2-9
First Build
S2-8 B6
I1-8 E2-1 S2-2
Second Building D2-8
5 Storey
T2-2
D1-3 D2-2
T1-3
Building
I1-5
T2-8
T1-4 D2-7 E2-4
I1-2 T2-3
D1-2 T1-8 E1-1 I2-5 S2-6 D2-5 I2-4
S2-5 D2-6 S2-7
E1-4
T1-10 T1-5 T2-7 E2-3
E1-5 T1-9 B5
I1-3 T2-4 B4 T2-5 E2-2 T2-6
B1
Historical
D1-8 0 6 10 20
Building
I1-4
(2 Storey) D1-2
T1-6 Legend
E1-3 Inclinometer in D-Wall Deep settlement plate (5.0m)
B2
Building Tilt plate VW Strain gauge
7 Storey Vertical beam sensor Earth pressure gauge
Surface settlement plate (1.0m) Pre-load location
Figure 1 Layout of excavation sites and instrumentation
Two basements are located 40m apart, separated by an existing 5 storey building. As shown in
Figure 1, construction site is located along the bank of Chao Phraya River and surrounded by
existing buildings including a historical building. Along the river, the distance between
diaphragm walls and existing old river wall is about 4m and 8m for the first and second buildings
respectively. Locations of the excavation sites themselves posed some constraints, which called
for the need of careful consideration in establishing the design principles and sequence of
construction. The major constraints imposed on sites are outlined below.
2.1 Loading condition

Thasnanipan et al. (1999) reported the presence of unbalanced loading condition. As the
excavation areas are located immediately next to the river and surrounded by existing buildings,
the walls alongside the river and the opposite walls were particularly expected to undergo an
unbalanced lateral loading condition. No case history data of similar condition in the area were
available at the time of design and construction of the basements.
2.2 Risk of damage to the existing buildings

The existing buildings located in the proximity of the excavation areas are of old structures.
According to the available information, the buildings are supported by the piled foundation.
However, no detailed information such as type, size and length of the piles was available. The
most critical structure is the historical building which is a predominantly masonry and only 5m
away from the outer face of the diaphragm wall of the first excavation site. As the building is
important for its historical heritage, a damage category worse than negligible (as defined by
Rankin 1988) would not be acceptable.
302
GTE - 54
3. SUBSOIL INVESTIGATION
Two stages of site investigation were made. Preliminary stage of site investigation was
performed during the tendering stage. Second stage of soil investigation was conducted prior to
designing temporary bracing and basement excavation work. As subsoil properties obtained from
the boreholes and test data were well reported by Thasnanipan et al. (1999), they are not
presented again in this paper.
4. DIAPHRAGM WALL AND BORED PILES

For both buildings, 800mm thick cast in-situ concrete diaphragm walls of 28m toe depth with
two level temporary bracings were designed for basement excavation. The maximum excavation
depths for the first and second buildings were 12.7m and 9.7m respectively. Deep diaphragm
walls of toe depth 28m were necessary to ensure the overall stability considering the location of
the basement excavation at the bank of the river. Bored piles of diameter ranging from 800mm to
1500mm founded at 48m depth were constructed by wet process using bentonite slurry.
5. INSTRUMENTATION AND MONITORING PROGRAM

Instrumentation played major role in this project. The primary objective of the instrumentation
program was to monitor the performance of the excavation to ensure that diaphragm wall as well
as the entire excavation system were stable and that adjacent structures were not adversely
affected. Furthermore, the instrumentation program was established to provide the main feedback
in application of observational method. Types of instrumentation used in the project are
presented in Table 1. Layouts of instrumentation for the first and the second buildings are shown
in Figure 1.
Table 1 Quantity of Instrumentation and parameters measured

Type of Instrumentation First Building Second Building Monitored Parameters
Inclinometer 8 6 Wall deflections
Settlement Plate 10 20 Ground settlements
Tiltmeter 10 10 Tilting of Buildings
Vertical Beam Sensor 5 5 Tilting of buildings
VWSG - 1 location* Strain in rebars
Earth Pressure Gauge - 2x2 Strut forces
* A pair of VWSG in 5 layers along the depth of wall panel
6. EXCAVATION METHOD AND BRACING SYSTEM

Conventional bottom–up method with two levels of temporary bracing was applied for both
buildings excavation. As reported by Thasnanipan et al. (1999), to prevent adverse wall behavior
alongside the river due to the presence of unbalanced lateral loading condition, pre-loading was
applied at one end of struts on the land side wall with simple and efficient temporary bracing
system. As the excavation work of the first building was completed 1 year before the
commencement of the second building, monitoring results of the first excavation work provided
an ample opportunity to review the design assumption and fine tune the parameters used in the
analysis of the diaphragm wall for the second building. The major modification was to use single
strut for the first level of the second building (less rigid but economical compared to that of the
first building) with the provision of strut-force monitoring, according to Thasnanipan et al.
(1999). Summarized information of excavation and bracing systems of two buildings is tabulated
below.
Table 2 Summary of excavation and bracing systems
Building Excavatio Bracing Bracing Strut Design Strut Force
n Depth Level Elevation Sections kN/m
First -12.79m I -2.0 2 x WF350 x 350 484.0
II -6.5 2 x WF400 x 400 789.0
Second -9.70m I -2.0 1 x WF400 x 400 279.4
II -7.0 2 x WF350 x 350 332.1
303
GTE - 55
7. SETTLEMENT PREDICTION
Due to the presence of unbalanced loading condition, non-symmetric pattern of settlement
distribution around the perimeter of excavation could be expected. However, considering the fact
that pre-loading was to be used for the land side diaphragm wall but not for the riverside wall,
settlement distribution was assumed to be symmetrical. Available methods were reviewed for
predicting the ground movement with the consideration of two main factors such as, (1) simple
and practical in application (2) enable to correlate with the predicted and measured diaphragm
wall deflection.
7.1 Prediction of Surface Settlement

The method proposed by J. E. Bowels (1990) was selected as it meets the required criteria
mentioned above. Bowels suggested the ground settlement induced by excavation as a function
of ground loss due to the deflection of the retaining wall. Bowel demonstrated the calculation of
settlements at specified distance by assuming parabolic variations of settlement within the
influence distance. Using predicted diaphragm wall deflection, surface settlement behind the wall
was computed by empirical formulas proposed by Bowels.
7.2 Prediction of Sub-surface Settlement
A simplified prediction of sub-surface settlement was made based on the calculated surface
settlement described above. First, subsurface settlement influence line was constructed. As
shown in Figure 2, settlement influence zone is assumed to decrease with depth from “Do” at the
surface and zero at the wall toe. With the assumption of linear relationship between the volume
of deflected wall shape and the volume of settlement trough at any depth within settlement
influence zone, subsurface settlement at different depths were calculated.
X
W D O = HW+ W
Total deflected Surface settlement trough
shape volume, V O volume, V To
Sio
SWO
4V O
S WO =
D-wall DO
Dy
HE
2
 x 
S io = SWO *  
 DO 
HW
Sub-surface settlement D
trough volume, VTy DY = Y * O
Siy
Swy
HW
Settlement
Y
influence zone V
VTY = VY * TO
deflected volumn V y at depth (H w- Y) VO
Figure 2 Demonstration of settlement prediction from diaphragm wall deflection values
8. INSTRUMENTATION RESULTS
8.1 Inclinometer Monitoring
The predicted and observed diaphragm wall deflections of the first building and second buildings
are presented in Figure 3a and 3b respectively. As can be seen in the figures, observed wall
movements exceeded the predicted values at the top portion of the walls for both buildings. The
delay in installing the first level bracings, leaving the walls in cantilever conditions for the long
period was the main reason of the large movement induced by the excavation at that stage.
304
GTE - 56
Displacement (mm) Displacement (mm) Displacement (mm) Displacement (mm)
0 20 40 40 20 0 0 20 40 60 80 80 60 40 20 0
0.00 0.00
RIVER -1.50
-2.00
River
Day 67
5.00 5.00
-6.50
Day 133 -7.00 Day 67
Day 155 Day 133

Day 221
10.00 -9.70 Day 155
10.00 Day 221
-12.70
Depth (m)
15.00
Depth (m)
15.00 Prediction
PREDICTED
OBSERVATION
20.00
20.00
25.00 Day Description
25.00 67 Excavate to -2.5m

133 Excavate to -7.5m
Inclinometer Inclinometer 155 Excavate to -9.70m
221 Remove strut at -7.00m
30.00 No. I-7 No. I-2
30.00
Inclinometer Inclinometer
No. I-2 No. I-4
st
Figure 3a. Lateral wall movements (1 Building) Figure 3b. Lateral wall movements (2nd Building)
At Inclinometer No.I-6 of the second building, wall movement was found to reach the trigger
value at top portion when excavation reached 7.5m depth after the first level bracing installation.
Lateral wall movement reached a maximum rate of 6.5mm/day and a tension crack on the ground
was found about 8m away from the wall. A close examination of the bracing system indicated
that one strut had swayed slightly about 130mm. Immediate actions were taken by stacking
additional strut on the defective one and installing of the second level bracing carried out at that
area. The wall movements were found to cease after these actions.
The wall deflections of the first building excavation found unaffected by the unbalanced lateral
loading condition. However, for second building, monitoring results from inclinometer I-2 and I-
4 suggested that the wall alongside the river was pushed against the retaining soil by the opposite
wall.
8.2 Settlement and Building Movement

Figures 4a, 4b and 4c present the construction activities with time and corresponding monitoring
data. As can be seen in the figures, monitoring results reflected the construction activities in
general. It was also proved that performance and response of the diaphragm wall were largely
influenced by the construction activities. It can be observed that significant changes in the
responses were mainly occurred during the initial excavation stages in which walls were
unsupported for long period. Monitoring data particularly that of settlement and building tilt
measurement provided ample opportunity to assess the risk of the existing buildings damage. As
the measured data were far less than the critical values, the excavation works were proceeded
with confidence.
8.3 Strut Force Measurement

Daily monitoring was carried out for pressure gauges installed on struts. The comparison of
predicted and measure strut forces are presented in Table 3. In general, the measured values
indicated that the bracing system used was adequate.
Table 3 Comparisons of Predicted and Measured Strut Force

Bracing Level Predicted Force Measured Force (KN/m)
(KN/m)
First Level 354.5 331.3
Second Level 307.3 206.3
305
GTE - 57
Excav atio n R e cord Time , D ays Excavation Record Time, Days
0 50 100 150 200 250
0 50 100 150 200 250 300 350
0
2
Exc. De pth, m
4
6
Exc. Depth, m
8
10 Excav. near the river
12 Excava. aw ay from the river
14
Tilt M e te r D ata
200
100
Strut Load Data
Tilt, Se cond
300
0
250
Strut Load (KN)

T 1 - 1 (B 3 ) First Strut
T 1 - 6 (B 2 ) 200
-1 0 0 T 1 - 8 (B 1 )
150
100
-2 0 0
Second Strut
50
a) Tilt meter results with construction time 0
(1st Building) Surface Settlement Data

10
Excavation Record Time, Days
0 50 100 150 200 250 300 350 Settlement, mm 0
0.0
-10
2.0 S6
Exc. Depth, m
-20 S7
4.0
S9
6.0 S10
-30
8.0
10.0
-40
Tilt Meter and EL Sensor Data Inclinometer Data
200 80
Max. Hor. Movement (mm)
I-4
100 60
Tilt, Second
0 40
I-2
E2-2 (B4)
-100 T2-4 (B4) 20
T2-8 (B5)
T2-10 (B6)
0
-200
0 50 100 150 200 250 300 350
b) Tilt meter results with construction time (c) Strut load, settlement and max. wall movement
(2nd Building) with construction time (2nd Building)
Figure 4 Monitoring results of instruments with construction time for First and Second Buildings
9. BUILDING DAMAGE RISK ASSESSMENT

A simple approach was adopted for building damage assessment. Settlement and slope of the
buildings due to the excavation works were predicted assuming the “green filed” condition.
Neglecting restraints from foundation and structure, it was assumed that buildings follow the
ground settlement trough at foundation level (estimated pile tip level). Using predicted and
measured deflection values of diaphragm walls, surface and subsurface settlement profiles were
prepared as shown in Figures 5 and 6. Building slopes likely to be induced by differential
settlement were then predicted as presented in Table 4.
GTE - 58
306
Distance from Wall (m) Distance from Wall (m)
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
0 0
-5 -5
-10 -10
Settlement (mm)
Settlement (mm)
-15 -15
-20 -20
surface settlement
surface settlement settlement at depth 7m
-25 -25
settlement at depth 7m
-30 -30
a) From predicted deflection values b) From measured deflection values
Figure 5 Prediction of surface and subsurface settlement induced by 1st building excavation
Distance from Wall (m) Distance from Wall (m)
0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
0 0
-10 -10
-20 Settlement (mm) -20

Settlement (mm)
-30 -30
-40 -40
surface settlement surface settlement
-50 settlement at depth 7m -50 settlement at depth 7m
settlement at depth 14m settlement at depth 14m

-60 -60
a) From predicted deflection values b) From measured deflection values
Figure 6 Prediction of surface and subsurface settlement induced by 2nd building excavation
Table 4 Predicted Building Slope Calculated from D-Wall Deflection

Predicted Building Slope Calculated
Building Depth of Maximum Tilt Building Slope
from D-Wall Deflection
No. Settlement Measured by Tilt Criteria for
From From Inclinometer
Considered* Meter / EL Beam Negligible Damage**
Prediction Measurement
At surface 1/1778 1/2000
B1 1/2082 < 1/500
At depth 5m 1/1778 1/2133
At surface 1/1583 1/1827
B2 1/4125 < 1/500
At depth 7m 1/1583 1/2879
At surface 1/1563 1/1667
B3 1/2578 < 1/500
At depth 7m 1/1667 1/2778
At surface 1/880 1/880
B4 1/1412 < 1/500
At depth 7m 1/978 1/1257
At surface 1/944 1/1700
B5 At depth >1/5000 < 1/500
1/1771 1/4250
14m
At surface 1/926 1/926
B6 At depth 1/3437 < 1/500
1/2256 1/4400
14m
Note : (1) * assumed foundation level (2) ** unlikely to have superficial damage (Rankin 1988)
As can be seen in Table 4, the measured maximum tilt of the buildings reasonably agree with
predicted building slopes derived from inclinometer measurement, particularly if the buildings
were assumed to be settled by subsurface settlement at assumed pile tip levels. All buildings
307
GTE - 59
were predicted to be in negligible risk category as classified by W.J. Rankin (1988). Field
inspection confirmed that no damage was caused in all buildings from the minor movements
induced by excavation works.
10. CONCLUSIONS
Instrumentation played a major role in execution and successful completion of excavation works
adjacent to the existing sensitive buildings under unbalanced loading condition. Effective bracing
system, excavation sequence and adequate embedment of the retaining walls were the main
factors contributed in minimizing negative impact of the unbalanced loading condition.
A simple approach in prediction of ground settlement and assessment of the risk of building
damage has been presented. The measured maximum tilt of the buildings reasonably agreed with
the predicted values. It is necessary to use the rigid diaphragm wall in excavation work to
minimize the ground movement particularly where the work is to be carried out in close proximity
to the existing sensitive buildings.
REFERENCES
1. Bowels J.E., Foundation Analysis and Design, 4th Edition, McGraw-Hill, Singapore, 1990.
2. Rankin W.J., Ground Movement Resulting from Urban Tunnelling : Prediction and Effects,
Engineering Geology Special Publication No.5, p. 79-92, 1988.
3. Thasnanipan, N., Maung A. W., Tanseng P. & Teparaksa W., Behavior and Performance of
Diaphragm Walls under Unbalanced Lateral Loading along the Chao Phraya River , Field
Measurements in Geomechanics, Leung, Tan & Phoon (eds), p. 267-272, Balkema, 1999.
308
GTE - 60
Proceedings of the 12th Asian Regional Conference on Soil Mechanics
and Geotechnical Engineering
4-8 August 2003, Singapore
th
12 Asian Regional
Conference on Soil Mechanics
and Geotechnical Engineering
Performance of Buttress-Support Thin

Diaphragm Wall for Underground Car Park in
Bangkok
Narong Thasnanipan, Zaw Zaw Aye and Chanchai

Submaneewong
309
Performance of Buttress-Support Thin Diaphragm Wall for Underground
Car Park in Bangkok
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong

SEAFCO Company Limited, Bangkok, Thailand, seafco@seafco.co.th
Abstract : This paper presents the geotechnical aspect of the construction of two-level underground car park building located in the cul-
turally and historically significant area of Bangkok. Performance of buttressed-support diaphragm wall of 0.60m width is reported
based on the inclinometer monitoring results. Intensive modification of construction sequence in actual work execution with “value
engineering options” different from tender stage design is demonstrated along with application of observational method.
1 INTRODUCTION 3 MAJOR CONSTRAINTS AND DESIGN PRINCIPLE
Diaphragm walls have been utilized for underground permanent Location of the project site itself in the vicinity of sensitive
retaining structures in Bangkok for many years. The common buildings posed some constraints, which called for the need of
width of the diaphragm wall in Bangkok ranges from 0.80m to careful consideration in establishing the design principles and se-
1.20m. Thin diaphragm wall of width less than 0.80m is rarely quence of construction. The architectural and utility aspects of
used in this growing-metropolis. Performance of buttress-support the project called for the design of the basement with a number
diaphragm wall of 0.60m width for two level underground car of openings from the ground surface to the final basement slab
park building construction is discussed in this paper. level to facilitate the ventilation system as shown in figure 1.
Hence the roof slab can not be physically utilized as bracing in
most of the area where the diaphragm wall is to be acting as a
2 PROJECT OVERVIEW cantilever retaining wall in the permanent stage. It was analyzed
in the preliminary analyses that the deflection of the diaphragm
The project is a two-level underground car park located in the wall of 0.60 m width would be large if it was to be fully cantile-
center of Rattanakosin Island, the heart of an old established, his- vered. As the project is located in a sensitive area, ground
torically and culturally significant area of Bangkok. As shown in movement induced by large deflection was unfavorable. It was
Figure 1, construction site is surrounded by numbers of sensitive therefore decided to use buttress to minimize the diaphragm wall
structures - in the south, Wat Suthat, one of Thailand’s most im- deflection as shown in figure 2.
portant temples and the Historical Giant-swing, in the north, the
City Hall and at South-east corner, the Historical Brahmin Tem- CITY HALL
CITY HALL
N
ple. Rows of old shop-house buildings are closely located in the
MAHUNNOP ROAD
east and west boundaries of the project.
The project owner, Bangkok Metropolitan Authority (BMA)
OPENING
OPENING
awarded the semi-turnkey basis construction contract to
SHOP-HOUSES
SEAFCO Co., Ltd. as a contractor. The contract consists of 3 ma-

jor scope of works : (1) Construction of diaphragm wall, barrette
OPENING
DINGSOR ROAD
OPENING
and bored piles (2) Excavation works including temporary brac-

SIRIPONG ROAD
SHOP-HOUSES
ing design and installation (3) Construction of the entire base-

ment structure having car park area of 18,552 m2 and roof-level
park of 10,936m2 plus cut-and-cover tunnel, underpass access to
OPENING
OPENING
the City Hall. Since the project area was being used as a grade-
level car parking space prior to the award of the contract, the key
BRAHMIN TEMPLE
requirement was to construct the underground car park in two

phases – to construct Phase 1 while leaving space for car parking
in Phase 2, and to utilize semi-finished underground car park of OPENING OPENING
Phase 1 during construction of Phase 2.

Diaphragm wall having 600mm width founded at 16m below BAMRUNG MUANG ROAD
ground level (B.G.L) was constructed simultaneously with dry-
GIANT-SWING
processed bored piles of diameter 600mm with toe depth 20m
B.G.L. Barrettes having same toe depth as bored piles were in-
stalled at 8m spacing along with diaphragm wall panels. Sheet
pile wall (14m deep) was used as a temporary retaining wall at
the boundary of Phase 1 and 2 as shown in figure 2. Fig. 1. Layout plan of the project showing adjacent buildings
1
311
first strut and final excavation level for 600mm diaphragm wall)
N I-5 NORTH D-WALL due to large volume of excavation work involved.
6 POST TENDER DESIGN REVIEW AND MODIFICATION

I-4
PHASE 2 I-6 FOR PHASE 1
EAST D-WALL
WEST D-WALL
SHEET PILE
Prior to the commencement of excavation works, value engineer-
ing review of the temporary works was undertaken by the con-
124 m
tractor’s new team of design engineers together with the con-
See details in (b)
struction team. The main objectives were to minimize the
I-3 I-1
material and construction sequence involved in temporary works
PHASE 1
so as to accelerate the excavation time thereby saving overall
costs without compromising the safety aspect. After conducting a
series of re-analyses with different conditions major modifica-
tions were made : (1) To change first strut level to –1.8m from
I-2
SOUTH D-WALL
78 m the original tender stage design level –1.0m (2) To use only 1
temporary strut, omitting second level raking strut with the pro-
(a) vision of sloping soil berm against diaphragm walls. Soil berm
D IA P H R A G M W A L L A RC BUTTRESS
was to remove after completion of base slab construction in the
BARRETTE W ALE BEAM majority of area – minimizing the elapsed time of partially un-
0 .3 0 strutted diaphragm wall between temporary strut and the final
excavation level.
T IE B E A M
7 APPLICATION OF OBSERVATIONAL METHOD

P IL E C A P SLAB
In order to assess the most probable condition assumed in “value
(b) engineering design” and to take necessary actions if monitoring
+0.20 ROOF SLAB
0.00 results reveal most unfavorable conditions (i.e. actual deflection
G.W.L CAPPING
BEAM of diaphragm wall reaches maximum acceptable limit), observa-
-2.70 SLAB (B1)
RC BUTTRESS
tional method was implemented on the followings basis.
2
TIE PILE CAP • Reviewed the design parameters together with critical condi-
-5.60 (B2)
2
BEAM
tions posed on sites as well as most critical stage of excavation
-9m WALE SOFT CLAY Su = 18 kN/m2
Eu = 7000 kN/m2, γ = 16.5 kN/m3
• Designed and predicted the performance of diaphragm wall us-
BEAM
ing the “most probable” conditions and parameters.
DIAPHRAGM MEDIUM CLAY Su = 30 kN/m2
-12m WALL Eu = 19250 kN/m2, γ = 17.5 kN/m3 • Established the trigger criteria based on predicted diaphragm
STIFF CLAY Su = 60 kN/m2
wall deflection
FOUNDATION
PILES
Eu = 45000 kN/m2, γ = 19 kN/m3
• Predefined the practical contingency plan for “most unfavor-
-16m
DIA.600MM.
able” conditions where wall deflection reaches trigger levels
BARRETTE • Set out the instrumentation program with the consideration of
-20m.
above factors.
• Monitored the performance of diaphragm wall. Compared the
(c) monitoring results with the predicted and trigger values and re-
Fig. 2 (a) Layout of diaphragm wall and piles (b) plan of dia- assessed
phragm wall and buttress (c) sectional view of the structures • Implemented the contingency measures if monitoring result
reaches action level of trigger values
4 SUBSOIL CONDITION Comprehensive and robust monitoring program was set up as
a key element in application of the observational method and to
Similar to other localities in Bangkok a typical subsoil profile at ensure that modified construction sequence would not have ad-
the site is characterized by thick Bangkok soft clay layer at the verse effect in temporary stage and on permanent design. A total
top followed by thin layer of medium clay, and stiff clay layers of 6 inclinometers (3 in each phase) were installed in diaphragm
as depicted in figure 2 (c). The first sand layer is found below wall together with some survey points.
26m from the existing ground level. HORIZONTAL STRUT SLAB
5 TENDER STAGE DESIGN RAKING STRUT

D-WALL
The designers involved in the tender stage design made a fairly BERM
conservative design with two levels temporary bracing as shown
in figure 3. Uncertainty of the performance of a thin diaphragm D-WALL
BARRETTE
wall in soft clay layer was the likely reason to adopt the conser- -16m
vative design in the tender stage. It is not unreasonable to adopt
the conservative design considering rheological property of soft -20m BORED PILE
marine clay and likely long elapsed time of un-strutted dia-
Fig. 3 Tender stage bracing system – diaphragm wall was de-
phragm wall (relatively long un-strutted span of 5.6m between
signed to be supported by 2 struts (horizontal bracing and raker)
2
312
HORIZONTAL STRUT SLAB HORIZONTAL STRUT SLAB 8 PREDICTION AND PERFORMANCE OF PHASE 1 DIA-
PHRAGM WALL
BERM
RAKING STRUT Figure 5 (a) shows the maximum predicted diaphragm wall
D-WALL
deflection at different stages of excavation monitored by incli-

SHEET PILE nometer No.1 (I-1 at East wall of Phase 1) together with pre-
BARRETTE
-14m dicted maximum deflection profile of 3 different conditions -

-16m -16m
tender stage design (2 temporary struts), modified design with
-20m -20m BORED PILE buttress and modified design without buttress. It can be observed
EAST & WEST D-WALL SOUTH D-WALL vs SHEET PILE
from figure that measured lateral movement pattern of diaphragm
wall agreed well with that of prediction for modified design with
Fig. 4 Bracing system for Phase 1diaphragm wall buttress - meaning buttress-support has some influence on wall
deflection. Deflection profile of South diaphragm wall which
Cum. Displacement , mm. Cum. Displacement , mm.
braced against temporary sheet pile wall at Phase 1 and 2
0 20 40 60 80 100 0 20 40 60 80 100
0 0
boundaries is presented in figure 5 (b). As can be seen in figure 5
(b), South diaphragm wall deflection is significantly higher than
2 2
that of east wall, which is likely to be caused by the fact that
4 4
South diaphragm wall is braced with more flexible sheet pile
6 6 wall. Description of stages shown in the legend of figure 5 is
8 8 summarized in table 1.
Depth, m.
Depth, m.
10 10
12 Stage 1 12
Stage 1
Stage 2
9 MODIFICATION OF PHASE 2 BRACING SYSTEM
Stage 2 Stage 3
14 Stage 3 14 Stage 4 Monitoring results of the Phase 1 excavation work provided an
Stage 4 Stage 5
16 Stage 5
16 with buttress ample opportunity to review the design assumption, fine tune the
18
with buttress 18
without buttress parameters used in the analysis of the diaphragm wall for the
without buttress Tender Stage
Tender Stage 20
Phase 2 and made modification of construction sequence. The
20
major modifications are : (1) Removal of soil-berm at East and
(a) East diaphragm wall (b) South diaphragm wall West diaphragm wall in shorter duration than that of Phase 1, and
(2) Using raking struts instead of horizontal strut for North dia-
Fig. 5 Phase 1 - Monitored diaphragm wall deflection at differ- phragm wall a shown in figure 7.
ent stages with predicted maximum deflection profiles of differ-
ent condition (with buttress, without buttress and tender stage) 10 PERFORMANCE OF PHASE 2 DIAPHRAGM WALL
Figure 8 (a) depicts the measured deflection of east diaphragm

Table 1. Summarized description of stages shown in figure 5
wall. As can be observed in figure 8 (a) in comparison with Fig-
Stage East Diaphragm wall South Diaphragm wall ure 5 (a), the maximum deflection of east diaphragm wall in
1 Excavate to –2.2m and Excavate to –2.2m and Phase 2 is larger than that of Phase 1. The likely reasons of this
installed temporary strut installed temporary strut observation are;
• Un-strutted elapsed time for first temporary bracing in Phase
2 Excavate to –4m Excavate to –4m 2 was longer than that of Phase 1.
3 Excavate to –6.6m with Excavate to –6.6m with • In Phase 1, horizontal struts were installed in north-south di-
berm berm rection which temporary kingpost columns and strut were in-
4 Removal of berm Removal of berm tegrated in crisscross pattern with east-west struts - providing
complete-support more rigid bracing system. Whereas in
5 Removal of temporary Removal of temporary Phase 2, horizontal struts were installed only in east-west di-
strut after completion of strut after completion of rection without having crisscross pattern with north wall –
buttress buttress having less rigid bracing system than that of Phase 1.
With assurance of diaphragm wall performance from moni-

toring results of Phase 1, original plan of using horizontal struts
for North diaphragm wall was modified by using raking struts in-
stead of horizontal struts. As can be seen in figure 8 (b), deflec-
tion of North diaphragm wall (with raking strut support) is sig-
nificantly higher than that of East diaphragm wall (with
horizontal strut support). The main reason of larger movement of
north diaphragm wall is due to the fact that it was supported only
by the berm for the long period (about 52 days) before comple-
tion of raking struts so that soil-rheology effect took place during
the long elapsed un-strutted period. Time-dependent deflection
pattern due to soil-rheology effect can be observed in North dia-
phragm wall as illustrated in figure 9. North diaphragm wall
Fig. 6 Phase 1 excavation work in progress with Wat Suthat in moved progressively toward excavation before completion of
background (Phase 2 area was being used as car parking space) raking struts (at 52 days) as can be seen in figure.
3
313
HORIZONTAL STRUT SLAB COMPLETED PHASE 1
TIE BEAM CONSTRUCTED IN BAYS
STRUCTURE TEMPORARY WALER
BUTTRESS
RAKING STRUT DIAPHRAGM WALL
DIAGONAL STRUT
BERM SLAB CAPPING BEAM TEMPORARY KING-POST

BERM
RAKER
D-WALL
TEMPORARY
D-WALL BERM STEEL BEAM
BARRETTE
PILE
-16m -16m CAP
LAB
-20m -20m PLE
TED
S
COM
EAST & WEST D-WALL NORTH D-WALL with raker support
Fig. 7 Bracing system of Phase 2 diaphragm walls

0 20 40 60 80 100 0 20 40 60 80 100 120 140

Fig. 10 Perspective of North diaphragm wall constructed in bays
0 0
2 2
4 4
6 6
11 IMPLEMENTATION OF CONTINGENCY PLAN
8 8
Depth, m.
Depth, m.
10 10
Since deflection of North diaphragm wall (Phase 2) approached
12
Stage 1
Stage 2 12 Stage 1 trigger levels, monitoring frequency was increased and the fol-
14
Stage 3
14
Stage 2 lowing contingency measures were implemented on site.
Stage 4
16 Stage 5 16
Stage 3 • Poured 15cm thick 1m wide lean concrete on the top of the
with buttress Stage 4
18 without buttress 18
berm along North diaphragm wall to provide bearing-effect
Prediction
20 20
• Installed additional king-post and diagonal struts attached to
the raking struts to provide more rigid support against dia-
(a) East diaphragm wall (b) North diaphragm wall
phragm wall
Fig. 8 Phase 2 - Monitored diaphragm wall deflection at differ- • Soil-berm was removed locally in bays followed by construc-
ent stages with predicted maximum deflection profiles of differ- tion of wale beam, tie beam and buttress as shown in figure 10
ent condition
Cum. Displacement , mm.
Movement of diaphragm wall movement was observed to be
0 20 40 60 80 100 120 140 decreased by the above actions. No significant ground settlement
0 was observed in the vicinity of the North diaphragm wall.
2
6 12 COST AND TIME SAVING FROM VALUE ENGINEER-

ING OPTIONS
Depth, m.
10
0 days
Significant cost and time saving were achieved from the value
12
30 days engineering options coupled with observational method imple-
14
52 day (raker installed) mented for both Phase 1 and 2. The major savings were achieved
16
56 days by less operation and material utilized in the following elements
18 74 days of temporary works.
20 Prediction • Cancellation of 2nd level raking struts against diaphragm wall
Fig. 9 Phase 2 – Time dependent wall deflection of North dia- for both phases
phragm wall – diaphragm wall was supported only by soil-berm • Modification of bracing system – using raking struts instead of
for 52 days before completion of raker installation. horizontal struts for North diaphragm wall in Phase 2
Table 2 Summarized description of stages shown in figure 8

13 CONCLUSION
Stage East Diaphragm wall North Diaphragm wall
1 Excavate to –2.2m and Excavate to –2.2m at d- This research study reveals that a permanent diaphragm wall
installed temporary strut wall, and to –6.6m with coupled with effective design and construction method supple-
sloping berm mented by observational method could offer a logistically and fi-
nancially attractive solution in construction of underground car
2 Excavate to –4m Installed raking strut
park without disturbing the environment in the prominent histori-
3 Excavate to –6.6m with Removal of berm cal area of Bangkok.
berm
4 Removal of berm Removal of raker after
completion of buttress
5 Removal of temporary
strut after completion of
buttress
4
314

(15 SEAG)
Edited by:
Sambhandharaksa, S.
Bergado, D. T.
Boonyatee, T.
Stability of a Trial Trench Excavated under

Polymer Slurry in Bangkok Soft Clay
Narong Thasnanipan, Aung Win Maung, Thayanan

Boonyarak and Zaw Zaw Aye
315
Stability of a Trial Trench Excavated under Polymer Slurry in Bangkok Soft
Clay
Thasnanipan, N., Maung, A. W., Boonyarak, T., and Aye, Z. Z.
Seafco Public Company Limited, 26/10 Rarm Intra 109 Road, Bangchan, Khlong Sam Wah, Bangkok 10510,
Thailand.
seafco@seafco.co.th
Abstract: This paper presents findings on trench stability resulting from the use of polymer slurry. A trial trench 0.60x4.50x18.00m
was excavated for diaphragm wall construction in Bangkok soft clay, using polymer as an alternative to bentonite slurry. Under
consideration was its superior environmental advantages for soft clay soils typical of the region, notwithstanding a limited capacity to
suspend soil particles. The trench was excavated with a cable suspended mechanical grab, filled with polymer slurry and left open for
4 days. Field measurement included recording of slurry properties, settlement surveys, inclinometer monitoring, periodical checking of
the trench profile with Koden drilling monitoring equipment and trench depth sounding. Polymer slurry properties in the trench were
found to change sharply within 12 hours, with density, pH and viscosity gradually decreasing after 12 hours. Sedimentation occurred
within a few hours after completion of trench excavation and a 7.00m thick deposit was indicated at the bottom of trench after 10.5hrs.
The deposit was cleared with the excavation grab but sedimentation recurred. The trench profiles plotted by Koden drilling monitoring
equipment showed trench surface failure starting at depths between 7.00m and 7.50m on one of the long sides of the trench face 18
hours after trench excavation, with inclinometer readings indicating maximum soil yielding of about 6mm after 10 hours. A major
collapse of the trench surface was observed approximately 36 hours after trench completion, with ground settlement of up to 9mm.
Based on field measurement results and characteristics of the polymer used, the suitability of using polymer slurry is discussed.
1. INTRODUCTION 2. SITE CONDITIONS
Polymeric fluids have been in use as support slurries to provide A 25m deep borehole was drilled to investigate subsoil
stability during excavation for bored piles and diaphragm walls. conditions at the trial trench location prior to trenching. The
The merits of such polymeric fluids have been evaluated and investigation indicated that a soft clay layer extended up to
reported by Beresford. J.J and et al in 1989. Polymeric slurries 13.5m below the existing ground level, overlaying a 3.0m thick
are environmentally compactable and acceptable. The used stiff clay layer with SPT values ranging from 10 to 20.
polymer fluid can be disposed directly into the municipal sewer Sensitivity of soft clay ranged from 2.3 to 3.5. Below the stiff
without harmful consequences to water courses. Excavated clay layer was very stiff clay 4.5m thick with SPT values ranging
materials can also be safely dumped in unregulated landfill sites. from 17 to 27, underlain by a fine to medium sand layer at a
As the polymer slurry offers considerable advantages in terms of depth of 22.0m. SPT values obtained in the sand layer ranged
extending operation capabilities, logistics, environmental from 27 to 54.
compatibility and overall cost-saving, as well as enhancement to Undrained Shear Strength (kN/m2) Atterberg Limit
frictional capacity of bored piles by providing intimate contact Depth (m) Soil Description 0 20 40 60 80 100 120 0 20 40
%
60 80 100 120
between pile shaft and surrounding soil, particularly in 0
Crust
cohesionless soil without forming a filter cake, it has been used
in bored piling work in Bangkok and elsewhere in Thailand 5
Soft Clay
(Ref). For diaphragm wall work, bentonite slurry is still used as
the main supporting slurry in Thailand. Polymeric slurry usually 10
M. Stiff Clay
purges itself of soil particles and particle settling takes place
15 Stiff Clay
whilst in the trench. Thus a special process to remove the
particles deposited at the bottom of trench is mandatory as a 20
Very Stiff Clay
conventional mechanical grab will not effectively remove the

sediments in such cases. For some basement excavation projects 25
M. Dense Sand
in Bangkok subsoils with a maximum excavation depth of up to 0

UCS
10 20 30 40 50 60 Natural water content
Liquid Limit
SPT N (Blow/ft)
10.0m, diaphragm wall systems constructed predominantly in SPT Plastic Limit
clay layers are required. Hence the use of polymer was

considered a practical option for diaphragm wall excavation in Fig. 1. Subsoil conditions at the trial trench site (Bangchan).
clay. A trial trench (0.6x4.5x18m) was then excavated under
polymer slurry in soft clay at Bangchan, north-eastern Bangkok, Natural water content in the soft clay generally ranges from 60-
to evaluate the performance of the slurry. 80%. However, at depths of 5m and 8m the natural water
Performance of trial trenches under bentonite slurry in soft content is slightly above or about 100% respectively.
clay have been reported previously by Dibiagio and Myrvoll Groundwater level observed in the borehole was at about 2.5m
(1972) and Thasnanipan et al (1998). below the ground.
317
3. TRIAL TRENCH the equipment was in a range of +/- 0.2% (in this case
the accuracy was within 1-2mm).
The trial trench was located about 1.0m away from the • Slurry testing – this was carried out on the viscosity,
borehole. A concrete guide wall of 1.5m deep was constructed density and pH of the slurry sample taken from the
with continuous reinforcement and cast against firm ground. trench at a depth of 9.0m.
The trial trench was 0.6m wide, 4.5m long and 18.0m deep. For • Ground settlement monitoring at 6 settlement points -
field monitoring, all monitoring points, comprising 6 settlement Monitoring was carried out approximately every 2 hrs
survey points using one inclinometer tube were located on the during excavation and every 6hrs after excavation.
opposite site of the excavation equipment to avoid disruption • Checking trench depth – Depth sounding at 3
caused by the excavation activities during trenching operation. locations, the center and near both ends of the trench
Locations of settlement survey points and inclinometer tube are was carried out every 30minutes after excavation.
shown in Figure 2. • Monitoring lateral soil movement with inclinometer –
An inclinometer access tube was installed in the soil
investigation borehole to a depth of 24.0m.
Trial trench 0.60 x 4.50 m
Monitoring was carried out every about 2 hrs during
excavation and at 2-6hrs of elapsed time after
I1
excavation.
P1 P2 P3 P4 P5 P6
5. TESTING AND MONITORING RESULTS
1m
2m 2m 2m 2m
The trial trench excavation was carried out on 16th of June 1999
at 14:00hrs and completed at 23:30hrs. Sedimentation occurred
Legend
Settlement Point
within a few hours after completion of the trench and sediment
P5 built up to 11.0m in depth (i.e. deposit thickness was 7.0m) after
I1
Inclinometer
10.5hrs. The estimated rates of settlement of soil particles
ranged from 1.0m to 1.4m per hour in thickness for the first 4hrs
Fig. 2. Layout of field monitoring locations. after trenching and then gradually decreased. Cleaning of the
trench base was carried out 14hrs after trench completion.
However, sedimentation recurred and built up to a depth of
Trenching was carried out on 16th June 1999 at about 14:00hrs 13.0m after 12hrs. 24hrs after base cleaning (36hrs after
under commercially available polymer slurry and completed at excavation), sedimentation built up to a depth of 6.9m due to
about 23:30hrs on the same day. A cable suspended mechanical trench face failure.
grab (0.6mx2.0m) was used for trench excavation. 3 bites
excavation process was used to achieve the required trench
Time after excavation (hrs)
0 12 24 36 48 60 72 84 96
dimensions. 0.00
The polymer type used for trial trenching was a water-soluble
Trench Depth below ground level (m)
synthetic polymer supplied as a concentrated liquid. As per by

supplier recommendations, a mix ratio of 0.5kg of polymer per 1 5.00
cu.m. of slurry was used for slurry preparation. The properties
of slurry were tested for compliance values specified for
excavation prior to introducing it to the trenching process. 10.00
after trench face collapsed
The slurry level in the trench was maintained up to ground
level at all times. The rate of raising the grab during excavation
was controlled to minimize the suction action which would 15.00
impact on the trench stability. The trench profile was checked
with Koden drilling monitoring equipment when excavation after trench base cleaning
reached a depth of 9.0m. The polymer slurry was also tested for 20.00
its quality. Trenching was then continued to the final depth of
18.0m. The verticality of the trench indicated by the drilling Fig. 3. Plots of sedimentation with. time.
monitoring equipment was within 1:180.
The density of polymer slurry in the trench was found to
4. FIELD TESTING PROGRAM change sharply within 12 hrs from 1.30g/ml to 1.08g/ml and then
gradually decrease. 36hrs after trench completion, this reached
Field monitoring and slurry testing were generally carried out 1.0g/ml. Similarly, viscosity of the slurry changed from 43sec to
when trenching reached a depth of 9.0m and the final depth of 32sec within 12 hrs and decreased to 29sec after 36hrs. A Marsh
18.0m, and then every 6hrs after trenching was completed. The Funnel viscometer was used for indicative measurement of
following field testing was performed. relative changes in viscosity only, rather than to evaluate the
slurry performance. The pH value also changed from 8 to 7 after
• Checking trench profile with Koden drilling 12 hrs. After 36 hrs, the properties of the slurry remained
monitoring equipment – the equipment was stationed constant till the trench was backfilled. It is considered that after
in a fixed position on the guide wall to monitor the 36hrs the properties of slurry indicated that the slurry in the
changes in trench profiles. The recording accuracy of
318
trench turned to water, purging all polymer constituents with soil
particles. P1 P2 P3 P4 P5 P6
Distance from trench side (m)

50 0 1 2 3 4 5 6 7 8 9
0
40 1
Time after
2 trench
Viscosity (sec) excavation
30 3
2 Hrs
Settlement (mm)
4 12 Hrs
20 5 24 Hrs
3 36 Hrs
Density (kN/m ) 6
48 Hrs
10 7 60 Hrs
72 Hrs
pH 8
84 Hrs
0 9 96 Hrs
0 12 24 36 48 60 72 84 96
10
Elapse time (hrs)
Fig. 4. Plots of polymer slurry property changes with time.
Fig. 6. Plots of ground settlement readings.
Plots from the inclinometer readings indicated lateral

movements of soil up to 5.8mm at a depth of 3.5m. However, Excavation width 0.60m Excavation width 0.60m Excavation width 0.60m
actual soil movements at the face of the trench may have been
greater than those measured by the inclinometer located 1.0m
Depth below ground level (m)
away. A maximum ground settlement of 9.0mm was observed

2.0m away from the trench after major failure of the trench face
opposite site settlement points. No ground settlement was Trench base -6.8 m
observed beyond the distance of 8.0m. Trench base -11.9m
Soil Displacement (mm)
0.00 1.00 2.00 3.00 4.00 5.00 6.00
0.0
Trench base -18.0m

5.0
0 Hr after excavation 96.5 Hrs after excavation (end of test)
Depth below Ground Level (m)
36.0 Hrs after excavation
10.0
Fig. 7. Profiles of trial trench plotted by Koden drilling
monitoring equipment at different times.
Excavated to -9.0m
15.0 Excavated to -12.0m
Excavated to -15.0m 6. DISCUSSION
0 Hr after trenching
4 Hrs
20.0 6 Hrs Excavation of the trial trench was mainly in clay layer, which is
12 Hrs non-permeable. Moreover, clay normally maintains a relatively
79 Hrs
stable ground condition apart from elastic deformation due to
25.0
reduction of confining pressure. The movement can be
controlled by hydrostatic pressure conceptually in diaphragm
Fig. 5. Lateral soil displacements plotted from inclinometer
wall excavation. Compared to bentonite, polymer contributes
readings.
relatively little additional density to the water in which it is
prepared, and is just as effective in holding back the plastic
deformation of the ground.
The trench profile plotted by the monitoring equipment
Empirical calculations were carried out for trench stability
shows that on the opposite side of instrumented monitoring
analysis based on the available soil test data with planned trench
points (inclinometer and settlement survey points), trench
dimensions prior to excavation. The analysis indicated that
surface failure started at depths of 7.0-7.5m about 18hrs after
polymer fluid pressure with a density of 1.0 t/m3 (reduced to
trench completion. The surface failure progressively increased
density of water) was still adequate to maintain the stability of
and a major collapse was observed at depths between 2.0m and
the trench. However, high natural water content approximating
8.0m after 38hrs. The trench surface failure was found to be a
the liquid limit or exceeding moisture content of 100% in soft
wedge failure, extending more than 1.0m into the soil face.
clay at depths between 3.5m and 9.0m was a major concern for
Likewise, after 42hrs had elapsed, a localized collapse was also
trench stability. In the past, occurrence of localized failure or
observed on the instrumented monitoring side of the trench at
trench collapse in such soil conditions under bentonite slurry was
6.0m below ground level, increasing progressively to a depth of
experienced when pore pressure and/or density of slurry
about 5.0m below ground level till the monitoring was
significantly changed after heavy rain.
terminated at 96hrs after excavation.
319
Instability of trench was not encountered in the trial trenches It is considered that high natural water content of the soil,
which were excavated in soft clay and under-consolidated which approximated the liquid limit or exceeded 100% posed a
marine deposits under bentonite slurry as reported by Dibiagio & risk to trench stability and contributed to the failure of the trial
Myrvoll (1972) and Cowen et al (2001) respectively. The soil trench.
properties and movements of the trial trenches, including the An efficient cleaning mechanism is considered necessary to
trial trench reported herein are compared where possible and remove the soil particles settling at the bottom of the trench
presented in Tables 1 and 2 respectively. under polymer slurry.
Table 1. Comparison of Soil Properties at Trial Trench Sites

ACKNOWLEDGEMENT
Test Location
The authors wish to acknowledge the contribution of EDE
Depth Soil Studenterlunden Khlong Bangchan, Company limited for field monitoring and assistance in trial
m Properties Norway Prapa, North-East trenching at Khlong Prapa and Bangchan sites.
North Bangkok
Bangkok
0-10 γ t/m3 1.60-1.86 1.60-1.80 1.37-1.80 RERERENCES
Su t/m3 4.0-8.0 1.3-2.5 1.6-2.0
WC % 40-45 40-70 40-105 Beresford, J. J., Cashman, P. M., and Hollanby, R. G. 1989. The
LL % 45-50 60-70 50-85 Merits of Polymeric Fluids as Support Slurries.
10-20 γ t/m 3
1.82-1.91 Below 1.6-2.0 Proceedings of the International Conference on Piling and
Su t/m3 2.0-5.0 14.0m is 2.0-18.7 Deep Foundations, London, 15-18 May 1989. pp 3-10.
WC % 40-42 medium 21-71 Cowan, S., Littlechild, B. and Green, E. 2001. Trial Diaphragm
LL % 40-50 dense sand 43-84 Wall Panels in Under-Consolidated Marine Deposits. HKIE
20-30 γ t/m3 1.81-1.96 Below Annual Seminar Geotechnical Deformations and
Su t/m3 4.0-6.0 22.0m is Movements 2001. pp. 179-192.
WC % 40-50 medium Dibiagio, E. and Myrvoll, F. 1972. Full Scale Field Test of a
LL % 40-55 dense sand. Slurry Trench Excavation in Soft Clay. Proceedings of 5th
ECSMFE, vol.1, pp. 461-471. Madrid.
Thasnanipan, N., Maung, A. W. and Tanseng, P. 1998. Barrettes
Table 2. Comparison of Ground movements Founded in Bangkok Subsoils, Construction and
Performance. Thirteenth Southeast Asian Geotechnical
Item Studenterlunden Khlong Prapa, Bangchan, Conference, 16-20 November 1998, Taipei, Taiwan: pp
Norway North North-East 573-578.
Bangkok Bangkok
Trench 1.0x5.0x28.0 0.8x2.7x14.0 0.6x4.5x18.0
Dimension in
m
Max. Ground 8mm - 9mm
Settlement
Max. Soil 55 20-30 6
yield in mm (after 31 days) (after 24 hrs) (after 10hrs)
Supporting Bentonite, Bentonite Polymer
Fluid Water after 32days
Most of the monitoring points for Bangchan trial trench were

located on one side of the long trial trench face. The trench
profile as shown by drilling monitoring equipment indicated that
the collapse of the trench face occurred initially at the side
opposite the monitoring points. It is considered that the recorded
maximum values of results may not have represented the actual
maximum values of soil movements occurring around the trench,
particularly on the side of the trench without monitoring
instrumentation due to lack of monitoring data.
7. CONCLUSIONS
Based on the field monitoring results obtained from the trial

trench excavated under polymer in Bangkok soft clay, the
following conclusions were made.
The trial trench was found stable up to 18hrs after completion
of excavation under polymer fluid at this site and after that the
stability of trench deteriorated with time.
320
Proceedings of International Symposium on Underground Excavation
and Tunneling
2-4 February 2006, Bangkok, Thailand
Urban Tunnel Construction

for Protection of
Environment
Edited by:
Phienwej, N.
Application of Top-down Construction Method

for Deep Excavations in Bangkok
Thasnanipan N., Aye Z. Z., Boonyarak T. and Submaneewong
C.
321
APPLICATION OF TOP-DOWN CONSTRUCTION METHOD FOR DEEP
EXCAVATIONS IN BANGKOK
Narong Thasnanipan, Zaw Zaw Aye, Chanchai Submaneewong, Thayanan Boonyarak

ABSTRACT
Top-down construction method was used for construction of deep underground

structures of major projects in Bangkok, including 18 subway stations of the first Bangkok
MRTA. This paper presents authors’ experiences in the practical installation of the different
stanchion types in large diameter deep-seated bored piles and barrettes in Bangkok. The
problems encountered during the actual installation of the stanchions and solutions to them
are also discussed. Application of top-down method in recently completed deep basement
structure is also reported highlighting the benefits offered by the method in comparison with
conventional bottom-up method. This paper demonstrates that the top-down method is the
most appropriate solution in construction of the deep underground structures for protecting
the modern urban environment.
1. INTRODUCTION
Top-down construction method as the name implies, is a construction method, which

builds the permanent structure members of the basement along with the excavation from the
top to the bottom. Top-down method is mainly used for two types of urban structures,
underground structures such as car parks, underpasses and subway stations and tall buildings
with deep basements. Top-down method is the most appropriate solution in minimizing the
duration and disruption to surface traffic and other urban activities, for the construction of the
subway stations, underground car parks and underpasses where the structure is located
directly underneath the existing roads. For tall buildings with deep basements, a full
application of top-down method enables superstructure to build concurrently with excavation
and construction of the basement giving a significant advantage in reducing the overall
construction time.
2. TYPICAL SUBSOIL PROFILE OF BANGKOK
Bangkok subsoil is featured by alternating clay and sand layers of thick Quaternary
deposits. As characteristics and properties of Bangkok subsoil layers have been well
documented in many literatures, only a general description of Bangkok subsoil is presented in
Top layer consists predominantly of Fill-Materials, Clayey Sand or Silty Clay with
some cement block rubble and rock fragments, is commonly found up to 2 to 4m depth. Soft
to very soft, highly compressible dark gray marine clay lies beneath top layer and in some
areas it lies under weathered crust layers of about 2m thick. Depending on the location, Soft
Clay layer is extended up to 12-18m. About 2m thick Medium Clay layer can be observed
between Soft Clay and underlying Stiff Clay. Generally Stiff Clay layer occurs directly
underneath Medium Clay and its depth goes up to 22m. Below Stiff Clay layer, First Sand
layer 5-8m in thickness can be found. This First Sand layer, however, is absent in some areas.
Stiff to Hard Clay layer underlies First Sand and it is found to be about 5m thick. Second
323
Sand layer generally occurs at depths between 45 to 65m. Typical subsoil profile of Bangkok
area is shown in Figure 1.

0 Weathered Crust
0

Clay  t = 14-16kN/m 3
10 -
10
Medium to
Su= 50-14020- KPa
Depth (m)
20 Stiff Clay  t = 17-21kN/m 3
Act
ual
Medium
-
Pie
30
H yd
(First Sand)
 t = 20kN/m 3
zo
ros
me
tatic
tric
-
40
 = 21KN/m 3
Dra
Line
t
wdo
Dense to Very
wn
50 -
50
 t = 21KN/m 3 Lin
(Second Sand)
e
60 -
60
Figure 1. Typical subsoil profile of Bangkok piezometric drawdown condition.
3. BORED PILE AND BARRETTES CONSTRUCTION METHOD
Wet process or slurry displacement method is always used for construction of deep-seated
large bored piles and barrettes in Bangkok. For bored piles, temporary casing of 14-15m in
length is used as a support in soft clay layer. Soil inside the casing is normally excavated by
auger applying rotary drilling method and drilling is continued with the bucket under
polymer-based slurry from the top of sand layer to the final depth. A mechanical cable-hung-
grab or hydraulic grab mounted on crawler crane is used for excavation of trench with slurry
support in barrette construction. Guide walls having a minimum depth of 1.2m are commonly
used for initial guiding of grab excavation at the top of the trench. Different from bored pile,
bentonite slurry is commonly used for barrette excavation in Bangkok. Tremie concreting is
necessary for casting both bored piles and barrettes.
Stanchions are mainly installed in the large diameter deep-seated bored piles and barrettes
founded in first or second sand layer. Bored piles of diameter 1.50m to 1.8m and rectangular-
shape barrettes of 0.80x2.8m to 1.5x3.0m (thickness x width) founding at depth between 50m
and 60m are commonly used to accommodate the stanchions.
324
4. ADVANTAGES OF TOP-DOWN CONSTRUCTION METHOD
Application of top-down method offers three main advantages as outlined below.

 It allows early commencement of super-structure construction without the need to
wait for excavation reaching the bottom level. Overall construction duration can be
significantly less than that of conventional method thereby saving the cost.
 It does not require temporary bracing system so that time requirement and cost for
temporary works (both material and labor costs) are eliminated, in turn significantly
saving cost.
 Minimize soil movement induced by excavation works which is an important factor
for some sensitive locations.
5. STRUCTURAL MEMBERS REQUIRED FOR TOP-DOWN CONSTRUCTION
Design and construction principles for top-down method primarily call for two major
structural elements.
 Columns with sufficient capacity must be pre-founded in bored piles or barrettes to
sustain the construction load and to utilize as part of bracing system.
 Excavation for basement must be carried out with the support of permanent
retaining wall so that basement floor slabs can be utilized as lateral bracing.
Diaphragm wall of 0.8m to 1.2m in thickness with sufficient embedment in firm soil
layers is commonly used as a retaining wall whereas prefabricated steel columns known as
stanchions embedded in either large diameter deep-seated bored piles or barrettes are utilized
as structural columns. Figure 2 illustrates the top-down construction method with utilization
of stanchions and diaphragm wall for underground subway station.
Figure 2. Top-down construction with stanchion and diaphragm wall
325
6. TYPES OF STANCHION AND THEIR APPLICATION IN BANGKOK
Pre-fabricated “H” section steel columns and steel built-up-section columns are mainly
used as pre-founded structural columns or stanchions. Pre-cast reinforce concrete columns are
very seldom used (Manoharn S. & Aye Z. Z. , 1994). Commonly used stanchion types and
their application in Thailand can be viewed in published paper of Thasnanipan et. al (2000).
7. ALLOWABLE TOLERANCE FOR POSITION OF STANCHIONS
Allowable tolerance for stanchions is usually called by the designer and it is mainly
governed by the structural tolerance of the steel as well as required position of the finished
columns. In most projects, allowable vertical and horizontal deviation of stanchion are
specified as 25 to 50mm whereas verticality is required between 1:200 and 1:400. There are a
number of constraints to get highly accurate position of heavy and lengthy stanchions which
have to be installed in deep-seated foundation piles constructed by wet process in Bangkok
subsoil as described in the following sections.
8. STANCHION INSTALLATION METHODS
Stanchion installation method is usually selected by the piling contractor who take into
consideration three main factors such as installation depth, size of stanchion and size of bored
or barrette piles. Though installation details may be different from one contractor to another,
stanchion installation can be categorized under two main methods, pre-concreting installation
or placing stanchion prior to concreting or post-concreting or plunging installation.
8.1 Pre-concreting or pre-placing installation method
This method was applied in Thailand Cultural Center Station of the M.R.T Chaloem
Ratchamongkhon Line. Stanchion was installed prior to tremie concreting, immediately after
completion of barrette excavation and reinforcement lowering. The main reason for selection
of this method in the mentioned project was due to the relatively large size of the barrettes. In
some projects stanchion is attached to the last section of reinforcement and installed together
with the reinforcement. General construction steps involved in this method are demonstrated
in Figure 3.
Concrete
Guide frame
Stanchion Stanchion Stanchion

Slurry Slurry Tremie pipe
Backfill
Pile cut-off Pile cut-off Pile cut-off Pile cut-off

level level level level
concrete concrete
Concreting
Complete drilling and Stanchion installation Complete casing extraction
(1 tremie for bored pile and
reinforcement installation and backfilling
2 tremie for barrette)
326
Specially designed guide frame used in installation of H-section steel column of
419x407x390kg/m, 24.7m long, inserted into barrettes applying pre-concreting method in
Bangkok MRT project is shown in Figure 4. Effectiveness of guide frame plays one of the
important roles in achieving positional accuracy of stanchion.
Figure 4 Guide frame used for pre-concreting installation in Thailand Cultural Center Station
of the M.R.T Chaloem Ratchamongkhon Line
Figure 5 shows the installation of a 24m long built-up steel stanchion into the pile
borehole prior to concreting.
Figure 5 Stanchion length 24m attached to the top portion of reinforcement cage installed into
borehole prior to concreting
327
7. 2 Post-concreting installation or plunging method
In this method, stanchion is installed immediately after completion of bored pile

concreting process. General construction sequence involved in this method is demonstrated in
Figure 6. Guide frame is used to install the stanchion at the correct position.
Concrete
Guide frame
Tremie pipe Stanchion Stanchion

Slurry Slurry Slurry Backfill
Pile cut-off Pile cut-off Pile cut-off

level level level
Concrete
Concrete
Complete drilling and Concreting by tremie Installing stanchion by Complete casing extraction
reinforcement installation method plunging into concreted pile and backfilling
REVIEW OF THE STANCHION INSTALLATION METHODS USED IN OTHER PARTS

OF THE WORLD
The stanchion installation methods used in other parts of the world have been presented
in some published papers. Among these, Findlay (1989) reviewed a number of stanchion
installation methods used in construction of large diameter bored piles for top-down
construction particularly in UK. The author reported that steel columns can be placed with
better accuracy by dry process bored piling method than wet process method (with support
fluid).
Table 1. Summary of stanchion installation method from published literatures

Reported by Location Stanchion installation method
Hollingsworth (1991) London, UK Post-concreting
Ressi di Cervia and Tamaro Boston, USA Pre-concreting
(1991)
Crawley and Stones (1996) London, UK Pre-concreting
Arz (1989) Germany Pre-concreting
From the available literatures it is noted that pre-concreting method is more common in
use than post-concreting method.
PROBLEMS ENCOUNTERED IN POST-CONCRETING INSTALLATION METHOD
The major problem commonly encountered in this installation method is inability to

install the stanchion at the design position due to one or combination of the following causes.
 Inclination of the borehole / trench
328
 Stanchion can not be inserted up to the required depth as concrete becomes hard due
to the premature setting
 During lowering, stanchion is stuck by reinforcement cage and due to the hardening
of the concrete, extraction of stanchion becomes impossible for reinstallation
As the installation of the stanchion is often associated with many unforeseen problems,
it is likely in many cases that concrete becomes stiff or prematurely set during the installation.
In Bangkok, premature setting of concrete is usually found to be attributed by;
 long delivery time due to traffic conjunction
 inappropriate mix
 severe weather
 disruption in concreting / equipment breakdown
PROBLEMS ENCOUNTERED IN PRE-CONCRETING INSTALLATION METHOD
Using this method can minimize some problems encountered during installation
provided that adequate clearance is available between pile reinforcement and the stanchion for
tremie pipes lowering. Inclination of borehole is also important for this installation method.
Proper planning in concrete ordering is required to avoid unnecessary waiting of concrete
trucks for stanchion placing which may affect the total batching time limitation. As stanchion
has to be hung in the position until completion of concrete placement, appropriate and
sufficient capacity of the hanging and positioning device should be provided to avoid falling
and deviation of stanchion in the borehole. Temporary casing should not be used as a hanging
system without additional support for the large stanchion.
FACTORS EFFECTING THE POSITIONAL ACCURACY OF THE STANCHION
It is hardly possible to install a stanchion so that its position and verticality is always as
designed. As-built position of stanchion is found to be generally influenced by the installation
method. Skill and experience of the contractor plays a major role in achieving positional
accuracy of the stanchion provided that the installation is practical with the design elements
such as size of foundation piles in relation to size of stanchion. The factors affecting the
positional accuracy for two different methods were reported in details by Thasnanipan et. al
(2000).
Figure 7. View of stanchions embedded in bored piles at final excavation level
329
STANCHION POSITION EFFECTED BY EXCAVATION
In Bangkok, post-installation movement (horizontal movement) of the stanchion is

frequently encountered and it is mainly caused by excavation induced soil displacement
particularly at initial stage of excavation where diaphragm wall deflection is characterized by
large cantilever rotation within Soft Clay layer. In some projects, deviated stanchions had to
be pushed or jacked back to the original position and restrained by means of temporary
support until permanent slab was cast.
APPLICATION OF TOP-DOWN METHOD IN CONGESTIVE URBAN AREA
Top-down method was successfully applied in Bangkok for a fifteen-storey commercial

building with five-level underground car park located in an old established commercial area
of Bangkok. Construction site is bounded by rows of old shop-house buildings estimated to
be supported by short piles. Hence, ground movement induced by large deflection of retaining
wall was unfavorable. Limited space availability on site and congestive neighborhood also
posed another major constraint on construction logistic. By the local traffic regulation,
concrete trucks and other heavy vehicles entering to the site were restricted during rush-hours
of both morning and afternoon. These factors called for the need of careful consideration in
establishing the design principles and sequence of construction. A numbers of construction
sequences were considered. After having detailed review and conducting a series of analyses
with different conditions, decision was made to apply the top-down construction method.
The main objectives were ;

 to minimize the material and construction sequence involved in temporary works so as
to accelerate the construction time
 to minimize the logistic problem imposed on site due to limited space in congestive
neighborhood
 to minimize soil movement induced by excavation works which is important factor for
the project adjacent to the old buildings of weak foundation
DIAPHRAGM WALL, BARRETTES AND BORED PILE CONSTRUCTION
Diaphragm wall having 800mm width founded at 24m below ground level (B.G.L) was
constructed simultaneously with wet-processed bored piles of diameter ranges from 800mm
to 1800mm with toe depth 60m B.G.L. Barrettes having same toe depth as bored piles were
constructed along with diaphragm wall panels. Depending on the loading conditions, different
sizes of permanent steel stanchions were installed by post-concreting method. Table 2 shows
the detailed of foundation elements constructed in the project. Figure 8 shows layout of
diaphragm wall and bored pile with stanchion. Figure 9 shows the aerial view of excavation
works.
330
Table 2 Detail of foundation elements in each project
Type and Size of Allowable Pile Permanent Steel Stanchion Size Total
Foundation Capacity (ton) Number
Barrette 0.80 x 3.00m 1750 2 No. H-350x350mm 6
Bored pile dia. 1.80m 1600 H-400x400mm + 2 No. H-200x200mm 3
+ Steel Plate 80mm
Bored pile dia. 1.80m 1600 H-400x400mm + 2 No. H-200x200mm 15
Bored pile dia. 1.80m 1600 H-400x400mm 17
Bored pile dia. 1.80m 1600 No stanchion 3
Boundary Line
Bored Pile with Stanchion
D-Wall
Figure 8. Layout of bored pile with stanchion and diaphragm wall
Figure 9. Arial view of the top-down construction in progress – concrete batching plant was
set up on site to minimize the concrete supply interruption in congestive area with severe
331
COMPARISON BETWEEN TOP-DOWN AND BOTTOM-UP METHOD
Master Construction Schedule of the major activities of top-down and bottom-up method is
illustrated in a simple bar-chart form as shown in the figure below. Apparently, top-down
construction method offered significant time saving as can be seen in the figure. This
schedule comparison is based on the construction sequences by top-down and bottom-up
method as demonstrated in Figure 10. Figure 11 shows construction progress after excavation
reached to final level.
Method Major Activity Duration Construction Period (Month)

(month) 4 8 12 16 20 24 32 36
Top-down D-wall and piling 4.5
Sub-structure 10
Super-structure 12
Bottom-up D-wall and piling 4.5
Sub-structure 16
Super-structure 13
Figure 10 Master construction schedule of major construction activities of top-down and

bottom-up method
Apart from significant time saving, the cost associated with temporary works was
significantly minimized by using top-down method. If bottom-up method was utilized, it
would have required, 3 layers temporary bracing as shown in Figure 12. A total weight of
steel H-beam type temporary bracings required for 3 layers was estimated to be 1,070 ton.
Significant amount of scaffolding usage was also minimized by applying top-down method.
Figure 11 View of construction progress after excavation reached to final level
332
Floor level Floor level
1st Bracing
Soft Clay Soft Clay
B1 -3.92 B1 -3.92
2nd Bracing
B2 -8.20 B2 -8.20
Soft Clay 2 Soft Clay 2
B3 -11.29 B3 -11.29
D-wall 0.8 m. thk.
D-wall 0.8 m. thk.

rd
3 Bracing
Medium Stiff Clay Medium Stiff Clay
B4 -14.30 B4 -14.30
B5 -17.39 B5 -17.39
Stiff Clay Stiff Clay
-19.10 -19.10
-24.00 -24.00
Hard Clay Hard Clay
Top-down construction Bottom-up construction
Top-Down Method Bottom-up Method

Stage Construction Sequence Stage Construction Sequence
1 Excavate to -1.7m and install 1 Excavate to -1.7m and install first
temporary bracing at -1.5m level temporary bracing at -1.5m
2 Excavate to -4.5m and construct B1 2 Excavate to -7.2 m and install 2nd
Slab level temporary bracing at -7.0m
3 Remove 1st bracing at -1.5m 3 Excavate to -13.0m and install 3rd
4 Excavate to -11.8m and construct B3 level temporary bracing at -12.8.0m
Slab 4 Excavate to -19.1 m and construct
5 Excavate to -19.1m and construct mat foundation
Mat Foundation 5 Construct B4 Slab and columns
6 Construct B4, B2 and Ground Floor 6 Remove 3rd level bracing
Slab 7 Construct B3 Slab and columns
7 Construct above floors 8 Construct B2 Slab and columns
9 Remove 2nd level bracing
10 Construct B1 Slab and columns
11 Remove 1st level bracing
12 Construct Ground Floor Slab
13 Construct above floors
Figure 12 Summarized construction sequence of top-down and bottom-up method
333
COMPARISON ON DIAPHRGM WALL BEHAVIOR
Behavior of diaphragm wall with two different construction methods was studied. Figure 13
depicts the results of bending moment, shear force and displacement of diaphragm wall
compared for top-down and bottom-up construction methods. Followings were concluded
from the comparison.
 Reinforcement required for diaphragm wall with top-down method would be less than
that of bottom-up method which in turn offered cost saving on material and labour
 Diaphragm wall deflection for top-down method was significantly less than that of
bottom-up method which would minimize risk of damage to adjacent existing
buildings
Predicted and measured diaphragm wall deflection is shown in Figure 14 As can be seen in
the figure, actual measured deflection values are less than those of prediction. The shape of
the deflection curves of prediction and actual measurement at each construction stage is fairly
identical which suggests that design parameters and assumption used in analysis are of
applicable.
Bending moment (t-m / m) Shear force (t/m) Cumulative Displacement (mm)

200 100 0 -100 -200 100 50 0 -50 -100 0 20 40 60 80
0 0 0
-5 -5 -5
-10 -10 -10

Depth (m)
Depth (m)
Depth (m)
-15 -15 -15
-20 -20 -20
-25 -25 -25

Top-down Top-down Top-down
Bottom-up Bottom-up Bottom-up
-30 -30 -30
Figure 13 Bending moment, shear force and displacement of diaphragm wall
comparison for top-down and bottom-up construction methods
334
Cumulative Displacement , mm.
0 10 20 30 40
0 Construction Stage
1. Excavate to -1.70 m
2. Install 1st bracing and Exc. to
1-P -4.30 m
5 3. Construct B1 and remove
1-A 3-A 3-P 1st bracing
2-A 2-P 4. Excavate to -11.80 m and
construct B3
10 5. Excavate to -19.10 m
4-A
4-P
5-A
Depth, m.
15
20
25
A-Actual Movement
P-Predicted movement
30
Figure 14 Comparison between predicted and actual movement of diaphragm wall in

each stage of construction
CONCLUSION
Application of top-down construction method in Bangkok subsoil has been presented
with particular emphasis on the two different stanchion installation methods. According to the
author’s experience, pre-concreting installation method provided fewer problems in practical
installation and achieved better positional accuracy. The designers such as structural
engineers and architects should be aware of the accuracy achievable by practical installation
method and take it into consideration at the design stage. Stanchion installation contractor
should select the appropriate method, equipment as well as experienced personnel plus a well-
formed plan with the consideration of all potential problems to achieve the successful
construction of foundation structure, which is of primary importance for top-down technique.
Practical application of top-down method in recently completed project is presented
highlighting the key advantages over bottom-up method. Significant cost saving from shorter
construction period and minimum usage of material for temporary works offered by top-down
construction method was remarkable achievement. In short, top-down method is the most
appropriate solution in construction of the deep underground structures for protecting the
modern urban environment.
335
REFERENCES
Arz P. 1989. Foundations of the Ko-Galerie in Dusseldorf Using Top-down Building

Technologies, Proceedings of the International Conference on Piling and Deep
Foundations, London, 15-18 May 1989 : 185-192. Rotterdam: Balkema.
Crawley J.D. & Stones C. S. 1996. Westminister Station – Deep Foundation and Top-Down
Construction in Central London, Geotechnical Aspects of Underground Construction in
Soft Ground, London : 93-97. Rotterdam : Balkema.
Findlay J. D., Wren G.E. 1989. Review of the Methods Used to Construct Large Diameter
Bored Piles for Top Down Construction. Proceedings of the International Conference
on Piling and Deep Foundations, London, 15-18 May 1989 : 199-210. Rotterdam:
Balkema.
Finsted J. A. 1991. Royal Christiana Hotel : Basement with Permanent Sheet Pile Wall, “Up-
and-down” Method, 4th International DFI Conference : 387-392. Rotterdam: Balkema.
Hollingsworth J.R. 1991. Diaphragm Walls, Load Bearing Piles and Piled Soil
Reinforcement for a Deep Top-down Basement Construction, 4th International DFI
Conference : 79-84. Rotterdam : Balkema.
Manoharn S. & Aye Z. Z. (1994) , Bored Pile with King-post and Secant Pile Wall
Construction for Top-Down Method – Central Sukhothai Hotel Haddyai, Latest
Technology in Construction Technology Conference, Bangkok, Thailand
Thasnanipan N., Maung A. W. & Z. Z. Aye, (2000) Practical Installation of Stanchions for
Top-Down Construction in Bangkok Subsoil, Year 2000 Geotechnics, Geotechnical
Engineering Conference, The Imperial Queen’s Park Hotel, Bangkok, Thailand, 27-30
November 2000
Yong K.S., Tsen C.N. & Tang S.K. 1998. Performance of a Deep Excavation with Top-down
Construction, 13th Southeast Asian Geotechnical Conference, 16-20 November 1998.
Taipei, Taiwan ROC : 587-593
336
and Tunneling

for Protection of
Environment
Edited by:
Noppadol Phienwej
Construction of Diaphragm Wall for Basement

Excavation Adjacent to Tunnels in Bangkok
Subsoil
Narong Thasnanipan, Aung Win Maung, Zaw Zaw Aye,
Chanchai Submaneewong and Thayanan Boonyarak
337
International Symposium on Underground Excavation and Tunnelling
Construction of Diaphragm Wall for Basement Excavation

Adjacent to Tunnels in Bangkok Subsoil
Narong Thansnanipan1, Aung Win Maung1, Zaw Zaw Aye1, Chanchai Submaneewong1
Thayanan Boonyarak1
1
Seafco Public Company Limited, Thailand
ABSTRACT
In July 2005, an earth retaining structure with diaphragm wall was completed in the vicinity of
Bangkok’s first subway system, the MRTA Chaloem Ratchamonkhon Line. Diaphragm wall design
analysis and ground movements associated with diaphragm wall construction were reviewed. The
review was carried out in accordance with technical requirements for engineering works within the
specified protection zone of MRTA structures. Preliminary diaphragm wall analyses indicated that
lateral movements of the wall should not exceed 26.54mm. A 6.0mm maximum movement allowance
is specified for MRTA structures. The construction work was undertaken in compliance with
technical requirements for engineering works within the specified protection zone of MRTA
structures. Two inclinometer tubes with 20.0m in length were installed in the ground between the
planned diaphragm wall and the south-bound tunnel. Surface settlements markers were also
established in line with the inclinometer tube location on the pavement. Ground movements were
monitored during diaphragm wall panel construction. Based on the monitoring results, the estimated
maximum ground movement at the tunnel location due to wall panel construction was less than
0.5mm, indicating no impact on the tunnels by diaphragm wall construction.
1. INTRODUCTION
Deep excavation work adjacent to existing structures have become a common construction activity in
most major cities as utilization of underground spaces, including underground car parks, basements,
mass transit subway stations and tunnels increase. With mass transit systems in place, properties
located along the commuter lines are considered the most favorable areas to be developed with
maximum utilization of areas available. In turn, deep basement construction adjacent to the existing
subway tunnels is also often necessary.
At present in Bangkok, such deep excavation work in the vicinity of the subway tunnels is
considered relatively uncommon compared to major cities in developed countries. It is anticipated that
this phenomenon will change in the near future. This paper presents an unprecedented case in
Thailand of deep excavation adjacent to a subway tunnel. The proposed hotel tower reported herein
was planned to be constructed with 2 basements using cast in-situ diaphragm wall located 7.8m to
9.6m away from the south-bound tunnel of the Mass Rapid Transit Authority, MRTA Chaloem
Ratchamonkhon Line. The main foundation for the superstructure consists of 66 bored piles (1.5m
diameter) and 7 barrettes (1.0x3.0m) embedded to depths of about 57.0m in the second sand layer of
Bangkok Suboil. For construction of basements and foundation mat of 3.5m in thickness, diaphragm
walls (0.8mx20m deep) with 2 levels of temporary bracing are planned as a retaining system for a
general excavation to -10.50m. The perimeter diaphragm wall is supported by 24 bored pile of 0.8m
in diameter. Due to the presence of the MRTA tunnels in the close vicinity of the project, a review
339
371
was carried out on the expected performance of the retaining system, with regards to technical
requirements for the engineering works within the protection zone of the MRTA.
In this paper, predicted diaphragm wall deflections and associated expected tunnel movements
are presented. Pre-established geotechnical instrumentation program and recommended contingency
measures are also described. Monitoring results of ground movement during different stages of
diaphragm wall panel construction is also reported.
2. SITE CONDITIONS
The project site is located on the eastern side of Asoke Road, Sukhumvit, Bangkok, Thailand. On the
eastern side are two one storey houses, 3 storey office building and open space car parking. The site is
flanked by a 12 storey office building on the northern boundary and a 5 storey residential building on
the southern boundary.
The MRTA twin tunnels with an outside diameter of 6.30m are located in front of the site about
15.0m below Asoke Road. Sound bound tunnel is adjacent to the site being 9.623m to 7.815m away
from the planned diaphragm wall. The layout plan of the site is presented in Figure 1.
Figure 1. Layout plan of the project
3. SUBSOIL CONDITIONS
A site investigation was carried out with 3 boreholes drilled at the project site. The borehole
information indicated that a soft clay layer about 10.0m in thickness occurred below 2.0m thick fill. A
medium stiff clay layer with Su 2.9-3.9t/m2 was encountered at a depth of 12.0m, overlaying a stiff
clay layer at a depth of 15.0m. A 2.0-2.5m thick sand layer was intercepted at a depth of 19.5m
overlying very stiff clay layers. Two dense sand layers with thickness of about 7.5m and 12.0m
340
372
occurred at depths of about 42.5m and 54.0m respectively in hard silty clay layers. A summary of
subsoil conditions obtained from borehole data is presented in Table 1.1.
Table 1. Summary of Subsoil Conditions

Depth below Description of Material Thickness Su (UC) SPT N
Ground (m) (m) (t/m2) (blow/ft)
0-2.0 Pavement/top soil 2.0 - -
2.0-3.0 Soft Clay 1 1.0 2.0 -
3.0-12.0 Soft Clay 2 9.0 1.2
12.0-15.0 Medium Stiff Clay 3.0 3.5
15.0-17.0 Stiff Clay 2.0 5.0
17.0-19.5 Stiff to Very Stiff Clay 5.0 12.0 18
19.5-22.0 Medium Dense Sand 2.5 - 20
22.0-35.0 Very Stiff Clay 13.0 20.0 30
35.0-42.5 Stiff Clay 7.5 12.0 18
42.5-50.0 Dense Sand 7.5 - >52
50.0-54.0 Hard Clay/ Silty Clay 4.0 29.9 45
54.0-66.0 Dense to Very Dense Sand 12.0 - >60
66.0-80.0* Hard Silty Clay - >45
Note:* at the end of borehole
4. TECHNICAL REQUIREMENTS FOR ENGINEERING WORKS WITHIN PROTECTION

ZONE OF MRTA
Over 6 to 10 years, deep excavations with braced diaphragm walls have been carried out in Soft
Bangkok Clay for multi-level basements, underground car parks, subway stations and deep shafts for
ventilation and launching TBMs. After completion of the subway system, protection zones for the
MRTA subway structures were established and technical requirements for future works planned within
the protection zone were also issued. There are two protection zones, named A and B. Protection
zone A is a zone located immediately above the MRTA structure, plus 3.0m clearance distance
measured laterally from the outer faces of MRTA structure. Zone B is located outside the 3.0m
clearance distance from the MRTA structure, but within the influence zone based on a line drawn from
a point at the center level of MRTA tunnel on the 3m clearance boundary line at 1:1 slope to the
ground for MRTA structures with base level no deeper than 20.0m depth. With the base level of
MRTA tunnel located deeper than 21.0m below the ground, the influence zone below 21.0m is
determined using a line with a slope 2:1 vertical to horizontal ratio from the clearance boundary line at
the center level of the structure. Figure 2 illustrates the MRTA protection zones and Figure 3 shows a
typical section of the project reported herein together with the protection zone.
The technical requirements generally cover construction method and technique allowed and
approval, surcharge load limits, clearance distance to the structure and movement limits induced by the
future works.
The most critical requirement for working adjacent to the MRTA structure is the differential
movement resulting from the work shall not produce final distortion in the track or its plinth in excess
of 3mm in 6.0m (1:2000), or a total movement in the MRTA structure or track not exceeding 6mm in
any plan.
Due to the above mentioned requirements, the diaphragm wall design analysis was reviewed with
finite element computer modeling to predict the movement of the tunnels caused by basement
excavation work. Subsequently, impact assessment was carried out and necessary precautions and
safety measures were outlined.
An assessment of works associated with diaphragm wall and bored piling for the project to meet
the technical requirements for engineering works within the protection zone is summarized in Table 6.
341
373
Figure 2. Configuration of MRTA Protection Zone
Figure 3. A Typical Section of Project
5. DIAPHRAGM WALL ANALYSIS AND DESIGN
Diaphragm wall analysis and design was primarily carried out using the WALLAP computer program.
Stability of wall, bending stresses and wall deflections were computed in relation to planned basement
excavation sequence. Due to the presence of the MRTA subway tunnels about 7.8m away from the
planned excavation, 2 dimensional analyses using the PLAXIS program was carried out to compute
the response/displacement of the tunnels and ground settlements caused by the excavation works.
342
374
Based on the primary analysis with WALLAP, 0.8m thick diaphragm wall embedded to a depth of
20.0m with two levels of temporary bracing at -1.50m and 6.20m was required.
WALLAP has been used by the authors for analysis of braced excavations with diaphragm walls
and sheet piled walls in Bangkok subsoils and elsewhere. More than 12 projects have been completed
by the authors. Additionally 9 underground stations and a number of ventilation shafts of MRTA
southern part has been completed with diaphragm walls designed by using the program (Schulz C.
2001).
Prior to determination of tunnel movements using PLAXIS, the same basic model (without
foundation piles and tunnel) was used in analysis using both WALLAP and PLAXIS programs for a
comparison between analysis results to establish a correlation for further analysis using PLAXIS
program. The comparison of analyses indicated that the predicted maximum wall deflection computed
by PLAXIS is about 39.0mm which is close to the 43.0mm computed by WALLAP. Figure 4 shows
results of analysis from both WALLAP and PLAXIS programs.
Bending moment (t-m / m) Shear force (t/m) Displacement (m)
100 50 0 -50 -100 -150 60 40 20 0 -20 -40 0.00 0.02 0.04 0.06 0.08 0.10
0 0 0
-5 -5 -5
-10 -10 -10

Depth (m)
Depth (m)
Depth (m)
-15 -15 -15
-20 -20 -20
-25 -25
-25
Wallap Plaxis Wallap Plaxis Wallap Plaxis
Figure 4. Comparison of Diaphragm Wall Analysis Results from WALLAP and PLAXIS
5.1 Wall Deflection and Ground Movements
The characteristics of diaphragm wall movements in the completed projects in Bangkok have been
reviewed by Phienwej et al (1998). It was reported that in terms of the ratio of δhmax /H, the typical
range of the maximum wall movement δhmax fell within 0.3-0.8% of excavation depth H.
For the planned project reported herein, primarily predicted maximum wall movement of
34.00mm, about 0.31% of the excavation depth (11.0m) by WALLAP was considered in the
reasonable range and thus control of the wall deflection within the predicted range could be practically
achievable, with adequate lateral supports and proper construction sequence and time controls if there
is no existing subway tunnel nearby.
However, for calculation of lateral and vertical displacements of the ground, including those of
the existing MRTA tunnels, a two dimensional simulation is necessary and thus PLAXIS program was
selected. A number of computer simulations were carried out using the PLAXIS program to meet the
technical requirements, particularly tunnel displacement criteria.
343
375
1
B 1 11
1 st Temp. Bracing 3 10
2 Excavate to -1.8m
B2 9
2nd Temp. Bracing 5 8
4 Excavate to -6.5m
Mat/Base Slab 7
6 Excavate to -11.0m
Note: 8 10 Remove Temp. Bracing
Figure 5. Computer modeling with PLAXIS Figure 6. Basement construction stages
Initially computer modeling, using PLAXIS to estimate the displacement of the tunnels without soil
improvement to the soft clay layer inside the excavation, indicated that tunnel displacement exceeded
the limit of 6.0mm specified for the MRTA structure or tracks.
A series of computer models with different wall configurations such as increased wall thickness
and embedment were carried out. It was found that the wall system did not significantly control the
displacement of the tunnels within the specified limit, but the expected tunnel movement is generally
governed by the wall deflection. Moreover, the computer modeling indicated that excavation for
installation of the 2nd bracing layer and subsequent removal of the bracing after completion of base
slab influenced the overall ground displacement and wall deflection, due to the presence of a thick soft
clay layer between at depths of 3.0m to 12.0m. To effectively reduce wall deflection and ground
displacement, improvement of the soft clay layer was considered necessary. We considered also that
the site conditions allowed the soft clay to be improved to the strength of stiff clay with deep cement
mixing inside the excavation zone close to the south-bound tunnel. The improve soil block could act
as a counterfort to the wall during installation of the 2nd bracing layer. Additionally the improved soil
block could also be used as a berm or contingency support for the wall during final excavation depth.
Predicted lateral and vertical displacements of the ground and tunnel movements are summarized
and presented in Table 2.
Table 2. Summary of Maximum Movements Computed by PLAXIS Program

Maximum Movement of Maximum Movement of
ground (mm) Tunnel (mm)
Model Type
Horizontal Vertical @ Specified
Predicted
@ D-wall ground level Limit
Simple model without soil 40.57 38.22 6.88 <6.00
improvement
Improved Soft Clay layer @3.0- 26.54 25.79 5.97 <6.00
13.5m, 5.0m wide inside
excavation.
5.2 Comparision with Similar Projects elsewhere
Publication of other cases resembling that of the planned project (Case A) in Bangkok has not been
found and thus published papers on the similar/relevant projects elsewhere are reviewed and an
assessment of movements in the MRTA structure or tracks induced by the planned excavation is
carried out and summarized as below.
344
376
Table 3. Summary of Subsoil Conditions at Project Sites
Case
Project Name Location Subsoil Ref.
No.
A Planned Project on Bangkok 12.0m thick soft clay overlying stiff In this paper
Asoke Road clay.
B Inland Revenue Singapore 25.0m thick very soft marine clay Yang et al
Headquarter overlying dense to very dense (1998)
decomposed granite
C New Dhoby Ghaut Singapore 2-3m thick fill, 3m thick organic clay Niu et al
Station over 2m thick fluvial and 2m thick (2003)
residual soil overlying sandstone.
D Cheung Sha Wan Gov. Hong Kong 11.0m thick fill over 11.0m thick Chu et al
Office marine deposit, overlying alluvium and (2001)
decomposed granite
E People Park Station Shanghai 30.0-40.0m thick soft clay and silty Jianhang
clay interbeds (1998)
F New World Masion Shanghai Silty clay and clay layers
G Hong Kong Square Shanghai Silty clay and clay layers
H The George Green London 4-4.5m thick made ground over clay Higgins et al
Cut&Cover (2001)
Although, general thickness of clay deposits vary with the projects, the nature of the work is
similar, it is noteworthy that deep excavation work was carried out in soft soils adjacent to existing
subway tunnels.
Table 4. Summary of Excavation Configurations

Max. Distance to Depth to
Case
Project Name Excavation Outer Face of Tunnel Wall Type
No.
Depth (m) Tunnel (m) Crown (m)
A Planned Project on Asoke -11.0 7.8 15.0 0.8mx20m
Road D-Wall
B Inland Revenue -18.0 16.0 11.0 1.0mx24.5m
Headquarter D-Wall
C New Dhoby Ghaut Station -30.0 4.5 11.5 Φ0.8+1.0mx35.0m
Secant pile wall
D Cheung Sha Wan Gov. -18.5 <5.0 19.0-22.0 D-Wall
Office
E People Park Station -23.067 12.9 6.959 D-Wall
F New World Mansion -13.0 3.0 8.0 26m D-Wall
G Hong Kong Square -14 7.0 - 0.8x24.0m
D-Wall
H The George Green -9.0 1.0m to below 13.7 Φ1.0mx19.0m
Cut&Cover excavation Secant pile wall
345
377
Table 5. Summary of Performance of Walls
Max. Wall Max. Tunnel Displacement
Case Excavation
Project Name Deflection (mm) (mm)
No. Method
Predicted Actual Predicted Actual Specified
A Planned Project Bottom Up 26.54 - 5.97 - 6.0
on Asoke Road 2 level Strut +
improved soil berm
B Inland Revenue Top-down 30 25 11 8 15
Headquarter
C New Dhoby Bottom Up 24 15
Ghaut Station 7 level Strut <7@ track
D Cheung Sha Wan Top-down <25 <25 17 18 20
Gov. Office
E People Park Bottom Up 35 38 <15
Station 5 level Strut
F New World Bottom Up - <12 - <10 <10
Mansion
G Hong Kong Bottom Up <40 <40 <10 <10
Square & improved soil
berm + 3 level strut
H The George Top-down 11 12 -
Green
Cut&Cover
Review of the above completed projects, with the exception of the planned project (Case A),
indicates that the specified limit of tunnel displacement ranges from 10mm to 20mm. Details of tunnel
lining and specified limits of the railway track movement of these projects were not specifically
reported. Only in Case No. C (Ref. Niu et al 2003), maximum movements of 7mm and 15mm were
recorded for both the railway track and tunnel respectively.
Based on the projects mentioned above, controlling the maximum movements of the MRTA
tunnel and railway track induced by the planned excavation for the project, within the specified limits
is possible. However the excavation work must be carried out by competent contractors with
contingency plans such as that outlined previously, with the geotechnical instrumentation and
monitoring mentioned in the following section as a minimum requirement.
5.3 Geotechnical Instrumentation and Monitoring
The following instrumentation is considered essential and monitoring must be carried out throughout
the basement construction period.
Inclinometer - 2 inclinometers in the ground between the tunnel and planned diaphragm wall and
(for diaphragm wall construction) 4 to 5 inclinometers in the diaphragm wall around the excavation.
Convergent Bolts & Tape Extensometer – If necessary, adequate quantities of convergent bolts
may be installed on the tunnel segments to monitor the convergent movement with tape extensometer
at each construction stage throughout the basement excavation period.
Settlement Plates/markers – A minimum of 4 settlement points on the ground across the tunnels
needs to be installed in line with each inclinometer location and surveyed at each construction stage
throughout the construction period.
Survey – Movement of tracks must be surveyed at each construction stage throughout the
basement excavation period.
Contingency Plan and Actions - Since the predicted displacement of the tunnels is very close to
the specified limits, monitoring of adjacent structures and ground movements during the staged
346
378
excavation is necessary, with checking against the predicted movements at every stage of construction.
A contingency plan with trigger values of displacements being based on the predicted values must be
prepared and necessary action must be taken immediately if the measured movements of diaphragm
wall and tunnels exceed the trigger values set. It was suggested that a contingency plan including, but
not limited to, the following be prepared and carried out accordingly when required.
• Preloading of struts as necessary
• Localized excavation and monitoring
• Provision of an additional bracing layer
• Speeding up the construction.
6. SUMMARY OF ASSESSMENT
An assessment of works associated with bored piling and diaphragm wall work to meet Technical
Requirements for Engineering Works within the Protection Zone of the MRTA Chaloem
Ratchamonkhon Line is summarized in the table below.
Table 6. Summary of Assessment of Basement Excavation with Braced Diaphragm Wall

Specified
Item Description Provision Assessment/Remarks
Requirement
1 Diaphragm Wall Method Statements Yes Documents prepared &
Bored Piling Yes submitted for approval
Bracing and Excavation outlined
2 Calculations Signed with Yes Conformed
Chartered Engineer
2.1 Engineering Work in
Zone B
2.1.1 Surcharge at ground level Not > 50kPa (5t/m2) 15kPa Used in D-Wall analysis and
(1.5t/m2) Design
2.1.2 Surcharge on tunnel Not > 25KPa 15kPa Used in D-Wall analysis and
(2.5t/m2) (1.5t/m2) Design
2.1.3 Differential Movement < 1:2000 - Use observational approach
< 6mm in total < 6mm and instrumentation program
2.1.4 Clear distance between > 3m or (5) pile > 6m or (7) pile layout plan and drawings
pile/d-wall and MRTA diameter diameter conform.
structure
2.1.5 Additional surcharge load Not > 50kPa @ - Pile cutoff is below 11.0m
by piles designed GL and generally outside Zone B.
D-Wall is supported by pile
legs to 57-60.0m.
2.1.6 Piling Bored piling with Yes Method statement conforms.
auger under slurry
2.1.7 Use of vibratory method Not within 10m > 10m Method statement conforms.
2.1.8 Future additional Use 1:1 slope of load Yes Conformed
surcharge in Zone B distribution
2.1.9 Approval for excavation >3.5m near tunnels ~ 11.0m with Approval is obtained.
adjacent to MRTA or station braced D-Wall.
structures >1.5m near others Instrumentation
Suitable program Documents prepared &
instrumentation outlined submitted for approval
2.1.10 Site investigation Fully grouted - The site is about 8m away
boreholes for location of approved by the from the MRTA structure.
MRTA structure authority
2.1.11 Soil Improvement Approval submission Yes (by Main It will be inside the perimeter
contractor) D-wall as analysed.
347
379
Moreover we also concluded that the structural design of the diaphragm wall as analyzed and designed
with the WALLAP program was adequate for the planned excavation work. However the maximum
deflection of the wall adjacent to the existing tunnels had to be reduced to 20mm, to limit the
displacement of the tunnel within the specified displacement as indicated by computer modeling with
PLAXIS program. Additionally, soft clay adjacent to the south-bound tunnel inside the excavation at
depths between 3.0m and 13.5m must be treated to improve the shear strength of the soil to that of stiff
clay layer (about 10t/m2). We intended the soil improvement to minimize the wall movement and acts
as a contingency measure in excavation. The observational approach is strongly recommended for a
deep excavation in the proximity of exiting tunnels.
7. DIAPHRAGM WALL CONSTRUCTION
A review assessment of ground movements induced by diaphragm walling was also carried out during
diaphragm wall design and analysis. Ground movements associated with diaphragm wall trenching in
Bangkok clay has been studied and reported (Thasnanipan et al, 2004). It was reported that an
inclinometer (24m long) was installed 1.0m away from the trial trench face for monitoring of soil
movement during and after trenching. Additionally 6 surface settlement points with 2.0m spacing
were also installed and monitored. The trial trench 0.6x4.5x18.0m was excavated under polymer
slurry. The monitoring results indicated that vertical ground settlement did not extend beyond 8.0m
from the trench face. The maximum vertical settlement of 9mm was found to occur at about 2.0m
away from the trench face. The inclinometer readings which were recorded 1.0m away from the
trench face indicated that the maximum lateral soil movement of 5.8mm occurred at the depth of about
3.5m below the ground. At 15.0m depth, the lateral movement of soil 1.0m away from the trench was
about 1.0mm after 12 hours of trench excavation. Based on the above information, for the diaphragm
wall construction adjacent to the existing MRTA tunnels located about 7.8m away from the outer face
of planned diaphragm wall of the project site and at depths of about 15.0m below ground level, the
following statements were made;
1. Vertical displacement of ground due to diaphragm wall trenching was expected to be within
8.0m from the trench face and thus not affect the tunnels which were located beyond 7.8m
from the trenching.
2. Lateral soil displacement within 1.0m from the trench face at a depth of 15.0m depth was
about 1.0mm. It was expected that the displacement would decreasing distance from the
trench face. At the tunnel location, no lateral soil displacement was expected as the south-
bound tunnel was about 8.0m from the trench face.
3. For trench stability, bentonite slurry was used instead of polymer slurry.
Diaphragm wall construction was carried out with a conventional rope-suspended mechanical
grab under bentonite slurry in July 2005. Two inclinometer tubes were installed between the south-
bound tunnel and the planned diaphragm wall, about 4.0m from the outer face. Since the subway
tunnels were beneath Asoke Road, settlement marks were established on the pavement in line with the
inclinometer location and crossing the tunnel alignment. Monitoring of ground movements was
carried out during construction of diaphragm wall panels nearest to the inclinometer tubes. Figure 8
shows the layout of inclinometer and settlement marks. For construction of bored piles, 15m long
temporary casing was used. The piles located within protection zone B were constructed after
completion of adjacent diaphragm wall panels to protect from the ground vibration induced by
temporary casing installation and extraction.
7.1 Soil Movement during Diaphragm Wall Construction
Soil movement during diaphragm wall panel construction were monitored (Table 7). Results of
settlement survey at the settlement marks indicated no vertical movements of the ground.
348
380
Table 7. Summary of Monitoring Results (I-6 & I-7): Soil Movements at Inclinometer Locations
Movement @ 15.0m depth
Construction Activities (mm) Remarks
I-6 I-7
D-Wall panel excavation to -10.0m -0.87 -0.40 Top 2.0m in depth was
D-Wall panel excavation to -20.0m 1.04 -0.81 protected by concrete
After de-sanding 0.12 -0.38 guide walls.
After concrete pouring -1.91 -2.18
24hrs after concrete pouring -2.34 -1.62
36/48hrs after concrete pouring -1.13 -2.55
Note: “-“ movement towards tunnel (opposite to direction of excavation)
INCLINOMETER I-6 (A AXIS) INCLINOMETER I-7 (A AXIS)

-20 -15 -10 -5 0 5 10 15 20 -20 -15 -10 -5 0 5 10 15 20

0 0
5 5
MRTA Northbound Tunnel
N
Asoke Road
10 10
Depth, m.
Depth, m.
Tun nel
MRTA Southbound
9.625m
7.815m
15 15
I-6 I-7
20
I-4
20
LEGEND 12 Storey O
Inclinometer
25
D-wall excavate to -10 m
25
D-wall excavate to -10 m Settlement Point
D-wall excavate to -20 m D-wall excavate to -20 m D-Wall with Pile Leg
After de-sanding
After de-sanding
ilding D-Wall with Barrette Leg
After concreting After concreting I-3
24 hrs After concreting 24 hrs After concreting Barrette
48 hrs After concreting" 36 hrs After concreting"
Figure 7. Inclinometer Monitoring Results of I-6 and I-7 Figure 8. Layout of Instrumentation
for Diaphragm Panel Construction
The following conclusions were made based on the above monitoring results.
1. No vertical displacement was observed at the surface settlement points; vertical displacement
of ground due to diaphragm wall trenching seemed to be within 2.0m from the trench face and
would not affect the tunnels which were located 7.8m beyond the trenching.
2. Maximum lateral soil displacement at 4.0m from the trench face at 15.0m depth was about
1.0mm towards excavation. It was expected that the displacement would decrease with
distance increasing from the trench face. At the tunnel location, lateral soil displacement was
estimated to be less than 0.5mm based on the extrapolation of displacement data.
3. After concrete casting of the panels, lateral soil movements were found to be towards the
tunnels, possibly from pressure of wet concrete. Maximum movement at depth 15.0m at the
inclinometer location was about 2.18mm after concrete pouring. After 36hrs of concrete
casting, the maximum movement observed was about 2.55mm. At the tunnel location, lateral
soil displacement was estimated to be less than 1.6mm based on the extrapolation of
displacement data.
4. After the diaphragm wall panel was completed, residual movement of the soil towards the
tunnels was observed. Due to the direction of this movement, it could be excluded from
tunnel movement measurements induced by diaphragm wall deflection during basement
excavation work.
349
381
8. CONCLUSION
At the time of drafting this paper, soil improvement with soil-cement columns was completed and
excavation work has commenced with specified instrumentation and monitoring programs. An
independent third party consultant has been appointed by the main contractor to review, inspect and
supervise the excavation work.
Diaphragm wall and foundation bored pile construction was completed, in compliance with
technical requirements for engineering work within the MRTA protection zone. Review and
assessment of the deep excavation with diaphragm wall adjacent to the existing tunnels indicated that
the project is feasible with adequate monitoring and contingency plans as outlined.
It is also concluded that diaphragm wall construction in close vicinity of existing subway tunnels
was successfully completed with adequate instrumentation and monitoring.
Further research is recommended on the actual performance of the diaphragm wall when
basement construction data and complete monitoring results are available.
REFERENCES
CIRIA Report 185 (1999), “The observational method in ground engineering: Principles and
applications”, CIRIA, 6 Storey’s Gate, Westminster, London, UK.
Chapman T. J. P., Nicholson D. P and Luby D. (2001), “Use of the observational method for the
construction of piles next to tunnels”, Proceedings of the International conference on Response
of Buildings to Excavation Induced Ground Movements held at Imperial College, London, UK.,
July 17-18, 2001, pp 213-223.
Chu R. P. K., Yau P. K. F., Leung D. H. K., and Mok K. H. (2001), “Integrate approach for deep
excavation along soft ground (MRTC) tunnels”, Proceedings of the Third International
Conference on Soft Soil Engineering, Hong Kong, December 6-8, 2001, pp 391-396.
Higgins K G, Paterson J., Moriaty J., Potts D. M., and Jardine R. J. (2001), “The effect of an
excavation in a stiff fissured clay on existing tunnels in close proximity”, Proceedings of the
International Conference on Response of Buildings to Excavation Induced Ground Movements
held at Imperial College, London, UK., July 17-18, 2001, pp 313-324.
Jianhang L. (1998), “Design and construction of excavation for Shanghai Metro”, Big Digs Around
the World, Proceedings of Sessions of Geo-Congress 98, Geotechnical Special Publication No.
86, Boston, Massachusetts, USA, October 18-21, 1998, pp 103-123.
Mair R. J., Taylor R. N., and Bracegirdle A., (1993), “Subsurface settlement profiles above tunnels in
clays”, Geotechnique 43, No. 2, pp 315-320.
Niu J. X.,Wong I. H., and Makino M., (2003), “Effects of deep excavation on adjacent transit tunnels
and station”, Proceedings of the Twelfth Asian Regional Conference on Soil Mechanics and
Geotechnical Engineering, Singapore, August 4-8, 2003, pp 883-886.
Phienwej N., Hock G. C., and Balasubramnaiam A. B., (1998), “Ground movements in deep
excavations with concrete diaphragm walls in Bangkok soils”, Proceedings of the Thirteenth
Southeast Asian Geotechnical Conference, Taipei, Taiwan ROC, November 16-20, 1998, pp
637-642.
Schulz C., (2001), “Geotechnical challenges in the construction of Thailand’s first subway”,
Proceedings of the Fourteenth Southeast Asian Geotechnical Conference, Hong Kong,
December10-14, 2001, pp 413-418.
Thasnanipan, N., Maung, A. W., Boonyarak, T., and Aye, Z. Z. (2004), “Stability of a trial trench
excavated under polymer slurry in Bangkok soft clay”, 15th Southeast Asian Geotechnical
Conference, Bangkok, Thailand, November 22-26, 2004.
Yang K. S., Tsen C. N., and Tang S. K., (1998), “Performance of a deep excavation with top-down
construction”, Proceedings of the Thirteenth Southeast Asian Geotechnical Conference, Taipei,
Taiwan ROC, November 16-20, 1998, pp 587-593.
350
382
Proceedings of International Symposium on Tunnelling for Urban
Development
14-15 December 2007, Pattaya, Thailand
International Symposium on
Tunnelling for Urban
Development
(IS-Pattaya 2007)
Application of Observational Method in

Diaphragm Wall Support Excavation
Zaw Zaw Aye, Chanchai Submaneewong, Thayanan
Boonyarak and Chakkrit Chanchad
351
International Symposium on Tunnelling for Urban Development (IS-Pattaya 2007)
Pattaya City, Thailand
Application of Observational Method in Diaphragm Wall Support

Excavation
Zaw Zaw Aye1, Chanchai Submaneewong2, Thayanan Boonyarak3 and Chakkrit Chanchad4
1,2,3,4
Seafco Public Co., Ltd., Bangkok, Thailand
Abstract: Application of observational method in managing geotechnical risk in diaphragm wall

support excavation in Bangkok is presented in this paper. Intensive modification of construction
sequence in actual work execution with “value engineering options” different from tender stage design
along with application of observational method in Diaphragm-wall-support two level underground car
park building located in the historically significant area of Bangkok is firstly demonstrated. Simple
approach of observational method in risk assessment of buildings and structures backup by extensive
instrumentation in tunneling and deep excavation works of Contract No. 1 of the M.R.T Chaloem
Ratchamongkhon Line, the first underground mass transit system project of Bangkok (MRTA) is also
reported. Key Factors contributed in successful application of observational method are discussed.
1. INTRODUCTION
Observational Method (OM) is not a very popular term among geotechnical engineers in Thailand,
though the method has been applied in some major projects. The main reason of unpopularity of the
method in Thailand may be due to the fact that traditionally there is less option opened for design
modification as the project progresses. Geotechnical engineers may be more concerned on misused
and limitation of the method and do not prefer to take risk. It is important to discuss among
geotechnical engineers with regard to overall philosophy, key requirement, limitation and practical
approach in application of observational method.
Tunneling works and deep excavation works have a high degree of geotechnical risk that needs
to be identified, assessed, minimized and managed. Deep excavation and tunneling works generally
interact with natural ground or geo-material and incorporating their characteristics as structural
elements of their own stability. This interaction is more significant in underground excavation and
tunneling projects than in other civil engineering projects. In this paper, application of observation
method in managing geotechnical risk in two different projects is presented.
2. OBSERVATIONAL METHOD AND GEOTECHNICAL RISK MANAGEMENT
Historically, engineers used observational method to deal with uncertainties in the ground or
geotechnical risks in early age of civil engineering. Peck (1967) pointed out the advantages of
observational method by giving an example on a dam across the Svir River in Russia, from a
documented record of the work of Graftio (1936).
Professor R. B. Peck set out procedures for the observational method (OM) as applied in soil
mechanics in the Ninth Rankine Lecture (Peck, 1969). Peck described the limitation and drawbacks of
observational method. Powderham (1996) reviewed the main features of the observational method of
Peck and summarized the key requirements as follows;
(1) It must be possible to alter the design during construction
(2) The contractual condition must be compatible and allow design to be directly related to actual
construction method
(3) An acceptable level of risk must be identified and controlled. In particular this requires a
planned course of action for every foreseeable eventuality
353
(4) Critical observation must be identified and obtained

3. OBSERVATIONAL METHOD, RISK MANAGEMENT AND VALUE ENGINEERING
Figure 1 shows a generalized link between the elements of observational method and risk management
with integration of value engineering.
Geotechnical
Observational Method
Risk Management
Identify Risk Design & Plan
Value Engineering
Analysed Degree Construction Control

of Risk
Monitoring & Review
Trigger Criteria
Exceeds No
Minimize
& Control Implement
Yes Planned Action
Figure 1. The elements of observational method and risk management with integration of value
engineering.
4. VALUE ENGINEERING AND APPLICATION OF OM IN BASEMENT EXCAVATION

WITH BUTTRESS-SUPPORT THIN DIAPHRAGM WALL
This project demonstrated the benefits of value engineering option powered by effective application of
observational method in historically significant area of Bangkok. The project is a two-level
underground car park located in the center of Rattanakosin Island, the heart of an old established,
historically and culturally significant area of Bangkok. The project owner, Bangkok Metropolitan
Authority (BMA) awarded the semi-turnkey basis construction contract “Lam Kon Muang
Underground Car Park” to SEAFCO Co., Ltd. as a contractor. The contract consists of 3 major scope
of works: (1) Construction of building foundation and retaining structure - diaphragm wall, barrette
and bored piles (2) Excavation works including temporary bracing design and installation (3)
Construction of the entire two-level underground car park building having car park area of 18,552 m2
and roof-level park of 10,936m2 plus cut-and-cover tunnel, underpass access to the City Hall.
Geotechnical aspects highlighting the application of observational method and performance of
buttress-support diaphragm wall of 0.60 m width for two-level underground car park building is
discussed in this paper.
5. PROJECT REQUIREMENT AND MAJOR CONSTRAINTS
Since there is a limited availability of car parking space in the surrounding congested neighborhood
and the project site was being used as a grade-level car parking space prior to the award of the
contract, the key requirement was to construct the underground car park in two phases – to construct
Phase 1 while leaving space for car parking in Phase 2, and to utilize semi-finished underground car
park of Phase 1 during construction of Phase 2. This requirement posed the need of temporary
retaining wall between Phase 1 and 2.
354
Construction site is surrounded by numbers of sensitive structures as shown in Figure 2, - in the

south, Wat Suthat, one of Thailand’s most important temples and the Historical Giant-swing, in the
north, the City Hall and at South-east corner, the Historical Brahmin Temple. Rows of old shop-house
buildings are closely located in the east and west boundaries of the project. Location of the project site
itself in the vicinity of sensitive structures and buildings therefore posed some constraints, which
called for the need of careful consideration in establishing the design principles and sequence of
construction.
CITY HALL
N
MAHUNNOP ROAD
OPENING
OPENING
SHOP-HOUSES
OPENING
DINGSOR ROAD
OPENING
SIRIPONG ROAD
SHOP-HOUSES
OPENING
OPENING
SHOP-HOUSES
OPENING OPENING
BAMRUNG MUANG ROAD

GIANT-SWING
WAT SUTHAT (TEMPLE)
Figure 2. Layout plan of the project showing adjacent buildings
Location of the project site itself in the vicinity of sensitive structures and buildings therefore
posed some constraints, which called for the need of careful consideration in establishing the design
principles and sequence of construction. The architectural and utility aspects of the project called for
the design of the basement with a number of openings from the ground surface to the final basement
slab level to facilitate the ventilation system as shown in figure 3.
Hence the roof slab can not be physically utilized as bracing in most of the area where the
diaphragm wall is to be acting as a cantilever retaining wall in the permanent stage. It was analyzed in
the preliminary analyses that the deflection of the diaphragm wall of 0.60 m width would be large if it
was to be fully cantilevered. As the project is located in a sensitive area, ground movement induced by
large deflection was unfavorable. It was therefore decided to use buttress to minimize the diaphragm
wall deflection as shown in figure 3.
355
BARRETTE W ALE BEAM
0 .3 0
T IE B E A M
P IL E C A P SLA B
(a)
N I-5 N O R T H D -W A L L
+0.20 ROOF SLAB
0.00
CAPPING
G.W.L
BEAM
RC BUTTRESS -2.70 SLAB (B1)
I-4 I-6
PH A S E 2
2
TIE PILE CAP
-5.60 (B2)
EAST D-WALL
BEAM
WEST D-WALL
SOFT CLAY 2
S H E E T PIL E
-9m WALE SOFT CLAY
124 m
BEAM
S ee d etails in (b ) DIAPHRAGM MEDIUM CLAY

-12m WALL (tip=-16m)
I-3 I-1
STIFF CLAY
PH A S E 1 -15m
-16m
VERY STIFF CLAY
FOUNDATION PILES
BARRETTE DIA. 600mm (tip=-20m)
S O U T H D -W A L L I-2 -20m. (tip=-20m) -20m.

78 m
(b) (c)
Figure 3. (a) Plan of diaphragm wall and buttress (b) layout of diaphragm wall and piles (c) sectional
view of the structures
6. DIAPHRAGM WALLS BARRETTES AND BORED PILES CONSTRUCTION
Diaphragm wall having 600mm width founded at 16m below ground level (B.G.L) was constructed
simultaneously with dry-processed bored piles of diameter 600 mm with toe depth 20 m below ground
level. Barrettes having same toe depth as bored piles were installed at 8m spacing along with
diaphragm wall panels. Sheet pile wall (14m deep) was used as a temporary retaining wall at the
boundary of Phase 1 and 2 as shown in Figure 4.
7. SUBSOIL CONDITION AND DESIGN PARAMETERS
Typical subsoil profile at the site is characterized by thick Bangkok soft clay layer at the top followed
by thin layer of medium clay, and stiff clay layers.
356
2
Undrained Shear Strength (kN/m )
1 0 50 100 150 200
0 0.00
5 Soft clay 5.00
10 Medium stiff clay 10.00

Depth from ground surface (m)
Stiff clay
15 15.00
20 Very Stiff clay 20.00
25.00
25
30.00
30 Medium Dense
Sand 35.00
35
40.00
40
Dense sand 45.00

45
50.00
50
0 20 40 60 80
SPT N-Value (blows / ft)
Su BH-1 Su BH-2
Su BH-3 Design (Su)
SPT BH-1 SPT BH-2
SPT BH-3 Design (SPT)
Figure 4. Plotted Su and SPT N-value with design lines
8. ORIGINAL TENDER STAGE DESIGN
The designers involved in the tender stage design made a fairly conservative design with two levels
temporary bracing as shown in Figure 5. Uncertainty of the performance of a thin diaphragm wall in
soft clay layer was the likely reason to adopt the conservative design in the tender stage. It is not
unreasonable to adopt the conservative design considering time-dependent consolidation property of
soft marine clay and likely long elapsed time of un-strutted diaphragm wall (relatively long un-strutted
span of about 6.0 m between first strut and final excavation level for 600 mm diaphragm wall) due to
large volume of excavation work involved.
HORIZONTAL STRUT SLAB
RAKING STRUT
D-WALL
BERM
D-WALL
BARRETTE
-16m
-20m BORED PILE
Figure 5. Tender stage bracing system – diaphragm wall was designed with soil-berm and 2 struts
support (horizontal bracing and raker)
357
9. VALUE ENGINEERING OPTION FOR PHASE 1
Value engineering review of the temporary works was undertaken by the contractor’s new in-house
design engineering team prior to the commencement of Phase 1 excavation works. Rigorous attention
to detail of the design concept and constructability was made in the pre-construction discussions
between design engineers and construction team.
The main objectives for value engineering options were to minimize the material and
construction sequence involved in temporary works so as to accelerate the excavation time thereby
saving overall costs without compromising the safety aspect. After conducting a series of re-analyses
with different conditions major modifications were made: (1) To lower the first strut level to –1.8 m
from the original tender stage design level –1.0 m (2) To use only 1 temporary strut, omitting second
level raking strut with the provision of sloping soil berm against diaphragm walls. Soil berm was to
remove after completion of base slab construction in the majority of area – minimizing the elapsed
time of partially un-strutted diaphragm wall between temporary strut and the final excavation level.
Modified bracing system of Phase 1 as an outcome of value engineering review is illustrated in Figure
6.
HORIZONTAL STRUT SLAB HORIZONTAL STRUT SLAB
BERM
RAKING STRUT
D-WALL
SHEET PILE
BARRETTE
-14m
-16m -16m
-20m -20m BORED PILE

Figure 6. Value engineering option - Modified Bracing system for Phase 1
10. IMPLEMENTATION OF THE OBSERVATIONAL METHOD
During the review of tender stage design for Phase 1, it was recognized that two-phase excavation
works in this project was ideally suitable for application of the observation method. The flexibility of
the contractual requirement in temporary design which allowed the contractor to modify the design
and construction method also provided the favor for the observational method.
In order to assess the most probable condition assumed in “value engineering design” and to
take necessary actions if monitoring results reveal most unfavorable conditions (i.e. actual deflection
of diaphragm wall reaches maximum acceptable limit), the observational method was implemented on
the followings basis.
 Reviewed the design parameters together with critical conditions posed on sites as well as most
critical stage in excavation and basement construction work
 Predicted the performance of diaphragm wall with “most probable” as well as “most unfavorable”
conditions and parameters.
 Established the trigger criteria based on predicted diaphragm wall deflection
 Predefined the practical contingency plan for “most unfavorable” conditions where wall deflection
reaches trigger levels
 Set out the instrumentation program with the consideration of above factors
358
 Monitored the performance of diaphragm wall. Compared the monitoring results with the predicted
and trigger values and reassessed
 Implemented the contingency measures if monitoring result reaches action level of trigger values
Process of OM in Phase 1 is illustrated in a flowchart in figure 7.
OM in Phase 1
Value
Design & Plan
Engineering Review &Modification of Tender Stage Design
Option
Construction Control Reassessment

Process
Monitoring & Review
Trigger Criteria
Exceeds No
Implement
Yes
Planned Action
Figure 7. Flowchart showing the OM process involved in Phase 1
Figure 8 shows the predicted diaphragm wall lateral displacement or deflection of Phase 1 (east,
west and south diaphragm wall) at two conditions together with trigger levels and tender stage
prediction. It should be noted that the diaphragm wall deflection was predicted to be maximum or
most critical after removing the horizontal temporary strut.
Most probable condition was established for the predicted deflection of diaphragm wall with
full influence of buttress support – assuming buttress effectively supports as permanent strut in
diaphragm wall analysis model. Most unfavorable condition was set out for the predicted deflection of
diaphragm wall without considering influence of buttress – buttress was excluded in the model.
Diaphragm wall reinforcement was designed based on the most unfavorable condition.
0 20 40 60 80 100
0
NORMAL ALARM ACTION
2 MOST
PROBABLE
PREDICTION
6
MOST
8 UNFAVORABLE
PREDICTION
Depth, m.
10
12
14
Tender Stage
16 Prediction with buttress :

Alarm Trigger Level
Prediction without buttress
18
Action Trigger Level
20
Figure 8. Prediction of diaphragm wall performance in pre-construction stage with trigger criteria in
comparison with tender stage prediction
359
In establishing the criteria for trigger values, it was necessary to consider the broad context in
which diaphragm wall exists, design assumption and concept, likely behaviour of the wall itself or its
predicted performance and effectiveness of selected temporary bracing system. Trigger levels were
established to provide the design team and construction team, an opportunity for early review and
resetting of the monitoring frequency as well as for implementing the contingency measure as
necessary.
In general, exceeding alert trigger levels must initiate a review of design data, construction
progress and monitoring frequencies with the consideration of possible measures to limit further
deflection. Exceeding action trigger levels must initiate further review of above mentioned points and
if necessary to initiate a planned course of action or contingency measures.
Effective and good communications between the design team and construction crew were made
all along the excavation stages with clear responsibilities in construction control. The contingency
measures included immediate backing filling of excavated soil and installing of temporary struts in
the critical area.
11. MONITORING PROGRAM
In planning the monitoring system and program, it is important to consider the parameters to be
measured which reflect the actual performance of the diaphragm wall support excavation. It is also
necessary to take account the practical measurement applicable for established trigger criteria, number
and frequency of measurement required to carry out a meaningful interpretation of wall behaviour
which would be integrated in the implementation of observational method.
Comprehensive and robust monitoring program was set up as a key element in application of the
observational method and to ensure that modified construction sequence would not have adverse effect
in temporary stage and on permanent design. A total of 6 inclinometers (3 in each phase) were
installed in diaphragm wall together with some survey points. In order to make effective use of the
established trigger levels, an adequate number of measurement were carried out at appropriate
frequency. Typical monitoring frequencies for inclinometer set up as guideline for the project is
outlined below.
• Measured immediately before commencing excavation in the vicinity of instrument
• Minimum readings of 2 times a week while excavation in progress
• Minimum readings of 1 time a week when no excavation the vicinity of instrument
• Minimum readings of 3 times a week when measured deflection values exceeded alert trigger
levels
• Minimum reading 1 time a day when measured deflection values reached action trigger levels
As the most critical stage was predicted at the time horizontal temporary bracings were removed,
a full attention was paid to the inclinometer monitoring with the following special criteria and
frequencies.
• First temporary strut removal was to carry out at the diaphragm wall panel where inclinometer
was located
• Measured immediately before removal of temporary strut at the closest distance to the instrument.
• Measured every 6-8 hours immediately after removing the first strut
• Second strut to be removed must be the one located immediately adjacent to first strut which had
removed
• Not to remove the second strut until inclinometer measured deflection values had stabilized
• Not to remove more struts unless measured deflection values were stabilized and within alarm
trigger levels
In addition to inclinometer measurement, diaphragm wall movement was also monitored by the
survey points strategically marked on the wall panels. Ground settlement and surface cracks behind
the diaphragm wall were also visually checked by the construction team as daily basic.
360
12. PERFORMANCE OF PHASE 1 DIAPHRAGM WALL
After carrying out the comprehensive desk studies and establishment of systematic monitoring
program presented above, Phase 1 excavation work was carefully commenced. Figure 9 shows the
maximum accumulated diaphragm wall deflection at different stages of excavation monitored by
inclinometer No.1 (I-1 at East wall of Phase 1) together with trigger levels and predicted maximum
deflection profile of 3 different conditions - tender stage design (2 temporary struts), modified design
with buttress and modified design without buttress. It can be observed from figure that measured
lateral movement pattern of diaphragm wall agreed well with that of prediction for modified design
with buttress - meaning buttress-support has significant influence on wall deflection.

0 20 40 60 80 100
0
NORMAL ALARM ACTION
MOST PROBABLE PREDICTION

2
6 MOST
UNFAVORABLE
PREDICTION
8
Stage 1
Depth, m.
10 Stage 2
Stage 3
12
Stage 4
14 Stage 5
Tender Stage
16
Alarm Trigger Level
18 Action Trigger Level

20
Figure 9. Phase 1 East Wall-Monitored diaphragm wall deflection at different stages with trigger
levels
0 20 40 60 80 100
0
NORMAL ALARM ACTION
2 MOST PROBABLE PREDICTION
6
MOST UNFAVORABLE
8 PREDICTION
Depth, m.
Stage 1
10
Stage 2
12 Stage 3
Stage 4
14 Stage 5
Tender Stage
16
Alarm Trigger Level

20
Figure 10. Phase 1 South Wall - Monitored diaphragm wall deflection at different stages with trigger
levels
361
Deflection profile of South diaphragm wall which braced against temporary sheet pile wall at
Phase 1 and 2 boundaries is presented in Figure 10. As can be seen in Figure 10, South diaphragm
wall deflection is significantly higher than that of east wall, which is likely to be caused by the fact
that South diaphragm wall is braced with more flexible sheet pile wall. description of stages shown in
the legend of Figure 9 and 10 is summarized in Table 3.
Table 3. Description of stages shown in Figure 9 & 10

Stage Description
1 Excavate to –2.2m and installed temporary strut
2 Excavate to –4m
3 Excavate to –6.6m with berm
4 Removal of berm
5 Removal of temporary strut after completion of buttress
Figure 11. Phase 1 excavation work in progress with historical Buddhist temple Wat Suthat in
background (Phase 2 area was being used as car parking space)
13. OM IN PHASE 2 - MODIFICATION OF BRACING SYSTEM
Monitoring results of the Phase 1 excavation work provided an ample opportunity to review the
design assumption, fine tune the parameters used in the analysis of the diaphragm wall for the Phase 2
and made modification of construction sequence. The major modifications are: (1) Removal of
soil-berm at East and West diaphragm wall in shorter duration than that of Phase 1, and (2) Using
raking struts instead of horizontal strut for North diaphragm wall a shown in Figure 12.
362

STRUCTURE
RAKING STRUT
BERM SLAB
BERM
D-WALL
D-WALL
BARRETTE
-16m -16m
-20m -20m
Figure 12. Bracing system of Phase 2 diaphragm walls
Process involved in application of OM in Phase 2 is shown in figure 13.
OM in Phase 2
Value
Engineering
Option Design & Plan
Modification of Design
Based on Actual Performance of Phase 1
Reassessment
Used data & Process
knowledge
gained from Construction Control
Phase 1
Calibrated Monitoring & Review

design parameters
Trigger Criteria
Exceeds No
Implement
Yes
Planned Action
Figure 13. Flowchart showing the OM process involved in Phase 2
14. PERFORMANCE OF PHASE 2 DIAPHRAGM WALL
Figure 14 depicts the measured deflection of east diaphragm wall of Phase 2. As can be observed in
Figure 14 in comparison with Figure 9, the maximum deflection of east diaphragm wall in Phase 2 is
larger than that of Phase 1.
363
The likely reasons of this observation are;

 Un-strutted elapsed time for first temporary bracing in Phase 2 was longer than that of Phase 1.
 In Phase 1, horizontal struts were installed in north-south direction which temporary kingpost
columns and strut were integrated in crisscross pattern with east-west struts - providing
complete-support more rigid bracing system. Whereas in Phase 2, horizontal struts were installed
only in east-west direction without having crisscross pattern with north wall – having less rigid
bracing system than that of Phase 1.
0 20 40 60 80 100
-2
NORMAL ALARM ACTION
6
MOST UNFAVORABLE
PREDICTION
8
Depth, m.
Stage 1
10
Stage 2
12 Stage 3
Stage 4
14
Stage 5
16 Alarm Trigger Level

18
20
Figure 14. Phase 2 East Wall - Monitored diaphragm wall deflection at different stages with trigger
levels
With assurance of diaphragm wall performance from monitoring results of Phase 1, original
plan of using horizontal struts for North diaphragm wall was modified by using raking struts instead.
As can be seen in Figure 15, deflection of North diaphragm wall (with raking strut support) is
significantly higher than that of East diaphragm wall (with horizontal strut support). The main reason
of larger movement of north diaphragm wall is due to the fact that it was supported only by the berm
for the long period (about 52 days) before completion of raking struts so that soil-berm became soften
during the long elapsed un-strutted period.
Table 4. Description of stages shown in Figure 14 & 15

1 Excavate to –2.2m and installed Excavate to –2.2m at d-wall, and to –6.6m with
temporary strut sloping berm
3 Excavate to –6.6m with berm Removal of berm
4 Removal of berm Removal of raker after completion of buttress
5 Removal of temporary strut after
364

0 20 40 60 80 100 120 140
-2
NORMAL ALARM ACTION
0
6
MOST UNFAVORABLE
PREDICTION
Depth, m.
10
Stage 1
12 Stage 2
Stage 3
14
Stage 4

18
20
Figure 15. Phase 2 North Wall - Monitored diaphragm wall deflection at different stages with trigger
levels
Time-dependent deflection pattern due to softening and deformation of soft clay can be
observed in North diaphragm wall as illustrated in Figure 17. North diaphragm wall moved
progressively toward excavation before completion of raking struts (at 52 days) as can be seen in
Figure 17.
Figure 16. View of buttress-support diaphragm wall prior to temporary bracing removal
365
0 20 40 60 80 100 120 140

0
6
Depth, m.
8
10
0 days
12
30 days
14 52 day
56 days
16 74 days
Prediction
18
20
Figure 17. Phase 2 – Time dependent wall deflection of North diaphragm wall – diaphragm wall was
supported only by soil-berm for 52 days before completion of raker installation.
15. IMPLEMENTATION OF CONTINGENCY PLAN
Since deflection of North diaphragm wall (Phase 2) approached action trigger levels, monitoring
frequency was increased and the following contingency measures were implemented on site.
 Poured 15cm thick 1m wide lean concrete on the top of the berm along North diaphragm wall to
provide bearing-effect
 Installed additional king-post and diagonal struts attached to the raking struts to provide more rigid
support against diaphragm wall
 Soil-berm was removed locally in bays followed by construction of wale beam, tie beam and
buttress as shown in Figure 18.

TEMPORARY WALER
BUTTRESS
DIAGONAL STRUT
DIAPHRAGM WALL
CAPPING BEAM TEMPORARY KING-POST
RAKER
TEMPORARY
BERM STEEL BEAM
PILE
CAP
B
SLA
TED
PLE
COM
Figure 18. Perspective of North diaphragm wall constructed in bay
366
Movement of diaphragm wall was observed to be decreased and eventually stabilized by the
above actions. No significant ground settlement was observed in the vicinity of the North diaphragm
wall.
Figure 19. View of raker and soil-berm support – soil-berm was removed locally bay by bay
Figure 20. View of buttress-support diaphragm wall after removal of temporary bracing
367
16. TIME AND COST SAVING FROM VALUE ENGINEERING OPTIONS AND THE
APPLICATION OF OBSERVATIONAL METHOD
Significant cost and time saving were achieved from the value engineering option coupled with
observational method implemented for both Phase 1 and 2. The major savings were achieved by less
operation and material utilized in the following elements of temporary works.
 Cancellation of 2nd level raking struts against diaphragm wall for both phases
 Modification of bracing system – using raking struts with soil-berm support instead of horizontal
struts for North diaphragm wall in Phase 2
17. KEY FACTORS CONTRIBUTED IN SUCCESSFUL APPLICATION OF

OBSERVATIONAL METHOD
Outcome of a through desk study at post-tender stage provided an effective value engineering
option which offered significant cost and time saving for overall construction program. Effective and
good communications between the design team and construction crew played a key role in successful
application of observational method to completion of the project. Systematic monitoring program with
clear defined trigger criteria was also the important element in implementing the observational method.
This case study reveals that a thin permanent diaphragm wall coupled with effective design and
construction method supplemented by the observational method and robust monitoring program could
offer a logistically and financially attractive solution in construction of underground car park without
disturbing the environment in the prominent historical area of Bangkok.
18. REFERENCES
Clayton C. R. I., 2001. Managing Geotechnical Risk. Improving Productivity in UK Building and
Construction, Institute of Civil Engineers and Thosmas Telford Ltd. UK
Clough, G.W., and O’Rourke, T.D., 1990. Construction Induced Movements of In-situ Walls. ASCE
Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. O’Rourke, 1990. Deep Excavation and Tunneling. 7th ICSME, Mexico City
Graftio H., 1936. Some Features in Connection with the Foundation of Svir 3 Hydro-electric Power
Development,” Proceedings of the first international conference in Soil Mechanics, Cambridge,
Mass., 1, pp. 284-290
O’Reilly M.P and New B.M., 1982. Settlements above tunnels in United Kingdom – their magnitude
and prediction. Tunneling ’82, pp. 173-181, London, IMM
Peck R.B., 1969. Deep Excavation and Tunneling in Soft Ground. 7th ICSME, Mexico City, State of
the Art 3, 225-290
Peck, R. B., 1969. “Advantages and limitations of the observational method in applied soil mechanics”,
Geotechniques 19, No.2, pp. 171-187
Powderham, A. J., 1996. “An overview of the observational method : development in cut and cover
and bored tunneling projects”, The Observational Method in Geotechnical Engineering, The
institute of civil engineers, Thomas Telford Publishing, London, pp. 7-24
Zaw Zaw Aye, Dhiraj Karki and Christian Schulz, 2006. “Ground Movement Prediction and Building
Damage Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil,
Proceedings of the International Symposium on Underground Excavation and Tunneling, Urban
Tunnel Construction and Protection of Environment, EIT, Bangkok, Thailand, pp. 281-297
368

Conference 2012
Diaphragm Wall Support Deep-Excavations in

Bangkok
Narong Thasnanipan, Zaw Zaw Aye, Aung Win Maung,
Chanchai Submaneewong
Thayanan Boonyarak
369
Diaphragm Wall Support Deep-Excavations
In Bangkok Subsoil
Narong Thasnanipan1, Zaw Zaw Aye1, Aung Win Maung1, Thayanan Boonyarak2
Phongpat Kitpayuck1, Nutthapon Thasnanipan1
1
2
ABSTRACT
Diaphragm wall support deep excavation works of some major projects including Bangkok first subway
project are presented in this paper. Available literatures related to diaphragm wall and deep excavation works
in Bangkok are summarized together with recent research works. Problems and difficulties in construction of
diaphragm walls and deep excavation works in early days are reported. The areas of development in
diaphragm wall construction and main factors contributed to these are also presented. An attempt is also made
to briefly discuss the future trend of diaphragm wall construction and deep excavation works in Bangkok. This
paper is intended to serve as a source of reference for the practitioners in the construction industry with regard
to the application of diaphragm walls in deep excavation works in urban area of Bangkok.
INTRODUCTION
As in other major cities, growing land price and need of underground space for commercial and infrastructure
developments necessitated deeper underground excavation works in Bangkok. Diaphragm wall as embedded
earth-retaining structure is one of the most suitable solutions to facilitate the deep excavation works in urban
area of the growing metropolis. The common thickness of diaphragm wall in Bangkok ranges from 0.80m to
1.20m. Though not very common, in some cases, thick diaphragm wall of 1.50m and thin diaphragm wall of
0.50 to 0.60m were also constructed. Toe depths of diaphragm walls are in the range of 16m to over 30m
depths depending on the final elevation of the excavation. This paper aims to provide the summaries of the
technical data on diaphragm wall construction and deep excavations in Bangkok with particular focus on the
progressive development in past decades. Literatures published throughout past 20 years are summarized
along with the recent research works. The information and data presented in this paper are based on the
author’s personal experience, observation from other completed projects and from published literatures.
SUBSOIL AND PIEZOMETRIC PROFILES IN BANGKOK
Subsoil profile and the present piezometric drawdown condition of Bangkok are presented in Fig. 1 below. A
typical subsoil profile is relatively consistent in different localities in Bangkok. It is characterized by
alternating layers of clay and sand deposits as shown in the Fig. 1.
371
0 Weathered Crust
0

Clay  t = 14-16kN/m 3
10 -
10
Medium to
Su= 50-14020- KPa
Depth (m)
20 Stiff Clay
 t = 17-21kN/m 3
Act
ual
Medium
-
P
30
Hyd
 t = 20kN/m 3
iezo
(First Sand)
ros
me
ta
40 - tric
tic L
Hard Clay Su> 200 KPa 40
 = 21KN/m 3
Dra
t
ine
wdo
Dense to Very
wn
50 -
50
Lin
(Second Sand)  t = 21KN/m 3

e
60 -
60
Fig. 1, Subsoil and piezometric profile of Bangkok
THE NEEDS OF DIAPHRAGM WALL FOR DEEP EXCAVATIONS IN BANGKOK
Generally, requirement and selection of basement retaining walls are based on the following factors.
• Depth of final excavation of basement – shallow or deep excavation

• Subsoil condition
• Construction sequence and permanent condition of basement structure
• Environmental conditions and constraints – effect of retaining wall installation and basement
excavation to neighbouring properties and structures
• Constructability – availability of space and physical condition of existing site
(Peck et. al. 1977) defined that deep excavations are those whose depths are deeper than 6m. Sheet
piles, contiguous-pile walls, secant pile walls and sinking caisson are alternative retaining structures available
in Bangkok for deep excavations. Prevailing subsoil (existence of very soft to soft clay) and ground water
condition are the main factors for using diaphragm walls for deep basement excavation in Bangkok. For most
of the deep excavation works in urban environment, advantages offered by diaphragm walls weigh more
favourably for both technical and economical reasons while other methods have distinct limitations. Major
advantages of diaphragm wall are listed below.
• Can be used as permanent structural wall
• Water retainable
• Can be installed to deeper depths and for load bearing element
• Less temporary propping needed
• Can be applied for top-down construction method
• Rigid structure so that ground movement induced by basement excavation is less than other flexible
retaining wall
• Vibration and noise generated from installation of diaphragm wall is less than other methods
Key limitations of diaphragm wall can be summarized as;

• Diaphragm wall itself is water retainable structure but not 100% water-proof. Additional measures are
required if basement required high-grade water-proof wall
• Availability of limited numbers of specialist diaphragm wall contractors
• Finished wall surface is influenced by the subsoil
372
DIAPHRAGM WALL CONSTRUCTION METHOD
The diaphragm wall construction method allows for the formation a reinforced concrete wall beneath the
ground surface. The construction of the diaphragm wall is carried out from ground level, panel by panel
forming series of cast-in-place underground RC walls. RC Guide walls of 1m to 1.5m deep are mainly used to
assist the trenching operation. Each panel is excavated by using cable operated mechanical clamshells
suspended from crawler cranes or by more modernized hydraulic grabs. Throughout the excavation Bentonite
slurry is maintained at the top of the excavated trench for wall stability. On completion of excavation, the
Bentonite slurry which may have become contaminated with soil, is cleaned by recycling through de-sanding
equipment. The reinforcement cages are then lowered into the slurry filled trench, with each unit spliced to
the other, to form a continuous cage to the required depth. Tremie pipes are then installed to the base of the
panel and concrete is cast from the panel toe up to the required cut-off. During casting the displaced Bentonite
slurry is drawn off and stored for reuse. Adjacent panels are then excavated using the same procedure. Stop-
ends are used to provide the formation shear-key and continuity between adjoining panels. Fig. 2 is a
summarized demonstration of process involved in diaphragm wall construction. The selection of panel width
is mainly depending upon the available grab size, ground condition and geometric constraints of the basement
footprint. Panel width of 3m to 6m are commonly use in Bangkok. Where practical, it is advisable to use same
panel width for the ease of preparation and handling of the reinforcement cage.
CONCRETE
GL PULL OUT
GUIDE WALL
WL STOP-END TUBE
BENTONITE
REINFORCEMENT
FINISHED PANEL
PANEL EXCAVATION
BENTONITE
EXCAVATED PANEL REINFORCEMENT FINISHED PANEL

GUIDE WALL
EXCAVATED PANEL
GUIDE WALL
STOP-END TUBE
STAGE 1 STAGE 2 STAGE 3 STAGE 4

EXCAVATION REINFORCEMENT CONCRETING WITHDRAW STOP-END TUBES
Fig. 2 – Diaphragm wall construction process
OVERVIEW OF DIAPHRAGM WALL PROJECTS IN BANGKOK
The first diaphragm wall in Bangkok was believed to have been constructed in the late 1970s for the basement
retaining walls of the Bangkok Bank Head Office Tower on Silom Road. Table 1 summarized the diaphragm
wall projects constructed in the early days.
373
Table 1, List of the early diaphragm wall projects in Bangkok
Diaphragm Wall
Excavation Year of
Project Name Dimension (m)
Depth (m) construction
Thickness Depth
Bangkok Bank HQ 0.50 14.00 7.00 1977
International Trade Center 0.82 26.00 18.50 1985
Srivara Hightech Tower 1.02 26.00 17.20 1991
Silom Precious Tower 1.02 28.00 20.00 1992
Jewelry Trade Center 0.82 28.00 15.60 1992
Oriental Tower 0.82 28.00 15.50 1992
Sukhumvit 33 Tower 0.82 26.00 14.00 1992
Tharmasat University Library Building 0.80 28.00 12.70 1994
Fig. 3 illustrates the embedded length of diaphragm wall in relation to the excavation depth of the
projects listed in Table 1. It is to be noted that except shallow excavation required for Bangkok Bank Head
Office, diaphragm wall embedded deeper into hard clay layer in all projects constructed in the early days. Fig.
4 shows the embedded length of diaphragm wall in relation to the excavation depth of the projects listed in
Table 2.
From 1991, with the booming construction industry, usage of diaphragm walls for deep basements
of the high-rise buildings in Bangkok has been significantly increased. In year 1997, due to the economic
crisis, with significant decline of property sector, diaphragm wall construction was concentrated only for
infrastructure projects.

DEPTH BELOW EXISTING GROUND (m)
Fig. 3, Diagram showing the excavation depths in relation to embedded length of diaphragm walls in early
days (1977 to 1994)
374

Fig. 4, Diagram showing the excavation depths in relation to embedded length of diaphragm walls from 1997
to 2008
Table 2, List of some diaphragm wall projects constructed from 1997 to 2008 in Bangkok (Basement
Excavations Deeper than 10m)
Diaphragm Wall Dimension (m) Excavation Year of
Project Name Depth (m) Construction
Thickness Depth
Charoen Pump Station 0.80 20.00 16.50 1997
Sathorn Pump Station 0.80 20.00 16.20 1997
MWA GMC-7A 0.80 21.00 16.00 2002
Bank of Thailand 0.80 20.00 15.20 2003
Hampton Condominium 0.80 20.00 12.20 2003
Central World Plaza 1.00 18.00 9.00 2004
Boe Bae Bumrung Meung Plaza 0.80 24.00 19.00 2004
Esplanade 0.80 18.00 10.00 2005
Por Teck Tung Office 0.80 16.00 8.00 2006
Noble Remix 0.80 20.00 13.00 2007
Life@Sathorn 0.80 18.50 10.50 2008
Royal Maneeya 1.00 30.00 2005
Sathorn Square 0.80 21.00 15.50 2007
Renaissance 2008
The Sukhothai Residences 1.00-1.20 20.00-22.0 17.80 2008
Bangsue Wastewater Treatment 1.20-1.50 24.70 2009
Grand Rama 9 Square 1.00-1.50 22-25.00 20.15 2010
MahaNakhon Hill & Tower 0.6-0.80 16.00-18.0 15.40 2011
MahaNakhon Retail Cube 0.80 22.0 18.10 2011
UBC III & EM2 0.80-1.00 18.0-22.0 8.00-18.00 2011
375
Extensive application of diaphragm wall for deep excavations can be observed between 1997 and
2001 for the first Bangkok subway project where 18 deep stations and associated structure plus some sections
of cut-and-cover tunnels were constructed by diaphragm wall. (Phienwej et. al., 2006) discussed the difference
between the length of diaphragm wall to depth of excavation between two contracts of the first Bangkok
subway, Chaloemratchamongkhon Line. The authors cited that it is partly due to the difference in design
criteria adopted. Fig. 5 illustrates the embedded length of diaphragm wall in relation to the excavation depth of
MRT Stations. Excavation depth over 30m carried out in Silom Station set the deepest excavation ever done in
Bangkok subsoil.

en
ph
it
et
/m
Ph
am
ng
n
am
thi
po
B
ng
vit
ri
e
am
rat
Ru
lyo
rao
ada
hea
IX
Su
san
bu
ini
it
um
an
ai
Ch
cha
cha
aL
t
aho
dP
m
ma
ng
iam
mp
thi
tch
mp
nk
iki
my
kh
o
Mo
Ba
Pra
Bo
Hu
Su
Pet
Sir
La
Ra
Sil
Ra
Lu
Ph
Ka
Su
Sa
Th
Fig. 5- Diagram showing the excavation depths in relation to embedded length of diaphragm walls of 18
subway stations of the first Bangkok Mass Rapid Transit Project
DIAPHRAGM WALL CONSTRUCITON PROBLEMS IN THE EARLY DAYS
Basic but extensive problems were experienced in early stages of diaphragm wall construction as summarized
below.
6
376
• Lack of skills in operation of equipment
• Limited knowledge and less advanced techniques in the control of bentonite slurry
• Limited knowledge and experience in construction method and related negative impact
• Quality of concrete used for tremie concreting
• Improper construction and quality control specifications and guidelines for local soil
• Improper design for constructability
• Limited availability of performance monitoring equipment (e.g. efficient and reliable inclinometer and
other instruments)
DEVELOPMENT IN DIAPHRAGM WALL CONSTRUCTION
Over the past three decades, along with the development of diaphragm wall construction technology in other
parts of the world, equipment, construction technique and design methods as well as better understanding of
construction impact on the performance of embedded retaining wall have significantly improved in Thailand.
Table 3 summarizes the areas of improvement in diaphragm wall construction and main factors contributed to
these developments.
Table 3, Summary of development in diaphragm wall construction in Bangkok

Area of development/ Main factor contributed to development
improvement
Speed of construction Better equipment, operating skills as well as improved-knowledge in
construction method, management and local soil condition
Size and depth of Better equipment, operating skills as well as improved-knowledge in
diaphragm wall construction method, management and local soil condition
Slurry management Experience from past projects and research studies
Construction impact on Experience from past projects and research studies
quality and performance
Quality control in Experience from past projects and research studies
construction process
Quality control test method Advance equipment, experience from past projects and research studies
and interpretation
In some projects where headroom is limited, special diaphragm wall grabs and equipment were
employed. Fig. 6 shows the specially-designed diaphragm wall equipment used in the project with limited low
headroom.
Fig. 6, Specially designed baby-grab with short-boom for limited low-headroom area
377
Diaphragm wall is an ideal solution for deep circular shafts. Though alternative method such as
sinking caisson may offer cheaper direct cost, significant time saving and maximum safety can be achieved by
using diaphragm wall in construction of deep shafts. View of diaphragm wall support deep shaft is presented
in Fig. 7.
Fig.7. Deep circular shaft constructed by diaphragm wall
Diaphragm walls were only solution for deep excavations required for subway stations of the first
Bangkok MRT. (Thasnanipan et. al., 2000) presented the construction of diaphragm wall for one of the largest
subways stations of the MRT. Construction parameters and technical details of the diaphragm wall were
reported by the authors.
DEVELOPMENT IN DEEP EXCAVATION SUPPORTED BY DIAPHRAGM WALLS
In Bangkok, most of the diaphragm wall support deep excavations are carried out by multi-propping or of
braced-excavation. Depending on the excavation depths and sequence, bracing span-lengths commonly ranges
from 4m to 8m. Improvements in diaphragm wall support deep excavation works are obvious and significant
in past decade. Developments in the following areas are notable.
• Application of top-down construction method in diaphragm wall support deep excavation works
• More complex excavation works to deeper depths
• More understanding on excavation induced ground movement and risk of damages to adjacent
structures
• Application of ground improvement method to minimize the ground movement
• Application of observational method integrated with value engineering option and comprehensive
instrumentation program
• More understanding on water-retainable capacity of diaphragm wall
Application of top-down construction method in urban area of Bangkok was presented by

(Thasnanipan et. al., 2006). Along with technical aspects of the top-down application for 19.10m deep
excavation works, the authors highlighted the cost saving gain from shorter construction period and minimum
usage of material for temporary bracing works. View of excavation process with permanent slabs having
braced the excavation can be seen in Fig. 8.
378
Fig.8. Application of top-down method for deep excavation work in Bangkok (Thasnanipan et. al., 2006)
All 18 deep underground stations of the first Bangkok MRT Subway were constructed between 1997
and 2000 by top-down method using diaphragm walls as permanent structural support. Fig. 9 shows the view
of subway station construction with application of top-down method. Skeleton slabs provided large openings
for which offered logistically significant advantage in deep excavation works.
Fig.9. View of subway station construction with application of top-down method
Complex and challenging deep excavation works at Silom MRT Station was reported by (Hall et. al.
2001). The authors demonstrated the complicated construction sequence involved in the deepest excavation
works in Bangkok subsoil supported by NS-Box Diaphragm Wall. The process involved in underpinning of
existing flyover was also described by the authors.
Ground movement prediction and building risk damage assessment for the deep excavation
works can be found in the works of (Aye et. al., 2006). With demonstration of the procedure used in staged
assessment on risk of damage by excavation induced ground movement, the authors proposed a simple method
9
379
to predict both vertical and horizontal surface and subsurface ground movement based on the deflection
profile of diaphragm wall. Fig. 10 shows the proposed method for prediction of subsurface ground movement.
X
W DDoO= 2.5 H
= HW+ W E 4VO
SWO 
Total deflected Surface settlement trough DO
shape volume, V O volume, V To 2
 x 
Sio  SWO *  
Sio
D-wall S WO  DO 
Dy D
HE
DY  Y * O
HW
HW
Sub-surface settlement VTO

trough volume, VTy VTY  VY *
Siy
VO
Settlement
Swy
influence zone
Y
deflected volumn V y at depth (H w- Y)
Fig.10. Demonstration of subsurface settlement prediction from diaphragm wall deflection values proposed by
(Zaw Zaw Aye et. al., 2006)
Fig.11. Typical section of diaphragm wall support deep excavation adjacent to MRT tunnel showing
protection zone, (Thasnanipan et. al., 2006)
(Thasnanipan et. al. 2006) reported an unprecedented case on use of soil-cement columns as ground
improvement to minimize ground movement induced by diaphragm wall support deep excavation adjacent to
existing Bangkok MRT tunnels. The authors presented the technical requirement of the deep excavation work
in the protection zone of MRT tunnels along with actual monitoring results of similar projects in other parts of
the world. This integration of ground improvement application to diaphragm wall deep excavation works is
expected to be more popular in the future as need of basement facilities are increasing along the existing MRT
tunnels. The performance of diaphragm wall and impact on tunnels meeting all technical criteria requireed
imposed by the Metropolitan Rapid Transit Authority of Thailand (MRTA) was subsequently reported by
(Teparaksa et. al., 2006)
10
380

Geotechnical
Risk Management
Identify Risk Design & Plan
Value Engineering
Analysed Degree Construction Control

of Risk
Monitoring & Review
Trigger Criteria
Exceeds No
Minimize
& Control Implement
Yes Planned Action
Fig.12. The elements of observational method and risk management

with integration of value engineering in deep excavations,( Aye et. al., 2006)
Application of observational method integrated with value engineering and risk management for
deep excavation works was demonstrated by (Aye et. al., 2006). The authors concluded that well organized
plan with systematic approach of observational method backup by extensive instrumentation and close
cooperation among the involved parties were the key factors contributed to successful completion of the
reported projects.
(Thasnanipan et. al., 2008) presented the comprehensive study of water-resisting deep basement in
Bangkok soil. The authors highlighted the preventive measures for water leakage commonly occurred in
basement with reference to BS8102. The authors also recommended the designers to consider additional
measures to prevent water leakage and dampness for diaphragm wall support excavation as diaphragm wall
itself is not fully water-proof.
(Pongrujikorn, 2006) reported the development of underpass construction in urban area of Thailand
with emphasis on the projects in Bangkok. The author stated that diaphragm wall has been proved to be a
reliable option for most cases in underpass construction.
IMPROVEMENT IN DESIGN PARAMETERS
In the initial stage of introducing diaphragm wall in Thailand, the design concepts and parameters were mainly
based on the available literature from research carried out in other parts of the world. With the passage of time,
design method and selection of parameters for local subsoil were improved as a result of research works
carried out in 1990s. Differences between the behaviour of flexible pile walls and diaphragm walls were well
realized from these studies. Development in powerful computer facilities and commercially available FEM
programs are key factors contributed to the advancement in analysis of diaphragm wall support deep
excavations.
Detailed advance FEM analyses of diaphragm wall support deep excavations in Bangkok subsoil can
be found in the work of (Balasubramaniam et. al., 1992). The authors demonstrated the FEM simulation for
both top-down and conventional braced excavation with temporary support. The authors concluded that
behavior of supported excavation in soft clay is mainly controlled by the cantilever mode of deformations and
it is 80% of the maximum deflection of diaphragm wall.
11
381
With the peak of construction-boom in Thailand, particularly in Bangkok, large numbers of
diaphragm wall support deep excavations were systematically monitored by geotechnical instrumentation
throughout 1990’s which provided better understanding on behavior of this type of embedded retaining wall.
Researches focused on the design parameters and methods were published based on these data. With improved
construction and design methods, diaphragm wall came to be regarded as reliable excavation support retaining
structure in the construction industry of Thailand. With more confidence on soil parameters selection and
better understanding on behaviour of diaphragm wall, significant cost savings were achieved by more efficient
design. This can be seen by shorter embedded depths of diaphragm walls in recent projects than those in early
days as shown in Fig. 4 in comparison with Fig. 3.
(Teparaksa et. al., 1999), back-analysed the soil stiffness parameters of the soft and stiff Bangkok
clay for diaphragm wall support excavation based on three cases studies. The authors proposed the soil
stiffness values for soft and stiff clay in terms of Young modulus in relation to undrained shear strength which
provided useful reference for the design engineers involved in diaphragm wall and deep excavation works.
IMPROVEMENT IN APPLICATION OF INSTRUMENTATION
In early days, use of instrumentation in diaphragm wall support excavations was limited. Increasing needs of
deep excavation works adjacent to sensitive buildings and structures called for more comprehensive
instrumentation and monitoring program.
(Thasnanipan et. al., 1998) reported the use of inclinometer, tiltmeter, vertical beam sensor, vibrating
wire strain gauge, settlement plate and earth pressure gauge in monitoring of diaphragm wall support deep
excavations and adjacent sensitive buildings at the bank of Chao Phraya River. The important role of
instrumentation for deep excavation in presence of unbalance loading condition was emphasized by the
authors.
Ignorance of instrumentation data resulted in catastrophe collapse of deep excavation for Inlet
Pumping Station was discussed by (Teparaksa et. al., 1999). The authors presented the findings of
investigation into the causes of failure showing excessive movement recorded by inclinometer prior to the
collapse.
Extensive application of geotechnical instrumentation was more evident in construction of 18 deep
subway stations for the first Bangkok MRT between 1997 and 2001. (Aye et. al., 2006) summarized the
instrumentation program used in monitoring of ground movement induced by deep station excavations. The
authors discussed the systematic approach of damage risk assessment of adjacent existing buildings and
structures with application of extensive instrumentation.
(Thasnanipan et. al., 2004) reported the effective application of inclinometer monitoring data in
construction of underground car park in the historical area of Bangkok. The authors described that thin
diaphragm wall with buttress support (Fig. 13) applied in this project was the first of its kind in Bangkok as
described by the authors.
Fig.13. View of thin diaphragm wall supported by buttress (Thasnanipan et. al., 2004)
12
382
The authors presented a good data set of diaphragm wall movement profile showing time-dependent
progressive deflection pattern due to softening and deformation of soft clay caused by delay in installation of
raker support as shown in Fig. 14. The authors demonstrated by comparing actual monitoring results with pre-
establish trigger levels contributed to the understanding of the condition of the excavation as work progressed.
This approach served as the basis to judge whether contingency actions are to be implemented. To ensure the
safety of the excavation, the contingency measures exercised upon diaphragm wall deflection reaching preset
trigger levels were reported by the authors.
The actual monitoring data of different type of instrumentation presented in above literatures served
as valuable information and technical references for all parties involved in diaphragm wall and deep
excavations industry.

0 20 40 60 80 100 120 140

0
6
Depth, m.
10
0 days
12
30 days
14 52 day
56 days
16 74 days
Prediction
18
20
Fig.14. Time dependent wall deflection of diaphragm wall – diaphragm wall was supported only by soil-berm
for 52 days before completion of raker installation (Thasnanipan et. al., 2004)
OBSERVATION OF GROUNDWATER CONDITION IN RECENT PROJECTS
Recently, deep excavations deeper than 18.0m in depth at two projects for basement constructions being
5.25km apart in Bangkok subsoil encountered presence of groundwater after reaching about 17.5m. Subsoil
information obtained by soil boring in these projects indicated a series of stiff silty and sandy clay between
15m and 24m, overlying a dense sand layer. Construction progress was significantly affected by presence of
groundwater in silty clay. Fully saturated silty clay turned liquid when disturbed by excavation activities,
causing difficulties in casting a concrete blinding layer for foundations construction. Dewatering and ground
improvement to the subgrade layer below foundation mat and footings needed to be carried out. Bearing
capacity of disturbed silty clay with high water content is very low so that cobble size concrete blocks and
crushed rocks to be compacted in to the soil for improving the subgrade. Then a lean concrete blinding layer
was cast on the subgrade part by part while dewatering.
A general piezometric profile in Bangkok shown in Figure 1 indicates a drawdown at -20m. Additional
peizometric readings from the piezometers installed at depths in the range of 28.0m to 33.0m and at 40m from
other sites in Bangkok indicate that piezometrice levels are around -15.0m and -23.0m respectively.
Piezometers installed at depths 28.0m to 33.0m shows piezometric levels increase to -12.0m within six
months. However, the groundwater encountered in the above mentioned sites was considered to be perched
water in semi permeable soils layer being vertically cut off by diaphragm walls, with an assumption that the
groundwater level was reduced gradually by pumping out.
13
383
Fig.15. Groundwater encountered in basement excavations deeper than -17.50m.
CONCLUSIONS
According to the authors’ experience as deep-foundation contractors and observation as researchers,

development in both construction and design aspects of diaphragm wall and deep excavations in Bangkok in
past decades were significant. With recognition of technical and economical advantages of using this type of
embedded retaining walls by local practitioners, it is expected that they will be more popular in the future
construction industry of Thailand. However, in the authors’ opinion, there is much work to be done with
particular focus on constructability issues, concrete technology for diaphragm walls, reliable but cost-effective
quality control testing and value-engineering. Starting from the planning stage, site investigation, design,
construction and inspection should be integrated so that designers, contractors and construction inspectors can
participate as a team with a common goal. Appropriate and practical specifications should be established
jointly by these parties for local soil conditions and construction methods. As a commercial and political
center of the country with a rapidly growing population, Bangkok will continue to witness more underground
construction for its infrastructure development in which diaphragm wall support excavation will clearly play a
great role.
REFERENCES
Balasubramaniam, A.S, Longanathan, N., Fernando, G.S., Indraranta, B., Phien-wej, N., Bergado, D.T.,
Hongjo, Y. (1992) Advance Geotechnical Analysis, Finite Element Analysis Coupled with Critical State
Theories Using CRISP Computer Program, Asian Institute of Technology, Thailand
Hall, J.T. , Munson, S.R., Sakai, K. (2006) Silom Underground Station-Bangkok Design for Unusual Planning
Challenges, Proceedings of the Symposium on Underground Construction of Bangkok MRT
Chaloemratchamongkhon Line, EIT, MRTA, Bangkok, Thailand, pp. 58-68
Peck, R.B., Hanson, W. E. and Thornburn, T.H. (1977). Foundation Engineering, John Wiley & Sons, New
York
Phienwej, N., Photayanuvat, C., Suwansawat, S. and Mcconochie, D., (2006) Geotechnical Aspects in Design
and Construction of Chaloemratchamongkhon Underground MRT, Proceedings of the International
Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for Protection of
Environment, 2-4 February 2006, Bangkok, Thailand, pp. 171-180
14
384
Pongrujikorn, T., (2006) Development and Use of Underground Strucutre as Underpass in Urban Area in
Thailand, Proceedings of the International Symposium on Underground Excavation and Tunneling, Urban
Tunnel Construction for Protection of Environment, 2-4 February 2006, Bangkok, Thailand, pp. 441-448
Teparaksa, W., Thasnanipan, N., Maung, A. W. and Tanseng, P. (1999), Lessons from the collapse during
construction of an inlet pumping station: Geotechnical instrumentation aspect, 5th International Symposium
on Field Measurements in Geomechanics, Singapore, p.p. 247-252.
Teparaksa, W., Thasananipan, N., and Tanseng, P. (1999), Analysis of Lateral Wall Movement for Deep
Braced Excavation of Bangkok Subsoils, Civil and Environmental Engineering Conference, Bangkok,
Thailand, p.p. II-67 to II-76.
Teparaksa W., Boonsong C., Submaneewong C., Wisuttipat N., (2006) MRT’s Tunnel Deformation due to
Deep Basement Excavation in MRT’s Protection Zone, Proceedings of the 11th National Convention on Civil
Engineering, 20-22 April 2006, Phuket, Thailand (in Thai), pp. GTE-095
Thasnanipan, N., Maung, A.W., Tanseng, P., Wei, S.H., (1998), Performance of a Braced Excavation in
Bangkok Clay, Diaphragm Wall Subject to Unbalanced Loading Conditions, Thirteenth Southeast Asian
Geotechnical Conference, Taipei, Taiwan, ROC, p.p.655-660.
Thasnanipan, N., Maung, A.W. and Ganeshan, B., (2000), Diaphragm Wall and Barrette Construction for
Thiam Ruam Mit Station Box, MRT Chaloem Ratchamongkhon Line, Bangkok, An International Conference
on Geotechnical & Geological Engineering, Melbourne, Australia, 19-24 November 2000, pp. 157
Tanseng P., Aye, Z. Z. and Submaneewong, C. (2001), Monitoring of Diaphragm Wall Displacement and
Associated Ground Movement, Braced Excavation Adjacent to Historical Building at the Bank of Chao
Phraya River, Proceedings of the Seventh National Convention on Civil Engineering, Bangkok, Thailand, pp.
GTE-53-60
Thasnanipan, N. , Aye, Z. Z. and Submaneewong, C. (2004), Construction of Diaphragm Wall Support

Underground Car Park in Historical Area of Bangkok Proceedings of Fifth International Conference on Case
Histories in Geotechnical Engineering, April 13-17 2004, New York, U.S.A. pp. 5.66
Thasnanipan, N. , Aye, Z. Z., Submaneewong, C. and Boonyarak, T. (2006) Application of Top-down

Construction Method for Deep Excavations in Urban Area of Bangkok, Proceedings of the International
Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for Protection of
Environment, 2-4 February 2006, Bangkok, Thailand, pp. 383-396
Thasnanipan, N., Maung, A.W., Aye, Z. Z., Submaneewong, C. and Boonyarak, T. (2006) Construction of
diaphragm wall for basement excavation adjacent to tunnels in Bangkok subsoil, Proceedings of the
International Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction for
Protection of Environment, 2-4 February 2006, Bangkok, Thailand, pp. 371-382
Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006]. “Ground Movement Prediction and Building
Damage Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil, Proceedings of
the International Symposium on Underground Excavation and Tunneling, Urban Tunnel Construction and
Protection of Environment, EIT, Bangkok, Thailand, , pp. 281-297
15
385

Conference 2012
Underpass Construction in Congestive Area of

Bangkok
Zaw Zaw Aye, Phongpat Kitpayuck

Sedtha Mahasantipiya, Nutthapon Thasnanipan
Thayanan Boonyarak
387
Underpass Construction in Congestive Area of Bangkok
Zaw Zaw Aye1, Thayanan Boonyarak2, Phongpat Kitpayuck1
Sedtha Mahasantipiya1, Nutthapon Thasnanipan1
1
2
Diaphragm wall has been used as permanent retaining wall for a vehicle underpass at major
intersection in urban area of Bangkok besides deep basement structures of high-rise buildings,
subway stations, deep shafts and other forms of underground structures,. This paper presents the
key technical features of recently completed vehicle underpass along one of the busiest roads in
Bangkok. The design aspects of earth-retaining structure integrated with top-down construction
method are reported. Predicted and measured deflections of diaphragm wall at different stages of
construction are presented. Practical construction method adopted under the existing flyover is
highlighted. Difficulties encountered during the construction are briefly described. This project
marked the first application of post-tension roof slabs along with top-down construction
technique for vehicle underpass construction by cut-and-cover method in Thailand.
1. Introduction
To combat appalling traffic congestion, Bangkok Metropolitan Authority has been implementing
various measures in both urban and sub-urban areas. Flyovers are frequently used to improve the
traffic flow along the road at many major intersections. Though underpass is not commonly
used, it is the only solution in some critical traffic jam areas. The first underpass in the form of
depressed-road for rail-road junction was believed to be constructed about 35 years ago at
Bangsue, northern outskirt of Bangkok. Several years later, vehicle underpass was constructed
using cast-in-place diaphragm wall by Bangkok Metropolitan Authority (BMA) at Din Daeng
intersection in 1994. Since then more than 20 underpasses have been constructed by BMA and
other agencies in Bangkok and other major cities in the country.
This paper presents the design and construction of recently completed vehicle underpass along
one of the busiest roads in Bangkok. The presenting project in this paper demonstrated the
effective use of underground space in congestive urban areas of Bangkok and exemplified good
solutions for environmental preservation.
2. Project Requirement and Major Constraints
The project works involved construction of a vehicle-underpass and associated utility diversion
and road expansion at one of the major intersections in the eastern part of Bangkok. Figure 1
shows the alignment and location of the underpass overlaid on aerial photo of the site (modified
from GoogleEarth). Four-lane underpass was to construct in the form of 758 m long cut-and-
cover tunnel along six-lanes existing Srinakarin Road. As can be seen in the figure, at the
intersection, tunnel section of underpass was to construct under existing flyover of Udomsuk
Road.
389
As the main six-lanes Srinakarin Road is one of the major roads linking eastern parts and inner
Bangkok, traffic density along this road is high. Presence of two major shopping malls in the
close proximity to the intersection is also another factor contributed to congestive traffic
condition in the area.
Location of the project in the heavy traffic area and need of construction under the existing
flyover with limited headroom therefore posed major constraints, which called for the need of
careful consideration in establishing the design principles and sequence of construction. As the
construction is to be carried out along the major public road, ground movement induced by cut-
and-cover tunnel excavation must be minimized.
Udomsuk Rd Srinakarin Rd
Flyover 4-Lanes Vehicle Underpass
Total length of underpass 758 m
Figure 1 Alignment and location of vehicle underpass (aerial photograph modified from
GoogleEarth)
3. Sub-soil Condition
Similar to other localities in Bangkok a typical subsoil profile at the site is characterized by the
alternating layers of clay and sand deposits as shown in Figure2 . Undrained shear strength (Su)
and SPT N-value obtained from 5 SI boreholes were plotted and design line was derived as
depicted in Figure 2. Typical subsoil profile at the site is characterized by thick Bangkok soft
clay layer at the top followed by thin layer of medium clay underlying by alternative of stiff clay
and dense sand layers. Undrained shear strength of soft clay is obtained by using unconfined
compressive strength test from Shelby tube sample. In stiff to hard clay layer, correlation of
undrained shear strength with SPT N-value is adopted where undrained shear strength is equals
to 6.85N (kPa) for high plasticity clay (Sambhandaraksa & Pitupakorn. 1985). Angle of internal
friction (φ’) of dense sand is estimated to be 32˚ by using SPT correlation (Peck et al., 1974).The
design soil parameters are tabulated in Table 1.

390

Undrained Shear Strength, Su (kN/m2)
1 0 50 100 150 200
0 0.00
Fill

5 5.00
Soft clay
10 10.00

15 15.00
Medium stiff clay
20 20.00
Very stiff
clay
25 25.00

30 Dense sand 30.00

35
Very stiff to
35.00
hardclay
40 40.00

45 Very dense sand 45.00

50 50.00
0 20 40 60 80


Su BH-1 Su BH-2
Su BH-3 Su BH-4
Su BH-5 Design Su
SPT BH-1 SPT BH-2
SPT BH-3 SPT BH-4
Figure 2 Soil profile, undrained shear strength and SPT N-value

Table 1 Design soil parameters
Soil type Depth γt Su φ’
(m) (kN/m3) (kPa) (degree)
Fill 0-2 18.5 25 0
Soft clay 1 2-10 16.5 17 0
Soft clay 2 10-14 17.5 22 0
Medium stiff clay 14-18 19.0 40 0
Stiff clay 18-20 19.0 80 0
Very stiff clay 20-25 19.5 120 0
Hard clay 25-27 20.0 180 0
Dense sand 27-32 20.0 0 32
Hard clay 32-40 20.0 180 0
Very dense sand 40-45 20.0 0 32

391
4. Review of Tender Stage Design and Modification
Review of the practical construction method was undertaken by the contractor’s in-house design
team. Rigorous attention to detail of the design concept and constructability was made in the pre-
construction discussions among the client, consultants, design engineers and construction team.
Original tender called for the conventional construction method – to complete the excavation to
final excavation depth and construct the roof-slab in the tunnel section of the underpass by using
precast pre-stressed concrete panels. Installation of 2.5m x 18.8m heavy precast panels
underneath the existing flyover was considered difficult and high risk involved. After reviewing
different options with consideration of the constraints imposed on site and constructability, top-
down construction with post-tension slabs as roof-slab of the tunnel section was selected.
4.1 Numerical analysis for diaphragm wall design

Figure 3 illustrates an example of finite element mesh of deep approach with sump pit. In other
zone, basic geometry of the mesh follows that of excavation depth, wall depth and construction
sequence of each zone. Numerical analyses are carried out by using PLAXIS V7.2 (Brinkgreve
and Vermeer, 1998) finite element program. Plane strain total stress analyses using elasto-plastic
with Mohr-Coulomb failure criterion constitutive model are adopted. Undrained shear strength
of clay (Su) and internal friction angle of sand (φ’) are selected according to subsoil properties as
previously discussed. Young’s modulus ratio (Eu/Su) of each clay ranges from 250 to 750 which
is estimated based on expected shear strain of diaphragm wall. Boundary condition consists of
roller support applied at both vertical sides and pin support applied at the bottom of the mesh.
Numerical procedure follows the construction sequence in the field conditions. The primary
objective of numerical analysis is to verify structural capacity and overall stability during
construction. It should be noted that time dependent behavior of clay is not simulated.
10
Depth (m)
25
45
Figure 3 Finite element mesh of deep approach with sump pit
392
5. Construction of Diaphragm Wall, Barrettes and Bored Piles
Figure 4 shows the diaphragm walls, bored piles and barrettes layout plan. Diaphragm wall was
designed as earth-retaining wall for the entire section of the underpass. Diaphragm wall having
800mm thick founded at 25m below ground level (B.G.L) was constructed in three phases. Both
Phase 1 (north section) and Phase 2 (south section) were constructed in two stages (different
working stages for east and west diaphragm walls) to meet the traffic diversion requirement.
North section of the underpass diaphragm wall was constructed in Phase 1 in the initial stage of
the project. In Phase 2, south section of diaphragm wall was constructed prior to construction of
cross-wall for a deep sump-pit - diaphragm wall with 6m deep cut-off-level perpendicular to
longitudinal alignment of the main underpass diaphragm wall.
Diaphragm wall
Barrette 0.8 x 2.7 m x 29 m deep
Centerline of 0.8m x 25m
constructed underneath flyover
Udomsuk Rd Flyover
Bored pile φ 1.0 m x 29 m deep
Phase 2 Phase 1
Phase 3
(under flyover)
Note: only diaphragm wall in tunnel zone is shown for clarity
Figure 4 Diaphragm wall, bored pile and barrette layout plan of vehicle underpass
As the tunnel section of the underpass at the intersection was to be constructed under limited
headroom due to the presence of existing flyover, careful planning was needed for Phase 3
diaphragm wall construction. Main factors were the limited time available for traffic diversion at
the intersection and the need of special equipment to practically construct the diaphragm wall
under the flyover where headroom was only 6m. Specially-designed mechanical grab was used
to construct the diaphragm wall in this low-headroom section as shown in Figure 5. Another
major challenge involved in construction of diaphragm wall under the existing flyover was the
obstruction of existing water supply pipe at 8m below ground level along the diaphragm wall
alignment. Removing the pipe buried 8m below ground level by open-cut or sheet-pile supported
excavation option was considered time consuming, unpractical as well as risky, taking the
consequences involved into account (unpractical and risky to have 8m open-cut excavation in
soft clay, unpractical to install 16m deep sheet pile low headroom. After reviewing different
options to remove the obstructed water supply pipe, decision was made to remove the pipe by
sending a man inside the pipe - to cut the obstructed part in smaller pieces and close the pipe to
stop the flow of bentonite into the pipe during diaphragm wall construction. Risk assessment and
safety measures with contingency plan were made before actual pipe removal was done. The
obstructed section of the buried pipe was successfully removed.
393
Figure 5 Special-design diaphragm wall grab working under existing flyover
Figure 6 demonstrates the 3D perspective of stanchions embedded in bored pile and barrettes.
Total 43 bored piles of 1000mm diameter founded at depth 29m below ground level were
designed to support roof-slab of tunnel section. Bored piles were constructed by wet-processes
method using polymer-based slurry. Steel stanchions were installed in bored piles to facilitate
top-down construction. As it was not practical to utilize bored pile drilling machine underneath
the flyover, within the low-headroom section of the underpass, three barrettes were installed to
replace bored piles.
Figure 6 Three-Dimensional perspective of stanchions embedded in bored pile and barrettes
394
6. Application of top-down method integrated with post-tension slab
Figure 7 shows the plan of cut-and-cover tunnel constructed by using top-down method
integrated with post-tension slab. Total length of tunnel zone is approximately 180 m along the
longitudinal profile of underpass or about 25% of overall underpass length.
As discussed in the previous section, top-down construction requires modifications from the
original design. Instead of using temporary prop to support the diaphragm wall during
excavation of tunnel section, roof-slab was used as temporary support. Four openings were
provided at an evenly spacing of about 40 m for access of soil excavation underneath the roof-
slab.
Post-tension
roof slab
y40 m y40 m
Slab opening
during construction
Figure 7 Plan view of tunnel zone showing post-tension roof slab constructed prior to
excavation and openings for excavation works
Construction sequences of tunnel zone are shown in Figure 8. Major construction activities in
each stage are given as follows.
Stage 1: Construct diaphragm wall and bored pile with embedded temporary stanchion
Stage 2: Construct upper part of median wall (spine-beam) on the top of stanchion
- Construct roof slab and apply post-tension force
- Excavate to formation level
Stage 3: Construct base slab
- Construct median wall between temporary stanchions
- Remove temporary stanchion and construct the remaining median wall
395
Stage 1 Stage 2 Stage 3
Figure 8 Construction sequences in tunnel zone
6.1 Comparison of roof slab between original tender design and alternative design
Comparison between tender design and alternative design of roof slab are shown in Figure 9.
Figure 9a shows precast prestress slab in the tender design. The width of each precast slab is 2.5
m. A gap of 0.2 m between each precast slab is filled with concrete after placing of precast slab.
Figure 9b shows post-tension slab used as alternative in this project. Each construction joint is
located at a distance of about 12.5 m. By using alternative design, numbers of joints were
significantly reduced. A major benefit of reduction of number of joint is possibility of water
leakage from roof slab through joint is dramatically reduced. In addition, using the roof slab as
permanent prop provide significant larger axial stiffness compared with when temporary steel
prop is used. The evidence of increasing axial stiffness on performance of diaphragm wall is
discussed later in the following section.
Tendon
Grouting hose
CONSTRUCTION JOINT CONSTRUCTION JOINT CONSTRUCTION JOINT
(a) (b) (c)
Typical tendon
(d)
Figure 9 (a) Precast prestress slab; original design (b) post-tension slab; actual application (c)
typical tendon profile (d) tendon and grouting hose
396
Before starting excavation underneath the roof slab, tension is applied to tendon embedded
inside the roof slab (Fig. 9c). Subsequently, grout is injected via grouting hose. Figure 9d shows
a typical tendon profile which is similar in both original and alternative design. A profile of
tendon was designed to resist loading condition of a beam with end and middle supports. In
order to apply roof slab as a permanent prop for excavation, rigorous structural capacity
checking for combined bending moment and axial force in the slab is required. Figure 10a and
10b shows a typical construction of post-tension roof slab.
(a) (b)
Figure 10 (a) Tendon installation for roof-slab (b) grouting after post-tensioning the slab
6.2 Excavation underneath the roof slab

Figure 11a and 11b show slab opening and temporary prop in tunnel zone. Spacing between each
opening was determined by optimizing efficiency of excavation and providing space to facilitate
the construction. Excavation was carried by using a small excavator inside the tunnel (Fig. 11a)
or using a long reach boom excavator on the roof slab (Fig. 11b). Presence of roof slab provides
additional space and increase efficiency of excavation works in comparison with the original
design where temporary working platform is to be used.
(a) (b)
Figure 11 Slab opening with temporary support (a) using a small excavator inside the tunnel; (b)
using a long reach boom excavator on the roof slab
397
6.3 Temporary support of the roof slab
Figure 12a shows temporary support of roof slab. Load from the roof slab is transferred to the
upper part of the median wall which works as a spine-beam. Additional reinforcement in the
upper part of median wall was required during construction stage. The upper part of the wall was
supported by stanchions made of steel H 350x350-137 kg/m embedded in φ1.0 m diameter bored
pile. The remaining part of median wall was constructed after completion of base slab (Fig. 12b).
After completion of median wall, temporary stanchions were removed since vertical load from
roof slab was designed to be transferred through the median wall.
Stanchion
Upper part of
median wall
Stanchion
Median wall
(a) (b)
Figure 12 (a) Temporary stanchions supporting post-tension roof slab (b) median wall and
stanchion
7. Performance of diaphragm wall
Figure 13 shows location of instrumentation in this project. Instrumentation consists of

inclinometer, surface settlement point, strain gauge and piezometer. Instrumentation in
diaphragm is installed in three zones which are tunnel, deep approach and deep approach with
sump pit. Each zone of monitoring is selected based on its construction sequences. Table 2
summarizes construction sequence in each zone according to the location of instrumentation
which is T1, T2, D1, D2 and DS1. In this paper, only some of measured results are reported for
conciseness of paper.
T1 D1
SO-73 SO-74 SO-75 SO-76 SO-77 SO-78 SO-79 SO-80 SO-81 SO-82 SO-83 SO-84 SO-85 SO-86 SO-87 SO-88 SO-89 SO-90 SO-91 SO-92 SO-93
CLE-69 SO-70 SO-71 SO-72 SO-94 SO-95 SO-96 SO-97 SO-98
NO-43 NO-64 NO-65 NO-66 NO-67 CLE-68
NO-44 NO-45 NO-46 NO-47 NO-48 NO-49 NO-50 NO-51 NO-52 NO-53 NO-54 NO-55 NO-56 NO-57 NO-58 NO-59 NO-60 NO-61 NO-62 NO-63
Sonic Logging Test G I

G I Sonic Logging Test
Sonic Logging Test
CWN-1
3.000
3.000
CWS-6
CWN-2
4.500
CWS-7
4.500
CWN-3
3.350
3.350
CWS-8
CWN-4
3.350
CWS-9
CWN-5 Sonic Logging Test

3.000
Sonic Logging Test

3.000
CWS-10
Sonic Logging Test G I
G I G I
S1 P
S1 P S1 P SI-76 SI-77 SI-78 SI-79 SI-80 SI-81 SI-82 SI-83 SI-84 SI-85 SI-86 SI-87 SI-88 SI-89
SI-71 SI-72 SI-73 SI-74 SI-75 SI-90 SI-91 SI-92 SI-93 S2 SI-94 SI-95 SI-96 SI-97
NI-45 NI-67 CLW-68 CLW-69 SI-70 SI-98
NI-43 NI-44 NI-63 NI-64 NI-65 NI-66
S2 NI-46 NI-47 NI-48 NI-49 NI-50 NI-51 NI-52 NI-53 NI-54 NI-55 NI-56 NI-57 NI-58 NI-59 NI-60 NI-61 S2 NI-62
S3
S3 S3
S4
S4 S4
S5
S5 S5
S6
S6 S6
DS1 T2 D2
Figure 13 location of instrumentation
398
Table 2 Summary of construction sequence in each zone
Tunnel (T1 & T2) Deep approach (D1 & D2) Deep approach with sump pit
(DS1)
Construct roof slab Excavate to -1.5 m Excavate to -1.5 m
Excavate to final depth Install temporary bracing at Install temporary bracing at
(-6 m) -1.0 m -1.0 m
Construct base slab Excavate to final depth (-6 m) Excavate to -5.5 m
Construct base slab Install temporary bracing at -5 m
Remove temporary bracing Excavate to final depth (-9.5 m)
Construct sump pit slab
Construct base slab
Remove temporary bracing
7.2 Lateral movement of diaphragm wall

Lateral movement of diaphragm wall in three zones are shown in Figure 14. Measured results
from three major construction stages which are excavation to final depth, base slab construction
and four months after completion of base slab are shown. Computed results at the final stage of
construction in each zone are compared with the measured results. Maximum lateral
displacement of diaphragm wall is 24, 41 and 30 mm in tunnel, deep approach and deep
approach with sump pit zone, respectively.
By comparing diaphragm wall performance in two different construction sequences with the
same diaphragm wall geometry and excavation depth, it is apparent that lateral movement in
bottom-up construction (Fig. 14b) is about 2 times larger than that in top-down construction
(Fig. 14a). Not only magnitude of wall movement is larger in bottom-up excavation compared
with that of top-down, but also the rate of movement. Rate of wall movement after final
excavation depth reached in deep approach is about one order of magnitude larger than that in
tunnel. The major reason of this observation is considered to be contributed by higher stiffness
of prop at the top of diaphragm wall. Prop stiffness in tunnel section which is the post-tension
roof slab is about 16 times higher than that of temporary steel prop used in deep approach
section. A major difference in wall deflection was observed at the top of the wall. In tunnel
section, only a small deflection was observed at the top of the diaphragm wall and maximum
wall deflection was observed at 9 m below ground surface. In deep approach section, movement
at the top of the wall was about 24 mm due to shortening of the bracing. Subsequently, earth
pressure is transferred to the formation level (6 m below ground surface) causing maximum
movement at depth 7 m.
In deep approach with sump pit (Fig. 14c), although it was constructed by using bottom-up
method and excavation depth (9.5 m) is deeper than that of deep approach - maximum wall
movement is about 73% of that in deep approach. Two major reasons are identified to explain
this observation. First, there are two levels of temporary steel prop in deep approach with sump
pit compared resulting in 25% smaller average prop spacing (H/number of prop) compared with
that in deep approach. Second, excavation length of sump pit is only 18 m in longitudinal
direction of the wall which is almost the same as excavation width (17 m). Thus geometry
399
effects may contribute to reducing of lateral wall movement compared with plane strain
condition in deep approach.
Computed result overestimates wall movement in tunnel zone while it underestimates that in
deep approach zone. The differences are expected to be due to two major causes. First, constant
soil stiffness is adopted. The reduction of soil stiffness with increasing shear strain is not
simulated. Second, time dependent soil properties are not modeled. It is apparent that time
dependent behaviour in deep approach is the major component of wall movement. Nevertheless,
computed results reasonably agree with measured result in terms of overall trend.
Displacement (mm) Displacement (mm) Displacement (mm)
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
0
5
0.34
mm/day
0.14
0.03
mm/day
mm/day
10
Depth (m)
15
Measured [Before Measured [Before

Measured [Before
base slab] base slab]
base slab]
Measured [Base slab Measured [Base slab Measured [Base slab
20 construction] construction]
construction]
Measured [4 months Measured [4 months Measured [4 months
after base slab] after base slab] after base slab]
Computed [Final Computed [Final Computed [Final
25 stage] stage] stage]
(a) (b) (c)
Figure 14 Lateral wall movement in (a) tunnel zone ; top-down method (T1); (b) deep approach
bottom-up method with 1-layer bracing (D2); (c) deep approach with sump pit bottom-up
method with 2-layers bracing (DS1)
7.3 Comparison of measured wall movement with other case histories

Figure 15 shows maximum measured wall movement by using two different construction
methods and sequences reported previously in other case histories (Phienwej et al., 1995).
Previous study conducted by Phienwej et al., 1995 suggests that maximum lateral wall
movement using bottom-up construction can be up to 0.4% of excavation depth (H). On the
other hand, wall movement in top-down construction is estimated to be only 0.2%H.
In this study, movement in bottom-up construction in deep approach and top-down construction
in tunnel exceeds 0.4%H and 0.2%H line, respectively. Wall movement using top-down
construction is closer to 0.27%H line proposed by Wang et al., 2010 based on measured field
results in Shanghai soft clay. Formation level of excavation in this study is in soft clay and only
one prop is provided. Thus, passive resistance is mainly from soft clay causing larger lateral
movement compared with other case histories. Another possible reason of large lateral wall
400
movement is mainly due to the construction time or time-dependent movement as construction
of underpass naturally takes more time due to traffic diversion and other factors.
60
Max. lateral movement, δmax (mm)
50
40
30
20
Bottom-up [Phienwej et al., 1995]
Top-down [Phienwej et al., 1995]
10
Soft clay Bottom-up [This study]
depth Top-down [This study]
0
0 5 10 15 20 25
Max. excavation depth, H (m)
Figure 15 Measured lateral wall movement compared with other case studies in Bangkok
[Modified from Phienwej et al., 1995]
7.4 Bending moment of diaphragm wall

Figure 16 shows comparison of measured and computed bending moment of diaphragm wall.
Measured bending moment is from vibrating wire strain gauges (VWSG) installed at
reinforcement of diaphragm wall at depth of 6 and 17.5 m where bending moment is expected to
be large. Measured strain is larger than 150 με thus wall flexural stiffness reduction factor of
80% is adopted. Deduced bending moment is from two times differentiation of measured lateral
wall movement (Fig. 16) using the same stiffness reduction factor as measured from strain
gauge.
In tunnel zone (Fig. 16a), measured bending moment at depth 6 m is slightly smaller than
deduced bending moment suggesting that deduced bending moment is reasonable. It should be
noticed that deduced bending moment exceeds computed bending moment from 8 to 13 m below
ground surface due to smaller stiffness in soft to medium clay in numerical analysis.
In deep approach zone (Fig. 16b), measured bending moment at 6 m from strain gauge is about
30% larger than that in tunnel zone. Deduced bending moment in deep approach also confirm
the measured result. Bending moment from depth 3 to 7 m significantly exceeds the computed
bending moment due to lacking of time dependent properties of clay in numerical analysis.
401
At depth 17.5 m, measured, deduced and computed bending appears to be in a good agreement.
In addition, deduced bending moments in both tunnel and deep approach zones are still within
allowable bending moment capacity of diaphragm wall.
Bending moment (kN.m/m) Bending moment (kN.m/m)

-500 0 500 1000 -500 0 500 1000
0
10
Depth (m)
15
Measured [VWSG] Measured [VWSG]

Deduced [INC] Deduced [INC]
20 Computed [FEM] Computed [FEM]
Allowable Allowable
25
(a) (b)
Figure 16 Bending moment of diaphragm wall in (a) tunnel zone (T1); (b) deep approach zone
(D2)
7.5 Excess pore water pressure

Measured excess pore water pressure normalized with initial pore water pressure is shown in
Figure 17. Pore water pressure is measured from standpipe piezometer installed at a distance of 1
m away from diaphragm wall at sump pit with tip of piezometer at 10 m. Initial pore water
pressure is about 70 kPa which is considered to be reasonable if hydrostatic pore water
distribution is assumed given ground water table is at 3 m below ground surface (Teparaksa,
1999). Wall displacement at depth of 10 m is shown to compare with response of piezometer.
During the first 120 day after initial reading, although there is constant increasing of wall
movement, there is no significant change of pore water pressure. This observation may be due to
unloading of horizontal stress causes negative excess pore water pressure and shearing in soft
clay causes positive excess pore water pressure. Consequently, two components counteract each
other causing only a slight change of excess pore water pressure. From day 120 to 150, lateral
402
wall displacement increase significantly when excavation reaches to final depth of 10 m.
Subsequently, there is a sharp increment of excess pore water pressure from day 160 to 200. This
is likely to be caused by shearing in compression mode dominates effect of horizontal stress
unloading. After base slab is constructed (at day 210), wall movement and excess pore water
pressure appear to be stable.
0.6 30
Excess pore water pressure
Wall movement
0.5 25
Lateral displacement (mm)

0.4 20
Δu/u0
0.3 15
0.2 construction 10
Base slab
Excavate
0.1 5
to 10 m
0.0 0
0 50 100 150 200 250
Construction time (day)
Figure 17 Normalized excess pore water pressure at 10 m deep outside of diaphragm wall at
sump pit (DS1)
8. Conclusion
Design and construction of vehicle underpass in the congestive area of Bangkok is presented.
Inclinometer measurement confirmed that top-down construction with permanent slab (stiffer
bracing) induced less deflection than that temporary steel bracing with lower stiffness. This
project reveals that a permanent diaphragm wall coupled with effective design modification of
associated structure members and construction method could offer a logistically and technically
attractive solution in construction of vehicle underpass in the congestive traffic area of Bangkok.
Careful consideration in post-tender review of technical options and selecting the most practical
construction method backup by sound design input and good performance of construction team
are important factors contributed to successful completion of the project.
403
References
Brinkgreve, R.B.J., and Vermeer P.A. 1998. PLAXIS v7.2 Manual, A.A. Balkema, Rotterdam,
Netherlands
Peck, R.B., Hanson, W.E., and Thornburn, T.H. 1974. Foundation Engineering. New York: John
Wiley&Sons.
Phienwej, N., Akawanlop, K and Balasubramaniam, A.S. 1995. Comparative evaluation of
ground movements associated with braced - excavation in Bangkok soft clay, 10th Asian
Regional Conference on Soil Mechanics and Foundation Engineering, Beijing, 341-344
Sambhandaraksa, S., and Pitupakorn, W. 1985. Prediction of prestressed concrete pile capacity
in stiff clay and clayey sand. Proc. of 8th Southeast Asian Geotechnical Conference,
Kuala Lumpur, Malaysia, (3), 58-63
Teparaksa, W. 1999. Principle and Application of Instrumentation for the First MRTA Subway
Project in Bangkok. 5th International Symposium on Field Measurements in
Geomechanics, Singapore, 411-416
Wang, J. H., Xu, Z. H. and Wang, W. D. 2010. Wall and ground movements due to deep
excavations in Shanghai soft soils. Journal of Geotechnical and Geoenvironmental
Engineering, 136 (7), 985-994
404

Engineering
Sonic integrity test on piles founded in Bangkok

subsoil – Signal characteristics and their
interpretations
Thasnanipan N. , Maung A. W. and Ganeshan B.
SEAFCO Co., Ltd.
Rolla, Missouri
405
1086
Proceedings: Fourth International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri,
March 9-12, 1988.
SONIC INTEGRITY TEST ON PILES FOUNDED IN BANGKOK SUBSOIL

SIGNAL CHARACTERISTICS AND THEIR INTERPRETATIONS
Narong Thasnanipan Aung Win Maung Ganeshan Baskaran Paper No. 11.02
SEAFCO Co., Ltd. SEAFCO Co., Ltd. SEAFCO Co., Ltd.
Bangkok 10320 Bangkok 10320 Bangkok 10320
Thailand Thailand Thailand
ABSTRACT
Sonic integrity testing is a common method used to test the integrity of piles in Bangkok. An acoustic technique equipped with
computer software, developed by TNO, PDI, IFCO, etc., is commonly used to test the piles. This paper presents the signal
characteristics and their interpretations from sonic integrity test performed on mainly bored piles varying in length, size, construction
method and founded soil strata. Toe reflection of small bored piles (diameter of 0.35m to 0.60m) constructed in Bangkok subsoil can
be observed for pile length of up to 20m to 24m . Beyond 35m below the ground , visibility of toe reflection is uncommon for large
bored piles (diameter of 0.80m and larger). Commonly predicted defects and frequency of occurrences are tabulated. Data sets have
been selected from the Data Bank of a foundations specialists’ organization to which the authors are associated with. Recently
completed and current projects in Bangkok have been selected so that the data set could be well represented by over 8000 piles. 285
piles (3.3%) out of 8,689 piles have been interpreted for poor concrete or inclusions, cracks and size reduction. Among these, cracks
contribute the major portion (2.2% or 194 piles). Cracks are mainly induced by either improper excavation adjacent to the piles or
trimming the pile head by ill fated ways. Small diameter piles are the most suffered by cracks. Prominent sectional variations have
been indicated by the signals at depth where temporary casing ends. Intermittent variations also are common at the interface of soft
clay and medium to stiff clay layers.
KEYWORDS
Integrity, Sonic signal, Toe reflection, Pile impedance, Discontinuity
407
1087
INTRODUCTION
For large bored piles of 0.80m to 1.50m in diameter, piles are
Recent economic growth is being embraced by high-rise constructed to depths of 30m to over 60m. Wet process is
buildings in Bangkok, the city of Thailand, supplemented by commonly used and piles are founded in sand layers.
other fast developing structures such as elevated highways and
railways, bridges etc. As the Bangkok subsoil suffers a lot of Compressive cylindrical strength of concrete used in bored
constraints especially by the presence of thick soft clay on the piles in Bangkok soil is in a range of 240-280 ksc (24-28 Mpa)
top, almost every structure relies on pile foundation. Tens of
thousands of piles are being constructed and tested annually in
PRINCIPLES OF SONIC INTEGRITY TESTING
this city. For old structures, short friction piles (6-8m long)
floating in the soft clay layer were commonly used. In modern The test method is based on the time-domain, one dimensional
practice, driven concrete piles with tips extending to the first wave theory. The compressive wave is propagated axially by
sand layer or to stiff clay layer at a depth about 20-30m are
the blow of hand-held impact device on to the pile head. The
normally employed for light to medium sized structures. For
wave travels down along the pile shaft and reflects upward at
heavy structures or high-rise buildings deep large diameter
the pile toe. In case where pile impedance variations occur due
(0.80m-1.5m) bored piles with tip extending to more than 45m
to changes in pile cross sectional area or properties of pile
are commonly used. Safety and overall stability of any finished
materials, or presence of discontinuities, part or whole of down
structure highly depend on the foundation and assessing the ward traveling wave reflects at the impedance variations and
quality of foundation is an inevitable aspect at the early stages returns to the pile top before the first reflection from the pile
of the structure. An acoustic technique equipped with computer toe. The reflected wave is detected by an accelerometer
softwares, developed by TNO, PDI, IFCO, etc., is commonly mounted on the pile top. Signal traces of time-velocity are
used to test the pile integrity. usually recorded. As piles are installed in the ground the
reflected signal is usually influenced by shaft friction
TYPICAL SUBSOIL CONDITION IN BANGKOK contributed by the surrounding soil layers.
The Bangkok subsoil generally consists of 12-18m soft marine The formulae commonly used in analysis of integrity are;
clay deposit underlain by medium to stiff clay. The medium
dense to dense first silty fine sand is found below the stiff clay L= c. T/2 (1)
layer at depths 20m to 35m. The dense second sand layer
Where L is pile length, c is velocity of stress wave and T is
occurs at about 50m depth. Between these sand layers, stiff to travel time of the wave from pile top to toe and reflect back to
hard clay layers are present. the top.
CONSTRUCTION METHOD Impedance of pile z (pile resistance to change in stress wave

velocity) is given by
Two types of construction method, dry and wet process are
commonly practiced in bored piling in Bangkok Soil. z = E.A/c (2)
where A is cross-sectional area of piles and E is Young’s

Dry process uses steel temporary casing (about 15m in length)
Modulus of the pile material
to protect the soft clay from caving in. The bore is drilled
either auger or using bucket. For small piles, shells with drop E = ρ.c2 (3)
weight is used to make the bore. Concrete is poured into
from Equation (2) and (3)
borehole which is in dry condition after placing the
reinforcement cage. z = ρ c.A (4)
Wet process too normally uses steel temporary casing of 14- where ρ is density of pile.
15m in length. A borehole is made by rotary method using The length of pile is determined from the travel time of the
auger or auger bucket under bentonite slurry. Concrete is reflected wave from the pile toe and the wave velocity in
poured with tremie method. concrete. The velocity wave can be estimated from the
concrete strength of piles or from the reference pile with
Dry process is employed mainly for small piles of 0.35m to known length. In the same way the location of discontinuities
0.60m in diameter. Length of the piles are up to 25m and and irregularities in piles can be also determined. Equation (4)
extends further if necessary, depending on subsoil condition, shows that the change in pile impedance could be due to
absence of sand layer with groundwater, etc. variation in pile cross-sectional area or in concrete quality.
408
1088
CHARACTERISTICS OF SONIC TEST SIGNAL IN In some cases, even though pile toe is extended beyond 35m in
BANGKOK SUBSOIL depth, the toe reflection can be observed in the test signals of
toe grouted piles having L/D within 40.
Soil influence Figure 3 is an example for such case where the bored piles
were base grouted and of 40m in length (toe at 55m, trim level
Influence by different soil strata on signals have been observed at -15m). However, it should be noted that these reflections are
in many cases. Stiff clay layer at depth 15-25m (undrained
prominent only for few piles. Influence on toe reflection by
shear strength of 10-22 t/m2.) influences prominently and
grouted toe also depends on stiffness of the soil, underneath
higher degree of signal reflections have been observed in small the toe, which has been grouted.
bored piles than that of large bored piles. Figures 1 and 2
explain these conclusions.
Soft Clay Very Stiff Clay Hard Clay Dense Sand

Soft Clay Med. Very Stiff Dense
Clay Clay Sand
5.3cm/s
0 5 10 15 20 25 30 35 40 45 50 V2-88c
10.6cm/s Pile G30 4000 m/s f:8 Sr
0 3 6 9 12 15 18 21 24 27 30 V2-88c 27 Jan 97 exp: 20

Pile 49 4000 m/s f:8 Sr
19 Sep 95 exp: 20
Fig. 3 Toe reflection observed from a toe grouted pile .(pileφ

Fig. 1. Influence of soil friction on sonic signals (a pre-cast 1.00m x 40.0m , cutoff –15.0m, toe depth 55.0m)
driven pile 0.40x0.40x25.0m in clay layers)
Size variations
Toe reflections Size variations interpreted by signals are often consistent with
casting records and drilling monitoring results available.
Toe reflections are clear when the toe depth is 20-22m for
piles of 0.35m-0.50m in diameter and up to 24m for 0.60m Figure 4A and 4B show the comparison of borehole profiles
piles. In general, when pile length over diameter ratio (L/D) is and sonic test records.
within 40, toe reflections are clear and hence the pile length is
predicted. However, for piles founded in depth deeper than 30-
35m the toe reflection is uncommon as damping of signals is
caused by deeper soil layers. Further, magnitude of toe
reflection is prominent for larger piles compared to small piles 5.4cm/s
with same length and constructed at same site. Figure 2 shows 0 5 10 15 20 25 30 35 40 45 50 V2-88c
the signal characteristics of different size piles constructed just Pile 322
10 Jan 95
4000 m/s
exp: 20
f:8 Sr
a few meters apart with same length.
23.2m
11.0cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
Pile 6 4000m/s f:8 Sr
12 Sep 95 exp: 20
22.0m
7.1cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c Soft Clay Med. Stiff to Dense Stiff Dense Sand
Pile 24 4000m/s f:8 Sr Clay Hard Clay Sand Clay
12 Sep 95 exp: 20
Fig 4A Size increase in pile shaft at 13m is shown by both
Fig. 2 Toe reflections from different pile sizes, pile no. 6 (φ sonic integrity test and drilling monitoring records. (Pileφ
0.50m x 22.0m) and pile no. 24 (φ 0.35m x 22.0m) 1.0m, cutoff –5.00m, toe depth –45m.)
409
1089
Figure 6 shows bored pile sections formed at the lower end of
temporary casing. These bored piles were cast together with a
steel column for construction of deep basements with
4.3cm/s
0 5 10 15 20 25 30 35 40 45 50 V2-88c diaphragm walls using top-down method. These piles were
Pile 99 4000 m/s f:8 Sr exposed when excavation reached to 15m below the ground
25 Nov 94 exp: 20
level.
Soft Clay Med. Stiff to Hard Dense Stiff Dense Sand

Clay Clay Sand Clay
Fig 4B Uniform pile size along the pile shaft is shown by both
drilling moitoring record and sonic integrity test record. (Pile
φ 0.80m, cutoff 3.0m, toe depth -45m.)
Common causes for size variations in bored piles are;

Fig. 6 The photograph illustrates the pile size changes caused
1. Effects from temporary casing by difference in diameter of temporary casing and auger used
- slight dimension changes of casing and bucket or auger in pile construction.
- inadequate length to protect the soft clay
- surface intact change at transition at the casing end The test signal in Fig. 7 shows size decrease caused by
2. Delayed feeding of slurry and improper slurry level inadequate casing length used in soft clay layer. Such defect is
maintenance for wet process found often in piles constructed by dry process using casings
3. Long time duration of maintaining open bore of inadequate length to protect the entire soft clay layer of
4. Non continuous concrete pouring varying thickness.
5. Soil conditions such as inclusion of sand lenses in clay
layers.
Very Stiff Clay Med. Stiff Clay Dense

Figure 5 shows reduction of pile size at about 14m. A Clay Sand
completed bored pile would have the size of outer casing at the
top portion while the lower portion would be of bucket/auger
size which is slightly less than the casing size. This size 11.0cm/s
variation has been indicated by the signals at the transition 0 3 6 9 12 15 18 21 24 27 30 V2-88c
zone. However dimensional variations by improper concrete Pile 66

25 Jun 97
4000 m/s
exp: 20
f:8 Sr
slumping upon extracting the temporary casing and gradual

Fig. 7 The signal shows decrease in pile size at 10.6m (pile φ
dimensional changes at the transition zone are hardly indicated
by the signal. 0.35m x 26.0m)
Very Soft Clay Med. Stiff Clay Dense Sand

Soft Clay Med. Very Stiff Clay Clay
Clay
10.0 cm/s
10.5cm/s
0 4 8 12 16 20 24 28 32 36 40 V2-88c
0 3 6 9 12 15 18 21 24 27 30 V2-88c
Pile 168 4000 m/s f:8 Sr
Pile 30 4000 m/s f:8 Sr 11 Nov 95 exp: 20
2 Jul 97 exp: 20
Fig. 8 Pile with localize size increase at13.5m indicated by the
Fig. 5 Pile size decrease or change at 14m shown by the sonic signal, (pileφ 0.80m, cutoff 1.4m, toe depth 35.85m))
signal (Pile φ 0.80m x 21.0m)
410
1090
In contrast to the size reduction, caused by smaller
auger/bucket size than the casing size, relative size increase
also have been observed. Casting records for piles indicated in Clay Sand
Fig. 8 noted about 30% over break of concrete volume.
Discontinuities/cracks 8.9 cm/s

0 3 6 9 12 15 18 21 24 27 30 V2-88c
Pile 50 4000 m/s f:8 Sr

Common causes attributed for these are; 4 Apr 97 exp: 20
1. disruption in concrete pouring
2. contamination of concrete with bentonite slurry or soil Fig. 10A Sonic test signal showing size decrease and or a
3. lifting up of insufficiently workable concrete or hardened discontinuity at 7.5-9.0m and poor quality pile section near
concrete upon extracting the temporary casing the to (Pile 0.35m x 20.0m)
Figure 9A shows a discontinuity in pile shaft caused by

disruption during concrete pouring. Concrete supply was
interrupted by poor coordination between the piling contractor Clay Sand
and concrete suppliers. By the time to resume the concreting
the lower section has been hardened. The discontinuity found
was later closed by coring and cement grouting. Figures 9A 3.5 cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
and 9B show the signal interpretation before and after closing Pile TEST 4000 m/s f:8 Sr
the discontinuity. 25 Jun 97 exp: 20
Fig. 10B Sonic signal on the failed test pile showing a

construction joint between pile cap and pile top at 1.40m. Size
decrease or a discontinuity is also indicated.
14.2 cm/s
0 4 8 12 16 20 24 28 32 36 40 V2-88c
Pile 1 4000 m/s f:8 Sr Figure 11A shows presence of poor quality concrete at the pile
24 Apr 97 exp: 20 top due to contamination of concrete with bentonite slurry.
Such defect is often found in piles with shallow cutoff level
Fig 9A Sonic signal showing a discontinuity at 11.0m caused
and hence adequate over cast length to have a sound concrete
by interruption in concrete pouring(pile φ 0.80mx38.8m) and
at cut off level, sometimes, cannot be made. Normally 1.0m to
multiple reflections from the discontinuity.
2.0m over-casting of concrete is practiced.
After trimming back the pile top, sonic test was performed and
the test signal indicates the sound integrity of the same (see
Fig. 11B)
7.6 cm/s
0 4 8 12 16 20 24 28 32 36 40 V2-88c
Pile 1 # 4000 m/s f:8 Sr
2 Jun 97 exp: 20
22.8 cm/s
Fig. 9B Sonic integrity test shows the discontinuity at 11.0m 0 3 6 9 12 15 18 21 24 27 30 V2-88c

Pile 28 4000 m/s f:8 Sr
has been closed by grouting. 15 Jan 97 exp: 20
Fig. 11A The signal showing poor quality concrete at pile top
Figures 10A and 10B show the characteristics of signals
observed at a site where dry process was adopted. Many piles section (Pile φ 0.60m x 18.0m)
at this particular site were interpreted for poor integrity
especially at the bottom portion. From the construction records
it was noted that boring was done in water bearing sand layer
8.8 cm/s
without casing or bentonite slurry causing concrete
0 3 6 9 12 15 18 21 24 27 30 V2-88c
contamination with water and/or segregation. Pile 28 4000 m/s f:8 Sr
18 Jan 97 exp: 20
Further, static pile load test results indicate the test pile (Fig. Fig. 11B The signal acquired after trimming top portion. The
10B) failed well before designed maximum test load. Figure reflections from the lower part of the pile and toe become
10B also shows the joint between pile cap and pile. visible.
411
1091
Physical pile damage due to excavation or improper chipping Type of Defect Causes %
to trim level
Poor concrete at cutoff level near the ground 0.1
Most discontinuities or cracks in bored piles after installation pile top (0-3m) or not enough overcast
of piles can be caused by construction activities associated Cracks/ Excavation works for pile 2.2
with basement excavation adjacent to them and improper discontinuities trimming and Construction
trimming to design cut-off level. activities
Inadequate casing length or 1.0
The test signal in Fig. 12 shows a crack or discontinuity in the Size reduction soft clay layer variation in
thickness
pile caused by soil movements occurred by adjacent
excavation work. From the above assessment, most of the defects are found to be
in small piles with a diameter of 0.60m or smaller. Study of
piling and excavation plans show that piles with a crack
Soft Clay Med. Very Stiff Dense Sand Dense Sand indicated by the sonic integrity were located in the vicinity of
Clay Clay
excavation boundaries.
6.2 cm/s
A total of 1491 piles were found with localize increments in
0 6 12 18 24 30 36 42 48 54 60 V2-88c pile size. All of size variations including size reductions are
Pile 681
8 Apr 96
4000 m/s
exp: 20
f:8 Sr
found to be at the transition interface of temporary casing and
soil layer, where soil strata changes also occur.
Fig. 12 Sonic signal showing a discontinuity (crack) at about
7.8m, caused by lateral soil movements.
CONCLUSION
29-Apr-96
1500 CAPWAP (R) Version 1 994 -1
Ton
For Msd
Vel Msd
Sonic integrity test is an effective method to assess the pile
750
integrity. Computerized acoustic technique developed by
10 20 30 40 50 ms
TNO, PDI and IFCO is commonly used in Bangkok for both
0
0 1 2 3 L/c
-750
Pile
driven piles and bored piles. About 3.3% of 8689 piles have
30
Ton / m
Shaft Resistance
Distribution been interpreted for either size reduction or poor quality of
(Toe 178.8 Ton)
Pile Forces at Rut
concreting or cracks/discontinuities.
1000
Ton
Fig. 13 High strain dynamic load test signal showing a crack Toe reflections are generally clear for small piles with
in the pile at about 8.0m. slenderness ratio of 40. But only in few cases, especially toe
grouted, toe reflection has been observed for longer piles with
An inadequate retaining system for excavation work has toe embedded below 30-35m.
caused the lateral movement to the pile top about 80cm and a
crack has developed in this pile. The crack was also detected Signal interpretations have been compared with soil
by the high strain dynamic load test performed to determine investigation results and construction records. Causes for
the pile capacity of the cracked pile (see Fig. 13). defects could be identified and conformed to casting records.
From Fig. 12 and Fig. 13 the duration of pulse in the sonic Minor defects, such as small cracks can be predicted in many
integrity test signal is about half of that in dynamic test signal. instances and nature of them need to be verified by further
The short duration (high frequency) pulses can resolve visual inspection.
reflections from narrow impedance changes. In such case
The predicted defects and possible causes verified from the
presence of a crack in the pile is more clearly indicated by
construction records would help the piling and excavation
sonic integrity test signal (low strain) than the dynamic test
contractors to be aware of the causes of defects associated with
signal (high strain).
construction practice.
An assessment of pile defects based on the sonic integrity test
ACKNOWLEDGEMENT
records on 8,689 bored piles in Bangkok subsoil and their
construction records including soil conditions was carried out.
The authors express their appreciation to E.D.E. Co. Ltd. in
The assessment indicates that integrity of 285 piles (3.3%) out
supplying test records and to Mr. Sujit Onsri and Mrs. Piyamitr
of 8689 piles has been suspected. They are summarized in the
Sura-auggool for their assistance in the preparation of this
table below;
paper.
412
1092
REFERENCES
CIRIA PG72 [1977] “Review of problems associated with the

construction of cast-in-place concrete piles” Construction
Industry Research & Information Association.
CIRIA PG4 [1977] “Integrity Testing of Piles: A Review”

Construction Industry Research & Information Association.
Fleming, W. G. K., Reiding, F., and Middendorp, P. [1985]

“Faults in Cast in place Piles and Their Detection”,
Proceedings of the Second International Conference on
Structural Faults and Repair, The Institution of Civil
Engineers, Westminster, London, pp. 301-310.
Rausche F, Likins E. G and Hussein M. [1988] “Pile Integrity

by Low and High Strain Impacts”, Application of Stress-Wave
Theory to Piles, Third International Conference. Ottawa,
Canada, pp. 44-55.
Seitz, J.M. [1984] “Low Strain Integrity Testing of Bored

Piles”, Application of Stress-Wave Theory to Piles, Second
International Conference, Stockholm, Swedin, pp. 94-102.
Sliwinski, Z. J. and Fleming, W.G.K. [1983] “The Integrity and

Performance of Bored Piles”, Advances in Piling and Ground
Treatment for Foundations, The Institution of Civil Engineers,
London, UK., pp. 153-165.
413
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON THE APPLICATION
OF STRESS-WAVE THEORY TO PILES / SAO PAULO / BRAZIL/ 11 – 13 SEPTEMBER 2000
Application of Stress-Wave
Theory to Piles
Quality Assurance on Land and Offshore Piling
Edited by
Sussumu Niyama
Institute for Technological Research – IPT, Sao Paulo, Brazil
Jorge Beim
PDI Engenharia, Rio de Janeiro, Brazil
Sonic Integrity test of piles-integrity effected by basement excavation

in Bangkok soft clay
N. Thasnanipan, A.W. Maung, P. Tangseng & Z. Z. Aye
415
Sonic integrity test of piles-integrity effected by basement excavation in
Bangkok Soft Clay
N. Thasnanipan, A. W. Maung, P. Tanseng & Z. Z. Aye
Seafco Company Limited, Bangkok, Thailand
ABSTRACT: Modern buildings in Bangkok nowadays are designed with basement facility and different ele-
vation of pile foundation which require excavation works in the immediate vicinity of the constructed bored
piles. Adverse effects on piles from adjacent excavation works in terms of excessive movement of piles and
pile damages have been observed in some projects. Assessment of bored pile damages using sonic integrity
test is focused in this paper with demonstration of test results from three case histories. Method of remedial
works for damaged piles and additional testing for pile capacity justification where available are also briefly
discussed. Stress induced in piles due to the soil movement caused by excavation works is also examined
with the aid of computerized finite element program. Location of cracks detected by sonic integrity tests is
correlated with the analyzed bending moment in pile, which exceeds the cracking moment capacity of pile.
1 INTRODUCTION 2 GENERAL DESCRIPTION OF BANGKOK

SUBSOIL
Due to the prevailing condition of subsoil (existence
of considerably weak stratum at top part) piles are Bangkok subsoil consisted of alternating clay and
commonly used for foundation of buildings and sand layers of Quaternary deposits extending down
structures in Bangkok. Deep-seated large diameter to about 550m depth where bedrock generally exists
bored piles of over 35 m depth with 0.8 –1.8m in di- (Balasubramanium, 1991). Made-Ground consists
ameter have been mainly used for heavy structures predominantly of Fill-Materials, Clayey Sand or
such as high rise buildings, elevated expressways, Silty Clay with some cement block rubble and rock
flyovers or overpass bridges and more recently for fragments, is commonly found up to 4m depth. Soft
underground subway stations of the first Bangkok to very soft, highly compressible dark gray marine
MRTA system. As a part of pre-planned quality as- clay lies beneath Made-Ground and in some areas it
surance regime and retrospective investigation, in- lies under weathered crust layers of 2m thick. De-
tegrity of the piles are tested to obtain the informa- pending on the location, this layer is extended up to
tion with regard to the potential deficiencies of the 12-18m. About 2m thick Medium Clay layer can be
constructed piles which may have formed during ac- observed between Soft Clay and underlying Stiff
tual pile construction process or may have been at- Clay. Generally Stiff Clay layer occurs directly un-
tributed by other activities after construction of the derneath Medium Clay and its depth goes up to 22m.
piles. Designs of the modern buildings in Bangkok Below Stiff Clay layer, First Sand layer 5-8m in
frequently call for the basement facility and for the thickness can be found. This First Sand layer, how-
different elevation of pile foundation, which require ever, is absent in some areas. Stiff to Hard Clay
excavation work mainly in top soft clay layer. In layer underlies First Sand and it is found to be about
some projects, adverse effect from excavation works 5m thick. Second Sand layer generally occurs at
adjacent to constructed piles has been observed. depths between 45 to 65m. The Generalized Fence
This paper presents the assessment of bored pile Diagram of Subsoil covering the central area of
damages which were induced by adjacent excavation Bangkok is illustrated in Figure 1.
using sonic integrity testing method.
163
417
the technical requirement of the particular projects
ranging from minimum 10% to maximum 100% of
total constructed bored piles. A sonic integrity tester
with built-in computer having high quality signal
acquisition system and computer programs devel-
oped by TNO, PDI, IFCO etc. are commonly used
for pile integrity testing.
5 OVERVIEW OF EXCAVATION INDUCED

PILE DAMAGE IN BANGKOK
In general, pile defects are caused at two stages, dur-

ing the pile construction process and after construc-
Figure 1. Generalized fence diagram of Bangkok subsoil
tion of the piles (post-pile-construction).
Thasnanipan et al. (1998b) reported that integrity
of 285 bored piles (3.3 % of 8,689) were found to be
3 BORED PILE CONSTRUCTION IN of doubtful quality according to an assessment on
BANGKOK
the results of sonic integrity test with the additional
information obtained from the pile construction re-
In Bangkok, bored piles are constructed mainly by
cords of bored piles in Bangkok subsoil. The results
two methods, dry and wet process. Dry process is
of the findings are summarized below.
applied commonly for small piles of diameter 0.35
to 0.60m with relatively shorter lengths. In dry Table 1. Summary of pile assessment on 8,689 bored piles
process, a pile is bored to the required depth within (Thasnanipan et al. 1998b)
impermeable layers (usually up to 25m) by either ro-
Type of Defect Causes %
tary or percussion method whilst temporary casing
(about 15m in length) is installed in soft clay layer Poor concrete at Cutoff level near the ground 0.1
pile top (0-3m) or inadequate overcast
for protection of soil caving. After installing rein-
forcement cage, concrete is poured into the borehole Size reduction Insufficient casing length or 1.0
under dry condition. Tripod method, shell with drop soft clay layer variation in
thickness
weight is also commonly used in dry process.
Wet process method as its name implies makes Cracks/discontinuit Excavation works for pile 2.2
the pile under wet condition by using drilling slurry ies trimming and construction
activities
(bentonite or polymer). Temporary casing of 14-
15m in length is also used as a support in soft clay As can be seen in the above table, higher percentage
layer and soil inside the casing is normally exca- of defects is caused by post-pile-construction activi-
vated by auger applying rotary drilling method. ties and more commonly by movement of surround-
Drilling is continued with the bucket under the ing soils induced by excavation works in the vicinity
slurry from the top of sand layer to the final depth. of the piles.
Tremie concreting is necessary for pile installed un- Modern buildings in Bangkok frequently require
der wet process. Bored piles constructed by wet the excavation works for basement facility and for
process are generally of large diameter (0.8m to the use of different elevations of pile foundation.
1.8m diameter) deep-seated (30m to over 60m) and Excavation works are mainly carried out by using
are normally founded in either first or second sand cut slope, sheet pile wall, diaphragm wall and secant
layer. Bored piles having compressive cylindrical pile wall. Since most of the bored piles are designed
strength of concrete in a range of 240-280 ksc (24- mainly to carry an axial load, lateral and tensile
28 MPa) with reinforcement ranging 0.5% to1.2% of force which may impose on pile during basement
pile sectional area, are commonly used in Bangkok. excavation are sometimes not considered or over-
looked. The lateral displacement of ground caused
by excavation naturally induced an additional lateral
4 SONIC INTEGRITY TESTING IN BANGKOK force and bending moment in the pile. Once the in-
duced bending moment exceeds the cracking mo-
Sonic integrity test is widely used for integrity test- ment capacity, piles are subject to be cracked. Un-
ing of both driven and bored piles in Bangkok. It is expected tensile force acting upon the pile due to an
employed as a part of quality control and or as a ret- excessive heave or uplift force induced by excava-
rospective investigation when some problem be- tion can also cause cracking of pile. In Bangkok,
comes apparent. For a quality control regime, the pile cracks caused by external forces are normally
number required for integrity testing depends upon found to be at the level where Soft and Medium to
164
418
Stiff Clay boundary is present. Figure 2 illustrates Basically, following formulae are used in analysis of
the typical features of external forces, which cause a sonic integrity test.
crack in the piles. Pile damage associated with sheet
pile wall excavation is presented in this paper. L = c T/2 (1)
where, L is a length from pile head to reflected sur-

TYPICAL SOIL PROFILE face, c is a velocity of stress wave and T is a travel
time of the wave for length L.
SOFT CLAY From this formula, length L, can be computed
from the travel time of reflected wave and the wave
MEDIUM CLAY velocity of concrete which may be estimated from
STIFF CLAY concrete strength of piles or from the reference pile
of known length.
SAND
z = E A/c (2)
crack induced crack induced crack induced E = ρ c2 (3)
by large lateral by sliding by uplift
movement failure force
where, z is an impedance of pile, E is Young
Figure 2. Typical features of external forces causing pile
Modulus of pile material, A is cross-sectional area of
cracks in Bangkok subsoil pile, c is the propagation velocity of the stress wave
and ρ is density of pile.
6 APPLICATION OF SONIC INTEGRITY TEST From Equations (2) and (3)

IN PILE-INTEGRITY CHECKING
z = ρcA (4)
Sonic integrity test is simple, rapid, efficient and
cost effective in examination of the pile integrity. It As can be seen from Equation (4), the change in pile
is reasonably reliable for the integrity testing of the impedance could be due to variation in pile cross-
pile especially for the upper part where potential of sectional area or in concrete quality.
damage is higher as far as defect caused by the ex-
ternal effects (lateral displacement of soft clay due 6.2 Interpretation of sonic integrity test
to adjacent excavation works, unexpected force from
improper pile head trimming) is concerned. In general the typical pile features detectable by
sonic integrity test includes;
6.1 Basic principles of sonic integrity test • Pile size variation
Sonic integrity testing also called as low-strain in- • Pile toe reflection (for pile length verification)
tegrity testing examines the response of a pile to a • Pile material variation
light external impulse force. Either time-domain or • Soil influence
frequency-domain can be used in analysis of meas- • Discontinuity and/or cracks
ured data by sonic integrity test. In Bangkok, time- The typical sonic integrity signals of above fea-
domain method is mainly applied in interpretation of tures have been reviewed and presented by Thas-
test data. nanipan et al. (1998a). As some features have a
Sonic integrity test is undertaken by striking the similar pattern to each other, it is always important
head of the pile with a light hand-held hammer and to check and compare the test data of pile in ques-
recording the response of the pile to this impulse tion with those of other piles within the same job site
blow by means of a sensor or accelerometer placed to establish “site signature”. In some cases correla-
in good contact with the pile head. Stress wave the- tions should be made with similar projects where
ory is a basic principle of this low-strain integrity damaged piles were investigated with firmed evi-
testing. The hammer blow induces a compressive dence (coring or excavation to detect damage level).
stress-wave into the pile, which propagates axially Interpretation should also be made with sound
along the pile shaft and reflects back toward the pile knowledge of pile construction technique, subsoil
head at a change of impedance within the pile. In condition and other factors (problem encountered
the case where pile impedance variation occurs due during and post pile construction) which may influ-
to changes in pile cross sectional area or properties ence the sonic test signals. Pile construction records
of pile materials or presence of discontinuities, part are also useful in interpretation of piles with de-
or whole of down ward wave reflects at the imped- tected anomaly. Conclusion should be made after
ance variations and returns to the pile head before careful and thorough review of test results in con-
the first reflection from the pile toe. junction with other available information. Further
419
165
investigation (if applicable excavate down to level of movement as well as settlement of ground including
detected crack for visual inspection) may require be- tension cracks of more than 300mm in width were
fore final conclusion is made. observed at some locations. Minor damages were
As a main focus of this paper, the typical features also investigated at the adjacent properties. Tempo-
associated with the pile damage (discontinuity/ rary raking struts were immediately installed in re-
cracks) detected by sonic integrity test is discussed sponse to these excessive ground movements caused
in detail with the data obtained from three cases as by sheet pile failure. A typical scheme showing ex-
presented in the following sections. cavation works of Case I is presented in Figure 3.
6.3 Pile damage assessment

From the results of sonic integrity test, degree of
damage or level of crack may be assessed. The as-
sessment of crack generally involved – setting of
amplification in the same range for all piles and
comparison of input impulse, amplitude of anomaly
and intensity of reflection from the anomaly (i.e.
multiple reflection). Additional information obtained
from further investigation as described in above sec-
tion is also used in pile damage assessment.
7 CASE STUDIES
7.1 Case I – Sheet pile excavation supported by soil Figure 3. A picture showing slope excavation with sheet pile
support – Case I
berm
7.1.4 Sonic integrity test and pile damage assess-
7.1.1 Subsoil condition
ment
A 12m thick soft clay layer occurs beneath fill mate-
More than 50% of piles were tested with sonic integ-
rial and weathered crust of 1.0 to 2.5m thick. Uncon-
rity test. Test results were thoroughly reviewed and
fined shear strength (Su) of soft clay layer ranges
1t/m2 at top and 2.5t/m2 at bottom whereas the natu- “site signature”, most common features of the
majority of piles, is established. It is to be noted that
ral water content is about 60-80%. Medium clay
test signals from piles of less potential for damage
layer of 1 to 2m in thickness, having shear strength
(least effect from excavation) are selected in estab-
of 2.5t/m2 to 3.5t/m2 underlies soft clay layer. Be-
lishment of “site signature” or reference signal of
low medium clay, at the depth between 18 to 27m,
good piles. Based on the “site signature” of good
stiff clay with traces of fine sand having SPT N
piles, defect piles were identified and assessed. Ta-
value of 15 to 37 is observed. Thick layer of fine
ble 2 shows the summary of pile damage assessment
sand interbedded with clayey sand occurs under-
by sonic integrity test.
neath stiff clay layer.
Table 2. Summary of pile damage assessment by sonic integ-
7.1.2 Pile foundation rity test – Case I
Pile foundation of eight high rise buildings in this
Pile No. of piles with defect Defect depth
project comprises 904 cast-in-place bored piles of
Dia. below ground
diameter 1.0m and 1.5m, founded at 59m below Least Less Prominent
(m) level (m)
ground level. Pile tips are embedded in dense sand prominent prominent crack
layer and the entire length of piles are reinforced crack crack
with steel reinforcement of 0.72% of pile sectional 1.0 4 8 7 6.0 – 13.8
area for the top part and reduced gradually to 0.23% 1.2 - 10 2 6.1 – 7.6
for the bottom section. To support the underground 1.5 - 1 - 9.9
water tanks, 88 pre-cast concrete piles (I-30cm) hav-
ing 21m length were also driven down to stiff clay The typical sonic integrity test results of undamaged
layer. pile and damaged pile classified as least prominent
crack, less prominent crack and prominent crack are
7.1.3 Excavation method and observed failure shown in Figure 4. Majority of piles with crack de-
Excavation was initially carried out by using sheet tected were located in the periphery of the excava-
pile (Type FSP III) with soil berm support or slope. tion zone. Piles, particularly with prominent cracks,
Before reaching to –8.1m maximum depth of exca- were observed to be deviated from their original lo-
vation, sheet pile failure and associate severe lateral cation up to 600mm towards excavation.
166
420
7.2 Case II – Pile damage caused by sheet pile
excavation with one level temporary support
7.00cm/s
0 6 12 18 24 30 36 42 48 54 60 v2-88c
7.2.1 Subsoil condition
Pile 890 4000m/s f : 12 sr Soft Clay layer extends to 15m below ground sur-
10 Apr 96 exp: 20
face with undrained shear strength increases from
1t/m2 at the top and 2t/m2 at the bottom. Medium
5.8cm/s
Clay layer of 2 to 3m in thickness is found between
0 6 12 18 24 30 36 42 48 54 60 v2-88c Soft Clay and the underlying Stiff Clay. Stiff Clay
layer occurs up to depth 42m with the SPT ‘N’ val-
Pile 680 4000m/s f : 12 sr
8 Apr 96 exp: 20
ues ranges between 14 and 40.
5.2cm/s
0 6 12 18 24 30 36 42 48 54 60 v2-88c
7.2.2 Pile foundation
Pile 894 4000m/s f : 12 sr 402 bored piles of diameter 1m and 1.5m, founded at
10 Apr 96 exp: 20
55m below ground, being embedded in dense sand
layer were used as foundation in this project. Rein-
8.8cm/s
forcement steel bars of 0.5% pile cross sectional area
0 6 12 18 24 30 36 42 48 54 60 v2-88c were provided for the entire length of bored piles.
Pile 681 4000m/s f : 12 sr
9 Apr 96 exp: 20
7.2.3 Excavation method and observed failure
Sheet pile wall of 14m in length with 1 level bracing
was utilized for maximum 8m excavation. Though
Figure 4. Seismic test results of good pile, pile with least the design required for installation of struts at 1m
prominent, less prominent and prominent crack (top to bottom below ground level, the actual excavation was car-
in order) – Case I
ried out up to 3m depth without installing the tempo-
rary struts causing sheet piles deflected in large
7.1.5 Remedial measures and pile capacity justifi-
magnitude.
cation
Bored piles with cracks were cored to the depth be-
7.2.4 Sonic integrity tests and pile damage assess-
low the crack and grouted with non-shrink cement.
ment
Sonic integrity test was conducted few days after
Sonic integrity test was performed for all piles after
rectifying the pile. Sonic integrity signals of the de-
trimming to the designed cutoff level. It is evident
fected pile tested before and after the remedial work
that piles with crack detected by sonic integrity test
is presented in Figure 5. After the rectification
were mostly located in the vicinity of the excavation
works, high strain dynamic load test was performed
boundary.
on two piles with large horizontal deviation and sus-
pected prominent cracks. The dynamic test results
indicate that tested piles could be capable of carry-
ing the design load with factor of safety higher than
1.5.
3.5cm/s
Soft Clay Med. Very Stiff Dense Sand Dense Sand
0 6 12 18 24 30 36 42 48 54 60 v2-88c
Clay Clay
Pile 165 4000m/s f: 12 sr
13 May 97 exp. 20
6.2cm/s
2.2cm/s
0 6 12 18 24 30 36 42 48 54 60 v2-88c
0 6 12 18 24 30 36 42 48 54 60 v2-88c
Pile 681 4000m/s f:8 sr Pile 193 4000m/s f: 12 sr
8 Apr 96 exp : 20 13 May 97 exp. 20
4.1cm/s
0 6 12 18 24 30 36 42 48 54 60
Figure 6. Sonic integrity test results of good pile and pile with
v2-88c
4000m/s
prominent crack – Case II
Pile 681A f:8 sr
29 May 96 exp : 20
A summary of the number of damaged piles and
Figure 5. Sonic integrity test results on defected pile before their degree of damage assessed are presented in Ta-
and after remedial work – Case I ble 3.
167
421
Table 3. Summary of pile damage assessment by sonic integ-
rity test – Case II

Dia. Least Less Prominent below ground
(m) prominent prominent crack level (m)
crack crack
1.0 - - 5 7.5 - 9.4
1.5 - - 14 8.4 - 21.0
7.2.5 Remedial measures

Damaged piles were cored to the level below the de-
Figure 7. Layout plan showing part of bored piles, sheet pile
fected crack and grouting was performed. wall and bracing system – Case III
7.3 Case III – Pile damage caused by sheet pile 7.3.4 Sonic integrity tests and pile damage assess-
excavation with two level temporary support ment
Though sonic integrity test was employed as a part
7.3.1 Subsoil condition of quality control regime in the original plan, it was
Underneath the top 2m of weathered crust, Soft Clay eventually applied for a retrospective investigation
layer can be observed up to 15m below the ground after encountering many piles with defect. Hence all
surface. Medium Clay layer underlies Soft Clay piles were undergone sonic integrity test. According
with the thickness of 2m. Undrained shear strength to the test results, 84% of piles located within the
of Soft Clay ranges between 1t/m2 and 3t/m2 excavation zone were identified to be with crack.
whereas it is about 50t/m2 in Medium Clay. Beneath Sonic integrity test results of good pile, pile with
Medium Clay, Stiff Clay layer is found up to 25m less prominent crack, and pile with prominent crack
where the boundary of Hard Clay is investigated. are illustrated in Figure 8. As can be seen in Figure
Within Stiff Clay stratum softer zone of Medium 8, for pile with prominent crack, a distinct and sharp
Clay is identified at the depth between 19 and 21m reflection is evident at depth about 5.5m below pile
characterized by reduction in SPT values which ap- top. Repetition of reflection at multiples of this
pear consistent with increases in moisture content. depth can also be observed for this pile. These fea-
tures are comparatively not distinct for the pile with
7.3.2 Pile foundation less prominent crack. An excessive basal heave, in-
A total of 143 bored piles of 0.6m diameter were insufficient reinforcement to resist the tensile stress
stalled with pile tip at about 26m below the ground and relatively small pile size are considered to be the
to support a residential apartment. Dry process was main reasons causing the cracks in this case. Table
employed to construct the piles. Cut-off level of 4 shows a summary of pile damage assessment by
bored piles ranges from 0.85m to 5.95m depth. En- sonic integrity test.
tire lengths of all piles are fully reinforced with steel
bars grade SD40 of 0.35% of pile sectional area.
20.2cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
7.3.3 Excavation method and observed failure Pile 103 4000m/s f:8 sr
Temporary retaining wall using FSP III sheet pile 17 Feb 93 exp : 20
type of 14m in length with two level bracing placed

at 1m and 3.5m below ground level was employed
13.3cm/s
for 6m depth excavation. A part of pile layout 0 3 6 9 12 15 18 21 24 27 30 V2-88c
within excavation area showing sheet pile wall and Pile 48 4000m/s f:8 sr
17 Feb 93 exp : 20
bracing system is illustrated in Figure 7. No major
failure in terms of large tension cracks and ground
settlement was observed at the surface adjacent to 18.8cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
the excavation though some excessive sheet pile de- Pile 50 4000m/s f:8 sr
17 Feb 93
flection was visually inspected at few locations. exp : 20
Hence, it was not expected by the visual inspection

at site that excavation work has effected the integrity Figure 8. Case III - Seismic test results of good pile, pile with
of piles. less prominent and prominent crack (top to bottom in order)
168
422
Table 4. Summary of pile damage assessment by sonic integ- Stiffness of foundation piles per linear meter (E.∑I
rity test – Case III per pile spacing in row) was used as a parameter for
plate elements, where E and I is Young Modulus and
moment of inertia of pile respectively. The soil be-
Dia. below ground
Least Less Prominent tween piles in a row is ignored. Mohr-Coulomb con-
(m) level (m)
prominent prominent crack stitutive model is adopted using undrained soil pa-
crack crack rameters derived from soil investigation at site and
0.60 - 38 33 5.5 – 13.8 associated lab test results.
7.3.5 Remedial measures and pile capacity justifi- 9 CORRELATION BETWEEN RESULTS FROM
cation FINITE ELEMENT AND SONIC INTEGRITY
Coring was carried out up to the depth below the de- TEST
tected crack to verify the pile damage and then it
was grouted with non-shrink cement. To justify the The analysis results indicate that bending moment in
capacity of damage piles, high strain dynamic load piles induced by excessive ground movement due to
test was carried out on 25 piles in which 24 piles excavation works are higher than that of cracking
with detected cracks and 1 pile without crack detec- capacity of pile in all cases. Table 6 presents the
tion. Summary of dynamic load test is presented in summary of the analyses.
Table 5.
Table 5. Summary of dynamic load test results of damaged Table 6. Moment capacity of piles and induced moment in
piles – Case III piles due to excavation simulated by finite element method
Case Pile Rebar Cracking Ultimate Moment by
Mobilized capacity / Design ultimate No. of
size (%) moment moment FEM model
capacity of 240 ton pile
(m) (t.m) (Whitney) (t.m)
Less than 50% 1 (t.m)
Less than 10% 2 I 1.0 0.75 30 79 149.7
More than 100% 22 II 1.5 0.50 102 188 117.2
255.7*
III 0.60 0.35 6.5 9.2 13.3
8 FINITE ELEMENT ANALYSIS WITH Note: (*) moment in pile located outside of sheet pile wall
COMPUTER MODEL SIMULATION within the active side of soil mass.
Finite element analysis was carried out using two- As can be seen in Table 7, the crack location in pile
dimensional modeling. With the aid of PLAXIS indicated by sonic integrity test is generally consis-
computer program staged excavation was simulated tent with the location of computed bending moment,
and pile/soil movements as well as bending stress which exceeds the cracking moment capacity of
induced in the piles were examined for above pre- piles for all three cases.
sented projects. Model profile of Case III used in Table 7. Locations of crack and computed bending moments
FEM analyses is presented in Figure 9. in piles in the vicinity of excavation boundary
Case Integrity test Depth of maxi- Depth of bending
0 1.52.5 5
17.00
Su (t/m2) SPT (blows/ft)
depth of crack in mum bending moment exceed-
piles near excava- moment by ing pile’s cracking
0.0 1.0 2.0 3.0 4.0 0 20 40 60
0.0
Crust
tion boundary (m) FEM (m) moment from

Sheet pile
5.0
FEM (m)
Soft CLAY
I 12.7 & 13.7 18.0 11.0 & 13.0
10.0
II 12.7 &13.7 15.0 12.8 & 14.5
15.0
Medium
CLAY
III 11.5 & 12.0 15.5 13.0 & 14.0
Very Stiff
CLAY
20.0 Medium It is evident that detected cracks are located near or

at the depths where boundary of Soft and Medium or
CLAY
25.0 Hard
CLAY
Stiff Clay is present.
In case III, FEM analysis results suggest that
30.0
piles are subject to some degrees of basal heave as
illustrated in Figure 10. This is found to be caused
by inadequate embedded length of sheet pile. As can
Figure 9. Finite element model profile of Case III be seen in Figure 11, the results also show that
169
423
bending moment induced by excavation exceeds the Observation of actual construction works on sites
ultimate moment capacity of pile . Sonic integrity suggests that over excavation prior to installation of
test results of some piles indicate the presence of support and inadequate retaining system are the
more than one crack at some depths, suggesting piles main causes leading to excessive ground movement,
would have experienced both bending and tensile a primary reason of pile damage. Excavation con-
stress due to the excavation. tractors should be well aware of the consequences of
excessive soil movement induced by improper con-
struction practice.
The designer of the foundation should also be
aware of the potential problem on site and should
take the actual construction practice into considera-
tion in the design. For instance, if piles are expected
to experience the excessive tensile or bending stress
from the most practical excavation method, suffi-
cient reinforcement should be provided in the design
at the first place. The foundation designer should
closely be involved in the design of the retaining
system and construction control on site during the
excavation period so that any possible negative ef-
fect can be minimized if not entirely avoided.
Figure 10. Finite element analysis results showing large degree

of basal heave caused by excavation with inadequate sheet pile
REFERENCES
embedded length - Case III
Balasubramanium, A.S. 1991, Inaugural Lecture on Contribu-
tions in Geotechnical Engineering, Soil Mechanics and
Foundation Engineering, AIT, Bangkok, March 1991
Bending Moment
(T-m)
Displacement
(mm)
EXCAVATION AND BRACING SYSTEM Turner, M. J. 1997. Integrity Testing in Piling Practice, CIRIA
-10 0 10 20 0 20 40 Report No. 144, Construction Industry Research and Asso-
0.00
1ST STRUT ciation, London.
Thasnanipan, T., Maung, A. W. and Tanseng, P. 1998a. Dam-
ages to Piles associated with Excavation Works in Bangkok
2ND STRUT
Soft Clay, 6th International Conference on Problems of Pile

5.00
SHEET PILE
Foundations Building, Ufa, Russia. September 14-18, 1998.

pp. 91-98
BORED PILE
10.00
SOFT TO MEDIUM
Thasnanipan, T., Maung, A. W. and Baskaran, G. 1998b.
Depth
(m) Cracking
CLAY Sonic Integrity Test on Piles Founded in Bangkok Subsoil –
15.00
Moment of
Pile Signal Characteristics and Their Interpretations, 4th Interna-
Interface
between Soft STIFF CLAY
tional Conference on Case Histories in Geotechnical Engi-
and Stiff
Clay
neering, St. Louis, Missouri. March 9-12, 1998. pp. 1086-
20.00 MEDIUM CLAY 1092.
STIFF CLAY
25.00
HARD CLAY
30.00
Figure 11. Computed bending moment and deflection of pile

located closest to sheet pile - Case III
10 CONCLUSIONS
With the demonstration of three case studies, effec-

tiveness of sonic integrity test in pile damage as-
sessment in Bangkok is presented in this paper. This
research study also proves that external forces due to
both lateral and uplift movement of the ground in-
duced by adjacent excavation can cause considerable
damage to constructed bored piles.
170
424
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON THE APPLICATION
OF STRESS-WAVE THEORY TO PILES / SAO PAULO / BRAZIL/ 11 – 13 SEPTEMBER 2000
Application of Stress-Wave
Theory to Piles
Quality Assurance on Land and Offshore Piling
Edited by
Sussumu Niyama
Institute for Technological Research – IPT, Sao Paulo, Brazil
Jorge Beim
PDI Engenharia, Rio de Janeiro, Brazil
Non-destructive integrity testing on piles founded in Bangkok subsoil

N. Thasnanipan, A.W. Maung, T. Navaneethan & Z. Z. Aye
425
Application of Stress-Wave Theory to Piles, Niyama & Beim (eds) © 2000 Balkema, Rotterdam, ISBN 90 5809 150 3
Non-Destructive integrity testing on piles founded in Bangkok subsoil

N. Thasnanipan, A. W. Maung, T. Navaneethan & Z. Z. Aye
ABSTRACT: Pile foundation is compulsory in Bangkok, Thailand, to support almost all types of structure
due to prevailing subsoil condition. Large bored piles of 600mm to 1800mm in diameter and barrettes
(0.8mx2.7m-1.5mx3.0m) with toe level up to 60m depth are commonly used to support heavy structures cov-
ering high-rise buildings, subway stations, elevated highways, etc. In order to test or check the quality of
such high load carrying foundation piles constructed in multi-layered soil conditions, three main types of non-
destructive integrity tests such as sonic integrity, sonic logging and high strain dynamic load tests are usually
employed. Sonic integrity test is the most common method in integrity testing on both driven and bored piles.
Sonic logging test is occasionally employed as a part of pre-planned site quality control system. High strain
dynamic load test is commonly used for pile capacity evaluation. This paper presents an overview of these
three commonly used non-destructive tests on piles in Bangkok subsoil. Some findings from the author’s in-
volvement in research works are also presented.
1 INTRODUCTION 38m. Second Dense to Very Dense Sand layer oc-

curs at about 46m depth. Between these sand layers,
Three techniques namely sonic integrity, sonic log- Hard Clay layer is present.
ging and high strain dynamic load tests are the
Engineering Pore Pressure (t/m2)
commonly used non-destructive integrity tests ap- Parameters
plied in Bangkok, Thailand. The effectiveness and 0 0
0 20 40 60
suitability in terms of both cost and method of appli- Weathered Crust
cation itself are the primary reasons of the greatest Bangkok Soft Su= 10-20KPa
γ t = 14-16kN/m3
Clay (BSC)
growth in use of these tests. An overview of sonic 10 (CH) -10
integrity, sonic logging and high strain dynamic load

tests conducted on piles particularly bored piles Medium to
Depth (m)
Su= 40-140KPa
founded in Bangkok subsoil is presented in this pa- 20 Stiff Clay -20
γ t = 17-21kN/m3
(CH)
per.
Act
u
Dense to Very
al P

-30
Hyd
γ t = 20kN/m3
iez
(SM)
2 TYPICAL SUBSOIL CONDITION IN
om
ros
BANGKOK
etri
tati
40 Hard Clay Su= 250KPa -40

cD
cL
(CL) γ = 21KN/m3
t
ine
raw
Subsoil profile is relatively consistent at different

Do
localities in Bangkok. A typical subsoil profile is 50 Dense to Very -50

wn

characterized by the alternating layers of clay and (SM) γ t = 20KN/m3
Lin
sand deposits as soil succession shown in Figure 1.

e
60
Weathered Crust of 1m to 3m thick occurs as upper
-60
most layer underlain by Very Soft to Soft Clay

commonly known as Bangkok Soft Clay (BSC). Figure 1. Typical soil profile of Bangkok plain with pie-
Thickness of Soft Clay layer varies from 12m to zometric draw down conditions (after Thasnanipan et al.,
18m depending on the location. Underneath Soft 1998a).
Clay, lies Medium to Stiff Clay. Dense to Very
Dense Sand layer also known as First Sand layer is The depth of bedrock is not well determined in
found below Stiff Clay layer at depths about 25m to the Bangkok area but has been reported to be at least
171
427
550m deep (Balasubramaniam, 1991). Existing 5 THE SELECTION OF SUITABLE METHOD
pore water pressure conditions in upper part of BSC
are hydrostatic from nearly 1m below the ground In general, non-destructive tests are conducted to
level. Then the hydrostatic conditions change to pie- verify the quality of the pile. Turner (1997) high-
zometric draw down near bottom level of BSC. lighted the factors in selection of pile testing in
CIRIA Report 144 as follows;
• The perceived nature of possible features or de-
3 COMMON CONSTRUCTION METHOD
fects within the pile
The common practice of bored pile construction • The ability of the test method to detect the fea-
method in Bangkok can be divided into wet and dry ture or defect under investigation
processes. • The cost of testing and examination
For the wet process, bored piles are constructed • The ease of use and interpretation
using bentonite or polymer slurry as drilling fluid
over the full length of the piles, to ensure the stabil- Like other indirect measurement of pile integrity,
ity of the excavation at all stages. Top 15m-thick sonic integrity and sonic logging tests are only capa-
soft clay was temporarily cased. Drilling for the ble of identifying structurally significant features.
piles was done with a conventional drilling auger or Rejection or acceptance of individual piles should
auger bucket. To clean the sediments at the bore- not rely only on the results of these tests. Further
hole base the airlift technique or cleaning bucket is investigation, engineering evaluation and judgement
applied. After lowering the rebar cage, concrete is are highly recommended to confirm the defect de-
poured with tremie method. tected by these indirect techniques. Both methods
Generally the dry process is employed for small however are very useful for cost-effective screening
piles (0.35m to 0.6m in diameter and 25m in length). test to identify piles with potential defect.
This method uses a steel temporary casing of about Sonic integrity test is usually selected for both
15m in length to protect the soft clay from caving in. quality check (control test) and retrospective inves-
Boring is carried out by rotary drilling or percussion tigation. It is the cheapest in terms of cost and the
tripod rig. simplest in terms of testing process. The main ad-
Bored piles are commonly constructed with rein- vantage of this test is that since no particular meas-
forcement of 0.5-1.2% of sectional area of pile. ures/preparation is necessary during the pile con-
struction phase it is more flexible to select which
pile is to be tested. However, interpretation of sonic
4 COMMON DEFECTS CAUSED IN BORED integrity testing needs considerable experience and
PILES knowledge in testing, subsoil condition and con-
struction method.
Common pile defects such as size reduction/necking, Sonic logging test is relatively expensive. It is
discontinuity, soil/slurry inclusions, etc. may be mainly employed as a pre-planed site quality control
caused during pile construction. Most of the cracks testing. The major advantage of this method is that
in bored piles caused at post construction stage are
test can be carried out shortly after the pile construc-
usually induced by construction activities associated
with basement excavation and improper trimming of tion. Hence, rectification measures can be imple-
pile head to design cut-off level. Causes of defects mented while the foundation contractor is on site.
in bored piles during construction are; However, this method is generally not applicable if
pile integrity is in question due to post construction
• Inadequate length to protect the soft clay layer activities as access tubes are usually grouted after
• Delayed feeding of slurry and improper slurry completion of the test.
level maintenance for wet process High strain dynamic load test is usually selected
• Long time duration of maintaining open bore- to verify the load carrying capacity of piles. Pile in-
hole tegrity can also be determined by high strain dy-
• Non continuous concrete pouring or disruption namic load test. The major advantages of this test in
in concreting comparison with static pile load test are those of
• Lifting up of insufficient workable concrete or cost, time and space requirement. In some projects,
hardened concrete upon extracting the temporary dynamic load test was applied to justify the capacity
casing
of pile after rectifying its defect detected by other in-
• Contamination of concrete with drilling slurry or
soil tegrity tests. Table 1 shows the suitability of test
• Soft pile toe due to improper base cleaning method for different type of common defects.
428
172
Table 1. The suitability of test method for different type of de- fully developed so that damping effect or attenuation
fects. of the sonic signal by shaft friction is relatively low.
Type of defects Test method and suitability
for defect detection 52.0m
Sonic Sonic High strain
integrity logging dynamic 16.0cm/
s
load test 0 6 12 18 24 30 36 42 48 54 60 v2-88c
Crack Suitable Possible Suitable pile 30 4000m/s f:8 Sr

14 Jan 91 exp: 20
Pile size Suitable Possible Suitable (if
change severe de- Figure 2. Sonic signal showing a toe reflection of bored pile at
fect) 52.0m tested few days after concreting.
Poor quality con- Possible Suitable Suitable (if
crete or Contami- severe de- Use of temporary casing usually causes larger size at
nated concrete fect) the top portion (from pile top to casing bottom) for
Soft pile toe - Not usually Suitable completed piles in Bangkok subsoil. This is caused
by the size difference between casing and the drill-
ing tool (bucket/auger) as the top section of pile
6 SONIC INTEGRITY TEST formed by the outside diameter of the casing is usu-
ally 5-10cm larger than the lower section of pile
with nominal diameter formed in the borehole made
6.1 Overview of the test in Bangkok by drilling tool (Figure 3). Test signal shown in
Sonic integrity test, also known seismic test is the Figure 3 indicates a clear reflection at approximate
most common method of integrity testing for both 11m from pile top (about 15m below ground level)
driven and bored piles in Bangkok. In many projects with a repeat or multiple reflection at about 2 times
it is a part of the contractual requirement to conduct of that distance (approximate 22m from pile top). As
sonic integrity test. Minimum 10 % to maximum the polarity of the reflection is negative, the first in-
100 % of production piles are commonly tested. It is terpretation is that this pile has decrease in cross sec-
also a reasonably acceptable method for applying as tion or crack at 15m below ground level. However,
a retrospective investigation in determining integrity with the consideration of pile construction method, it
of the pile. is concluded that negative reflection is caused by
variation in pile cross section at the bottom end of
15m long temporary casing. The sonic signals with
6.2 Interpretation of the test results anomaly acquired on those piles are often misinter-
As has been reported in various literatures, it is to be preted as indicative of defect piles, leading to an ar-
emphasized that sonic integrity test is highly de- gument.
Med.
pendent on the experience in both field testing and Very soft clay clay Stiff clay Dense sand
interpretation. The signal characteristics and their
interpretations of sonic integrity test on piles Pile Cut-off level Temporary casing
bottom level
founded in Bangkok subsoil were reported by Thas-
nasipan et al. (1998b). The authors presented the in-
terpretation of various features of sonic integrity test
signals in comparison with soil investigation data Outer diameter of Nominal dia./dia. of
and construction records. casing collar drilling tool
6.3 Some findings from sonic integrity test in 11.0 cm/s

Bangkok Pile 66
0 3 6 9 12 15 18
4000 m/s
21 24 27
f: 8
30 v2-88c
sr
25 Jun 97 exp : 20
Toe reflections are generally clear for the piles with
length to diameter ratio (L/D) within 40, but toe re- Figure 3. A typical pile shape constructed in Bangkok soil and
flection was hardly detected for piles founded at resulted sonic integrity test signal profile.
depth deeper than 30-35m as reported by Thasnani-
pan (1998b). However, it was found that toe reflec- For the analysis of sonic integrity test, velocity of
tion is detectable for some deep-seated bored piles sonic wave in pile material (concrete) is typically as-
of about 50m in length if pile was tested at early age. sumed 4000m/s in most cases. Basically, lower
Figure 2 shows the significant toe reflection of strength of concrete gives lower wave propagation
52.0m-long pile tested few days after concreting. velocity. The trend of sonic wave velocity incre-
This is considered mainly due to the fact that at early ment with pile age derived from a number of 600mm
stage the bonding between soil and pile has not yet diameter bored piles with known lengths is pre-
sented in Figure 4.
429
173
4300
overcome by drilling through the obstruction in the
4100 worst case (but not possible if tube is bent), it should
be eliminated in the first place by careful fabrica-
Velocity (m /s)
3900 tion, installation and protection of the tubes.

3700
3500
7.2 Interpretation of test results
The results of a series of laboratory tests to deter-
3300
mine the effects of various defects or inhomogenei-
0 9 18 27 36 45 54 63
Days ties within a concrete section were reported by Stain
and Williams (1991). The test results were carried
Figure 4. A curve of sonic wave velocity vs. pile age in Bang-
out on small panels constructed to model various
kok sub soil (fc’ ≈ 35 MPa). pile construction defects and anomalies. The results
from these panels were related to tests on control
panels formed from homogeneous concrete.
7 SONIC LOGGING TEST Faiella and Superbo (1998) reported the analysis
of the sonic logging test results of over 6800 piles
collected from 37 sites in Italy. Based on the results
7.1 Overview of the test in Bangkok of analyses, the authors presented the defect classifi-
Use of Sonic logging test for checking pile integrity cation criteria for piles monitored by different num-
has increased in Thailand in recent years. It is used ber of tubes as illustrated in Tables 3 (a) and 3 (b).
for cast-in-place foundation elements such as bored
Table 3 (a). Defect classification criteria monitored by means
piles, barrettes, diaphragm walls and caissons. Both of 2 tubes (Faiella and Superbo, 1998)
PVC and steel tubes can be used as access tubes.
Ta/Ts Tk/D Type of defect
Though cost of steel tube is higher than that of PVC,
<1.15 - Non-homogeneous concrete
steel tubes are considered more suitable than PVC
<3 Light
tubes for two reasons; (1) less potential for bending 1.15-1.45
>3 Probably serious
and damage especially for deep-seated bored piles <1 Light
and barrettes with heavy steel reinforcement and (2) 1.45-2.0
>1 Probably serious
better bonding with concrete. These access tubes are <0.5 Probably serious
sometimes also used for pile base grouting. The ac- >2.0
>0.5 Light
cess tubes are attached to the reinforcement cage,
which is installed prior to concreting as a normal Table 3 (b). Defect classification criteria monitored by means
practice for cast-in-place piling. of 3 and 4 tubes (Faiella and Superbo, 1998).
Two access tubes are required as a minimum for Ta/Ts Pa/Pt Tk/D Type of defect
sonic logging test. For a good coverage of the test Pa<Pt
<1.15 - Non-homogeneous concrete
pile, recommended number of tubes for different Pa=Pt
sizes of bored piles is shown in Table 2. <3 Non-homogeneous concrete
Pa<Pt
>3 Light
Table 2. Recommended number of tubes for different pile size 1.15-1.45
<3 Light
Pile Diameter Minimum No. of Tube spacing Pa=Pt
>3 Serious
(mm) tubes (degree)
<1 Non-homogeneous concrete
D ≤ 750 2 180
Pa<Pt 1-3 Light
750<D ≤ 1000 3 120
1000<D≤ 1500 4 90 1.45-2.0 >3 Serious
1500< D ≤ 2500 6 60 <1 Light
Pa=Pt
2500 < D 8 45 >1 Serious
<0.5 Light
A typical problem encountered in sonic logging test Pa<Pt 0.5-3 Serious
is the access tube blockage which often comes to >2.0 >3 Serious
know only at the time sonic testing is conducted one <0.5 Light
Pa=Pt
>0.5 Serious
week or more after pile construction. The blockage
of access tube is mainly caused by intrusion of con-
where:
crete from leakage at either bottom cap or tube con-
Ta = Travel time of sonic waves in the anomalous zone
nection and intrusion of soil or other material from
Ts = Travel time of sonic waves in the sound concrete
the top cap of the tube. Sonic test instruments
Tk = Thickness of the anomalous zone
(transducer and receiver) sometimes may not be able Pa = Number of measurement paths affected by sonic anomalies
to insert down to the bottom of the access tubes due Pt = Total number of travel paths for each pile
to the bend of the tube itself. Though the problem D =Pile diameter
associated with the access tube blockage can be
430
174
Srivanavit et al. (1999) reported the interpretation of Case Concrete & ma- Profile Discussion of
test signals in comparison with the actual integrity of terial properties signal
9 model tests conducted in Bangkok. Model piles 6 Concrete mixed Sonic logging
with bentonite profile shows
are of 800mm diameter with 1m length. The sum- slurry. non-uniform
mary of the model test results including the descrip- signal, not clear
tion of model piles as well as discussions by authors signal due to
are illustrated in Table 4. poor concrete
and inhomoge-
Table 4. The summary of sonic logging model test results neous concrete.
Case Concrete & ma- Profile Discussion of
terial properties signal 7 Thin smear of Sonic logging
1 Standard good Sonic logging 2mm. bentonite profile shows
concrete pile profile shows slurry on tubes
signal loss from
with clean tubes continuos sig- (50 cm. from
top of concrete). depth
and clean water. nal, no anomaly. 0-0.5m (more
travel time).
2 Good concrete Sonic logging

pile with dirty profile shows 8 Thick smear of Sonic logging
tubes and clean defect at pile 5mm. bentonite profile shows
water. head due to slurry on tubes continuous sig-
bleeding of con- (50 cm. from nal loss and sig-
crete. top of concrete). nal delay at
depth 0-0.5m,
similar to case
3 Good concrete Sonic logging 7.
pile with dirty profile shows
tubes and dirty
defect at pile
water (contami-
nated with ben- head due to
tonite slurry) bleeding of con- 9 Weak concrete Though weak
filled. (high w/c ratio). (low strength)
crete, similar to
concrete is used,
case 2. sonic logging
profile shows
4 Good concrete Sonic logging continuous sig-
with Layer of profile shows nal, because of
signal loss from homogeneity of
bentonite
0.4m to 0.7m the tested me-
slurry, sand, due to delay of dia.
gravel fill sonic wave.
(30cm.Thickness
whole Section) at
middle of the 7.3 Factors to be considered in interpretation of
pile. sonic logging test
Turner (1997) pointed out the need of both theoreti-
cal knowledge and practical experience in interpreta-
tion of the test results in CIRIA Report 144. Sug-
5 Good concrete Sonic logging gestion is also made in this report that the anomaly
with Clay on profile shows shown in sonic test signals can be caused not only
surface of one by changing in physical properties of the materials,
signal delay at
tube (50 cm.
from top of con- 0.5m depth. but also by factors within the measuring system it-
crete). self. These factors are;
• Free movement of the probes within access tubes

• Mismatched probe positions especially at pile toe
• Measurement resolution
• Incorrect position of access tube
• Air gaps or different material around access
tubes
• Aggregate variation (in the case of base grout-
ing)
431
175
It is to be noted from the author’s suggestion based defect by other integrity tests. Figure 5 shows the
on the above factors that only variations of transmit photo of dynamic load test carried out on barrette of
time more than 15 to 20% of the norm should be re- size (1.0x2.7m) with toe depth 48.94m located at the
garded as warranting further investigation. bank of water supply canal. A 20 ton hammer was
used to activate the 2050 ton test load in this project.
8 HIGH STRAIN DYNAMIC LOAD TEST
8.1 Overview of the test in Bangkok

High strain dynamic integrity test has become a
well-accepted method especially for evaluating the
pile capacity in the foundation industry today. A
large number of related technical papers and case
histories of the test have been published and it is a
part of standards and specifications such as ASTM
D4945-89 (Standard Method for High-strain Dy-
namic Testing of Piles). In Bangkok, dynamic load
test is applied for both driven and bored piles. A
summary of the number of dynamic load test carried
Figure 5. Dynamic load test carried out on a barrette
out during 1991 to 1997 is shown in Table 5. It is to (1.0m x 2.7m x 48.94m).
be noted that quantities shown are collected from the
available sources and the actual tested numbers are
likely to be slightly more than those indicated in the 8.2 Comparison and correlation of dynamic and
table. As can be seen in the table, application of dy- static pile load test
namic load test increases year by year in Thailand. A comparison between dynamic and static load test
The pile driving analyzer (PDA) with computer results has been reported by various researchers.
software by Pile Dynamic INC, USA is mainly used Seidel and Rausche (1984) reported the results of
in Bangkok. dynamic and static load tests performed on drilled
shafts of the West Gate Freeway in Melbourne, Aus-
Table 5. Yearly minimum quantity of dynamic load test con-
ducted in Thailand during 1991 to 1997 tralia. A 20 ton hammer with drop heights between
Year No. of dynamic load test done 1.6 and 2.5m was used for the dynamic tests of 12
1991 84 shafts ranging from 1100mm and 1500mm in diame-
1992 134 ter and 35m to 64m in length. The authors reported
1993 242 that dynamic activation of static pile resistance
1994 369 forces exceeded 3000 ton for some 1500mm diame-
1995 464 ter shafts. Skin friction predictions from dynamic
1996 473 load tests and values obtained from instrumented
1997 695 shafts under static load tests were remarkably simi-
lar and pile head load-movement relationships ob-
For driven pile it is usually performed at two stages, tained from both test methods were comparable as
during driving (initial driving monitoring) and some reported by the authors.
period after installation of pile (restrike test). For Prebaharan et al. (1990) reported the results of
bored piles, a specially designed pile cap is normally dynamic and static tests conducted on bored piles at
required as an integral part of the pile head to avoid the Marina Bay Station of the Singapore Rapid
pile head damage. Ram weights of 4 ton to 20 ton Transit System. According to the report, dynamic
are commonly used in Bangkok. The common ram load tests were carried out for bored piles of
weights used for different pile sizes are shown in 1000mm diameter and 25m to 50m length founded
Table 6. in Old Alluvium of Singapore Island. Agreement
between dynamic and static load tests results permit-
Table 6. Common ram weights used for different size of piles ted replacement of 44 out of 51 tests with dynamic
Pile size (m) Ram weight (ton) load test from originally planned static load tests as
≤ 0.50 4.00 stated by the authors.
0.50 – 0.80 8.00 Vasinvarthana and Kampananon (1997) presented
1.00 – 1.50 20.00 the efficiency and reliability of dynamic load test
carried out in Bangkok. Figures 6 (a) and 6 (b) show
The dynamic load test is also occasionally used as the load settlement curves of dynamic and static load
an additional investigation when pile is found with tests conducted in Bangkok on large and small di-
176
432
ameter piles respectively. As can be seen in the fig- ACKNOWLEDGEMENT
ures, load-settlement characteristics obtained from
dynamic load test agree well with those of static load The authors wish to express their appreciation to
test for both large and small diameter bored piles. EDE Co., Ltd. for providing the test data. Thanks
0 are also given to Mr. Veera Vasinvarthana and Mr.
Natomon Kampananon STS Engineering Consult-
ants Co., Ltd. for giving permission for the use of
-5 their previous works in this paper.
Settlem ent (m m )
-10 REFERENCES
American Society for Testing and Materials 1989. ASTM

-15 D4945-89 Standard method for high-strain dynamic testing
Dynamic load test of piles, Philadelphia: American Society for Testing and
Static load test Materials.
-20
Balasubramaniam, A. S. 1991. Inaugural lecture on contribu-
0 200 400 600 800
tion in geotechnical engineering soil mechanics and foun-
Load (ton)
dation engineering, AIT, Bangkok, 14 March 1991.
Faiella, D. & Superbo, S. 1998. Integrity non destructive tests
Figure 6 (a). Load-settlement curves of dynamic and static load of deep foundations by means of sonic methods - Analysis
tests on large diameter piles (φ 0.80m x 50.0m). of the results collected on 37 sites in Italy, Van Impe &
Haegeman (eds), Proc. 3rd Int. Geotechnical Seminar on
0 Deep Foundations on Bored and Auger Piles, BAP III,
Ghent, Belgium, 19-21 October 1998: 209-213. Rotterdam:
Balkema.
Prebaharan, N., Broms, B., Yu, R. & Li, S. 1990. Dynamic
-5
testing of bored piles, Proc. of the Tenth Southeast Asian
Settlem ent (m m )
Geotechnical Conference, Volume 1, Taipei, ROC, 16-20

April 1990: 373-378.
-10 Seidel, J. & Rausche, F. 1984. Correlation of static and dy-
namic pile tests on large diameter drilled shafts, Proc. 2nd
Int. Conf. on the Application of Stress Wave Theory on
-15 Piles, Stockholm, Sweden, 27-30 May 1984: 313-318.
Stockholm: Swedish Pile Commission.
Dynamic load test
Static load test
Srivanavit P., Hiranoon J. & Pimsarn S., 1999. Integrity tests
on deep bored piles constructed in Bangkok subsoil by
-20 sonic logging method & interpretation, The engineering
0 20 40 60 80 100 technology exhibition and symposium, Organized by Engi-
Load (ton) neering Institute of Thailand, Bangkok, 1-2 November
1999: 1-8. (in Thai).
Figure 6 (b). Load-settlement curves of dynamic and static Stain R. T. & Williams, H. T. 1991. Interpretation of sonic
load tests on small diameter piles (φ 0.35m x 18.91m). coring results: A research Project, Proc. 4th Int. DFI Conf.
on Piling and Deep Foundations, Stresa, Italy, 7-12 April
9 CONCLUSIONS 1991: 633-640. Rotterdam: Balkema.
Thasnanipan, N., Baskaran, G., & Anwar, M.A. 1998a. Effect
of construction time and bentonite viscosity on shaft capac-
Use of sonic integrity, sonic logging and high strain ity of bored piles, Deep Foundations on Bored and Auger
dynamic tests in Bangkok has been presented. Piles, Van Impe & Haegeman (eds), Proc. 3rd Int. Geotech-
Though sonic integrity test is a simple and cost nical Seminar on Deep Foundations on Bored and Auger
effective method, the reliability of the test results is Piles, BAP III, Ghent, Belgium, 19-21 October 1998: 171-
highly dependent on the experience of the person in 177. Rotterdam : Balkema.
Thasnanipan, N., Maung, A. W. & Baskaran, G. 1998b. Sonic
both field-testing and interpretation. integrity test on pile founded in Bangkok subsoil – signal
The results from sonic logging test conducted on characteristics and their interpretations, Proc. 4th Int. Conf.
model piles in Bangkok helps to extend the knowl- on Case Histories in Geotechnical Engineering, St. Louis,
edge of the signal characteristics and interpretation. Missouri, 9-12 March 1998: 1086-1092.
High strain dynamic test has become a well- Turner, M. J. 1997. Integrity testing in piling practice, CIRIA
Report 144, London: Construction Industry Research and
accepted method especially for load testing of piles. Information Association.
Agreement between static and dynamic load test re- Vasinvarthana, V. & Kampananon, N. 1997. Efficiency and
sults enhances the confidence in using the dynamic reliability of dynamic load test, Seminar on Foundation’
load test as shown by rapid increment in number of 97, Organized by Engineering Institute of Thailand, Bang-
tests conducted in seven years in Thailand, mainly in kok: 79-94. (in Thai).
Bangkok.
177
433
THE SIXTH INTERNATIONAL CONFERENCE ON PROBLEMS OF PILE FOUNDATIONS
BUILDING, RUSSIA, SEPTEMBER 14-18, 1998
Problems of Pile
Foundations Building
Damages to piles associated with excavation works

in Bangkok soft clay
N. Thasnanipan, Maung A. W. and Tanseng P.
ISBN 5-88151-161-1
435
The 6th International Conference on Problems of Pile Foundations Building, Russia, September 14-18, 1998
DAMAGES TO PILES ASSOCIATED WITH EXCAVATION WORKS IN BANGKOK SOFT CLAY
Narong Thasnanipan Aung Win Maung Pornpot Tanseng

SEAFCO Co., Ltd. SEAFCO Co., Ltd. SEAFCO Co., Ltd.
Bangkok 10320 Bangkok 10320 Bangkok 10320
Thailand Thailand Thailand
ABSTRACT
In Bangkok, Thailand, pile foundations are commonly used to support all type of structures. The excavation for construction of
pile caps and mat foundations are carried out mostly in the soft clay layer. Sheet pile walls with temporary bracing systems are
commonly used for excavation works (depth ranging form 4.0-11.3m) in the built-up areas of central and outskirts of Bangkok
metropolis. Sometimes open cut excavation is adopted for shallow excavation (2.0-6.0m) in the areas where no buildings exist in
the close vicinity of the site. Lateral movements of soils take place in such excavations causing pile deviation and cracks to the
piles, especially located in the vicinity of boundary of excavation. Such defects have been found more frequently in smaller
diameter piles than the larger diameter piles. These tension cracks in piles are commonly observed above the level of interface of
soft and stiff clay layers where excavation induced bending moment in pile exceeds the pile’s cracking moment capacity. Sonic
integrity testing is usually employed to test the piles’ integrity. High strain dynamic load test is sometimes employed to check the
performance of piles with such defects. This paper presents four cases in which damages caused to piles by soil movement
associated with excavation works. Pile damages, especially tension cracks indicated by integrity testing, analysis on stresses in
pile due to the soil movements and behavior of retaining systems are also examined. Correlation between location of crack and
bending moment in pile which exceeds the pile cracking moment capacity also made.
KEYWORDS
Pile deviation, Lateral movement, Sonic integrity test, Dynamic load test, Pile damage, Bending moment, Cracking moment
91
437
INTRODUCTION TYPES OF EXCAVATIONS
Almost every structure in Bangkok sits on pile foundations. Common type of excavation found in Bangkok subsoil are
Excavation works for preparing the pile cap, footing and/or tabulated as below,
mat are usually within the soft clay commonly found in
Bangkok. However for structures requiring many basement Table 1. Common Types of Excavations in Bangkok Subsoil
levels and underground facilities, the excavation works are Type of Excavation Excavation Site
undertaken to below the soft clay more than 20m from the below G.L. Conditions
ground. Various soil retaining systems are adopted in such (m)
excavation works to suit the site and soil conditions and Cut Slope 3.0-6.0 open area, no
depth of excavation. Analysis and design for the retaining adjacent buildings
systems sometimes are overlooked by the practicing Sheet Piles Wall 4.0 to 11.3 adequate site
engineers for the possible movements, both vertical and clearance
lateral, of soil adjacent to the excavation boundaries. Jet Grouted Wall 5.0 to 10.5 adequate site
Moreover monitoring of soil movements, with exception of clearance
existing structure near by, is often ignored. Such ignorance Secant or 4.0 to 20.5m in built-up area
leads to damage the piles already constructed or installed in Contiguous Pile
soft soil stratum. As a result secondary defects found in the Wall
piles after completion of piling work, including pile top Diaphragm Wall 7.0 to 23.0m in built-up area
deviations far from allowable tolerance become an argument
between the piling contractor, excavation subcontractor and
First two types of excavations are found to induce higher
the employer or supervising engineer.
lateral and vertical soil movements compared to the
remaining types (Ref. 2) and sometimes cause a localize
failure of the soil mass retained due to improper
PILE FOUNDATIONS
construction practice. In this paper pile damage associated
with the first two types in four cases are studied and
Both bored and driven piles are commonly used for
presented.
foundations of structures in Bangkok subsoil. The subsoil
generally consist of a 12-18m thick very soft clay layer
overlying medium to stiff clay layer. A sand layer occurs at
INTEGRITY OF PILES
depth of 20m to 35m below the stiff clay layer. A second
layer of dense sand is generally ound below 50m where the
Sonic integrity test is usually employed to test the integrity
toe of deep bored piles are founded for high rise or heavy
of piles after piles have been trimmed to the design cutoff
structures. Stiff to hard clay layers occurs intermittently
level. Generally a minimum of 10% to a maximum of 100%
between the sand layers. Large diameter bored piles having
of piles are tested. A sonic integrity tester with built-in
diameter of 800mm to 1500mm are extended down to 35m
computer having high quality signal acquisition system has
to 60m respectively being embedded in the sand layers. For
been commonly used for pile integrity testing. However,
low rise buildings driven or small diameter bored pile of
severity of the defect/crack or any irregularity in piles is
35mm to 600mm are commonly used with footing for pile
usually determined comparatively based on magnitude of
groups or for individual piles. The length of pile is
reflection amplitude of individual piles from such a
generally to 20 to 28m embedded in stiff clay or first sand
defect/crack or irregularity, with reference to the input
layer.
pulse. Figure shows the sonic test records of a good pile
and piles with crack in different severity from Cases IV.
The working load capacity of bored piles is usually limited
by the allowable stress of the concrete. Cutoff level of such
piles generally varies from about 2.0m to 6.0m below
CASE I - CUT SLOPE (Bangna-Trad Highway KM. 26)
ground surface. For the structures with deep basements,
bored piles are trimmed down to 20.0m.
Subsoil Condition - A 17m thick soft clay layer occurs with
unconfined average shear strength (Su), ranging from 0.69
Generally bored piles are constructed with reinforcement of
t/m2 at the top to 2.20 t/m2 at the bottom of the layer. The
0.5-1.2% of sectional area of pile and concrete having cube
field vane shear test indicates in-situ shear strength of that
strength of 30-32 MPa. Wet process, using bentonite slurry
layer is 1.40t/m2 to 3.80t/m2 with sensitivity of 2.7 to 3.7.
is normally employed for drilling in water-bearing sand
The natural water content of soft clay layer exceeds 100%.
layers and concrete pouring is done under the slurry. Dry
Below the soft clay layer is medium clay layer with Su value
process normally used for construction of small diameter
of 4.0 t/m2 to 6.0t/m2. Then a stiff clay layer having Su
bored piles with depth not greater than 26m.
value of 13.0t/m2 is found at the depth of 21m-22m
overlying the first sand layer which extends to the depth of
about 34m to 37m. The second sand layer occurs at the
92
438
depths of about 52.0m - 61.0m. Between sand layers a Most piles with large crack detected were located in the
thick hard clay layer is present, but sporadically. outer row of foundation piles ( Fig. 1) and located under the
traffic ramps of temporary working platform. It was found
Pile foundation - A total of 428 concrete bored piles of that the crack in these piles were caused by localized soil
diameters 1.0m, 1.2m and 1.5m were installed with pile tip sliding which was occurred by the heavy traffic load. Pile
at 53m to 65m below the ground, being embedded in second deviations in order of 50cm to 100cm towards excavation
dense sand layer to support a high rise structure. Cutoff were observed.
level of piles ranges from 4.6m to 5.9m in depth. Entire
length of piles were fully reinforced with steel An inspection after the soil sliding revealed that king posts
reinforcement of 1% and 0.5% of pile sectional area in top were not installed to support parts of the ramps behind the
33m and lower section of pile shaft respectively. slope on the ground. Apart from the traffic ramp area, a
number of localized soil slidings up to 3m to 4m occurred in
Soil Retaining System - Cut slope excavation with two some areas. They are found to be caused by loose materials
levels of 1:3 slope and a berm was adopted to achieve the filled in bored hole above pile casting level during casting.
maximum excavation to -5.9m below the existing ground Less Prominent crack detected in some piles, particularly
level as the site is a large open space being located outside within excavation zones are considered to be caused by
built-up area. Slope stability analysis using effective stress improper pile trimming method. A backhoe was used to
soil parameters indicated that a safety factor of 1.2 for break off the over cast section by hitting the pile head after
temporary condition can be achieved by driving 6m long I- chipping of the concrete cover at the cutoff level around the
12 RC soldier piles at a spacing of 1.5m at the toe of the pile shaft and separating the dowel bars form the overcast
bottom slope. Moreover, the ground level around the section.
perimeter of excavation was reduced down to 0.5m.
Excavation was carried out on temporary working platforms Soil around the suspected pile with prominent crack were
supported by 24m long steel king posts (H300x300) driven excavated down to detected crack level. These piles were
into the ground. trimmed down to below the crack and re-constructed to the
cut off level. Tie beams were provided for the piles with
Pile Damages - All piles were tested with sonic integrity crack and horizontal deviation more than allowable
test. Test results are summarized in the Table below; tolerance.
Table 2. Summary of Pile Damage Detected by Sonic

Integrity Test - Case I CASE II - SHEETPILE EXCAVATION SUPPORTED
Pile No. of Piles with Defect BY SOIL BERM (Huay Kwang Area)
Dia. Least Less Prominent Depth
Prominent Prominent Crack below Subsoil Condition - A 12m thick soft clay layer occurs with
(m) crack Crack GL(m) unconfined average shear strength (Su), ranging from 1.0
1.0 4 8 7 6.0-13.8 t/m2 at the top to 2.50 t/m2 at the bottom of the layer. The
1.2 10 2 6.1-7.6 natural water content of soft clay layer is about 60-80%.
1.5 1 - 9.9 Below the soft clay layer is medium clay layer with Su value
of 2.5 t/m2 to 3.5t/m2. Then a stiff clay layer with traces of
fine sand having SPT N value of 15 to 37 is found at the
depth of about 18m-overlying sand layers which occur
below 27m in depth.
Pile foundation - A total of 904 concrete bored piles of

diameters 1.0m and 1.5m were installed with pile tip at
about 59m below the ground, being embedded in dense sand
layer to support a total of eight high rise buildings located at
the particular size. Cutoff level of bored piles ranges from
5.1m to 8.1m in depth. Entire length of piles are fully
reinforced with steel reinforcement of 0.72% of pile
sectional area for the top and reduced gradually to 0.23% for
bottom section of pile shaft. A total of 88 pre-cast concrete
(I-30) piles with 21m in length was also driven to support
the underground water tanks. The cutoff level of the pre-
cast piles was 3.1m below the ground level.
Soil Retaining System - The foundation structure is about

Figure 1. Excavation Plan and Location of Damaged Piles 10m to 20m away from the adjacent property boundaries.
Initially excavation with cantilever sheet pile wall (FSP III
93
439
with 14m in depth) supported by a soil berm or slope was
adopted to achieve the maximum excavation to -8.10m
below the existing ground level. However after excavation
to some depths, due to sheet pile failure at some locations
and excessive vertical and lateral soil movements,
temporary raking struts were then immediately installed.
Tension cracks wider than 300mm were observed on the
ground about 4.0m away from the excavation zone. Minor
damages also occurred to the adjacent properties due to soil
movement.
Pile Damages - More than 50 % of piles were tested with

sonic integrity test.
Figure 2.1. A Typical Scene showing excavation works
Table 3 Summary of Pile damage Detected by Sonic
Case II
Integrity Test - Case II.
Pile No. of Piles with Defect
Dia. Least Less Prominent Depth CASE III SHEETPILE EXCAVATION WITH ONE
Prominent Prominent Crack below LEVEL TEMPORARY SUPPORT (Yannawa)
(m) crack Crack GL (m)
I-30 - 49 14 6.2-8.0 Subsoil Condition - In this project site, a 14-15m thick soft
1.0 1 23 4 13-16.6 clay layer occurs with unconfined shear strength (Su)
1.5 - 1 1 15.6 ranging from 1.0 t/m2 to 2.00 t/m2 at the top and bottom of
the layer respectively. Below is stiff to very stiff clay layers
Large numbers of piles with crack detected were located in extend down to about 42m in depth. The subsoil condition
the periphery of the excavation zone (within the soil berm) . below 36m from the ground level is variable, having a 13m
thick sand layer sporadically. A stiff to hard dark grey clay
is found at the depth between 40m and 49m. Generally a
thick dense sand layer occurs below 50m in depth.

diameters 1.0m and 1.5m were installed with pile tip at
about 55m below the ground, being embedded in dense sand
layer to support a building structure. Cutoff level of bored
piles ranges from 0.75m to 7.80m in depth. Entire length of
all piles are fully reinforced with steel reinforcement of
0.5% of pile sectional area.
Soil Retaining System - Due to different cutoff levels of

piles, sheet pile (14m) with one level of temporary bracing
at about 1m below the ground level was used for each
excavation zone. Figure 3 shows a portion of pile layout
Figure 2. Location of Damaged Piles (only a quarter of
which includes an excavation zone with braced sheet pile
Project Site shown here) Case II
wall. Prior to installation of temporary struts, excavation
was carried out to about 2.5 to 3m in depth. The over
Figure 2 shows a portion of the piling plan affected by soil
excavation without support caused excessive movements of
movements. Pile deviations of up to 600mm, particularly in
soil and sheet pile wall along grid line C10 (Fig.3).
pile with prominent crack towards excavation were
Deflected sheet piles were re-aligned and struts were
observed.
installed immediately with pre-loading.
Bored piles with crack were cored to the depth below the
Pile Damages - All piles were tested with sonic integrity
crack and grouted. High strain dynamic load test was
test after pile were trimmed to the designed cutoff level.
carried out on two piles suspected with crack and large
Piles with crack detected were mostly located in the vicinity
horizontal deviation after remedial work. The dynamic test
of the excavation boundary (Fig. 3 and Ref. 6) . Moreover,
results indicate that these piles could be capable of carrying
lateral pile deviation in a range of 100-840mm towards
the design load with factor of safety higher than 1.5.
excavation were also observed in those piles located in the
vicinity of the sheet pile wall ( Fig. 3).
94
440
Table 4. Summary of Pile damage Detected by Sonic
Integrity Test - Case III.
Dia. Least Less Depth
Prominent Prominent Prominent below
(m) crack Crack Crack GL (m)
1.0 - - 5 7.5-9.4
1.5 - - 14 8.4-21.0
Damaged piles were cored to the depth below the crack and
grouted. Modification of reinforcement for the mat was
done for defect piles with large horizontal deviation.
Figure 4. Plan of Excavation and Location of Damaged

Piles (a Portion of Project Area) Case IV
Pile Damages - All piles were tested with sonic integrity

test. In this particular project, 84% of piles within the
excavation zone were found to be cracked (Fig. 4 ).
Table 5. Summary of Pile damage Detected by Sonic

Integrity Test - Case IV.
Dia. Least Less Prominent Depth
Prominent Prominent Crack below
(m) Crack Crack GL (m)
Figure 3. Plan of Excavation and Location of Damaged
0.6 - 38 33 6.5-13.8
Piles and Pile Movements (a Portion of Project Area) Case
III
CASE IV - SHEETPILE EXCAVATION WITH TWO

LEVEL TEMPORARY SUPPORT (Sukhumvit Area)
Subsoil Condition - A 14-15m thick soft clay layer occurs Figure 5. Sonic Integrity
with unconfined shear strength (Su), ranging from 1.0 t/m2 Test Records showing
at the top to 3.00 t/m2 at the bottom of the layer. The natural Good pile, Pile with less
water content of soft clay layer is generally about 37 - 87%. Prominent Crack and
Below is very thick stiff to very hard clay layers inter- Prominent Crack
bedding with a 3m thick medium clay layer at the depth of respectively - Case IV.
19m.
Five bored piles with crack suspected were cored to verify
the crack and grouted later. High strain dynamic load test
0.6m diameter were installed with pile tip at about 26m
was also carried on a total of 25 piles in which 24 piles with
below the ground to support a residential apartment. Dry
crack and 1 pile without crack. Summary of dynamic load
process was employed to construct the piles. Cutoff level of
test results is presented in the Table 5.
bored piles ranges from 0.85m to 5.95m in depth. Entire
length of all piles are fully reinforced with steel
Table 6. Dynamic Load Test Results of Damaged Piles -
reinforcement of 0.35% of pile sectional area.
Case IV
Soil Retaining System - Type FSP III sheet pile (14m) with Mobilized Capacity/Design Ultimate No. of
two level of temporary bracing at about 1.0m and 3.5m Capacity of 240 ton Pile
below the ground level was used. Figure 4 shows a portion less than 50% 1
of pile layout which includes an excavation zone with less than 10% 2
braced sheet pile wall. More than 100% 22
95
441
From the results, only a few piles lead to low safety factor to 17.00
Su (t/m 2) SPT (blows/ft)
carry the designed load and also connected by mat.

0 1.5 2.5 5
0.0 1.0 2.0 3.0 4.0 0 20 40 60
0.0
Crust
Modification of reinforcement for the mat was done

according to redistributed load. 5.0
Sheet pile
Soft CLAY
10.0
FINITE ELEMENT ANALYSIS WITH COMPUTER

MODEL SIMULATION 15.0
Medium CLAY
Very Stiff
CLAY
Two dimensional finite element computer modeling was 20.0

Medium CLAY
carried out using PLAXIS computer program (ref. 7) for

simulation of staged excavation process, pile/soil 25.0
Hard CLAY
movements and to analyze the associated bending stress in

the piles of the above projects. Stiffness of foundation piles 30.0
per linear meter (E.ΣI/pile spacing in row) was used as a Figure 8. Model Profile of Case III
parameter for plate elements, where E and I is Young
Modulus and moment inertia of pile respectively. The soil 40.50 9.50
Su (t/m 2) SPT (Blows/ft)
between piles in a row is ignored. Mohr-Coulomb

1
0 3 5
0 CL 0. 2. 4. 6.
0
2 4
0. 0 0 0 0 0 0
0
constitutive model is adopted with soil parameters derived Crust
from the soil borehole data at the site. A plane-strain

5. Sheet pile
0
Soft to Vsoft
analysis approach was adopted with undrained soil 10.

0
CLAY
condition. 15.
Medium CLAY
0
Model profiles are presented in Figures 6 to 9. Figure no. 20.

0
Stiff to Vstiff
CLAY
10 shows computed bending moments in pile and lateral 25.

0
deflection of pile in Case IV.

Figure 9. Model Profile of Case IV
Center
Line
DISCUSSION
6.00 32.00 9.00 18.00 Su (t/m2) SPT (Blows/ft)
0 3 5 10
0 4.0 8.0 0 50 100
Slope
0.0
1:3 Slope Crust
1:3 Mat Zone
5.0
10.0
Soft CLAY The modeling results indicate that bending moments in piles
15.0
Medium
induced by excavation are higher than the cracking moment
20.0
CLAY
Stiff to Very capacity of the piles in all cases. Summary of analysis is
presented in the Table below;
Stiff CLAY
25.0
Dense to
30.0 Very Dense
SAND
35.0
Table 6. Moment Capacity of Piles and Excavation Induced
Moments in the Piles Closer to Excavation Boundary.
Figure 6. Model Profile of Case I
Ultimate Moment
Su (t/m2) SPT (Blows/ft)
Case Pile Rebar Cracking Moment by FEM
0.0
0 3 5 10 Weathered
0.0 2.5 5.0 0 20 40 60
Size Moment (Whitney) Model
CLAY
(m) (%) (t.m) (t.m) (t.m)
I 1.0 1.0 30 100 86.4
5.0 MAT FOOTING
Soft to Very
Sheet pile
Soft CLAY
10.0
109.5*
14.0
II 1.0 0.75 30 79 149.7
III 1.5 0.5 102 188 117.2
Medium
15.0
CLAY
20.0
255.7*
Stiff to Hard
CLAY IV 0.6 0.35 6.5 9.2 13.3
25.0
Note: 1. (*) moment in pile located in active side of soil
mass either outside of sheet pile wall or cut slope.
Figure 7. Model Profile of Case II
In Cases I and III, piles located in the active side of soil
mass are found to be under the bending stress about 10%
and 36% higher than the ultimate bending capacity of piles
respectively.
In Case II, the computed bending moments in the pile

closest to the sheet pile wall and piles in second row is 89%
and 27% respectively higher than the ultimate bending
96
442
capacity of piles. The sonic integrity test also detected Bending Moment (T-m) Displacement (mm)
presence of crack in 35% of the piles which are located in -20

0.00
-10 0 10 20 40 20 0
first two outer rows (see Fig. 2).
In Case IV, the modeling results show that excavation 5.00
induced bending moments in pile exceed the ultimate

moment capacity of piles. This is confirmed by the
observed pile damages (see Fig. 5). Moreover the model 10.00
results suggest that these piles are subject to some degrees
Depth (m)
of basal heave due to flexibility of and soil condition at Cracking
Moment of Pile
embedment of sheet pile wall. Sonic integrity test results 15.00

Figure 10. Computed
shows presence of more than one crack in some piles at Interface
between Soft Bending Moment in
depths, suggesting piles would have experienced combined Pile and Pile’s
and Stiff Clay
stresses. Comparison between Case III and Case IV 20.00

Deflection Closest to
illustrates that cracks are found more frequently in small Sheet Pile Wall - Case
piles than large piles in a case of soil movements due to 25.00
IV
excavation.
The crack location in piles indicated by sonic integrity test

generally is found to be consistent with the location of
computed bending moment which exceeds the cracking
moment capacity of piles for all cases.
Table 7. Locations of Crack and Computed Bending

Moments in Piles in the Vicinity of Excavation Boundary
Integrity Test Depth of B.M
Depth of Crack Depth of by FEM
Case in Piles Max. B.M exceeding
near Excavation by FEM Pile’s Cracking
Boundary Moment
(m) (m) (m)
I 6.0 & 13.8 19.5 13.5 Figure 11. Section Showing Displacements of Soil - Case
IV
II 12.7 & 13.7 18.0 11.0 & 13.0
III 12.4 &19.0 15.0 12.8 & 14.5
Instrumentation such as installation of inclinometer and
IV 11.5 & 12.0 15.5 13.0 &14.0 strain gauges in piles and active side of soil mass is
suggested for the future projects to analyze behaviour and
FEM analysis indicates the maximum bending moment in performance of piles subject to lateral loads induced by
piles occurs at the level close to the interface between soft excavation.
clay and stiff clay layers. For lateral movements of piles
and soil, a direct comparison can not be made between
model and observed data as the accurate field monitoring CONCLUSION
data are not available. Performance of soil retaining
systems by model analysis are summarized in the Table This paper presents four case studies on bored piles
below; damaged by excavation induced lateral soil movements in
Bangkok soft clay layer. Performance and response of pile
Table 8. Summary of Lateral Movements of Retaining are studied by FEM modeling and confirmed with pile
System integrity test results. Causes of pile damages found are also
Type of Level Excav. Max. Lateral discussed.
Case Retaining of Depth Movement
(m) Strut (m) (mm) In addition to FEM analysis, the followings were found
I Cut Slope - 5.4 117 from the actual site construction activities,
II Sheet Pile 1 5.7 175 1. Over excavation prior to installation of support
III Sheet Pile 1 8.0 56 2. Inadequate retaining system which can effectively
IV Sheet Pile 2 6.0 52 control soil movement.
3. Inadequate support for construction traffic load (Ref. 4)
For Case III, model analysis also confirmed that without 4. Improper pile trimming method (Ref. 4)
temporary bracing, sheet pile would fail when soil berm is 5. Inadequate pile design for lateral load during
removed for trimming pile adjacent to the wall. construction
6. Inadequate reinforcement in piles to resist soil heave
97
443
In designing of bored piles, the above mentioned causes, to New Delhi 1936-1994), Bangkok, Thailand, pp 5.31 -
including deviation of piles due to excavation induced soil 5.42
movement are to be taken into account.
4. Sliwinski Z. J. and Fleming G.K. [1983]. “The Integrity
and Performance of Bored Piles” Advances in Piling and
ACKNOWLEDGEMENT Ground Treatment for Foundations, London, UK, pp 153-
165.
especially to Mr. Ganeshan Baskaran, Mr. Muhammad 5. Phienwej N and Kanok-nukulchai W. [1989].
Ashfaq Anwar, Mr. Tossaporn Sree-iam and Mr. Chetsadar “Excavation Induced Pile Movement and Damage - A Case
Plongkrathok for their assistance in the preparation of this History in Bangkok, Piletalk - International 1989, Kuala
paper. Lumpur, Malaysia. pp 234-244.
6. Chiruppapa P. [1995]. “A Case Study of Pile Failure

REFERENCES Created by Excessive Lateral Deflection of Braced Sheet
Pile Wall During Basement Excavation’. Annual Technical
1. Phienwej N., Balasubramaniam A.S., and Sutabutr T. Conference 1995, Engineering Institute of Thailand. pp CE-
[1996]. “Performance of Retaining Structures for Deep 81 - CE94.
Excavation in Bangkok Soft Clay” Short Course on
Deformation of Soils & Rocks and Displacement of 7. PLAXIS Manual [1995]. “PLAXIS - Finite Element
Structures, Ground Movment in Deep Excavation - Case Code for Soil and Roack Analyses” Version 6, Edited by
Histories in Bangkok Subsoil, Bangkok, Thailand. pp.14-20. Vermeer P.A. and Brinkgreve R.B.J.
2. Phienwej N., Akewantop, K. and Balasubramaniam 8. Poulos H.G. and Chen L.T. [1997]. “Pile Response due
A.S., [1996]. “Comparative Evaluation of Ground to Excavation - Induced Lateral Soil Movement”, Deep
Movements Associated with Braced Excavation in Foundations, Excavations, Ground Improvements &
Bangkok Soft Soil” Short Course on Deformation of Soils & Tunneling, 30th Year Anniversary Symposium of The
Rocks and Displacement of Structures, Ground Movement Southeast Asian Geotechnical Society, Bangkok, Thailand.
in Deep Excavation - Case Histories in Bangkok Subsoil, pp 5.24-5.47.
Bangkok, Thailand. pp.21-35.
9. Nilson, A.H. and Winter G. [1991]. “Design of Concrete
3. Teparaksa W. [1994]. “Prediction of Movement and Structures, eleventh edition”, McGraw -Hill, Inc.
Behavior of Sheet Pile Wall for Deep Braced Cut
Excavation in Soft Bangkok Clay”. Symposium on
Developments in Geotechnical Engineering (From Havard
98
444
Proceedings of the 7th International Conference on Application of
Stress-Wave Theory to Piles
9-11 August 2004, Kuala Lumpur, Malaysia
7th International Conference

on Application of Stress-
Wave Theory to Piles
Quantifying Pile Head Condition before

Basement Excavation by Cross-Hole Sonic
Logging Tests
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak
Natamon Kampananon
STS Instruments Co., Ltd.
445
QUANTIFYING PILE HEAD CONDITION BEFORE BASEMENT
EXCAVATION BY CROSS-HOLE SONIC LOGGING TESTS
Narong Thasnanipan1, Zaw Zaw Aye2, Thayanan Boonyarak3

and Natamon Kampananon4
ABSTRACT
In wet process bored piles construction, it is necessary that concrete is continued to be placed until good
quality of concrete is formed above the design pile cut-off level to avoid contamination of slurry within
the design pile length. Controlling the excessive overcast length to maintain the appropriate pile top level
is one of the major problems particularly for the piles with low cut-off level. It is also impossible to check
the quality of concrete at the top part of the piles prior to exposing the pile head for pile cap construction
or basement excavation. If the pile head condition of recently cast bored piles could be justified in the
initial stage of the project, it would be better and easier for the piling engineers to efficiently decide the
optimum overcast length of the pile which in turn would provide effective saving of unnecessary concrete
consumption. Application of sonic logging tests in justifying overcast pile length is demonstrated in this
paper.
Keywords: cross-hole sonic logging test, wet-processed bored piles, cut-off level, contaminated
concrete, pile integrity
1. INTRODUCTION
The first sonic logging test in Thailand was believed to be conducted in 1982 for the wet-processed bored
piles of the Memorial Bridge Project where bad concrete zones were detected at two levels, depth about
20m and 1m from the pile top as reported by Ng (1983). In the past, sonic logging test was only employed
as a pre-planed site quality control testing for large infrastructure projects in Bangkok. The primary reason
of using sonic logging test in only large scale projects was due to its relatively expensive cost since cost of
testing the entire length of deep-seated bored piles are high – higher costs involved in installation of
lengthy access tubes and the test itself. With the passage of time, the application of this quality control
test is being extended. As interpretation skills relevant to local soil condition and construction method of
this test have been significantly improved in local industry, the piling engineers begins to use sonic
logging tests for further assessment of constructed bored piles.
2. WET-PROCESSED BORED PILE CONSTRUCTION METHOD
Due to the prevailing subsoil and groundwater conditions, deep-seated bored piles of toe depth over 24m
are constructed by wet-process or slurry displacement method. In the rotary drilling method, a temporary
casing of appropriate length (12 to 18m is commonly used in Bangkok depending on the thickness of soft
clay) with required diameter (internal diameter not less than that of design bored pile diameter) is first
installed to ensure the stability of the borehole in the top soft or loose soil layers.
1
President, 2Senior Vice President, 3Geotechnical Engineer SEAFCO Public Co., Ltd. ,
Bangkok, Thailand, seafco@seafco.co.th
4
Senior Geotechnical Engineer, STS Instruments Co., Ltd., Bangkok, Thailand
447
Drilling is commenced by auger to drill out the soil inside the temporary casing. Auger-drilling is
commonly continued up to the top of the first water-bearing sand layer or bottom of the casing. Drilling
slurry or supporting fluid is then supplied to the borehole and drilling is proceeded with a bucket down to
the design final depth of the pile. Before lowering the reinforcement cage, a special cleaning bucket is
used to clean the base of borehole. If bentonite slurry is used, recycling method by air-lift or pump is
applied as the base cleaning process. Reinforcement cages are then lowered into the borehole and
concreting is carried out by tremie method.
3. CLASSIFICATION OF THE SONIC LOGGING TEST ANOMALY
Based on the experience of testing specialist on over 3000 deep-seated bored piles in past 15 years in
Thailand coupled with the comparison between sonic logging test results and direct examination from
cored samples of zones detected with anomaly, the category and criteria for interpretation of cross-hole
sonic logging test results have been established as shown in Table 1. Figure 1 depicted the sonic signal
profile of 4 categories classified in Table 1.
Table 1. Category of sonic logging signals and criteria

Category Criteria Integrity
A No or slight increase in transmission time not more than 10% Homogeneous concrete
B Minor increase in transmission time >10% but <25% Minor anomaly
C Transmission time more than Category B but transmission Major anomaly
time can still be defined
D Transmission time can not be defined Defect
Figure 1. Sonic profile showing different category (A, B, C and D) as classified in Table 1
4. PILE HEAD CASTING LEVEL OF WET-PROCESSED BORED PILES IN BANGKOK
In order to ensure the sound concrete at the specified pile cut-off level, wet-process bored piles are
commonly cast well above the cut-off level. Though most of the piling engineers wish to minimize the
overcast length, their primary concern is a formation of contaminated concrete at the design pile cut-off
level which would need the piling contractor to come back to the site to carry out the remedial work while
the superstructure contractor has commenced his work. There could be a claim from the superstructure
448
contractor for having delay to rectify the imperfect pile heads. Hence, most of the bored piling engineers
prefer to cast the pile significantly higher than specified pile top level as shown in Figure 2.
Figure 2. View of exposed diameter 1.20m bored piles for mat foundation - Most of the piles were
constructed with excessive overcast length (over 4m above design pile cut-off level)
The need of using the long temporary casing in Bangkok subsoil also makes it difficulty to maintain the
top of overcast at the desired level. Table 2 shows the casting tolerance above cut-off levels according to
The Institution of Civil Engineers (1996) United Kingdom.
Table 2. Casting tolerance above cut-off levels for wet-processed bored piles (ICE, UK, 1996)
Cut-off level below Casting tolerance above cut-off level, m Condition
commencing surface, H, m
0.15 – 10.00 1.0 + H/12 + C/8 Piles cast under water or
Where C = length of temporary casing support fluid
below the commencing surface
5. COMMON DEPTH OF ANOMALY FORMED BY CONTAMINATED CONCRETE IN

BORED PILES
As construction method and quality control process for wet process bored piles are being well established
in local construction industry of Thailand, defect (discontinuity or contaminated concrete) in lower part of
pile shaft rarely occurs in these days. Hence investigating the top part of pile is more important than
testing the entire length of the deep-seated piles in Bangkok subsoil. Moreover, if sonic access tubes are
installed only in the top part of bored piles, the costs associated with both the material and test process
itself would be significantly reduced and integrity testing by sonic logging method would be more
feasible. According to the authors’ experience from over 5,000 exposed bored piles in Bangkok subsoil, to
achieve good pile head condition within design cut-off level, bored piles constructed under polymer-based
slurry require shorter overcast lengths than those constructed under pure bentonite slurry. In authors’
opinion, the tolerable overcast lengths of bored piles constructed under polymer-based slurry should be
less than those specified in Table 2. Two cases were studied to determine the common depth of significant
anomaly (Category C or D) of bored piles in Bangkok subsoil.
5.1 Case 1
According to the sonic logging test results obtained from 400 piles of diameter 1.50m with depth over
50m constructed under polymer-based slurry in Bangkok, only 2 piles (0.5% of total contract piles) were
detected with C Category in the design pile length. The level of the detected C category for these 2 piles
449
was found to be within 2m below cut-off level of the pile. The overcast length above cut-off level of 95%
of piles were within 2.5m.
5.2 Case 2
The sonic logging test conducted on 174 bored piles of diameter 1.0m with depth over 50m constructed
under polymer-based slurry reveals that no pile was detected with C category below the cut-off level. The
overcast length of 76% of piles were within 2.2m above design cut-off level.
Table 3. Summary of sonic test results from 2 cases – bored piles constructed under polymer-based slurry
in Bangkok
Case Total no. Pile diameter Number of piles detected with C Maximum overcast
of piles / depth category below design cut-off level length of majority of pile
by sonic logging test from sonic logging test
(pile) (%) result
Case 1 400 1.50m / 52m 2 0.5 2.5m (95%)
Case 2 174 1.0 m / 50m 0 0 2.2m (76%)
This finding suggests that contaminated concrete rarely occur below the cut-off level if overcast length of
the pile is about 2.5m for polymer-based bored piles. Quality control of the construction process such as
slurry mixing process and composition, properties of slurry, base cleaning, maintaining adequate tremie
length, concrete quality and concreting procedure however must be according to the standard practice to
achieve this result. Practical application of polymer-based slurry for deep-seated wet-processed bored
piles in multi-layered soil of Bangkok has been reported by Thasnanipan et. al. (2003).
Contaminated
concrete only
found at top 1m
Figure 3. View of polymer-based bored pile diameter 1.8m with overcast length about 2m – contaminated
concrete was found only at the pile top
6. QUANTIFYING THE PILE HEAD CONDITION BY SONIC LOGGING TEST PRIOR TO

BASEMENT EXCAVATION
Designs of the modern buildings in Bangkok frequently call for the basement facility which require design
cut-off level of the piles well below the existing ground level. In wet process bored piles construction, it is
necessary that concrete is continued to be placed until good quality of concrete is formed above the design
pile cut-off level to avoid contamination of slurry within the design pile length. Controlling the excessive
overcast length to maintain the appropriate pile top level is one of the most difficult task particularly for
the piles with low cut-off level. Checking the condition of concrete at the top part of the piles prior to
exposing the pile head for pile cap construction or basement excavation by direct inspection is impossible.
450
If the pile head condition of recently cast bored piles could be justified in the initial stage of the project, it
would be better and easier for the piling engineers to efficiently decide the optimum overcast length of the
pile which in turn would provide effective saving of unnecessary concrete consumption. Cross-hole sonic
logging test as an early indicator may be the ideal method to check the optimum overcast length of
recently constructed bored piles in controlling the excessive overcast length.
Sonic logging test could also be used as a performance-indicator for the piling contractor. In one major
project, during the excavation works for 8 to 10m deep basement construction, the main contractor
claimed that over 50% of exposed bored piles from total 174 contract piles were found to be with
excessive overcast length (over 4 to 6m from design pile cut-off level). All bored piles in this project were
constructed under polymer-based slurry and sonic logging test was conducted on every pile. The piling
contractor carefully reviewed all the sonic logging test results and found the followings;
• 76% of piles were within 2.2m overcast length
• 18% of pile were with overcast length between 2.2m and 3m
• Only 10 piles (5%) exceeded overcast length of 3m.
Over 4m to 6m overcast length claimed by the main contractor was caused by the solid-like mixture of
backfill material mixed with highly contaminated concrete at the top of the piles as boreholes were
backfilled with crushed-rocks. Figure 4 shows the typical sonic logging test results of one of the piles
claimed by the main contractor as excessively overcast piles. As can be seen in the figure, homogenous
concrete was detected only up to 2m above design cut-off level whereas non-homogeneous (Category C or
D) was observed above 2m. Field observation also confirmed that majority of piles were with highly
contaminated at top 2m above design cut-off level. The main contractor finally accepted the proof of sonic
logging test result to justify the overcast length.
Level of overcast pile

top claimed by main
contractor - 4.5m
Top level of homogeneous
concrete -6.8m
Design cut-off
level -8.9m
Figure 4. Typical sonic logging test profile of bored pile claimed as excessively overcast (overcast length
4.4m above cut-off level) by the main contractor
7. RECOMMENDED PROCEDURE OF QUANTIFYING THE PILE HEAD CONDITION BY

SONIC LOGGING TEST
As presented in earlier section, if the construction method strictly follows the quality assurance and
control procedure (QA & QC), contaminated concrete is rarely formed within the design bored pile length.
It should be sufficient to test the pile integrity of the top 10 to 15m by sonic logging test so that cost
associated with both access tubes and the test itself can be significantly reduced. It is also not necessary to
cast the concrete with excessive overcast length as most of the modern bored piles are constructed under
polymer-based slurry. Followings are the recommended process in minimizing the formation of
contaminated concrete within the design pile length for wet-processed bored piles in Bangkok subsoil.
451
• Carefully check and use the required temporary casing length considering the thickness of soft
clay layer (common thickness of Bangkok soft clay layer is 15 to 18m)
• Pre-install the sonic logging access tubes covering only the top 15m for majority of the piles.
Small percentage of piles may have full length of sonic access tubes to the pile tip to randomly
check the integrity of the entire pile shaft
• Check the borehole profile of each pile by drilling monitor (koden equipment) to roughly assess
the required concrete volume (to check the size and shape of borehole by drilling monitor to
estimate required concrete volume)
• Strictly follow and control the slurry mixing process, slurry properties and quality check test
• Use appropriate concrete mix with good quality and strictly follow and control the tremie
concreting process
• Carefully withdraw the temporary casing
• Construct the pile of higher design cut-off level first (if possible) and perform the sonic logging
test 10 days after concreting
• If possible (for high cut-off level piles or cut-off level within 3m below existing ground)
excavate with sufficient support to visually inspect the pile head and compare with sonic logging
test. This process may not be necessary though it will be very useful to assess the sonic logging
test result to correlate with actual condition of pile top at early stage of pile construction
• The piling engineer must carefully review the borehole profile, actually poured concrete volume
and sonic logging test of each completed pile to assess and justify the construction method, site
characteristic and quality assurance procedure for further improvement
8. CONCLUSION
Cross-hole sonic logging test is a useful quality control test to examine the integrity of the wet-process
bored piles in early stage of the project. Discontinuity in the form of contaminated concrete in the lower
part of pile shaft rarely occurs in these days as construction method and quality control process for wet
process bored piles have been well established in local construction industry in Thailand,. Therefore, it is
recommended that sonic logging test should be mainly used to investigate the integrity of the pile at the
top portion and a random check is performed only for few piles to examine the integrity of entire pile
shaft. An extended application of cross-hole sonic logging test in quantifying pile head condition prior to
exposing the pile has been presented. The piling engineers should pay more attention to control the
excessive overcast length to minimize the cost associated with cutting extra length of pile head and
disposal of the waste. The piling engineers should be trained to effectively apply the sonic logging method
not only as integrity-proof test but also as a tool to quantify the pile head condition at the early stage of the
piling work.
REFERENCES
Institution of Civil Engineers, 1996. Specifications for piling and embedded retaining walls, The
Institution of Civil Engineers and Crown, Thomas Telford Services Ltd. London, United Kingdom
Ng, K. C. 1983. The construction problem and performance of large bored piles in second sand layer, M.
Eng. Thesis No. GT 82-26, Asian Institute of Technology, Bangkok, Thailand
STS Instruments Company Limited, 2004. Proposal for sonic logging test (unpublished), Method
Statement for Performing and Interpretation of cross-hole sonic logging test submitted to SEAFCO Public
Company Limited, Bangkok, Thailand
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak, 2003. Behaviour of Polymer-based Slurry for
Deep-seated Bored Piles in Multi-layered Soil of Bangkok, Proceedings of the Fourth International
Geotechnical Seminar on Deep Foundation on Bored and Auger Piles, Ghent, Belgium.
452
Proceedings of the 29th Annual Conference on Deep Foundations
29 September – 1 October 2004, Vancouver, British Columbia,

Canada
Emerging Technologies
Advantages and Limitation of Integrity Tests

for Large Diameter Bored Piles in Bangkok
Subsoil
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak
Natamon Kampananon
STS Instrument Co., Ltd.
453
ADVANTAGES AND LIMITATION OF INTEGRITY TESTS FOR LARGE DIAMETER
BORED PILES IN BANGKOK SUBSOIL
Narong Thasnanipan, Zaw Zaw Aye, Thayanan Boonyarak, Seafco Public Co., Ltd., Bangkok, Thailand
Natamon Kampananon, STS Instruments Co., Ltd., Bangkok, Thailand
In Bangkok due to the prevailing condition of subsoil (existence of considerably weak

stratum at top part) deep-seated large diameter bored piles of over 50 m depth with 0.8 -
1.8m in diameter have been mainly used for heavy structures such as high rise buildings,
elevated expressways, flyovers and more recently for underground subway stations of the
first Bangkok MRTA system. As a part of pre-planned quality assurance regime and
retrospective investigation, integrity of the piles are tested to obtain the information with
regard to the potential deficiencies of the constructed piles which may have formed during
pile construction process or may have been attributed by other activities after construction
of the piles. An overview of three commonly used non-destructive tests such as sonic
integrity, sonic logging and high strain dynamic load tests on deep-seated piles in Bangkok
subsoil is presented in this paper. Advantages and limitations of each integrity test are
discussed.
INTRODUCTION
Due to the prevailing condition of subsoil deep- 0 0
Weathered Crust
seated large diameter bored piles of over 50 m
depth with 0.8 - 1.8m in diameter are mainly Bangkok Soft Su= 10-25 KPa
γ t = 14-16kN/m 3
Clay
used for heavy structures such as high rise 10 -
10
buildings, elevated expressways, flyovers or
overpass bridges and more recently for
Medium to
underground subway stations of the first 20 Su= 50-14020- KPa
Depth (m)
Stiff Clay γ t = 17-21kN/m 3

Bangkok MRTA system. As a part of pre-
planned quality assurance regime and Act
Medium ual
retrospective investigation, integrity of the piles 30 Dense Sand SPT-N= 20-40
-
30 Pie
Hyd
are tested to obtain the information with regard (First Sand)
γ t = 20kN/m 3 zom
rost
to the potential deficiencies of the constructed
etric
piles which may have formed during actual pile 40 atic

-
Hard Clay Su> 200 KPa
Dra
40
γ = 21KN/m 3 Line
construction process or may have been t
wdo
attributed by other activities after construction of

wn
Dense to Very
the piles. Three techniques namely sonic
-
50
Line
50
integrity, sonic logging and high strain dynamic (Second Sand) γ t = 21KN/m 3
load tests are the commonly used non-
destructive integrity tests applied in Bangkok, 60 -
60
Thailand. The effectiveness and suitability in Figure 1. Typical soil profile of Bangkok with
terms of both cost and method of application piezometric drawdown condition (Thasnanipan,
itself are the primary reasons of the greatest et. al., 2002)
growth in use of these tests.
SUBSOIL AND EXISTING PIEZOMETRIC Existence of soft to very soft highly compressible
PROFILE dark gray marine clay makes difficulties in foun-
dation construction works. This notorious soil
A typical subsoil profile is relatively consistent in layer lies beneath made-ground in most urban
different localities in Bangkok. It is characterized area and in some areas it lies under weathered
by alternating layers of clay and sand deposits of crust layers of 2m thick. Depending on the loca-
Quaternary age extending down to about 550m tion, this layer is extended up to 12-18m. About
depth where bedrock generally exists. 2m thick Medium Clay layer can be observed
455
between Soft Clay and underlying Stiff Clay. soil/slurry inclusions. The major causes of these
Generally Stiff Clay layer occurs directly under- defects are;
neath Medium Clay and its depth goes up to
22m. Below Stiff Clay layer, First Sand layer 5- • Inadequate length to protect the soft clay
8m in thickness can be found. This First Sand layer (length of temporary casing)
layer, however, is absent in some areas. Stiff to • Delayed feeding of slurry and improper slurry
Hard Clay layer underlies First Sand and it is level maintenance for wet process
found to be about 5m thick. Second Sand layer • Long time duration of maintaining open
generally occurs at depths between 45 to 65m. borehole
Typical subsoil profile and the present piezomet- • Non continuous concrete pouring or
ric drawdown condition of Bangkok up to 60m disruption in concreting
depth are presented in Figure 1. • Lifting up of insufficient workable concrete or
hardened concrete upon extracting the
CONSTRUCTION METHOD temporary casing
• Contamination of concrete with drilling slurry
Due to the prevailing subsoil and groundwater or soil
conditions, deep-seated bored piles of toe depth • Soft pile toe due to improper base cleaning
over 24m are constructed by wet-process or
slurry displacement method using rotary drilling Most of the defects or cracks in bored piles
machine. A temporary casing of appropriate caused at post construction stage are usually
length (12 to 18m in Bangkok depending on the induced by construction activities associated
thickness of soft clay) with required diameter (in- with adjacent excavation work with inadequate
ternal diameter not less than that of design earth retaining system and improper trimming of
bored pile diameter) is first installed to ensure pile head to design cut-off level.
the stability of the borehole in the top soft or
loose soil layers. In some projects where the vi- Thasnanipan et al. (1998b) reported that integrity
bration is strictly limited, a standard length of of 285 bored piles (3.3 % of 8,689) were found to
casing with oscillator or short temporary casing be of doubtful quality according to an
of 5 to 8 m is pushed down in combination with assessment on the results of sonic integrity test
the pre-boring process. Drilling is commenced by with the additional information obtained from the
auger to drill out the soil inside the temporary construction records of bored piles in Bangkok
casing. Auger-drilling is commonly continued up subsoil. The results of the findings are
to the top of the first water-bearing sand layer or summarized in Table 1. As can be seen in the
bottom of the casing when using the short casing table, higher percentage of defects is caused by
method. Drilling slurry or supporting fluid is then post-pile-construction activities and more
supplied to the borehole and drilling is pro- commonly by movement of surrounding soils
ceeded with a bucket down to the design final induced by excavation works in the vicinity of the
depth of the pile. Before lowering the reinforce- piles.
ment cage, a special cleaning bucket is used to
clean the base of borehole. If bentonite slurry is Table 1. Summary of pile assessment on 8,689
used, recycling method by air-lift or pump is ap- bored piles (Thasnanipan et al. 1998b)
plied as the base cleaning process. Reinforce-
ment cages are then lowered into the borehole Type of Defect Causes %
and concreting is carried out by tremie method.
Poor concrete Cutoff level near the 0.1
OCCURENCE OF POTENTIAL DEFECTS IN at pile top (0- ground or inadequate
DEEP-SEATED WET-PROCESSED BORED 3m) overcast length
PILE Insufficient casing length
Size reduction 1.0
or soft clay layer varia-
In general, pile defects can be caused at two tion in thickness
stages, during the pile construction process and
after construction of the piles (post-pile- Excavation works for
construction). Pile defects that may arise during Cracks/discont pile trimming and con- 2.2
the construction of wet-processed bored pile are inuities struction activities
size reduction/necking, discontinuity, and
456
SELECTION OF SUITABLE TEST METHOD defect detected by other integrity tests. Table 2
shows the suitability of test method for different
Turner (1997) highlighted the factors in selection type of common defects.
of pile testing in CIRIA Report 144 as follows;
• The perceived nature of possible features or SONIC INTEGRITY TEST
defects within the pile
• The ability of the test method to detect the Sonic integrity test, also known seismic test is
feature or defect under investigation the most common method of integrity testing for
• The cost of testing and examination both driven and bored piles in Bangkok. In many
• The ease of use and interpretation projects it is a part of the contractual requirement
to conduct sonic integrity test. Minimum 10 % to
Like other indirect measurement of pile integrity maximum 100 % of production piles are
testing methods, sonic integrity and sonic commonly tested. It is also a reasonably
logging tests are only capable of identifying acceptable method for applying as a
structurally significant features. Rejection or retrospective investigation in determining
acceptance of individual piles should not rely integrity of the pile.
only on the results of these tests. Further
investigation, engineering evaluation and Features Detectable by Sonic Integrity Test
judgment are highly recommended to confirm for Deep-seated Bored Piles
the defect detected by these indirect techniques.
Both methods however are very useful for cost- Following negative features of deep-seated
effective screening test to identify piles with bored piles can be commonly detected by sonic
potential defect. integrity tests.
• Discontinuity / crack
High strain dynamic load test is usually selected • Size reduction or necking
to verify the load carrying capacity of piles. Pile • Discontinuity / Contaminated concrete
integrity can also be determined by high strain The characteristic of sonic integrity signals of
dynamic load test. The major advantages of this above features have been thoroughly reviewed
test in comparison with static pile load test are and presented by Thasnanipan et al. (1998a).
those of cost, time and space requirement. As some features have a similar pattern to each
other, it is always important to check and
Table 2. The suitability of test method for compare the test data of pile in question with
different type of defects. those of other piles within the same job site to
Type of Test method and suitability establish “site signature”. In some cases
defects for defect detection correlations should be made with similar projects
Sonic Sonic High strain where damaged piles were investigated with
firmed evidence (coring or excavation to detect
Integrity logging dynamic
damage level). Interpretation should also be
load test made with sound knowledge of pile construction
Crack Suitable Possible Suitable technique, subsoil condition and other factors
Pile size Suitable Possible Suitable (if (problem encountered during and post pile
change severe de- construction) which may influence the sonic test
fect) signals. Pile construction records are also useful
in interpretation of piles with detected anomaly.
Poor Possible Suitable Suitable (if
Conclusion should be made after careful and
quality severe de- thorough review of test results in conjunction
concrete fect) with other available information. Further
or Con- investigation (if applicable excavate down to
taminated level of detected anomaly for visual inspection)
concrete may require before final conclusion is made.
Soft pile - Not Suitable
toe usually Discontinuity / Crack Detectable by Sonic
Integrity Test
In some projects, dynamic load test was applied
Modern buildings in Bangkok commonly require
to justify the capacity of pile after rectifying its
the excavation works for basement facility and
457
for the use of different elevations of pile
foundation. Excavation works are mainly carried
out by using cut slope, sheet pile wall, Soft Clay Med. Very Stiff Clay
diaphragm wall and secant pile wall. Since most Clay
of the bored piles are designed mainly to carry

an axial load, lateral and tensile force which may 10.5cm/s
impose on pile during basement excavation are
0 3 6 9 12 15 18 21 24 27 30 V2-88c
sometimes not considered or overlooked. Pile 30 4000 m/s f:8 Sr
2 Jul 97 exp: 20
TYPICAL SOIL PROFILE
Figure 3 Pile size reduction or necking detected

SOFT CLAY
by sonic signal
MEDIUM CLAY
STIFF CLAY
Figure 3 shows reduction of pile size at about
14m. According to the construction record the
SAND
cause of the necking was believed to be caused
by use of inadequate 12m long temporary casing
at the particular subsoil condition where the soft
crack induced crack induced crack induced
by large lateral by sliding by uplift to medium clay layer extends up to 16m.
movement failure force
Contamination of concrete or Discontinuity

of Pile Material
Figure 2. Typical features of external forces A discontinuity in pile shaft caused by disruption
causing pile cracks in Bangkok subsoil during concrete pouring can be seen in Figure 4.
Concrete supply was interrupted by poor
Excessive lateral displacement of ground caused coordination between the piling contractor and
by improper excavation naturally induced an concrete suppliers. As previously poured
additional lateral force and bending moment in concrete in the lower pile section had been
the pile. Once the induced bending moment hardened when concreting resumed, the tremie
exceeds the cracking moment capacity, piles are pipe was also blocked. Blockage was cleared
subject to be cracked. Unexpected tensile force after completely pulling out the entire tremie from
acting upon the pile due to an excessive heave borehole. Reinstallation of tremie pipe in the
or uplift force induced by excavation can also borehole filled with bentonite slurry eventually
cause cracking of pile. In Bangkok, pile cracks caused the formation of contaminated layer
caused by external forces are normally found at between old and new concrete.
the level where Soft and Medium to Stiff Clay
boundary is present (typically at 12 to 16m below
the existing ground level) as illustrated in
Figure 2.
14.2 cm/s
Size Reduction or Necking 0 4 8 12 16 20 24 28 32 36 40 V2-88c
Pile 1 4000 m/s f:8 Sr
24 Apr 97 exp: 20
Common causes for size reduction in bored piles
are;
Figure 4. Sonic signal showing a discontinuity at
• Insufficient length of temporary casing to 11.0m caused by interruption in concrete
protect the soft clay pouring - multiple reflections of signal can be
• Delayed feeding of slurry and improper slurry observed from the discontinuity
level maintenance for wet process
• Long time duration of maintaining open bore The detected discontinuity was later rectified by
• Non continuous concrete pouring coring and cement grouting. Figures 5 shows
• Soil conditions such as inclusion of sand the sonic test signal of the same pile after
lenses in clay closing the discontinuity.
458
Based on the experience of testing specialist on
over 3000 deep-seated bored piles in past 15
years in Thailand along with the comparison
7.6 cm/s between sonic logging test results and direct
0 4 8 12 16 20 24 28 32 36 40 V2-88c examination from cored samples of zones
Pile 1 # 4000 m/s f:8 Sr
detected with anomaly, the category and criteria
2 Jun 97 exp: 20
for interpretation of cross-hole sonic logging test
results have been established as shown in the
table below.
Figure 5. Sonic integrity test shows the Table 5 Category of sonic logging results and
discontinuity at 11.0m has been closed by criteria
grouting Category Criteria Integrity
A No or slight increase in Homogeneous
CROSS-HOLE SONIC LOGGING TEST transmission time not concrete
more than 10%
Cross-hole sonic logging test can mainly be ap-
B Minor increase in Minor anomaly
plied as a pre-planed site quality control testing.
transmission time
Use of Sonic logging test for checking pile integ-
>10% but <25%
rity has increased in Thailand in recent years. It
is used for cast-in-place foundation elements C Transmission time Major anomaly
such as bored piles, barrettes, diaphragm walls more than Category B
and caissons. Both PVC and steel tubes can be but transmission time
used as access tubes. Though cost of steel tube can still be defined
is higher than that of PVC, steel tubes are con-
sidered more suitable than PVC tubes for two D Transmission time can Defect
reasons; (1) less potential for bending and dam- not be defined
age especially for deep-seated bored piles and
barrettes with heavy steel reinforcement and (2)
better bonding with concrete. These access
tubes are sometimes also used for pile base
grouting. The access tubes are attached to the
reinforcement cage, which is installed prior to
concreting as a normal practice for cast-in-place
piling. Two access tubes are required as a mini-
mum for cross-hole sonic logging test. For a
good coverage of the test pile, recommended
number of tubes for different sizes of bored piles
is shown in Table 4.
Table 4. Recommended number of tubes for

different pile size
Pile Diameter Minimum No. Tube spacing
(mm) of tubes (degree)
D ≤ 750 2 180
750<D ≤ 1000 3 120
1000<D≤ 1500 4 90
1500< D ≤ 6 60
2500
Figure 6. Sonic profile showing A, B, C and D
2500 < D 8 45 categories
459
Turner (1997) pointed out the need of both 0
theoretical knowledge and practical experience
in interpretation of the test results in CIRIA
Report 144. Suggestion is also made in this -5
report that the anomaly shown in sonic test
Settlem ent (m m )
signals can be caused not only by changing in
physical properties of the materials, but also by -10
factors within the measuring system itself.
These factors are;
-15
• Free movement of the probes within access Dynamic load test
tubes Static load test
• Mismatched probe positions especially at pile -20
toe 0 20 40 60 80 100
• Measurement resolution Load (ton)
• Incorrect position of access tube
• Air gaps or different material around access
tubes Figure 7. Load-settlement curves of dynamic and
• Aggregate variation (in the case of base static load tests on large diameter piles (φ 0.80m
grouting) x 50.0m).
0
It is to be noted from the author’s suggestion
based on the above factors that only variations
of transmit time more than 15 to 20% of the
-5
norm should be regarded as warranting further
Settlem ent (m m )
investigation.
-10
HIGH STRAIN DYNAMIC LOAD TEST -15

Dynamic load test
High strain dynamic load test can be employed Static load test
as a part of pre-planned quality assurance -20
regime and retrospective investigation. High 0 200 400 600 800
strain dynamic integrity test has become a well- Load (ton)
accepted method especially for evaluating the
pile capacity in the foundation industry today. A Figure 8. Load-settlement curves of dynamic and
large number of related technical papers and static load tests on small diameter bored piles (φ
case histories of the test have been published 0.35m x 18.9m).
and it is a part of standards and specifications
such as ASTM D4945-89 (Standard Method for Pile integrity can also be determined by high
High-strain Dynamic Testing of Piles). strain dynamic load test. Figure 9 and 10
illustrate the discontinuity (crack) detected at the
Vasinvarthana and Kampananon (1997) same pile by sonic integrity and high strain
presented the efficiency and reliability of dynamic load tests respectively.
dynamic load test carried out in Bangkok.
Figures 7 and 8 show the load settlement curves It is to be noted from Figure 9 and 10 that the
of dynamic and static load tests conducted in duration of pulse in the sonic integrity test signal
Bangkok on large and small diameter piles is about half of that in dynamic test signal. The
respectively. As can be seen in the figures, short duration (high frequency) pulses can
load-settlement characteristics obtained from resolve reflections from narrow impedance
dynamic load test agree well with those of static changes. In such case presence of a crack in
load test for both large and small diameter bored the pile is more clearly indicated by low strain
piles. with sonic integrity test signal than the high
strain by dynamic test signal.
460
• Some defect features can be of similar signal
pattern and it can not detect the axial extent
Soft Clay Med. Very Stiff Dense Sand Dense Sand of anomaly. Therefore crack may be
Clay Clay misinterpreted as contaminated concrete or
void which can lead to an argument between
piling contractor and excavation contractor
6.2 cm/s
0 6 12 18 24 30 36 42 48 54 60 V2-88c
Sonic logging tests are much more expensive
Pile 681 4000 m/s f:8 Sr
8 Apr 96 exp: 20 than sonic integrity tests as the cost of access
tubes and test itself are comparatively high for
deep-seated bored pile. It is mainly applicable as
Figure 9. Sonic signal showing a discontinuity a pre-planed site quality control testing. The
(crack) at about 7.8m, caused by excessive major advantage of this method is that test can
lateral soil movement induced by adjacent be carried out shortly after the pile construction.
excavation work Hence, rectification measures can be
implemented while the foundation contractor is
29-Apr-96
on site. However, this method is not applicable
1500
Ton
CAPWAP (R) Version 1 994 -1 for retrospective investigation. It is less useful if
750
For Msd
Vel Msd pile integrity is in question due to post
construction activities as small horizontal cracks
ms
are hardly detectable by this test method.
10 20 30 40 50
0
0 1 2 3 L/c
-750
Pile The major advantage of high strain dynamic load
Shaft Resistance
30
Ton / m Distribution test is its capability to verify both the load
(Toe 178.8 Ton)
carrying capacity and integrity of the pile. The
1000
Pile Forces at Rut
main advantages of this test in comparison with
Ton
static pile load test are effectiveness in terms of
cost, time and space requirement. It can be also
Figure 10. High strain dynamic load test signal applied to justify the capacity of pile after
showing a crack in the pile at about 8.0m rectifying its defect detected by other integrity
tests. However, testing and interpretation of the
DISCUSSION results must be made by qualified testing
specialists who have adequate knowledge in
Sonic integrity test can be applied for both local soil conditions and construction process of
quality check (control test) and retrospective wet-processed bored piles.
investigation. It is the cheapest in terms of cost
and the simplest in terms of testing process. In
Bangkok a single testing crew can carried out CONCLUSIONS
the test over 70 piles within a day at one site if
all piles tested are easily accessible and readily Integrity tests commonly applied for large
prepared. The main advantage of this test is that diameter bored piles in Bangkok are presented
since no particular measures or preparation is highlighting the advantages and limitations of
necessary during the pile construction phase it is each method. It is to be emphasized that
more flexible to select which pile is to be tested. integrity test are highly dependent on the
However, interpretation of sonic integrity testing experience of the testing firm in both field testing
needs considerable experience and knowledge and interpretation. Interpretation should be
in testing, subsoil condition and construction made with sound knowledge of pile construction
method. The major limitations of sonic integrity technique, subsoil condition and other factors
test are; such as problem encountered during and post
• Signal characteristics is influenced by pile construction which may influence the test
damping effect due to the stiffer soil layers at results and interpretation. Pile construction
deeper depth. Hence toe reflection is hardly records are also useful in interpretation of the
detectable for piles founded at depth deeper piles with detected anomaly. Conclusion should
than 35m. Similarly signal reflections from be made after careful and thorough review of
anomalies are generally clear for the piles test results in conjunction with other available
with length to diameter ratio (L/D) within 40
461
information. Further investigation may require
before final conclusion is made.
REFERENCES
American Society for Testing and Materials

1989. ASTM D4945-89 Standard method for
high-strain dynamic testing of piles, Philadelphia:
American Society for Testing and Materials.
Thasnanipan, T., Maung, A. W. and Tanseng, P.

1998a. Dam-ages to Piles associated with
Excavation Works in Bangkok Soft Clay, 6th
International Conference on Problems of Pile
Foundations Building, Ufa, Russia. September
14-18, 1998. pp. 91-98
Thasnanipan, N., Maung, A. W. & Baskaran, G.

1998b. Sonic integrity test on pile founded in
Bangkok subsoil – signal characteristics and
their interpretations, Proc. 4th Int. Conf. on Case
Histories in Geotechnical Engineering, St. Louis,
Missouri, 9-12 March 1998: 1086-1092.
Turner, M. J. 1997. Integrity testing in piling

practice, CIRIA Report 144, London:
Construction Industry Research and Information
Association.
Vasinvarthana, V. & Kampananon, N. 1997.

Efficiency and re-liability of dynamic load test),
Seminar on Foundation’ 97, Organized by
Engineering Institute of Thailand, Bangkok: 79-
94. (in Thai).
462
Proceeding of the 8th International Conference on the Application of
Stress Wave Theory to Piles
8-10 September 2008, Lisbon, Portugal
8th International Conference

on the Application of Stress
Wave Theory to Piles
Edited by:
Jaime Alberto Dos Santos
Application of Low-strain Integrity Tests for

Quality Assurance of Deep Foundations in
Thailand
Aung Win Maung, Zaw Zaw Aye and Thayanan Boonyarak
463
Application of Low-strain Integrity Tests for Quality Assurance of Deep
Foundations in Thailand
Maung A. Win, Aye Z. Z. and Thayanan Boonyarak
SEAFCO Public Company Limited, Bangkok, Thailand.
Keywords: Sonic integrity test, Sonic logging, Parallel seismic test
ABSTRACT: Low-strain integrity tests have been widely used in Thailand since the 1980s, when large-
diameter bored piles were introduced for the earliest high rise buildings in Bangkok. The most commonly
used integrity test methods on deep foundations in Thailand are sonic integrity tests (locally known as seismic
tests) and sonic logging tests. The sonic integrity test is specified as a compulsory quality assurance test for
bored piles in most projects. Additionally it is also carried out on some reference piles soon after piling work
is completed, particularly in projects which require deep excavation for further construction stages by other
than the piling contractor. This practice is adopted to prevent contractual dispute between the piling
contractor and the other construction contractor regarding pile damage contributed by soil movement
associated with deep excavation work. Sonic logging tests are commonly specified for all large bored pile
with diameters up to 2.00m and barrette piles for bridges and elevated highways and viaduct foundations. In
addition parallel seismic tests are now often utilized to determine pile length of old bridges and flyover
foundations to check alignment clearance for tunnels designed along the existing roads and canals. This
paper presents a brief history of low-strain integrity tests used in Thailand and discusses interpretation of test
signals in conjunction with pile construction method and subsoil conditions.
conditions with relatively shallow bedrock is
present. In order to assess the quality of foundation
1 INTRODUCTION piles, a number of low-strain integrity test methods
have been applied according to pile type. A
Thailand has been one of the economically fast monograph of the low-strain integrity testing
growing nations in South East Asia since the 1980s, described in this paper has been published by Turner
with the economic growth fuelling a boom in civil M. J. (1997).
construction. However, the majority of large scale
construction work has been mainly in Bangkok, the
capital of Thailand, where over 10% of the total 2 FOUNDATION PILES
population resides. Bangkok is located in a
sediment basin comprising thick layers of soft clays In Thailand, two major types of deep foundation
and a series of inter-bedding sand and clay layers in piling are pre-cast driven piles and bored piles. Pre-
hundreds of meters of deep. The depth of bedrock is cast piles are generally used outside urban or less
not well determined in the Bangkok area but has built-up areas where there are no noise or vibration
been reported to be about 550m deep constraints for pile driving. In Bangkok, bored piles
(Balasubramaniam, 1991). Deep foundations with tips extending to the first sand layer or to stiff
consisting of large diameter bored piles and barrettes clay layer at a depth of about 20 -30 m are normally
reaching a depth of 70m in such subsoil support employed for light to medium-sized structures
heavy structures are therefore concentrated in located in vibration-restricted areas. For heavy
Bangkok. structures or high-rise buildings deep large diameter
With improvements in the network of highways (0.80m-2.00m) bored piles with tips extending to
connecting all provinces over the last 20 years, civil more than 45m are commonly used. Two types of
construction is now active in all major provincial construction method, dry and wet process are
cities and tourist destinations. In these areas, typically applied in bored piling.
foundation constructions are generally much The dry process uses steel temporary casing
shallower than those used in Bangkok as better soil (about 15m in length, especially in Bangkok) to
465
protect the soft soil from caving in. The bore is
drilled either using an auger or bucket. For small
piles, shells with drop-weight are used to make the
bore. Concrete is poured into a borehole which is in
dry condition after placing the reinforcement cage.
The wet process also uses steel temporary casing.
A borehole is made with rotary drilling using auger
or an auger bucket under bentonite or polymer
slurry. Concrete is poured with tremie method under
drilling slurry.
The dry process is employed mainly for small
piles of 0.35m to 0.60m in diameter. The length of
the piles are up to 25m and extends further if
necessary, depending on subsoil conditions, absence
of sand layer with groundwater, etc.
For large bored piles of 0.80m to 2.00m in
diameter and barrettes (up to cross sections of Figure 1. A photographic record of sonic integrity test result
1.5x3.0m), piles are constructed to depths of 30m to taken by Polaroid camera in 1986.
over 60m. The wet process is commonly used and
piles are founded in sand layers. Compressive In the period 1992-1997, with economic growth in
cylindrical strengths of concrete used in bored piles Asia, massive development in construction of multi-
is in a range of 240-280 ksc (24-28 Mpa.). storey buildings, elevated highways, bridges, etc.
took place in Bangkok. Large quantities of bored
piles and driven piles were installed annually. The
3 SONIC INTEGRITY TESTS sonic integrity test was then widely used for quality
assurance of these piles.
After 1997, the construction industry in the
3.1 Overview of sonic integrity testing in Thailand private sector was significantly affected by an
economic crisis. However infrastructure work such
The sonic integrity/echo test (locally known as as mass transit system and airport construction
seismic test), performed with an accelerometer and projects was still active.
hammer, has been used in Thailand since the 1980s The sonic integrity test is used mainly to detect
when large-diameter bored piles were introduced for defects, changes in the physical properties and
the earliest high-rise buildings in Bangkok. In length of the bored piles. For driven piles, it is used
addition to construction records and specification for for detecting pile defects such as cracks and joint
the bored pile construction for quality assurance, a splitting induced by over driving and/or pile
sonic integrity test was also employed. In the early inclination.
days of high-rise construction, the sonic integrity
test equipment showed a test record on the display 3.2 Signal characteristics and interpretation
unit only. The record was photographed by a
Polaroid camera for later interpretation. The quality The sonic integrity test indicates toe reflections of
of test records was visually poor and sometimes pile clearly for small piles with slenderness ratio of
comprehensive conclusions on test results were 40 . In few cases, especially toe grouted piles toe ref-
difficult to make. A few years later, advancements lection has been observed for longer piles with toe
in both computer hardware and software enabled the embedded below 30-35m.
user to obtain high resolution data with signal In some cases, even though pile toe is extended
enhancement and digital processing for beyond 35m in depth, the toe reflection can be ob-
interpretation of test data. Test results can thus be served in the test signals of toe grouted piles having
accurately interpreted, particularly for types of L/D within 40 .
defect or anomalies and their position, and pile Figure 2 is an example of such a case where the
length. bored piles were base grouted and of 40m in length
Other derivatives of low-strain integrity tests, such (toe at 55m, trim level at -15m). However, it should
as frequency response tests were introduced by a be noted that these reflections are prominent only for
few testing companies. However, the frequency- a few piles. Influence on toe reflection by grouted
based testing was not as popular as the time-based toe also depends on stiffness of the soil underneath
testing due to unfamiliarity with both analysis and the toe, which has been grouted.
interpretation of the test results.
466
23.2m
Soft Clay Very Stiff Clay Hard Clay Dense Sand

11.0cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
Pile 6 4000m/s f:8 Sr
5.3cm/s
12 Sep 95 exp: 20
0 5 10 15 20 25 30 35 40 45 50 V2-88c
Pile G30 4000 m/s f:8 Sr
27 Jan 97 exp: 20
22.0m
Figure 2. Toe reflection observed from a toe grouted pile. (pile

1.0x40.0m, cutoff -15.0m, toe depth 55.0m) 7.1cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
In a few cases, it was found that toe reflection is Pile 24 4000m/s f:8 Sr
detectable for some deep-seated bored piles of about 12 Sep 95 exp: 20
50m in length if the pile was tested at an early age.

Figure 5. Toe reflections from different pile sizes, pile no. 6
Figure 3 shows the significant toe reflection of
(0.50mx22.0m) and pile no. 24 (0.35mx22.0m)
52.0m-long pile tested a few days after concreting.
Note young pile age, the bonding between soil and
Typically, sonic integrity tests show size variation
pile has not yet fully developed so that damping
in bored piles constructed in Bangkok. All size
effect or attenuation of the sonic signal by shaft
variations including size reductions are found to be
friction is relatively low.
at the transition interface of temporary casing and
soil layer, where soil strata changes also occur.
52.0m
16.0cm/
s Soft Clay Med. Very Stiff Clay
0 6 12 18 24 30 36 42 48 54 60 v2-88c Clay
pile 30 4000m/s f:8 Sr
14 Jan 91 exp: 20
10.5cm/s
Figure 3. Sonic signal showing a toe reflection of bored pile at 0 3 6 9 12 15 18 21 24 27 30 V2-88c

52.0m tested a few days after concreting. Pile 30 4000 m/s f:8 Sr
2 Jul 97 exp: 20
Influence by different soil strata on signals has

been observed in many cases (Thasnanipan et al Figure 6. Pile size decrease or change at 14m as shown by the
1998). Stiff clay layer at depths of 15m-25m signal (Pile 0.8mx21m)
(undrained shear strength of 10-22 t/m2)
significantly influences signal reflections. This was Test signals shown in Figure 7 indicate a clear
in small rather than large bored piles. Figures 4 and reflection at approximately 11m from pile top (about
5 explain these conclusions. 15m below ground level) with a repeat or multiple
reflections at about 2 times that distance
(approximately 22m from pile top). As the polarity
of the reflection is negative, the first interpretation is
that this pile has decreased in cross section or crack
Soft Clay Med. Very Stiff Dense at 15m below ground level. However, based on the
Clay Clay Sand
pile construction method, it is concluded that
negative reflection is caused by variation in pile
cross section at the bottom end of the 15m-long
10.6cm/s temporary casing. The sonic signals with anomaly
Pile 49
0 3 6 9 12 15
4000 m/s
18 21 24 27
f:8
30 V2-88c
Sr acquired on those piles are often misinterpreted as
19 Sep 95 exp: 20 indicative of defective piles, leading to dispute.
Figure 4. Influence of soil friction on sonic signals (a pre-cast

driven pile 0.40x0.40x25.0m in clay layers)
467
Med.
Very soft clay clay Stiff clay Dense sand
Pile Cut-off level Temporary casing

bottom level
4.3cm/s
0 5 10 15 20 25 30 35 40 45 50 V2-88c
Pile 99 4000 m/s f:8 Sr
25 Nov 94 exp: 20
Outer diameter of Nominal dia./dia. of

casing collar drilling tool
11.0 cm/s
0 3 6 9 12 15 18 21 24 27 30 v2-88c
Pile 66 4000 m/s f: 8 sr
25 Jun 97 exp : 20
Figure 7. A typical pile shape constructed in Bangkok soil

with resulting sonic integrity test signal profile.
Soft Clay Med. Stiff to Hard Dense Stiff Dense Sand
Clay Clay Sand Clay
Figure 8 shows bored pile sections formed at the
lower end of temporary casing, confirming size Figure 9B. Uniform pile size along the pile shaft is shown by
changes occur below the casing. These bored piles both drilling monitoring record and sonic integrity test record.
were cast together with a steel column for (Pile 0.80mx42.0m, cutoff -3.0m, toe depth – 45.0m.)
construction of deep basements with diaphragm
walls using top-down method. These piles were Sonic integrity test can clearly reveal a crack or
exposed when excavation reached to 15m below discontinuity in piles. Most cracks or discontinuities
ground level. in bored piles after installation can be caused by
construction activities associated with basement
excavation adjacent to them and improper trimming
to design cut-off level. In the past, from the
assessment of pile defects, most of the defects are
found in small piles with a diameter of 0.60m or
less. Study of piling and excavation plans show that
piles with a crack indicated by the sonic integrity
were located in the vicinity of excavation boundaries
(Thasnanipan et al, 2000a). These piles were
cracked by soil movement due to both reinforcement
inadequate for resisting bending stresses and
improper soil excavation practice. Since
geotechnical experience has been gained from
several construction projects with deep basements,
Figure 8. The photograph illustrates pile size changes caused pile design has been improved over the years and
by difference in diameter of temporary casing and auger used pile defects are significantly reduced.
in pile construction. The test signal in Figure 10 shows a crack or
Size variations shown by signals are often discontinuity in the pile caused by soil movements
consistent with casting records and available drilling due to adjacent excavation work.
monitoring results. Figures 9A and 9B show the
comparison of borehole profiles and sonic test
records.
Soft Clay Med. Very Stiff Dense Sand Dense Sand
Clay Clay
5.4cm/s
0 5 10 15 20 25 30 35 40 45 50 V2-88c
Pile 322 4000 m/s f : 8 Sr
10 Jan 95 exp: 20
6.2 cm/s
0 6 12 18 24 30 36 42 48 54 60 V2-88c
Pile 681 4000 m/s f:8 Sr
8 Apr 96 exp: 20
Figure 10. Sonic signal showing a discontinuity (crack) at

about 7.8m, caused by lateral soil movements.
Figure 11A shows a discontinuity in the pile shaft

caused by disruption during concrete pouring.
Soft Clay Med. Stiff to Dense Stiff Dense Sand
Clay Hard Clay Sand Clay
Figure 9A. Size increase in pile shaft at 13m is shown by both Concrete supply was interrupted by poor
sonic integrity test and drilling monitoring records. (Pile coordination between the piling contractor and
1.0mx40m, cutoff -5.0m, toe depth – 45.0m.)
468
concrete suppliers. By the time the concreting was Due to cost effectiveness, simplicity of the
resumed the lower section had hardened. The method and ease of performing the test, the sonic
discontinuity was later closed by core-drilling and integrity test is now specified for testing 100% of
cement grouting. Figures 11A and 11B show the piles in most of the projects for quality assurance.
signal interpretation before and after closing the Additionally, a sonic integrity testing is also
discontinuity. carried out on some piles located in the boundary of
deep excavation zones soon after piling work is
completed. The test records use a reference pile to
document pile integrity prior to deep excavation
work to prevent dispute between piling contractor
and excavation contractor for any pile damage
14.2 cm/s
0 4 8 12 16 20 24 28 32 36 40 V2-88c caused by improper construction practice.
Pile 1 4000 m/s f:8 Sr
24 Apr 97 exp: 20
Figure 11A. Sonic signal showing a discontinuity at 11.0m 4 CROSSHOLE SONIC LOGGING
caused by interruption in concrete pouring (pile 
0.80mx38.80m) and multiple reflections from the
discontinuity. 4.1 Overview of Crosshole sonic logging test in
Thailand.
The first crosshole sonic logging test is believed to
have been conducted in 1982 for quality insurance
7.6 cm/s of bored piles based on the test results for the
0 4 8 12 16 20 24 28 32 36 40 V2-88c
Memorial Bridge Project in Bangkok as reported by
Pile 1 # 4000 m/s f:8 Sr Ng (1983). Usually the test is employed as quality
2 Jun 97 exp: 20 control testing for foundation piles and diaphragm
walls of infrastructure projects such as subways,
Figure 11B. Sonic integrity test shows the discontinuity at elevated highways and bridges. Due to the relatively
11.0m has been closed by grouting. high costs involved in installation of a number of
steel pipes along the entire length of piles as sonic
Figure 12 shows the characteristics of signals emitter and receiver access tubes, it is employed for
observed at a site where dry process was adopted. large-scale projects. To be cost effective, use of the
Sonic integrity test signals were considered access tubes for sonic logging tests is sometimes
indicative of poor pile integrity at the site. extended for pile base grouting and for checking soil
Construction records revealed that boring was done conditions below the pile base.
in water bearing sand layer without casing or During the 1990s, sonic logging tests were
bentonite slurry causing concrete contamination with extensively used for foundation piles of elevated
water and/or segregation. highways and bridges in Thailand. After 2000,
large bored piles up to 2.0m in diameter and
barrettes are now used for foundations of buildings,
Very Stiff Clay Med. Stiff Clay Dense viaducts and cable-stayed bridges. Whereas
Clay Sand previously, bored piles of 1.5m in diameter were
routinely used for construction of large buildings.
Corsshole sonic logging tests have become
compulsory for quality insurance for these large
8.9 cm/s
0 3 6 9 12 15 18 21 24 27 30 V2-88c
Pile 50 4000 m/s f:8 Sr piles.

4 Apr 97 exp: 20
Two access tubes are required as a minimum for
the sonic logging test. A recommended number of
Figure 12. Sonic test signal showing size decrease and or a
discontinuity at 7.5-9.0m and poor quality pile section near the
tubes for good coverage of different pile sizes is
toe (Pile �0.35x26.0m). shown in Table 1.
Table 1. Recommended number of tubes for different pile size
3.3 Current applications Pile Diameter Minimum No. of Tube spacing
(mm) tubes (degree)
The sonic integrity test has been accepted by
D  750 2 180
engineers and has been widely used in the deep 750D  1000 3 120
foundation works in Thailand for over 25 years. The 1000D 1500 4 90
test is also applied to testing of barrette piles in 1500 D  2500 6 60
conjunction with sonic logging. For barrettes, 2500  D 8 45
multiple testing points with signal stacking are used
for obtaining the test data.
469
4.2 Defect and Anomaly Classifications A review of sonic logging test results and pile
construction records of 148 large diameter bored
Faiella and Superbo (1998) published defect piles (dia. 2.0m, length 64.0m) was carried out to
classification criteria for piles monitored by different study the detected anomalies and its causes.
numbers of access tubes based on interpretation Anomalies were detected in 14% of the piles, with
from tests on over 6,000 piles from 37 sites in Italy. equal percentage of minor and major anomalies and
In Thailand, it was based on the experience of defects as classified in Table 2. Construction
testing specialists who performed sonic logging tests records indicated these anomalies were associated
on over 3,000 piles in the past 15 years, coupled with relatively high sand content in drilling slurry
with the analysis on both sonic logging test results (the sand content was still under the limit of
and core samples taken from the zones detected with specification). Actually the high sand content was
anomaly. The classification of anomalies was contributed by soil disturbance caused by removal of
established with signal transmission time criteria. underground obstruction – the existing old pile at the
Figure 13 shows the sonic signal profiles of 4 project pile location. Another major cause of
categories classified in Table 2. anomaly in pile was found to be due to congested
reinforcement arrangement. As the reinforcement
Table 2. Category of sonic logging signals and anomaly cage was considerably large, horizontal stiffener
classification, Thasnanipan et. al (2004). rings were included for lifting at levels of pile cutoff
Category Criteria Integrity and every cage section lapping. Such reinforcement
Classification with spiral bars of the pile reduced the spacing of
A No or slight increase in transmis- Homogeneous horizontal bars down to 32mm around the rebar cage
sion time not more than 10% concrete
B Minor increase in transmission Minor at the above mentioned levels for some piles.
time >10% but <25% anomaly Contaminated concrete and sediment were
C Transmission time more than Cat- Major considered to be trapped at these congested areas
egory B but transmission time can anomaly during concrete pouring. This was later confirmed
stilled be defined when piles were trimmed to cut-off level and
D Transmission time cannot be de- Defect trapping of slurry contaminated concrete was found
fined
below the horizontal bars.
Based on the review of both records of pile
construction and sonic logging tests, a summary of
causes of anomaly is presented in Table 3.
Table 3. Summary of causes of anomaly in concrete of large

diameter bored pile found in a typical project.
Causes Probability
High Medium Low
Reinforcement bar congestion X
Relatively high sand content in X
slurry
Concrete pouring time longer X
than usual
Interruption and concrete pouring X
Localized collapse of borehole X
surface
5 PARALLEL SEISMIC TEST

Figure 13. Sonic profiles showings different category (A, B, C
and D) as classified in Table 2, (after Thasnanipan et al 2004). Parallel seismic test is used in Thailand for checking
the length and integrity of old foundation piles
where both construction records and pile head were
4.3 Additional information from sonic logging test not accessible for investigation. Available records
results indicate that the earliest parallel seismic test in
As additional information, sonic logging test records Thailand was performed in 2004.
can also be used for quantifying overcast length and The test is adopted for quality assurance of tunnel
volume of hard concrete above pile cut-off. In design alignment since the tunnels are usually
project planning/bidding of prospective main planned along roads and canals. Bangkok has many
contractors this information gave the accurate canals with bridges crossing over and flyovers at
quantity of concrete to be trimmed (Thasnanipan et. road intersections. Frequency of parallel seismic
al 2004). tests is increasing, as tunnels for subway, water
supply, storm-water drainage system and power
470
transmission lines are being designed and for investigation of the subsoil condition in the close
constructed to pass adjacent to the foundation piles vicinity of the pile to be tested.
within the available corridor and due to subsoil
conditions. To date parallel seismic tests were
performed on a total of 190 piles for 50 projects. 6 CONCLUSIONS
Table 4. List of projects using parallel seismic test for pile Use of low-strained integrity tests in Thailand for
integrity and length determination quality assurance of deep foundation piles has been
Description of Site Number of Number of Piles
Project Tested
presented. Low-strain integrity tests are effective
methods to assess the pile integrity. Signal
Bridge 2 15 interpretations need to be compared with soil
Old building 47 145 investigation results and pile construction records.
Building near Pile 1 30
Driving
ACKNOWLEDGEMENT
The authors wish to express their appreciation to

EDE Co., Ltd. and STS Engineering Consultants
Co., Ltd. for providing the test data and project
information.
REFERENCES
Thasnanipan, N., Aye Z.Z., and Thayanan B. and Kampananon

N., (2004). “Quantifying Pile Head Condition before
Basement Excavation by Cross-Hole Sonic Logging Tests”,
Proceedings of 7th International Conference on Application
of Stress-Wave Theory to Piles, Kuala Lumpur, Malaysia.
9-11 August 2004. Pp. 233-238.
Thasnanipan, N., Maung A. W., Tanseng P., and Aye Z. Z.
(2000a). “Sonic Integrity Test of Piles – Integrity Effected
by Basement Excavation”, Proceedings of 6th International
Conference on Application of Stress-Wave Theory to Piles,
Balkema, Rotterdam. Pp. 163-170.
Thasnanipan, N., Maung A. W., Navaneethan T., and Aye Z. Z.
(2000b). “Non-destructive Integrity Testing on Piles
Founded in Bangkok Subsoil”, Proceedings of 6th
Figure 14. Parallel seismic test (after Williams and Stain, International Conference on Application of Stress-Wave
1987) Theory to Piles, Balkema, Rotterdam. Pp. 171-177.
Thasnanipan, N., Maung A. W., and Baskaran B. (1998).
“Sonic Integrity Test on Piles Founded in Bangkok Subsoil
– Signal Characteristics and Their Interpretations”,
Proceedings of 4th International Conference on Case
Histories in Geotechnical Engineering, St. Louis, Missouri.
9-12 March 1998. Pp. 1086-1092.
Faiella, D. & Superbo, S. (1998). “Integrity Non-destructive
Tests of Deep Foundations by Means of Sonic Methods -
Analysis of the Results Collected on 37 Sites in Italy, Van
Impe & Haegeman (eds), Proc. 3rd Int. Geotechnical
Seminar on Deep Foundations on Bored and Auger Piles,
BAP III, Ghent, Belgium, 19-21 October 1998: 209-213.
Rotterdam: Balkema.
Turner, M. J. (1997). “Integrity Testing in Piling Practice”,
Construction Industry Reasearh and Information
Figure 15. A typical test record of parallel seismic test (after Association, CIRIA Report 144. 6 Storey’s Gate,
Williams and Stain, 1987) Westminster London.
Balasubramanium, A. S. (1991). “Inaugural Lecture on
Contributions in Geotechnical Engineering, Soil Mechanics
Information on old pile length and integrity is and Foundation Engineering”, Asian Institute of
also used in carrying out geotechnical analysis for Technology, Bangkok, March 1991.
studying the impact of tunneling, deep excavation Ng, K. C. (1983). “The Construction Problem and Performance
and pile driving works on structure supported by old of Large Bored Piles in Second Sand Layer”, M. Eng.
Thesis No. GT 82-26, Asian Institute of Technology,
piles. To be cost effective, the parallel seismic test is Bangkok, Thailand.
carried out using the same borehole which is drilled
471
and Tunneling

for Protection of
Environment
Edited by:
Phienwej, N.
Ground Movement Prediction and Building

Damage Risk-Assessment for the Deep
Excavations and Tunneling Works in Bangkok
Subsoil
Zaw Zaw Aye
Dhiraj Karki
CPG Consultants Pte. Ltd., Singapore
Christian Schulz
Bilfinger Berger AG , Bangkok Thailand
473
International Symposium on Underground Excavation and Tunelling
Urban Tunnel Construction for Protection of Environment
Sponsored by International Tunnelling Association (ITA/AITES)
2-4 February 2006, Bangkok
Ground Movement Prediction and Building Damage Risk-

Assessment for the Deep Excavations and Tunneling Works in
Bangkok Subsoil
Zaw Zaw Aye1 , Dhiraj Karki2 and Christian Schulz3
1
SEAFCO Public Company, Limited, Bangkok, Thailand, (former employee of JV BCKT)
2
CPG Consultants Pte. Ltd., Singapore, (former employee of JV BCKT)
3
Bilfinger Berger AG , Bangkok Thailand, (former employee of JV BCKT)
ABSTRACT
This paper presents the methods adopted in prediction of excavation and tunneling induced
ground movement and building damage risk-assessment carried out for the Contract No. 1 of
the M.R.T Chaloem Ratchamongkhon Line, the first underground mass transit system project
of Bangkok. The measured ground and associated movement of the buildings and structures
were within predicted values. All deep excavation works and tunneling were successfully
completed without any significant damages to the adjacent buildings and structures.
1. PROJECT OVERVIEW
The M.R.T Chaloem Ratchamongkhon Line is the first underground mass transit system
project in Bangkok developed by the Mass Rapid Transit Authority (MRTA) of Thailand.
Total length of the underground structure is 21.5km, comprises of 16 km of twin single-track
bored-tunnel, a total of 4 km long18 cut- and-cover stations, a twin 1.5km long cut-and-cover
approach tunnel to Depot and other associated structures. The project was commenced in late
1996 and opened for the public in 2004. The Contract No. 1, Underground Structure South -
southern portion of this initial system comprising 9 km of twin 6m outside diameter bored
tunnels, nine underground stations and cut-and-cover depot approach tunnel, was awarded to
the Joint Venture BCKT consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch
Karnchang Public Company Limited, Kumagai Gumi Company Limited and Tokyu
Construction Company Limited. Figure 1 shows the full length of the Contract No. 1 in
simple schematic layout.
One of the major challenges of the Contract No. 1was the need of stacked-alignment
for a major portion of twin bored-tunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations, Lumpini, Si Lom and Sam
Yan. This requirement led the contractor to design and construct the deepest underground
structures of Bangkok at Si Lom Station - four-floor stacked-station of 32m deep right under
the existing flyover and adjacent to a number of sensitive structures. Deep excavation and
bored tunneling works in congested urban environment of Metropolitan Bangkok requires a
systematic process of ground movement prediction, building risk damage assessment and
protection. This paper presents the methods adopted in prediction of excavation and tunneling
induced ground movement and building damage risk-assessment carried out for the Contract
No. 1.
475
2. SUBSOIL CONDITIONS
Similar to other localities in Bangkok, a typical subsoil profile along the tunnel alignment is
characterized by the alternating layers of clay and sand deposits as shown in Figure 1.
Weathered crust of 2 m thick is commonly found as the top layer. In urban areas of Bangkok,
this layer is covered by fill material. Soft to very soft, highly compressible dark gray marine
clay lies beneath weathered crust or fill. Depending on the location, this layer extends up to
12-16 m. About 2 m thick medium clay layer can be observed between soft clay and
underlying stiff clay. Generally stiff Clay layer occurs directly underneath medium clay and
its depth goes up to 22 m. Below stiff clay layer, first sand layer of 5-8 m thickness can be
found. This first sand layer, however, is absent in some areas. Stiff to hard clay layer
underlies first sand and it is found to be about 5 m thick. Second sand layer generally occurs
at depths between 45 to 65 m.
Figure 1. Generalized soil profile along the tunnel alignment and stations
3. OVERVIEW OF GROUND MOVEMENT PREDICTION AND BUILDING

DAMAGE ASSESSMENT PROCESS APPLIED IN THE PROJECT
Staged Assessment for each station and tunnel sections was carried out according to the
requirement stipulated in the Outline Contract Design Specification (OCDS). Specified
references were used as basic literatures in ground movement prediction and building damage
assessment. For cut-and-cover excavation zone, the work of Peck (1969) and Clough and
O’Rourke (1990) were used whereas published papers of Burland et. al. (1977) and Boscardin
& Cording (1989) were applied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and Stage 3 Assessment.
476
“Greenfield Site” assumption was adopted in Stage 1 and 2 Assessment ignoring any
influence of the stiffness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assessment. In Additional Stage 2
assessment, the variation of settlement trough shape and magnitude with depth were
considered in identifying the risk of building damage. In Stage 3 assessment, a detailed
qualitative assessment of structure and vulnerability of each building or structure to damage
were taken into account, coupled with detailed structural analysis in assessing the buildings
falling in risk category 3 classified by Mair et. al. (1996).
Table 1: Building Damage Classification (after Burland et. al., 1977 and Boscarding and
Cording, 1989)
Risk Degree of Description of Typical Damage Approximate Max.
Category Damage Crack Width Tensile
(mm) Strain %
0 Negligible Hairline cracks Less than
0.05
1 Very Slight Fine cracks easily treated during 0.1 to 1 0.05 to
normal decoration 0.075
2 Slight Cracks easily filled. Several slight 1 to 5 0.075 to
fractures inside building. Exterior cracks 0.15
visible.
3 Moderate Cracks may require cutting out and 5 to 15 or a number 0.15 to
patching. Door and windows sticking. of cracks greater 0.3
than 3
4 Severe Extensive repair involving removal and 15 to 25 but also Greater
replacement of walls, especially over depends on number than
doors and windows. Windows and door of cracks 0.3
frames distorted. Floor slopes noticeably.
5 Very Major repair required involving partial or Greater than 25 but
Severe complete reconstruction. Danger of depends on number
instability. of cracks
4. SUMMARY OF GEOTECHNICAL INSTRUMENTATIONS
Instrumentation played a major role in the construction of this first underground mass transit
system project in congestive urban area with difficult soil condition. The primary objective of
the instrumentation program was to monitor the performance of the deep excavation and
tunneling operation to ensure safe execution of the construction works and that adjacent
structures were not adversely affected. Furthermore, the instrumentation program was
established to provide feedback in application of the observational method. Extensive sets of
instrumentation were installed around the predefined influence zones of station box
excavation, tunneling and associated building, structures and utilities. In addition to other
geotechnical instruments, inclinometers, extensometers, various types of settlement points,
tile-meter, crack meter and crack gauge were installed to monitor the ground and associated
building and structures movement.
5. BUILDING CONDITION SURVEY
477
Prior to commencement of the major works, all the existing buildings and structures within
the predefined influence zone were systematically surveyed. The primary objectives of the
surveys were;
• To record pre-construction status of all buildings and structures within the influence
zone. The records included were age of building, building type, storey height,
expected foundation types, visible defects supported by sketches and photographs as
appropriate
• To utilize the information collected from the building / structure condition survey in
the assessment of the buildings and evaluate sensitive structures with respect to
construction aspects and monitoring
• To establish a benchmark to monitor the possible effect of construction.
Expected foundation types of different buildings were classified based on the height,
number of floors and age of the buildings. Building survey information was also used to
identify the types of instrumentation required.
6. PREDICTION OF GROUND MOVEMENT INDUCED BY STATION EXCAVATION

WORKS AND BUILDING DAMAGE ASSESSMENT
6.1 Prediction of ground movement induced by station excavation
Surface settlement induced by station excavation was predicted based on the dimensionless
design settlement profile behind the earth-retaining structures. The dimensionless design
settlement profile was developed from the results of various prediction methods such as
numerical analysis carried out by CRISP program, method proposed by Bowels (1990) and
Clough & O’Rourke (1990).
6.2 Stage 1Assessment
Based on the calculated surface settlement described above, contours of settlement around the
station and associated underground structures were plotted in the topographic plan with
buildings and other structures. Figure 2 depicts the surface settlement contours prepared for
one of the stations.
ENTRANCE A ENTRANCE B
STATION
ENTRANCE C
40
30
20
10
478
Figure 2. Settlement contour in station excavation zone prepared for stage 1 assessment
Buildings falling outside 10mm settlement contour were eliminated from further
assessment. Remaining buildings with settlement greater than 10mm were short-listed for
Stage 2 Assessment.
6.3 Stage 2 Assessment
The dimensionless settlement profile for the station box construction were fitted using a
polynomial curve to identify the point of inflexion and divide the settlement profile into
sagging and hogging zone and calculated the tensile strain induced in the buildings from
excavation works using the method proposed by Mair et. al. (1996). The results were then
used to identify the category of risk to damage according to the classification of Building
Damage Classification presented in Table 1. Buildings classified within “Slight” risk category
of damage were eliminated from further assessment. Buildings falling into “Moderate” or
higher risk category of damage were further evaluated.
6.4 Additional Stage 2 Assessment
The settlement estimated in Second Stage Assessment presented above was based on surface
movement. For the buildings located in close proximity to the excavation founded on piles, it
was considered that subsurface settlement may be more critical. Stage 2 Assessment was then
further extended by Additional Stage 2 Assessment in which the variation of settlement
trough shape and magnitude with depth were considered in identifying the risk of building
damage. Settlement of the buildings due to the excavation works were predicted assuming the
“green filed” condition. Neglecting restraints from foundation and structure, it was assumed
that buildings follow the ground settlement trough at foundation level (estimated pile tip
level).
Available methods were further reviewed for predicting the surface settlement with
the consideration of two main factors; (1) simple and practical in application (2) enable to
correlate with the predicted and measured diaphragm wall deflection. The method proposed
by Bowels (1988) was selected as it meets the required criteria mentioned above. Bowels
suggested the ground settlement induced by excavation as a function of ground loss due to the
deflection of the retaining wall. Bowel demonstrated the calculation of settlements at
specified distance by assuming parabolic variations of settlement within the influence
distance. Using predicted diaphragm wall deflection, surface settlement behind the
wall was computed by empirical formulas proposed by Bowels.
6.4.1 Prediction of vertical sub-surface settlement
A simplified prediction of subsurface settlement was carried out based on the calculated
surface settlement described above. First, subsurface settlement influence line was
constructed. As shown in Figure 3, settlement influence zone is assumed to decrease with
depth from “Do” at the surface and zero at the wall toe. With the assumption of linear
relationship between the volume of deflected wall shape and the volume of settlement trough
at any depth within settlement influence zone, subsurface settlement at different depths were
calculated.
479
X
W DDoO== H2.5
W+ WHE 4V O
S WO =
⎛ x ⎞
S io = S WO * ⎜⎜ ⎟⎟
Sio
S WO
D-wall ⎝ DO ⎠
Dy D
HE
DY = Y * O
HW
HW

trough volume, VTy VTY = VY *
Siy
VO
Swy
Settlement
Y
influence zone
Figure 3. Demonstration of subsurface settlement prediction from diaphragm wall deflection

values
Based on the polynomial curve developed in Stage 2 Assessment, subsurface

settlement can be predicted by the similar approach described above.
6.4.2 Prediction of subsurface horizontal ground movement
Assuming a simple linear distribution of horizontal ground movement with distance,

prediction method for sub-surface horizontal ground movement was developed. The influence
distance of horizontal displacement behind the diaphragm wall was assumed to be 2.5 Hg at
ground level as shown in Figure 4. The influence depth of horizontal displacement is a depth
at zero diaphragm wall deflection, Hw, which can be obtained by extrapolating the diaphragm
wall deflection profile. Horizontal ground movement influence zone was constructed by
connecting the end points of Do and Hw as shown in the figure. Horizontal displacement is
assumed to decrease linearly with distance, maximum at wall face and zero at end of
influence zone.
Do = 2.5 Hg
Doi
Xi
Shwi Shi DO
D Oi = * H Wi
HW
z
Shwi Shi
Zi
H Wi
= HW − Zi
Hg
Doi
D Oi − X i
S hi = S hwi *
Hwi
Hw
D Oi
Limit of
D-wall influence zone
y Deflection curve
extrapolated to zero
Figure 4 Demonstration of subsurface horizontal movement prediction from diaphragm wall

deflection values
480
6.4.2.1 Risk of damage based on maximum tensile strain
Based on the polynomial curve developed in Stage 2 Assessment and predicted subsurface
ground movement using method described in section 6.3.1, maximum tensile strain of the
building was predicted applying the method proposed by Mair et. al. (1996). Risk category of
to the building was then classified according to Table 1.
6.4.2.2 Risk of damage specified by angular distortion and horizontal strain

The chart proposed by Boscarding and Cording (1989), was used to determine the potential
damage of the building. The chart is a simple tow-dimensional plot of horizontal strains
versus angular distortion in which zones of potential damage to buildings are mapped.
Angular distortion (β) was calculated based on the difference between predicted settlement of
building one edge to another or two adjacent foundation supports. Subsurface settlement at
pile toe level was used in determining angular distortion.
β = (SV1 – SV2) / L
SV1 = vertical displacement at one end of the building
SV2 = vertical displacement at other end of the building
L = building length
Horizontal strain induced by excavation was predicted assuming horizontal ground

movements at surface level decrease linearly with distance, maximum at excavation wall and
zero at end of influence zone. Average horizontal strain of the building can be computed by ;
∈h = (Sh1 – Sh2) / L
where, ∈h = average horizontal strain of building
Sh1 = horizontal displacement at surface at one end of the building
Sh2 = horizontal displacement at surface at other end of the building
L = building length
Table 2 shows the summarized results of risk category obtained from two different methods
described above and assumed foundation types (without and with pile)
Table 2 Summary of risk category obtained from two different methods and foundation type
Building No. of Building Estimated Distance Risk category by Risk category by
Block stories Type Pile Length to edge maximum tensile chart of Boscarding
of wall strain and Cording (1989)
Without With Without With
pile pile pile pile
A 2 Masonry / 7-14m 18m 0 0 1 2
Concrete
B 2 Masonry / 7-14m 17.5m 0 0 1 2
Concrete
C 2 Masonry / 7-14m 32m 0 1 1 2
Concrete
D 1 Masonry / < 6m 2m 1 1 1 2
Concrete
E 37 RC 50m 8m - 0 - 0
481
The predicted degree of damage of the buildings using simplified chart is more critical
than that of maximum tensile strain method as can be seen in above table. According to the
results obtained from of simplified chart, predicted degree of damage was higher if buildings
were assumed to be with piled foundation rather than without piled foundation.
6.5 Third Stage Assessment
Examples of the buildings and structures that required the stage 3 assessment as a result of
Stage 2 assessment are listed below.
Table 3 The buildings required Stage 3 Assessment in station excavation zones

No. of Storey / Preliminary Stage 2 Calculated Total Risk Category /
Building Type Estimated Pile Length Tensile Strain (%) Degree of Damage
4 / RC* 8m 0.121 % 2 / Slight
22 / RC >22m 0.159 % 3 / Moderate
4 / RC 7-14m 0.197 % 3 / Moderate
2 / Steel Frame 15-22m 0.288 % 3 / Moderate
5 / RC, Masonry 15-22m 0.166 % 3 / Moderate
2 / RC, Masonry <6m 0.29 % 3 / Moderate
3 / RC, Masonry 7-14m 0.242 % 3 / Moderate
1 / RC, Masonry ≤ 6m 0.191 % 3 / Moderate
* Stage 3 Assessment of this building was required due to the pre-existing damage of the building.
A process of Stage 3 Damage Assessment involved;

• Collection of the record of pre-construction defects/damages of the building from
Building Condition Survey Reports
• Conducting detailed structural Survey and assessing the current state of the building
• Condcuting damage assessment
6.5.1 Collection of building condition survey

The building condition survey records were collected and primary review of the state of the
building prior to commencement of construction works.
6.5.2 Detailed Structure Survey

In order to understand the building response to the ground movement and to carry out the
qualitative damage assessment of the building, details of the structure should be known in
advance. A structural survey was carried out by experienced structural engineer to obtain the
detailed information of structural members.
6.5.3 Damage Assessment
Structural survey information was the key factor in the selection of methodology for the Stage
3 Assessment. Based on the pre-construction state building condition survey and detailed
structural survery, the method for damage assessment was selected. Two main approaches
were applied.
482
• Conservative approach, mainly applied in additional stage 2 assessment with some
refinements
• Frame analysis using FEM structural software
6.5.3.1 Stage 3 Assessment by conservative approach

Conservative approach applied in Stage 3 Assessment involved consideration of two separate
criteria, namely ;
• The risk of damage to the building specified by angular distortion and horizontal strain.
• The risk of damage to the building in terms of magnitude of building slope with reference
to the limiting values specified by various researchers.
The following assumptions were made in the conservative approach of Stage 3

assessment.
• Buidling movement is influence by excavation induced settlement of both main station
box and associated underground structures (e.g. Entrance Building)
• Building settlement was casued by subsurface settlement at foundation level as buildings
are supported by piled foundation.
• No allowance was made for structural stiffness of the building.
• No allowance was made for possible effects of soil-structure interaction.
A summary of risk category determination in Stage 3 Assessment based on the predicted
and monitored settlement of the building adjacent to station excavation work is shown in the
Table 4 below.
Table 4 Summary of risk category determination based on Boscarding and Cording (1989)
Data Taken Horizontal Strain Angular Distortion
Risk Degree of
From Sh1 Sh2 L Sv1 Sv2 L
εh ΔS β Category Damage
(mm) (mm) (m) (mm) (mm) (mm) (m)
Settlement
31 24 18.6 0.0004 48 23 25 18.60 0.0013 2 slight
Prediction
Actual 0.0004
31* 24* 18.6 53 14 39 18.60 0.0020 2 slight
Monitoring Data *
*Predicted horizontal movement was used in computing horizontal strain for actual
miniotring data.
From the above table, following conclusions can be made.
• Predicted angular distortion broadly agrees with that of actual monitoring data. Hence,
settlement prediction and assumed effective depth of piled foundation is fairly reasonable.
• The settlement of actual monitroing data is slightly higher than that of prediction. This
may be due to the additional settlement caused by diaphragm wall panel construction
which influence zone might have extended to 15m from the diaphragm wall (equivalent
distance to bottom of soft clay level ).
Limiting values of the building slopes at two critical conditions specified by various
researchers as shown in the table below were also used in determining the potential building
damage of stage 3 assessment.
483
Table 5 Summary of Limiting Building Slope Specified by Various Researchers

Limiting Values by Various Researchers
Condition Type of Damage
Skempton and Meyerhof Polshin and Bjerrum
Macdonald Tokar
1 Cracking in walls and partitions 1/300 to 1/500 1/500 1/500 1/500
2 Structural Damage 1/150 1/250 1/200 1/150
6.5.3.2 Stage 3 Assessment by structural frame analysis
Some buildings of critical condition were assessed by structural analysis using detailed
structural survey information as basic data. By simulating the differential settlement of the
building predicted from Additional Stage 2 Assessment, development of moment and shear in
the main structural members were determined with the use of structural analysis software.
Computed moment and shear were assessed in comparison with the capacity of the structural
members estimated from collected building condition survey and detailed structural survey
data. The process involved in structural assessment can be summarized below.
• Factor of safety of each structural members were determined base on predicted
differential settlement of the buildings.
• A series of assessment were performed to determine the critical differential settlement
causing structural instability of the members or differential settlement at limiting
factor of safety.
• Trigger values were established based on above results.
• Based on the trigger values and the actual building condition and response in
conjunction with monitoring data, the most sensitive structural members were
protected, mainly by strengthening approach.
6.5.4 Requirement of Stage 3 Assessment for bridge structures
Due to the close proximity to the adjacent station, the predicted maximum settlement in the
Stage 1 Assessment of the flyover and elevated expressway structures was greater than
10mm. Therefore, further assessment using Stage 2 approach was required. However, since
flyovers and elevated expressways are bridge structure with piled foundations, the approach
of strain calculations used in Stage 2 Assessment was considered inappropriate and they were
placed under the requirement of the Stage 3 Assessment as listed in Table 6.
Table 6 List of Flyovers / Bridges required for Stage 3 Assessment

Name of the Bridge / Type of
Location Foundation Type
Flyover bridge
Hua Lamphong Second Stage Expressway Steel Deep-seated bored pile
Sam Yan Thai-Japan Flyover Steel Deep-seated bored pile
Si Lom Thai-Japan Flyover Steel Deep-seated bored pile
Lumphini Thai-Belgium Flyover Steel Deep-seated bored pile
Phetchaburi Asok Flyover Steel Pre-cast pile
484
Stage 3 Assessments of above listed bridges were carried out by soil-structure interaction
analysis using finite element method. FEM Analyses were mainly carried out by in-house
design team consisted of experience geotechnical engineers of Tokyu Construction Company
Limited. Two-dimensional non-linear soil model applying Duncan-Chang parameters was
employed. The process involved in soil-structure analysis is presented in diagram as shown in
Figure 5.
Assign Data and Conditions for Analysis

Structures
Ground Properties
Construction Sequence
Finite Element Analysis

Non-linear Elastic Analysis
Superstructure assessment Pile Assessment
Distribution of movement of the Distribution of horizontal ground
ground surface movement along piles
Calculation of vertical and horizontal

movement of the piers Structural analysis
Applied response
displacement
Assessment Assessment method
based on based on
differential horizontal Verification of RC section of
Settlement criteria movement criteria pile
Figure 5 Diagram showing process of stage 3 assessment for bridge structure
Among the bridges listed in Table 6, impact of excavation work on the structure of
Petchaburi flyover was considered the most critical due to its close proximity to the
diaphragm wall and having relatively less rigid and shallower foundation of pre-cast piles.
Detailed structural assessment of the existing pre-cast piles of the flyover was performed
using results of FEM and verified with structural capacities of the piles calculated from the
available as-built drawings. Predicted differential settlement of two adjacent piers (pier close
to diaphragm wall and pier away from diaphragm wall) caused by subsurface settlement at
pile toe level was used to calculate the angular distortion of the flyover superstructure and
compared with the tolerable angular distortion specified by for simple span bridge. Predicted
angular distortion (0.0003) was much lower than specified tolerable value of ASSHTO
(0.005). Predicted horizontal movement of flyover piers were 10mm and 12mm for left and
right pier (of Figure 6) respectively. Tolerable horizontal movement of flyover structure based
on the observations of actual bridge performance from the works of various researchers
(extracted from Xanthakos, 1995) are shown in Table 7.
485
Table 7 Tolerable horizontal movement of bridge recommended by various researchers
Horizontal Recommendation Recommended by
Movement (mm)
25 Not harmful Bozozuk (1978)
38 Tolerable in most cases Moulton et. al. (1985)
51 Harmful but tolerable Bozozuk (1978)
Extensive sets of instrumentation were installed including tilt meter, settlement points
and inclinometers. Monitoring was carried out in high frequency during critical stage of
excavation and compared with established trigger values.
Figure 6 Cross-section of Petchaburi Station diaphragm wall showing the position of

adjacent flyover foundations
7. PREDICTION OF GROUND MOVEMENT INDUCED BY TUNNELING AND RISK

ASSESSMENT
7. 1 Prediction of surface settlement
The Gaussian normal distribution curve developed by Peck (1969), O’Reilly and New (1982)
was used to estimate the ground settlement induced by a single tunnel.
First stage assessment of the risk of damage to buildings and structures in tunneling zones
involved preparation of settlement profile induced by tunnel excavation. Settlement was
computed at every 100m interval along tunnel alignment using the method based on Gaussian
normal distribution curve presented above. Settlement induced by each tunnel (north and
486
south bound) was predicted independently and then superimposed to determine the total
settlement induced by the twin tunnels. 10mm and 25mm settlement contours were plotted in
the topographic plan with buildings and other structures including flyover and khlong bridges
in the tunnel alignment. Buildings falling outside 10mm settlement contours were eliminated
from further assessment. Remaining buildings with settlement greater than 10mm were short-
listed for stage 2 assessment.
7. 3 Stage 2 Assessment
The buildings and structures short-listed in stage 1 were assessed in this stage. Stage 2
assessment involves calculation of maximum tensile strain induced in the building and
structures due to tunnel boring and classification of risk category. Risk category of the
buildings was determined applying the method proposed by Mair et. al. (1996), similar to the
building damage assessment made for excavation induced settlement presented in earlier
sections.
7. 4 Additional Stage 2 Assessment in tunnel zones
As most of the buildings and structures in Bangkok along tunnel alignment are supported by
piled foundations, it was considered that Stage 2 Assessment using surface settlement profile
might be insufficient. Hence, Additional Stage 2 Assessment considering the variation of
settlement trough shape and magnitude with depth (subsurface settlement) was carried out for
the critical buildings.
In general, subsurface settlement trough will be narrower and steeper than those of the surface
as shown in the sketch below. A simple approach was used in predicting the subsurface
settlement profile with following assumptions.
(1) The shape of the subsurface settlement profile caused by the tunnel construction is
characterized by a Gaussian distribution similar to surface settlement profile.
(2) The ground settlement influence line for tunnel in clay is the same for subsurface as
shown in the figure below:
Figure 7 Distribution of surface and subsurface settlement induced by tunnel boring
Therefore, substituting the distance above the tunnel axis Z0-Z for Z0, equations derived from
Gaussian normal distribution curve subsurface settlement profile was determined. In addition
to this method following equation can also be used to predict the subsurface settlement.
487
K’ = [0.14+0.26(1-z/z0)] / (1-z/z0)
By substituting K’ values for K in equation derived from Gaussian normal distribution curve
the subsurface settlement at required depth can be calculated.
7. 5 Stage 3 Assessment
7.5.1 Stage 3 Assessment for the buildings
The method used in Stage 3 Assessment of the buildings in tunnel zone was similar to that of
station excavation zone based on the superimposed predicted subsurface settlement trough
induced by twin tunnels.
7.5.2 Structures Required for Stage 3 Assessment
The most critical structure required for Stage 3 Assessment were MWA main water tunnel as
tunnel alignment crosses the water tunnels at two critical locations – the first near Rama IV
and Rachadaphisek road intersection and the second at Rama IV and Siphraya road
intersection near Sam Yan station. The position of water tunnels in relation to MRT tunnels
drawn in the subsoil profile at relevant locations are presented in Figure 8 and 9.
Figure 8 Position of MWA water tunnel in relation to MRT tunnels (clear distance are 3.38m
and 3.18m for North and South bound respectively) at the first crossing
According to the available information, MWA tunnel at the first crossing is of 2.5m
internal diameter steel lining with 150mm segmental ring and at the second crossing MWA
tunnel is of 2.5m diameter segmental ring with lightly reinforced concrete lining.
In preliminary stage 3 assessment, based on the predicted subsurface profile using the
method described in earlier section, structural assessment of water main was carried out and
results were compared with structural capacity of water tunnel calculated from the material
properties indicated in the construction drawings provided by MWA. As bending moment
induced in the water tunnel was considered critical parameter in assessing the damage which
is governed by both maximum settlement and curvature of the water tunnel further assessment
was proposed.
488
• To carry out back-calculation using monitoring data obtained from the settlement
arrays (settlement points, extensometer data) at first 3 locations before reaching to
MWWA water tunnel near Rama IV – Rachadapisek intersection, to determine the
volume loss and relationship between surface and subsurface settlement as well as
best estimation of subsurface K value and subsurface settlement profile at water
tunnel level.
• To perform beam-on-elastic foundation analysis using the subsurface settlement
profile obtained from above item (1) to determine bending moment likely to occur in
MWWA water tunnel.
• To compare the result obtained from above item with allowable values.
• To establish the final trigger values based on above comparison.
• To confirm the settlement profile of last array is within the alert level of final trigger
value.
Figure 9 Position of MWA water tunnel in relation to MRT tunnels (clear distance are 1.5m
and 2m for North and South bound respectively) at the second crossing
For the first location of MRT-MWA tunnel crossing, target ground loss from tunnel
operation was set at 0.5%, whereas ground loss based on allowable movement curvature of
water tunnel derived from predicted pattern of subsurface settlement was 0.65%. Back-
calculated ground loss derived from the monitoring data of rod-extensometer placed at the
crown of MWA water tunnel was within target value. In addition to the preliminary
assessment based on prediction of subsurface settlement using empirical formulas, a detailed
soil-structure interaction was analyzed by experience geotechnical engineers of Tokyu
Construction Company Limited using 2-dimensional FEM considering behavior of tunnel
boring by earth pressure balance (EPB) method. Stress relief from the effect of cutter-head,
tail void and over-cut from EPB tunnel boring was considered in the analyses. Adding ground
movement due to consolidation caused by disturbance from tunnel operation surface and
subsurface settlement were predicted by finite element analysis. Detailed parameters and
assumption made in analysis was presented in the published paper of Sakai & Sugden (2000).
After the successful passing of both tunnels at the first MWA crossing, another trial area
tunneling was planned to further determine the ground movement associated with various
tunneling parameters well before reaching to more critical crossing after Sam Yan station.
Tunnel section between Silom and Samyan station, in which soil profile similar to that of
second crossing was selected and comprehensive instrumentation program was established to
489
monitor surface and subsurface settlement as well as horizontal movement of the ground at
depth.
8. COMPARISON OF PREDICTION AND ACTUAL MONITORING RESULTS
The monitored ground deformation due to station box construction and bored tunneling were
compared with the predicted values for one of the station and tunnel sections to evaluate the
performance of the actual operation.
8.1 Ground Movement Induced by Excavation Works
The monitored wall deflections and ground settlement induced due to station constructions
works were within the design prediction. Adopted top-down method in excavation works of
station boxes utilizing rigid diaphragm wall extensively propped by stiff permanent slabs, was
the key factor contributed in minimizing associated ground movement and disturbance to
adjacent buildings and structures. Figure 10 shows the comparison of monitored wall
deflections with predicted values at final stages of station construction.
-10 0 10 20 30 40 50 60 70 -10 0 10 20 30 40 50 60 70
100 100
95 95
Reduced Level (m)
Reduced Level (m)
90 90
85 85
80 80
75 75
70 70
Deflection (mm ) Deflection (mm )
A mber_Co ncourse A mber_B ase
Red_Co nco urse Red_B ase
M o nitored_Conco urse M onito red_B ase
Figure 10 Diaphragm wall deflection showing predicted trigger levels and monitored data for
final excavation stages to concourse and base slab level
The comparison of monitored and predicted ground settlement profile resulting from
station excavation is presented in Figure 11.
S urf a c e S e t t le m e nt P ro f ile B e hind D / Wa ll R e s ult ing f ro m S t a t io n
B o x E xc a v a t io n
20
10
-10 -1
-4 -4 -2 -1
-2
-4 -4
-9
-20
-15 -17
-30 -22
-40 -30 -32
-50
DW_East DW_West
-60
-70
-60 -50 -40 -30 -20 -10 0 10 20 30 40 50 60
D is t a nc e , m
Initial Reading M easured-End o f Ro o f Exc.

M easured-End o f A ccess Exc M easured-End o f Co nco urse Exc
M easured-End o f B ase Exc P redicted-End o f Statio n B o x Excavatio n
Figure 11 The comparison of monitored and predicted ground settlement profile resulting
from station excavation works
490
8.2 Ground Movement Induced by Tunneling Works
The observed ground movement induced by tunneling works were within predicted values.
Well organized plan, systematic approach of observational method to achieve control values
backup by extensive instrumentation and close co-operation among relevant personnel were
the key factors contributed to successful completion of EPB tunneling particularly in the most
sensitive points at MWA water tunnel crossings. The monitored ground settlement profile in
the most critical section of parallel twin tunnel between Petchaburi and Sukhumvit station is
presented in Figure 12. As can be seen in the figure monitored settlement profile was within
that of prediction. The volume loss due to tunneling works in this tunnel zone was back
calculated using the monitored profile and found less than 2% and is typically in the range of
1.1% to 1.7%.
T ra ns v e rs e S e t t le m e nt T ro ugh due t o
T unne l B o ring, P e t c ha buri- S uk hum v it T unne l
20
10
-10
-20
-30
-40
-50 Monitored
-60 Predicted
NB SB
-70
-25 -20 -15 -10 -5 0 5 10 15 20 25
D is t a nc e , m
Figure 12 Predicted and observed ground settlement profile due to tunnel boring
Distance (m)
0 10 20 30 40 50 60 70 80 90
Subsurface Settlement (mm)
-10
Settlement _ NB Tunnel
-30
Settlement_SB Tunnel
Total Settlement
-50
491
Figure 13 Predicted subsurface settlement trough of assumed ground loss 1% calculated
from empirical formula and actual subsurface settlement (indicated by dark-circle) observed
from rod extensometer
9. CONCLUSION
The M.R.T Chaloem Ratchamongkhon Line, one of the most modern underground mass
transit systems in the region was successfully completed and has been in operation for over 1
year. Ground movement prediction and the detailed stage assessment carried out in the M.R.T
Chaloem Ratchamongkhon Line are presented. The measured ground and associated
movement of the buildings and structures were within predicted values. All deep excavation
works and tunneling were successfully completed without any significant damages to the
adjacent buildings and structures. Well organized plan, systematic approach of observational
method to achieve control values backup by extensive instrumentation and close co-operation
among relevant personnel were the key factors contributed to successful completion of the
project. The purpose of this paper however is not to extol success but to provide a set of
technical guidelines for ground movement prediction and building damage risk assessment of
deep excavation and tunneling works in Bangkok subsoil which may partly serve as a
reference for the future underground mass transit system projects in urban area of Bangkok.
Successful completion of the first MRT project in Bangkok marked the practicality of well-
executed deep excavation works and tunneling by EPB method in protecting the sensitive
urban environment.
10. ACKNOWLEDGEMENT
The authors wish to express their appreciation to Mass Rapid Transit Authority (MRTA) for
permission to publish this paper. The authors would like to thank Construction Supervision
Consultants (CSC 1 Consortium) and Project Management Consultant (MPMC) for their
cooperation. Valuable advice and support given by Dr. D. J. Maconochie (CSC Consortium)
and Mr. A. J. Polley (former geotechnical advisor, Sindhu Maunsell Consultants) are
gratefully acknowledge.
11. REFERENCES
Boscarding, M.D. and Cording, E.G. (1989), Building Response to Excavation-Induced

Settlement. Journal of Geotechnical Engineering, ASCE, 115;1;1-21
Bowels J.E (1988) Foundation Analysis and Design, 4th Edition, McGRAW-Hill International
Book Company, Singapore, 1988
Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings and Associated Damage.
SOA Review, Conference on Settlement of Structures, Cambridge, Pentech Press.
London, pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of Foundations and Structures,
SOA Report Session 2, Proceedings of 9th International Conference, SMFE, Tokyo,
2:495-546
Clough, G.W., and O’Rourke, T.D., (1990), Construction Induced Movements of In-situ
Walls, ASCE Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. O’Rourke (1990), Deep Excavation and Tunneling, 7th ICSME,
Mexico City
492
JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for MRTA Initial System -
UGS, Bangkok
Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of Ground Movements and
Assessment of Risk of Building Damage. Geotechnical Aspects of Underground
Construction in Soft Ground, pp. 712-718, Balkema, Rotterdam
O’Reilly M.P and New B.M (1982), Settlements above tunnels in United Kingdom – their
magnitude and prediction, Tunneling ’82, pp. 173-181, London, IMM
Peck R.B. , (1969) Deep Excavation and Tunneling in Soft Ground, 7th ICSME, Mexico City,
State of the Art 3, 225-290
Petros P. Xanthakos (1995) Bridge Substructure and Foundation Design, Prentice Hall
PTR, USA, 1995
493
Proceedings of the Observational Method and its Application in Hong
Kong Seminar
18 November 2006, Hong Kong
Observational Method and its

Application in Hong Kong
Key Factors in Application of Observation

Method – Bangkok Experience
Noppadol Phienwej
Zaw Zaw Aye

495
KEY FACTORS IN APPLICATION OF OBSERVATION METHOD
– BANGKOK EXPERIENCE
Noppadol Phienwej1 and Zaw Zaw Aye2
Abstract: Application of observational method in two different projects is presented in this

paper. Intensive modification of construction sequence in actual work execution with “value
engineering options” different from tender stage design along with application of
observational method in Diaphragm-wall-support two level underground car park building
located in the historically and culturally significant area of Bangkok is firstly demonstrated.
Simple approach of observational method in building damage assessment backup by
extensive instrumentation in tunneling and deep excavation works of Contract No. 1 of the
M.R.T Chaloem Ratchamongkhon Line, the first underground mass transit system project of
Bangkok (MRTA) is also presented. Key Factors contributed in successful application of
observational method are discussed.
INTRODUCTION
Observational Method is not a very popular term among geotechnical engineers in

Thailand, though the method has been applied in some major projects. The main reason of
unpopularity of the method in Thailand may be due to the fact that traditionally there is less
option opened for design modification as the project progresses. Geotechnical engineers may
be more concerned on misused and limitation of the method and do not prefer to take risk. It
is important to discuss among geotechnical engineers with regard to overall philosophy, key
requirement, limitation and practical approach in application of observational method. In this
paper, application of observation method in two different projects is presented.
CONSTRUCTION OF UNDERGROUND CAR PARK IN HISTORICAL

AREA OF BANGKOK
The project is a two-level underground car park located in the center of Rattanakosin
Island, the heart of an old established, historically and culturally significant area of Bangkok.
The project owner, Bangkok Metropolitan Authority (BMA) awarded the semi-turnkey basis
construction contract “Lam Kon Muang Underground Car Park” to SEAFCO Co., Ltd. as a
contractor. The contract consists of 3 major scope of works : (1) Construction of building
foundation and retaining structure - diaphragm wall, barrette and bored piles (2) Excavation
works including temporary bracing design and installation (3) Construction of the entire
two-level underground car park building having car park area of 18,552 m2 and roof-level
park of 10,936m2 plus cut-and-cover tunnel, underpass access to the City Hall. Geotechnical
1
Associate Professor and Coordinator, Geotechnical and Geo-environmental Filed of Study,
Asian Institute of Technology, Thailand
2
Senior Vice President, Seafco Public Company Limited, Thailand
497
aspects highlighting the application of observational method and performance of
buttress-support diaphragm wall of 0.60m width for two level underground car park building
is discussed in this paper.
Project Requirement and Major Constraints
Since there is a limited availability of car parking space in the surrounding congested
neighborhood and the project site was being used as a grade-level car parking space prior to
the award of the contract, the key requirement was to construct the underground car park in
two phases – to construct Phase 1 while leaving space for car parking in Phase 2, and to
utilize semi-finished underground car park of Phase 1 during construction of Phase 2. This
requirement posed the need of temporary retaining wall between Phase 1 and 2.
Construction site is surrounded by numbers of sensitive structures as shown in Fig. 1, -
in the south, Wat Suthat, one of Thailand’s most important temples and the Historical
Giant-swing, in the north, the City Hall and at South-east corner, the Historical Brahmin
Temple. Rows of old shop-house buildings are closely located in the east and west boundaries
of the project. Location of the project site itself in the vicinity of sensitive structures and
buildings therefore posed some constraints, which called for the need of careful consideration
in establishing the design principles and sequence of construction.
CITY HALL
N
MAHUNNOP ROAD
OPENING
OPENING
SHOP-HOUSES
OPENING
DINGSOR ROAD
OPENING
SIRIPONG ROAD
SHOP-HOUSES
OPENING
OPENING
SHOP-HOUSES
OPENING OPENING
BAMRUNG MUANG ROAD

GIANT-SWING
WAT SUTHAT (TEMPLE)
Fig. 1. Layout plan of the project showing adjacent buildings
498
N I-5 N O R T H D -W A L L
I-4 I-6
PH A S E 2
EAST D-WALL
WEST D-WALL
S H E E T PIL E
124 m
S ee d etails in (b )
I-3 I-1
PH A S E 1
I-2
S O U T H D -W A L L
78 m
(a)
BARRETTE W ALE BEAM
0 .3 0
T IE B E A M
P IL E C A P SLA B
(b)
+0.20 ROOF SLAB
0.00
CAPPING
G.W.L
BEAM
RC BUTTRESS -2.70 SLAB (B1)
2
TIE PILE CAP
BEAM -5.60 (B2)
SOFT CLAY 2
-9m WALE SOFT CLAY

BEAM
DIAPHRAGM MEDIUM CLAY

-12m WALL (tip=-16m)
STIFF CLAY
-15m
-16m
VERY STIFF CLAY
FOUNDATION PILES
BARRETTE DIA. 600mm (tip=-20m)
-20m. (tip=-20m) -20m.
Fig. 2 (a) Layout of diaphragm wall and (c)

piles (b) plan of diaphragm wall and buttress (c)
sectional view of the structures
Location of the project site itself in the vicinity of sensitive structures and buildings
therefore posed some constraints, which called for the need of careful consideration in
499
establishing the design principles and sequence of construction. The architectural and utility
aspects of the project called for the design of the basement with a number of openings from
the ground surface to the final basement slab level to facilitate the ventilation system as
shown in figure 2.
Hence the roof slab can not be physically utilized as bracing in most of the area where
the diaphragm wall is to be acting as a cantilever retaining wall in the permanent stage. It was
analyzed in the preliminary analyses that the deflection of the diaphragm wall of 0.60 m width
would be large if it was to be fully cantilevered. As the project is located in a sensitive area,
ground movement induced by large deflection was unfavorable. It was therefore decided to
use buttress to minimize the diaphragm wall deflection as shown in figure 2.
Diaphragm Walls Barrettes and Bored Piles Construction
Diaphragm wall having 600mm width founded at 16m below ground level (B.G.L) was
constructed simultaneously with dry-processed bored piles of diameter 600mm with toe depth
20m below ground level. Barrettes having same toe depth as bored piles were installed at 8m
spacing along with diaphragm wall panels. Sheet pile wall (14m deep) was used as a
temporary retaining wall at the boundary of Phase 1 and 2 as shown in Fig. 3.
Subsoil Condition and Design Parameters
Typical subsoil profile at the site is characterized by thick Bangkok soft clay layer at the
top followed by thin layer of medium clay, and stiff clay layers.
Undrained Shear Strength (kN/m2)
1 0 50 100 150 200
0 0.00
5 Soft clay 5.00
10 Medium stiff clay 10.00

Stiff clay
15 15.00
20 Very Stiff clay 20.00
25.00
25
30.00
30 Medium Dense
Sand 35.00
35
40.00
40
Dense sand 45.00

45
50.00
50
0 20 40 60 80
Su BH-1 Su BH-2
Su BH-3 Design (Su)
SPT BH-1 SPT BH-2
Fig. 3. Plotted Su and SPT N-value with design lines
500
Undrained shear strength (Su) and SPT N-value obtained from 3 SI boreholes were
plotted and design line was derived as shown in Fig. 3. The design soil parameters are
tabulated in Table 1.
Table 1. Design soil parameters

Depth γs Su Eu
Soil Type (kN/m2)
(m) (kN/m3) (kN/m2)
Soft clay 0-9 16.50 18 7000
Medium clay 9 – 12 17.50 30 19250
Stiff Clay 12 – 15 19.00 60 45000
Very stiff clay 15 - 26 19.50 80 60000
Original Tender Stage Design
The designers involved in the tender stage design made a fairly conservative design with
two levels temporary bracing as shown in Fig. 4. Uncertainty of the performance of a thin
diaphragm wall in soft clay layer was the likely reason to adopt the conservative design in the
tender stage. It is not unreasonable to adopt the conservative design considering
time-dependent consolidation property of soft marine clay and likely long elapsed time of
un-strutted diaphragm wall (relatively long un-strutted span of about 6.0 m between first
strut and final excavation level for 600mm diaphragm wall) due to large volume of excavation
work involved.
HORIZONTAL STRUT SLAB
RAKING STRUT
D-WALL
BERM
D-WALL
BARRETTE
-16m
-20m BORED PILE
Fig. 4. Tender stage bracing system – diaphragm wall was designed with soil-berm and 2
struts support (horizontal bracing and raker)
Value Engineering Option for Phase 1
Value engineering review of the temporary works was undertaken by the contractor’s
new in-house design engineering team prior to the commencement of Phase 1 excavation
works. Rigorous attention to detail of the design concept and constructability was made in the
pre-construction discussions between design engineers and construction team.
501
The main objectives for value engineering options were to minimize the material and
construction sequence involved in temporary works so as to accelerate the excavation time
thereby saving overall costs. without compromising the safety aspect. After conducting a
series of re-analyses with different conditions major modifications were made : (1) To lower
the first strut level to –1.8m from the original tender stage design level –1.0m (2) To use only
1 temporary strut, omitting second level raking strut with the provision of sloping soil berm
against diaphragm walls. Soil berm was to remove after completion of base slab construction
in the majority of area – minimizing the elapsed time of partially un-strutted diaphragm wall
between temporary strut and the final excavation level. Modified bracing system of Phase 1 as
an outcome of value engineering review is illustrated in Fig. 5.
HORIZONTAL STRUT SLAB HORIZONTAL STRUT SLAB
BERM
RAKING STRUT
D-WALL
SHEET PILE
BARRETTE
-14m
-16m -16m
-20m -20m BORED PILE

Fig. 5. Value engineering option - Modified Bracing system for Phase 1
Implementation of the Observational Method

Professor R. B. Peck set out procedures for the observational method (OM) as
applied in soil mechanics in the Ninth Rankine Lecture (Peck, 1969). Peck described the
limitation and drawbacks of observational method. Powderham (1996) reviewed the main
features of the observational method of Peck and summarized the key requirements as
follows;
(1) It must be possible to alter the design during construction
(2) The contractual condition must be compatible and allow design to be directly related
to actual construction method
(3) An acceptable level of risk must be identified and controlled. In particular this requires
a planned course of action for every foreseeable eventuality
(4) Critical observation must be identified and obtained
During the review of tender stage design for Phase 1, it was recognized that
two-phase excavation works in this project was ideally suitable for application of the
observation method. The flexibility of the contractual requirement in temporary design which
allowed the contractor to modify the design and construction method also provided the favor
for the observational method.
In order to assess the most probable condition assumed in “value engineering design”
and to take necessary actions if monitoring results reveal most unfavorable conditions (i.e.
502
actual deflection of diaphragm wall reaches maximum acceptable limit), the observational
method was implemented on the followings basis.
 Reviewed the design parameters together with critical conditions posed on sites as well as
most critical stage in excavation and basement construction work
 Predicted the performance of diaphragm wall with “most probable” as well as “most
unfavorable”conditions and parameters.
 Established the trigger criteria based on predicted diaphragm wall deflection
 Predefined the practical contingency plan for “most unfavorable” conditions where wall
deflection reaches trigger levels
 Set out the instrumentation program with the consideration of above factors
 Monitored the performance of diaphragm wall. Compared the monitoring results with the
predicted and trigger values and reassessed
 Implemented the contingency measures if monitoring result reaches action level of trigger
values
Figure 6 shows the predicted diaphragm wall lateral displacement or deflection of
Phase 1 (east, west and south diaphragm wall) at two conditions together with trigger levels
and tender stage prediction. It should be noted that the diaphragm wall deflection was
predicted to be maximum or most critical after removing the horizontal temporary strut.

0 20 40 60 80 100
0
NORMAL ALARM ACTION
2 MOST
PROBABLE
PREDICTION
6
MOST
8 UNFAVORABLE
PREDICTION
Depth, m.
10
12
14
Tender Stage
16 Prediction with buttress :

Alarm Trigger Level
18
20
Fig. 6. Prediction of diaphragm wall performance in pre-construction stage with trigger

criteria in comparison with tender stage prediction
Most probable condition was established for the predicted deflection of diaphragm
wall with full influence of buttress support – assuming buttress effectively supports as
permanent strut in diaphragm wall analysis model. Most unfavorable condition was set out for
the predicted deflection of diaphragm wall without considering influence of buttress –
buttress was excluded in the model. Diaphragm wall reinforcement was designed based on the
503
most unfavorable condition.
In establishing the criteria for trigger values, it was necessary to consider the broad
context in which diaphragm wall exists, design assumption and concept, likely behaviour of
the wall itself or its predicted performance and effectiveness of selected temporary bracing
system. Trigger levels were established to provide the design team and construction team, an
opportunity for early review and resetting of the monitoring frequency as well as for
implementing the contingency measure as necessary.
In general, exceeding alert trigger levels must initiate a review of design data,
construction progress and monitoring frequencies with the consideration of possible measures
to limit further deflection. Exceeding action trigger levels must initiate further review of
above mentioned points and if necessary to initiate a planned course of action or contingency
measures.
Effective and good communications between the design team and construction crew
were made all along the excavation stages with clear responsibilities in construction control.
The contingency measures included immediate backing filling of excavated soil and
installing of temporary struts in the critical area.
Monitoring Program
In planning the monitoring system and program, it is important to consider the
parameters to be measured which reflect the actual performance of the diaphragm wall
support excavation. It is also necessary to take account the practical measurement applicable
for established trigger criteria, number and frequency of measurement required to carry out a
meaningful interpretation of wall behaviour which would be integrated in the implementation
of observational method.
Comprehensive and robust monitoring program was set up as a key element in
application of the observational method and to ensure that modified construction sequence
would not have adverse effect in temporary stage and on permanent design. A total of 6
inclinometers (3 in each phase) were installed in diaphragm wall together with some survey
points. In order to make effective use of the established trigger levels, an adequate number of
measurement were carried out at appropriate frequency. Typical monitoring frequencies for
inclinometer set up as guideline for the project is outlined below.
• Measured immediately before commencing excavation in the vicinity of instrument
• Minimum readings of 2 times a week while excavation in progress
• Minimum readings of 1 time a week when no excavation the vicinity of instrument
• Minimum readings of 3 times a week when measured deflection values exceeded alert
trigger levels
• Minimum reading 1 time a day when measured deflection values reached action trigger
levels
As the most critical stage was predicted at the time horizontal temporary bracings
were removed, a full attention was paid to the inclinometer monitoring with the following
special criteria and frequencies.
• First temporary strut removal was to carry out at the diaphragm wall panel where
inclinometer was located
• Measured immediately before removal of temporary strut at the closest distance to the
instrument.
• Measured every 6-8 hours immediately after removing the first strut
• Second strut to be removed must be the one located immediately adjacent to first strut
504
which had removed
• Not to remove the second strut until inclinometer measured deflection values had
stabilized
• Not to remove more struts unless measured deflection values were stabilized and within
alarm trigger levels
In addition to inclinometer measurement, diaphragm wall movement was also
monitored by the survey points strategically marked on the wall panels. Ground settlement
and surface cracks behind the diaphragm wall were also visually checked by the construction
team as daily basic.
Performance of Phase 1 Diaphragm Wall

After carrying out the comprehensive desk studies and establishment of systematic
monitoring program presented above, Phase 1 excavation work was carefully commenced.
Figure 7 shows the maximum accumulated diaphragm wall deflection at different stages of
excavation monitored by inclinometer No.1 (I-1 at East wall of Phase 1) together with trigger
levels and predicted maximum deflection profile of 3 different conditions - tender stage
design (2 temporary struts), modified design with buttress and modified design without
buttress. It can be observed from figure that measured lateral movement pattern of diaphragm
wall agreed well with that of prediction for modified design with buttress - meaning
buttress-support has significant influence on wall deflection.

0 20 40 60 80 100
0
NORMAL ALARM ACTION

2
6 MOST
UNFAVORABLE
PREDICTION
8
Stage 1
Depth, m.
10 Stage 2
Stage 3
12
Stage 4
14 Stage 5
Tender Stage
16
Alarm Trigger Level

20
Fig. 7. Phase 1 East Wall-Monitored diaphragm wall deflection at different stages with
trigger levels
505
0 20 40 60 80 100
0
NORMAL ALARM ACTION
6
MOST UNFAVORABLE
8 PREDICTION
Depth, m.
Stage 1
10
Stage 2
12 Stage 3
Stage 4
14 Stage 5
Tender Stage
16
Alarm Trigger Level

20
Fig. 8. Phase 1 South Wall - Monitored diaphragm wall deflection at different stages with
trigger levels
Deflection profile of South diaphragm wall which braced against temporary sheet pile
wall at Phase 1 and 2 boundaries is presented in Fig. 8. As can be seen in Fig. 8, South
diaphragm wall deflection is significantly higher than that of east wall, which is likely to be
caused by the fact that South diaphragm wall is braced with more flexible sheet pile wall.
description of stages shown in the legend of Fig. 7 and 8 is summarized in Table 2.
Table 2. Description of stages shown in Fig. 7 & 8

Stage Description
1 Excavate to –2.2m and installed temporary strut
2 Excavate to –4m
3 Excavate to –6.6m with berm
4 Removal of berm
5 Removal of temporary strut after completion of buttress
506
Fig. 9. Phase 1 excavation work in progress with historical Buddhist temple Wat Suthat in
background (Phase 2 area was being used as car parking space)
Modification of Phase 2 Bracing System
Monitoring results of the Phase 1 excavation work provided an ample opportunity to

review the design assumption, fine tune the parameters used in the analysis of the diaphragm
wall for the Phase 2 and made modification of construction sequence. The major
modifications are : (1) Removal of soil-berm at East and West diaphragm wall in shorter
duration than that of Phase 1, and (2) Using raking struts instead of horizontal strut for North
diaphragm wall a shown in Fig. 10.

STRUCTURE
RAKING STRUT
BERM SLAB
BERM
D-WALL
D-WALL
BARRETTE
-16m -16m
-20m -20m
Fig. 10. Bracing system of Phase 2 diaphragm walls

507
Performance of Phase 2 Diaphragm Wall
Figure 11 depicts the measured deflection of east diaphragm wall. As can be

observed in Fig. 11 in comparison with Fig. 7, the maximum deflection of east diaphragm
wall in Phase 2 is larger than that of Phase 1.
The likely reasons of this observation are;

 Un-strutted elapsed time for first temporary bracing in Phase 2 was longer than that of
Phase 1.
 In Phase 1, horizontal struts were installed in north-south direction which temporary
kingpost columns and strut were integrated in crisscross pattern with east-west struts -
providing complete-support more rigid bracing system. Whereas in Phase 2, horizontal
struts were installed only in east-west direction without having crisscross pattern with
north wall – having less rigid bracing system than that of Phase 1.
0 20 40 60 80 100
-2
NORMAL ALARM ACTION
6
MOST UNFAVORABLE
PREDICTION
8
Depth, m.
Stage 1
10
Stage 2
12 Stage 3
Stage 4
14
Stage 5
18
20
Fig. 11 Phase 2 East Wall - Monitored diaphragm wall deflection at different stages with
trigger levels
With assurance of diaphragm wall performance from monitoring results of Phase 1,

original plan of using horizontal struts for North diaphragm wall was modified by using
raking struts instead. As can be seen in Fig. 12, deflection of North diaphragm wall (with
raking strut support) is significantly higher than that of East diaphragm wall (with horizontal
508
strut support). The main reason of larger movement of north diaphragm wall is due to the fact
that it was supported only by the berm for the long period (about 52 days) before completion
of raking struts so that soil-berm became soften during the long elapsed un-strutted period.
Time-dependent deflection pattern due to softening and deformation of soft clay can be
observed in North diaphragm wall as illustrated in Fig. 13. North diaphragm wall moved
progressively toward excavation before completion of raking struts (at 52 days) as can be seen
in Fig. 14.
Table 3. Description of stages shown in Fig. 11 & 12

1 Excavate to –2.2m and installed Excavate to –2.2m at d-wall, and to –6.6m
temporary strut with sloping berm
3 Excavate to –6.6m with berm Removal of berm
4 Removal of berm Removal of raker after completion of
buttress
5 Removal of temporary strut after

0 20 40 60 80 100 120 140
-2
NORMAL ALARM ACTION
0
6
MOST UNFAVORABLE
PREDICTION
Depth, m.
10
Stage 1
12 Stage 2
Stage 3
14
Stage 4

18
20
Fig. 12. Phase 2 North Wall - Monitored diaphragm wall deflection at different stages with
trigger levels
509
Fig. 13. View of buttress-support diaphragm wall prior to temporary bracing removal
0 20 40 60 80 100 120 140

0
6
Depth, m.
10
0 days
12
30 days
14 52 day
56 days
16 74 days
Prediction
18
20
Fig. 14. Phase 2 – Time dependent wall deflection of North diaphragm wall – diaphragm wall
was supported only by soil-berm for 52 days before completion of raker installation.
Implementation of Contingency Plan
Since deflection of North diaphragm wall (Phase 2) approached action trigger levels,
monitoring frequency was increased and the following contingency measures were
510
implemented on site.
 Poured 15cm thick 1m wide lean concrete on the top of the berm along North diaphragm
wall to provide bearing-effect
 Installed additional king-post and diagonal struts attached to the raking struts to provide
more rigid support against diaphragm wall
 Soil-berm was removed locally in bays followed by construction of wale beam, tie beam
and buttress as shown in Fig. 15.

TEMPORARY WALER
BUTTRESS
DIAGONAL STRUT
DIAPHRAGM WALL
CAPPING BEAM TEMPORARY KING-POST
RAKER
TEMPORARY
BERM STEEL BEAM
PILE
CAP
B
SLA
TED
PLE
COM
Fig. 15. Perspective of North diaphragm wall constructed in bay
Movement of diaphragm wall was observed to be decreased and eventually stabilized by the
above actions. No significant ground settlement was observed in the vicinity of the North
diaphragm wall.
Fig. 16. View of raker and soil-berm support – soil-berm was removed locally bay by bay
511
Fig. 17. View of buttress-support diaphragm wall after removal of temporary bracing
Time and Cost Saving from Value Engineering Options and the Application of
Significant cost and time saving were achieved from the value engineering option coupled
with observational method implemented for both Phase 1 and 2. The major savings were
achieved by less operation and material utilized in the following elements of temporary works.
 Cancellation of 2nd level raking struts against diaphragm wall for both phases
 Modification of bracing system – using raking struts with soil-berm support instead of
horizontal struts for North diaphragm wall in Phase 2
Key Factors Contributed in Successful Application of Observational Method
Outcome of a through desk study at post-tender stage provided an effective value

engineering option which offered significant cost and time saving for overall construction
program. Effective and good communications between the design team and construction crew
played a key role in successful application of observational method to completion of the
project. Systematic monitoring program with clear defined trigger criteria was also the
512
important element in implementing the observational method. This case study reveals that a
thin permanent diaphragm wall coupled with effective design and construction method
supplemented by the observational method and robust monitoring program could offer a
logistically and financially attractive solution in construction of underground car park without
disturbing the environment in the prominent historical area of Bangkok.
CONTRACT 1 OF MRTA CHALOEM RATCHAMONGKHON LINE PROJECT
Observational method was applied mainly in building and structure damage assessment
in this Bangkok very first mass transit system project. The M.R.T Chaloem Ratchamongkhon
Line is the first underground mass transit system project in Bangkok developed by the Mass
Rapid Transit Authority (MRTA) of Thailand. Total length of the underground structure is
21.5km, comprises of 16 km of twin single-track bored-tunnel, a total of 4 km long18 cut-
and-cover stations, a twin 1.5km long cut-and-cover approach tunnel to Depot and other
associated structures. The project was commenced in late 1996 and opened for the public in
2004. The Contract No. 1, Underground Structure South - southern portion of this initial
system comprising 9 km of twin 6m outside diameter bored tunnels, nine underground
stations and cut-and-cover depot approach tunnel, was awarded to the Joint Venture BCKT
consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch Karnchang Public Company
Limited, Kumagai Gumi Company Limited and Tokyu Construction Company Limited. The
project was commenced in late 1996, successfully completed and opened for the public in
2004.
One of the major challenges of the Contract No. 1was the need of stacked-alignment for
a major portion of twin bored-tunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations, Lumpini, Si Lom and Sam
Yan. This requirement led the contractor to design and construct the deepest underground
structures of Bangkok at Si Lom Station - four-floor stacked-station of 32m deep right under
the existing flyover and adjacent to a number of sensitive structures. Deep excavation and
bored tunneling works in congested urban environment of Metropolitan Bangkok requires a
systematic process of ground movement prediction, building risk damage assessment and
protection.
Subsoil Condition along Contract 1 Alignment
Similar to other localities in Bangkok, a typical subsoil profile along the tunnel
alignment is characterized by the alternating layers of clay and sand deposits as shown in
Figure 18. Weathered crust of 2 m thick is commonly found as the top layer. In urban areas
of Bangkok, this layer is covered by fill material. Soft to very soft, highly compressible dark
gray marine clay lies beneath weathered crust or fill. Depending on the location, this layer
extends up to 12-16 m. About 2 m thick medium clay layer can be observed between soft
clay and underlying stiff clay. Generally stiff Clay layer occurs directly underneath medium
clay and its depth goes up to 22 m. Below stiff clay layer, first sand layer of 5-8 m thickness
can be found. This first sand layer, however, is absent in some areas. Stiff to hard clay layer
underlies first sand and it is found to be about 5 m thick. Second sand layer generally occurs
at depths between 45 to 65 m.
513
Figure 18. Generalized soil profile along the tunnel alignment and stations
Overview of Ground Movement Prediction and Building Damage Assessment Process

Applied in the Project
Well accepted published literatures were used as basic reference and method in ground
movement prediction and building damage assessment. For the station boxes where
cut-and-cover methods were used for excavation, the work of Peck (1969) and Clough and
O’Rourke (1990) were used whereas published papers of Burland et. al. (1977) and Boscardin
& Cording (1989) were applied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and Stage 3 Assessment.
“Greenfield Site” assumption was adopted in Stage 1 and 2 Assessment ignoring any
influence of the stiffness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assessment. In Additional Stage 2
assessment, the variation of settlement trough shape and magnitude with depth were
considered in identifying the risk of building damage. In Stage 3 assessment, a detailed
qualitative assessment of structure and vulnerability of each building or structure to damage
were taken into account, coupled with detailed structural analysis in assessing the buildings
falling in risk category 3 classified by Mair et. al. (1996).
514
Summary of Instrumentation Program
Instrumentation played a major role in the construction of this first underground mass
transit system project in congestive urban area with difficult soil condition. The primary
objective of the instrumentation program was to monitor the performance of the deep
excavation and tunneling operation to ensure safe execution of the construction works and
that adjacent structures were not adversely affected. Furthermore, the instrumentation
program was established to provide feedback in application of the observational method.
Extensive sets of instrumentation were installed around the predefined influence zones of
station box excavation, tunneling and associated building, structures and utilities. In
addition to other geotechnical instruments, inclinometers, extensometers, various types of
settlement points, tile-meter, crack meter and crack gauge were installed to monitor the
ground and associated building and structures movement.
Building Condition Survey
Prior to commencement of the major works, all the existing buildings and structures
within the predefined influence zone were systematically surveyed. The primary objectives of
the surveys were;
 To record pre-construction status of all buildings and structures within the influence
zone. The records included were age of building, building type, storey height,
expected foundation types, visible defects supported by sketches and photographs as
appropriate
 To utilize the information collected from the building / structure condition survey in
the assessment of the buildings and evaluate sensitive structures with respect to
construction aspects and monitoring
 To establish a benchmark to monitor the possible effect of construction and feedback
information in applying observational method and if necessary to implement the
building protection measure
Expected foundation types of different buildings were classified based on the height,
number of floors and age of the buildings. Building survey information was also used to
identify the types of instrumentation required.
Building Damage Assessment and Observational Method
Process of Stage-Assessment is shown in figure 20. After carrying out the initial
assessment of potential building damage, it was recognized that without applying the
515
observational method, the result obtained from assessment process could be conservative for
some buildings and could be underestimated for others. For example, the settlement estimated
in Second Stage Assessment shown in above diagram was based on surface movement. For
the buildings located in close proximity to the excavation founded on piles, it was considered
that subsurface settlement may be more critical. Stage 2 Assessment was then further
extended by Additional Stage 2 Assessment in which the variation of settlement trough shape
and magnitude with depth were considered in identifying the risk of building damage.
Settlement of the buildings due to the excavation works were predicted assuming the “green
filed” condition. Neglecting restraints from foundation and structure, it was assumed that
buildings follow the ground settlement trough at foundation level (estimated pile tip level).
However, if the building is supported by deep foundations, assessment result could be too
conservative.
Stage 1 No further
Prediction and preparation of surface settlement < 10mm
assessment
(Green Field Site Condition)
>10mm
Stage 2 No further
Determination of building risk category < slight
assessment
(limiting tensile strain) based on surface settlement
> slight
Additional Stage 2 No further

Determination of building risk category (limiting tensile strain) < slight assessment
based on sub-surface settlement
> slight
Stage 3
Detailed Assessment
(building structural check, consider building stiffness, Soil-structure interaction)
Figure 19. Diagram showing process of Staged-Assessment
Available methods were further reviewed to predict the surface and subsurface
settlement with the consideration of two main factors; (1) simple and practical in application
(2) enable to correlate with the predicted and measured diaphragm wall deflection. The
method proposed by Bowels (1988) was selected as it meets the required criteria mentioned
above. Bowels suggested the ground settlement induced by excavation as a function of ground
loss due to the deflection of the retaining wall. Bowel demonstrated the calculation of
516
settlements at specified distance by assuming parabolic variations of settlement within the
influence distance. Using predicted diaphragm wall deflection, surface settlement behind the
wall was computed by empirical formulas proposed by Bowels.
A simplified prediction of subsurface settlement was carried out based on the
calculated surface settlement described above. First, subsurface settlement influence line was
constructed. As shown in Figure 20, settlement influence zone is assumed to decrease with
depth from “Do” at the surface and zero at the wall toe. With the assumption of linear
relationship between the volume of deflected wall shape and the volume of settlement trough
at any depth within settlement influence zone, subsurface settlement at different depths were
calculated.
X
Do = 2.5 HE
W D O = HW + W 4V O
S WO 
 x 
S io  SWO *  
Sio
S WO
D-wall  DO 
Dy D
HE
DY  Y * O
HW
HW

trough volume, VTy VTY  VY *
Siy
VO
Settlement
Swy
influence zone
Y
Figure 20. Demonstration of subsurface settlement prediction from diaphragm wall

deflection values
Based on the polynomial curve developed in Stage 2 Assessment, subsurface

settlement can be predicted by the similar approach described above. With predicted surface
and subsurface ground movement, risk of damage to the buildings were determined by two
methods (i) Risk of damage based on maximum tensile strain, the method proposed by Mair
et. al. (1996) and (ii) Risk of damage specified by angular distortion and horizontal strain
using the chart proposed by Boscarding and Cording (1989). Table 4 shows the summary of
building damage assessment, risk category obtained for different types of foundation
determined by two different methods.
517
Table 4 Summary of risk category obtained from two different methods and foundation
type
Building No. of Building Estimated Distance Risk category by Risk category by
Block stories Type Pile Length to edge maximum tensile chart of Boscarding
of wall strain and Cording (1989)
Without With Without With
pile pile pile pile
A 2 Masonry / 7-14m 18m 0 0 1 2
Concrete
B 2 Masonry / 7-14m 17.5m 0 0 1 2
Concrete
C 2 Masonry / 7-14m 32m 0 1 1 2
Concrete
D 1 Masonry / < 6m 2m 1 1 1 2
Concrete
E 37 RC 50m 8m - 0 - 0
The predicted degree of damage of the buildings using simplified chart is more critical
than that of maximum tensile strain method as can be seen in above table. According to the
results obtained from of simplified chart, predicted degree of damage was higher if buildings
were assumed to be with piled foundation rather than without piled foundation.
In order to justify the most probable condition assumed in building damage assessment
and to take necessary actions if monitoring results reveal most unfavorable conditions (i.e.
ground movement induced by deflection of diaphragm wall reaches maximum acceptable
limit), observational method was implemented on the followings basis.
 Reviewed the building condition survey record paying particular attention to the type of
foundation and existing visible defects
 Predicted the ground movement based on the calculated diaphragm wall deflection
 Performed structural assessment based on the predicted ground movement (for Stage 3
assessed buildings)
 Established the trigger criteria based on the magnitude of ground movement (maximum
settlement and differential settlement between foundations or structural columns)
 Predefined the practical contingency plan for “most unfavorable” conditions where
ground movement reaches trigger levels (ground movement was derived using simple
approach based on diaphragm wall deflection described in earlier section)
 Set out the instrumentation program with the consideration of above factors.
 Strictly controlled the construction sequence assumed and defined in design and if
necessary changed or modified the sequence
518
 Monitored the performance of retaining wall both diaphragm wall (for station boxes)and
sheet pile walls (for station entrance structures) . Compared the monitoring results with
the predicted and trigger values and reassessed
 Monitored the building settlement. Compared the monitoring results with the predicted
and trigger values and reassessed
 Implemented the contingency measures if monitoring result reaches action level of trigger
values
Monitoring results obtained from early stage of the project were used to reassess the potential
damage of the buildings and structures. Some buildings of critical condition were put in Stage
3 assessment requirement and assessed by structural analysis using detailed structural survey
information as basic data. By simulating the differential settlement of the building predicted
from Additional Stage 2 Assessment, development of moment and shear in the main structural
members were determined with the use of structural analysis software. Computed moment
and shear were assessed in comparison with the capacity of the structural members estimated
from collected building condition survey and detailed structural survey data. The process
involved in structural assessment can be summarized below.
 Factor of safety of each structural members were determined base on predicted
differential settlement of the buildings.
 A series of assessment were performed to determine the critical differential settlement
causing structural instability of the members or differential settlement at limiting factor of
safety.
 Trigger values were established based on above results.
 Based on the trigger values and the actual building condition and response in conjunction
with monitoring data, the most sensitive structural members were protected, mainly by
strengthening approach.
The following assumptions were made in the conservative approach of Stage 3
assessment.
 Buidling movement is influence by excavation induced settlement of both main station
box and associated underground structures (e.g. Entrance Building)
 Building settlement was casued by subsurface settlement at foundation level as buildings
are supported by piled foundation.
 No allowance was made for structural stiffness of the building.
 No allowance was made for possible effects of soil-structure interaction.
A summary of risk category determination in Stage 3 Assessment based on the predicted

and monitored settlement of the building adjacent to station excavation work is shown in the
Table 4 below.
519
Table 4 Summary of risk category determination based on Boscarding and Cording
(1989) – based on prediction and actual monitoring of settlement
Data Taken Horizontal Strain Angular Distortion
Risk Degree of
From Sh1 Sh2 L Sv1 Sv2 S L
h  Category Damage
(mm) (mm) (m) (mm) (mm) (mm) (m)
Settlement
31 24 18.6 0.0004 48 23 25 18.60 0.0013 2 slight
Prediction
Actual 0.0004
31* 24* 18.6 53 14 39 18.60 0.0020 2 slight
Monitoring Data *
*Predicted horizontal movement was used in computing horizontal strain for actual
miniotring data.
From the above table, following conclusions may be made.

 Predicted angular distortion broadly agrees with that of actual monitoring data. Hence,
settlement prediction and assumed effective depth of piled foundation is fairly reasonable.
 The settlement of actual monitroing data is slightly higher than that of prediction. This
may be due to the additional settlement caused by diaphragm wall panel construction
which influence zone might have extended to 15m from the diaphragm wall (equivalent
distance to bottom of soft clay level ).
Damage Assessment on Sensitive Structures
The most critical structure required for Stage 3 Assessment were MWA main water
tunnel as tunnel alignment crosses the water tunnels at two critical locations – the first near
Rama IV and Rachadaphisek road intersection and the second at Rama IV and Siphraya road
intersection near Sam Yan station. The position of water tunnels in relation to MRT tunnels
drawn in the subsoil profile at relevant locations are presented in Figure 21 and 22.
According to the available information, MWA tunnel at the first crossing is of 2.5m
internal diameter steel lining with 150mm segmental ring and at the second crossing MWA
tunnel is of 2.5m diameter segmental ring with lightly reinforced concrete lining.
In preliminary stage 3 assessment, based on the predicted subsurface profile using the
method described in published paper “Ground Movement Prediction and Building Damage
Risk-Assessment for the Deep Excavations and Tunneling Works in Bangkok Subsoil”, Zaw
et. al (2006) e, structural assessment of water main was carried out and results were compared
with structural capacity of water tunnel calculated from the material properties indicated in
the construction drawings provided by MWA. As bending moment induced in the water tunnel
was considered critical parameter in assessing the damage which is governed by both
maximum settlement and curvature of the water tunnel further assessment was proposed
coupled with observational approach.
520
 To carry out back-calculation using monitoring data obtained from the settlement arrays
(settlement points, extensometer data) at first 3 locations before reaching to MWWA
water tunnel near Rama IV – Rachadapisek intersection, to determine the volume loss
and relationship between surface and subsurface settlement as well as best estimation of
subsurface K value and subsurface settlement profile at water tunnel level.
 To perform beam-on-elastic foundation analysis using the subsurface settlement profile
obtained from predicted subsurface settlement bending moment likely to occur in
MWWA water tunnel.
 To compare the result obtained from above item with allowable values.
 To establish the final trigger values based on above comparison.
 To confirm the settlement profile from actual monitoring of last array is within the alert
level of final trigger value.
Figure 21. Position of MWA water tunnel in relation to MRT tunnels (clear distance are
3.38m and 3.18m for North and South bound respectively) at the first crossing
521
Figure 22. Position of MWA water tunnel in relation to MRT tunnels (clear distance are 1.5m
and 2m for North and South bound respectively) at the second crossing
For the first location of MRT-MWA tunnel crossing, target ground loss from tunnel
operation was set at 0.5%, whereas ground loss based on allowable movement curvature of
water tunnel derived from predicted pattern of subsurface settlement was 0.65%.
Back-calculated ground loss derived from the monitoring data of rod-extensometer placed at
the crown of MWA water tunnel was within target value. In addition to the preliminary
assessment based on prediction of subsurface settlement using simple empirical formulas, a
detailed soil-structure interaction was analyzed using 2-dimensional FEM considering
behavior of tunnel boring by earth pressure balance (EPB) method. Stress relief from the
effect of cutter-head, tail void and over-cut from EPB tunnel boring was considered in the
analyses. Adding ground movement due to consolidation caused by disturbance from tunnel
operation surface and subsurface settlement were predicted by finite element analysis.
Detailed parameters and assumption made in analysis was presented in the published paper of
Sakai & Sugden (2000).
After the successful passing of both tunnels at the first MWA crossing, another trial area
tunneling was planned to further determine the ground movement associated with various
tunneling parameters well before reaching to more critical crossing after Sam Yan station.
Tunnel section between Silom and Samyan station, in which soil profile similar to that of
second crossing was selected and comprehensive instrumentation program was established to
monitor surface and subsurface settlement as well as horizontal movement of the ground at
depth.
The observed ground movement induced by tunneling works were within predicted
values. Well organized plan, systematic approach of observational method to achieve control
522
values backup by extensive instrumentation and close co-operation among relevant personnel
were the key factors contributed to successful completion of EPB tunneling particularly in the
most sensitive points at MWA water tunnel crossings. The monitored subsurface settlement
data from extensometer and predicted settlement trough of assumed ground loss 1% ground
settlement profile of the first MWA tunnel crossing is presented in Figure 23.
Distance (m)
0 10 20 30 40 50 60 70 80 90
-10
-30
Total Settlement
-50
Figure 23 Predicted subsurface settlement trough of assumed ground loss 1% calculated

from empirical formula and actual subsurface settlement (indicated by dark-circle) observed
from rod extensometer – first MWA tunnel crossing
Key Factors in Successful Application of Observational Method
The M.R.T Chaloem Ratchamongkhon Line, one of the most modern underground mass
transit systems in the region was successfully completed and has been in operation for over 2
years. All deep excavation works and tunneling were successfully completed without any
significant damages to the adjacent buildings and structures. Well organized plan, systematic
approach of observational method to achieve control values backup by extensive
instrumentation and close co-operation among relevant personnel were the key factors
contributed to successful completion of the project. The purpose of this paper is to provide a
set of technical guidelines for ground movement prediction and building damage risk
assessment of deep excavation and tunneling works in Bangkok subsoil and application of
observational method which may partly serve as a reference for the future underground mass
transit system projects in urban area of Bangkok. Successful completion of the first MRT
project in Bangkok marked the practicality of well-executed deep excavation works and
tunneling by EPB method coupled with observational method in the sensitive urban
environment.
523
REFERENCES
Boscarding, M.D. and Cording, E.G. (1989), Building Response to Excavation-Induced

Settlement. Journal of Geotechnical Engineering, ASCE, 115;1;1-21
Bowels J.E (1988) Foundation Analysis and Design, 4th Edition, McGRAW-Hill International
Book Company, Singapore, 1988
Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings and Associated Damage.
SOA Review, Conference on Settlement of Structures, Cambridge,
Pentech Press. London, pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of Foundations and Structures,
SOA Report Session 2, Proceedings of 9th International Conference, SMFE, Tokyo,
2:495-546
Clough, G.W., and O’Rourke, T.D., (1990), Construction Induced Movements of In-situ
Walls, ASCE Geotechnical Special Publication No. 25, pp. 439-470
Clough C.W. and T.D. O’Rourke (1990), Deep Excavation and Tunneling, 7th ICSME,
Mexico City
JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for MRTA Initial System -
UGS, Bangkok
Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of Ground Movements and
Assessment of Risk of Building Damage. Geotechnical Aspects of Underground
Construction in Soft Ground, pp. 712-718, Balkema, Rotterdam
O’Reilly M.P and New B.M (1982), Settlements above tunnels in United Kingdom – their
magnitude and prediction, Tunneling ’82, pp. 173-181, London, IMM
Peck R.B. , (1969) Deep Excavation and Tunneling in Soft Ground, 7th ICSME, Mexico City,
State of the Art 3, 225-290
Peck, R. B. [1969]. “Advantages and limitations of the observational method in applied soil
mechanics”, Geotechniques 19, No.2, pp. 171-187
Powderham, A. J. [1996]. “An overview of the observational method : development in cut and
cover and bored tunneling projects”, The Observational Method in Geotechnical
Engineering, The institute of civil engineers, Thomas Telford Publishing, London, pp.
7-24
Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006]. “Ground Movement Prediction and
Building Damage Risk-Assessment for the Deep Excavations and Tunneling Works in
Bangkok Subsoil, Proceedings of the International Symposium on Underground
Excavation and Tunneling, Urban Tunnel Construction and Protection of Environment,
EIT, Bangkok, Thailand, , pp. 281-297
524
8-11 May 2007, Kuala Lumpur, Malaysia

Edited by:
Ooi Teik Aun, Yee Yee Weng and Ting Wen Hui
Ground Movements Associated with the

Underground Excavations of the
First Bangkok MRT Line
Noppadol Phienwej and Attasit Sirivachiraporn

Suchatvee Suwansawat
King Mongkut Institute of Technology-Lad Krabang, Bangkok, Thailand
Zaw Zaw Aye

525
Ground Movements Associated with the Underground Excavations of the
First Bangkok MRT Line
N. Phienwej & A. Sirivachiraporn

School of Engineering and Technology, Asian Institute of Technology, Pathumthani, Thailand
noppadol@ait.ac.th
S. Suwansawat
Department of Civil Engineering, King Mongkut Institute of Technology-Lad Krabang, Bangkok, Thailand
kssuchat@kmitl.ac.th
Z. Z. Aye
Seafco Public Company, Bangkok, Thailand
zaw@seafco.co.th
Abstract: Works of the first Bangkok Underground MRT project involved the largest diameter shield tunneling and the deepest dia-
phragm wall excavation ever attempted in Bangkok subsoils. The design-built contracts required the design and construction to avoid
damages to existing buildings, thus the Observational Method was adopted and large amount of instrumentation for monitoring of
ground movements was integrated as important component of the works. The ground and building movement data associated with the
shield tunneling and station excavation are interpreted and the findings are briefly presented. Tunnel volume loss in the EPB shield
tunneling was typically in the range of 0.5-2.5% while the maximum ground surface settlement above the tunnel was typically 20-60
mm. The magnitude of ground movement varied with the face pressure used in the shield driving. The settlement trough width parame-
ter was i = 0.5-0.6 Z for tunneling in clay layers and < 0.4Z for tunneling in the underlying sand layer. Maximum lateral deflection in
braced mode of the diaphragm wall in the station excavation was in the range of 0.05-0.4% of the excavation depth. A similar range
was observed for the maximum ground surface settlement adjacent to the wall. There were only limited cases of damages to adjacent
buildings, which occurred only to small sized buildings founded on short piles.
1 INTRODUCTION lined water main tunnels at two locations, and removal and un-
derpinning of piled foundations of four canal bridges.
The Bangkok MRT Initial System Project (ISP), the Chaloem Prior to the start of the MRT construction, there had been nu-
Ratchamongkhol Line, is the first underground MRT of Bangkok. merous shield tunneling and deep excavation works for both
Total length of the line is 21.5 km, comprising 16 km of twin building construction and utilities development. However, under-
single-track tunnels, 18 cut-and-cover stations and twin 1.5-km- ground works of the MRT project marked the largest tunnel di-
long cut-and-cover approach tunnels to Depot. The project was ameter and the deepest excavation ever attempted in Bangkok
commenced in late 1996 and opened for public use in 2004. The subsoils. Thus great caution must be paid to the design and im-
alignment mostly followed congested city roads. Construction plementation of the works for safety assurance of the works and
works of the underground structures were carried out in two fast success of the project development.
track design-build contracts, the South and North Sections. The Control of ground movement and preventing damages to adja-
tunnel tubes were bored at 6.3-m diameter by 8 EPB TBMs while cent structures from the excavation was the key for the success
the stations were excavated with an aid of diaphragm walls fol- for the project. The Observation Method design was adopted for
lowing the top down construction method. The depth of the tun- that purpose. A comprehensive instrumentation program for
nels was between 8 and 25 m below ground level. The general monitoring of ground movements and responses of concerned
layout of the twin tunnels is in side by side configuration (mostly structures was carried out as an important component of the con-
with clearance between tunnel walls of 1.0-1.5 times the exca- tracts. The monitoring data served as indices for control of works
vated tunnel diameter (5.0-9.0 m) but the smallest clearance was and safety assurance during construction as well as allowed for a
as smaller as 2.6 m). The vertically stacked configuration of the better understanding on the ground movement behavior induced
tunnels was adopted for the first 4.5 km of the alignment from the by EPB shield tunneling and deep excavation with diaphragm
terminal station at the main intercity railway station owing to ob- wall in the stratified Bangkok soils.
struction from foundations of elevated roadways and existence of
water main tunnels. Consequently, the stations in the section
2 SUBSOIL CONDITIONS
were excavated at depth as deep as 30 m. Two stations are lo-
cated beneath existing intersection overpasses that necessitated
underpinning works. Other design and construction issues in- Bangkok is situated on the low-lying plain. The subsoil within
volve consideration of the tunnels passing close to concrete- the first 20-30 m depth generally consists of layers of soft clay,
stiff to hard clay, and sand. Below are alternating layers of stiff to
527
hard clay and dense sand existing to a great depth. The subsoil 3 TUNNEL EXCAVATION
condition is relatively uniform through out the city area. The soft
clay layer generally extends from the ground surface to a depth of Eight earth-pressure-balanced (EPB) shields were used in the 16
12 to 15 m. The soft clay which is known as “Bangkok soft clay” km long twin tunnel excavation- 4 Kawasaki shields in North
has high water content (70-120%). The underlying stiff clay layer Contract and 2 Kawasaki +2 Herrenknecht shields in South Con-
has much better engineering properties than the soft clay and is tract. The tunnels were supported with 300-mm-thick bolted con-
an ideal medium for tunnelling within. Below the stiff clay layer crete segmental lining ring. To achieve effective performance of
is a layer of fine to medium, dense to very dense, silty to clayey the EPB shield operation, conditioning additives were added to
sand. The sand is normally saturated but the piezometric pres- excavated soils in the EPB chamber. The additives were ben-
sure has been considerably drawn down as a result of excessive tonite slurry, polymer, or organic foam. The four shields in each
deep well pumping in the past 60 years. Presently, the piezomet- of the contracts took about 14-15 months to excavate the tunnel
ric head in the sand layer is mostly at 21-24 m below the ground length. The average advance rate of excavation of each shield is
surface. The condition is beneficial to tunnelling and deep exca- 60-70 m/week. Despite some major downtime due to breakdowns
vation works. Fig. 1 shows subsoil stratigraphy along the MRT of main parts, all of the shields performed very satisfactory in all
alignment. Although most of the tunnel alignment is placed in the types of soils. The rates of excavation of the EPB shields in the
stiff clay layer, some shallow sections have roof in the overlying MRT project in all types of soil were essentially similar except
soft clay and some of the deep alignment was excavated in the for excavation in the stiff clay that the rate could be smaller due
underlying sand layer. to higher cohesion and stickiness.
Subsurface Condition of North Section

PRACHARAT BAMPHEN RATCHADA PHAHONYOTHIN KAMPHEANG PHET
THIAM RUAM MIT SUTTHISARN LAT PHRAO MO CHIT BANG SU
0 0
DEPTH BELOW GROUND LEVEL (m)
10 10
20 20
30 30
40 40
50 50
60 60
20+000 21+000 22+000 23+000 24+000 25+000 26+000 27+000 28+000 29+000 30+000
FILL DENSE CLAYEY SAND STATION

BANGKOK SOFT CLAY DENSE SAND TUNNELLING HORIZON
STIFF TO HARD CLAY
Fig. 1 Subsoil stratigraphy along the MRT alignment
528
4 STATION EXCAVATION
Excavation of the 18 stations was made by the top-down con-

struction using 1-m-thick concrete diaphragm walls. Tips of the
wall usually extend to the depth of 35 - 40 m below the ground
surface (Fig. 2) to satisfy load bearing and uplift resistance re-
quirements as well as to provide ground water cutoff in the first
sand layer. The distinct difference in depth of embedment of the
wall below the excavation level between the two contracts was
partly due to the difference in the adopted design criteria. Con-
struction period of each station from utility relocation to architec-
tural finishes was usually around 45 months- 6 months of piling
and diaphragm wall construction and 18 months of excavation
and casting of station floors and walls.
5 OBSERVATIONAL METHOD Fig, 2 Depth of station excavation and tip depth of diaphragm
walls
The contract required that the design and construction meth-
Station, Km
ods limit ground movements from tunneling and station exca- 10.000 12.000 14.000 16.000 18.000 20.000 22.000 24.000 26.000 28.000 30.000
vation to avoid damage to adjacent structures and buildings. 0.0
The Observational Method was then adopted for the control of -20.0
work. It involved the followings. -40.0
- Three stages of assessment on the effects of the ground -60.0

Settlement, mm
movement on structures depending upon the results of -80.0

-100.0
each stage.
-120.0
- In stage one, prediction of the green filed ground surface -140.0
settlement using approved empirical prediction methods -160.0
Stacked Tunnels Side-by-SideTunnels
was made. -180.0
- Building condition surveys of all structures within the

zone of influence determined by the first stage. Any struc- Fig. 3 Maximum ground surface settlement along the tunnel
tures with the predicted ground settlement less than 10 alignment
mm need not be subject to further assessment.
- Structures located in the zone of surface settlement were
checked for potential damages according to the limit ten-
sile strain criteria (Boscardin & Cording, 1989, Burland,
1995). Structures falling in damage category within slight
level should be acceptably safe thus need no further as-
sessment.
- Stage 3 assessment of structures exceeding slight level of
damage need stage assessment which involved detailed
structure analysis considering soil structure interaction).
Depending on the results of the analysis, pre-excavation
protection or preventive measures as well as contingency
plan to be implemented during construction may be
needed. Fig. 4 Maximum surface settlement versus shield face pressure
- Monitoring of settlements arising from works must be car-
maximum settlements were in the range of 5 mm to 120 mm,
ried out. Comprehensive instrumentation program was re-
mostly between 20 mm and 60 mm (Fig. 3). Larger settlement
quired as a part of the contract.
tended to occur in the break-ins and break-outs areas of the tun-
Relatively few incidents of damages to adjacent structures neling near the walls of the station boxes. The largest magnitudes
occurred. The most significant ones were damages to small were recorded in the early leg of tunneling of the North Contract
buildings on short piles. No cases of damages to tall buildings and in the areas of the vertically-stacked tunnels in South Con-
were experienced. tract where the tunnel were bored in sand layer below drawn-
down ground water level.
Studies were made to examine influencing factors controlling
6 GROUND MOVEMENTS IN TUNNELLING the magnitude of the ground movement in the EPB shield tunnel-
ing (Suwansawat, 2002, Timpong, 2002, Tavaranun, 2003, etc).
6.1 Surface Settlement It was found that the most important factor seems to be the ap-
Monitoring of ground surface settlements over the tunnel align- plied face pressure in the front chamber of the shield machine.
ment was made at 50 m intervals. The data showed that the The face pressure played an important role in maintaining stabil-
529
ity of the excavation and minimizing ground movement. Fig. 4 tion, the trough expanded and shifted toward the second tunnel.
shows the plot between the observed maximum surface settle- However the centerline of the trough was normally skewed to-
ment and the face pressure in the project. The plot clearly shows ward the first excavated tunnel.
the influence of the applied face pressure on the magnitude of In the Bangkok MRT tunneling, the shifted distance of the
ground settlement. The lower the face pressure used, the larger centerline of the trough from centerline of the twin tunnels (a)
the settlement occurred. In case of a very high face pressure, observed at 24 monitoring arrays ranged from 0.03-0.58 of the
heave occurred at the ground surface. On the other hand, when distance between the two tunnel centerlines (d). The mean a/d is
the shields were operated at very low face pressures (<100 kPa), 0.16. The width of the settlement trough observed after the exca-
large ground movement occurred. The effect of the face pressure vation of the twin tunnels in the project was about 6 to 12 m for
was observed in all soil settings except for the case of the full the vertically-stacked tunnels and 10 to 17 m for the side-by-side
face in sand layer. No trend was found between the face pressure twin tunnels.
and surface settlements when the tunnels were excavated entirely A plot is made between relationship between the settlement
in sand layers. trough width parameter (i) and tunnel depth (z), Fig.7. Most of
Out of the 39 locations of settlement trough monitoring arrays, the data points fall within the range of i = 0.5Z to 0.6Z for tun-
19 arrays showed the observed trough shapes reasonably fitted nels located in both the stiff clay and soft clay layers and i < 0.4Z
with the Gaussian curve (Peck, 1969), Figs. 5(a) and 5(b). The for tunnels in sand layer. This is in agreement with the observa-
tunnel ground loss was mostly in the range of 0.5-2.5%. In diffi- tion of O’Reilly & New (1982) for tunneling in UK.
cult sections, the ground loss was up to 3.5%.
Ground settlement array no. 26-AR-001
Zone 26
Ratchada - Lat Phrao
Soft clay
Soft clay
First tunnel
excavation (SB)
Stiff clay
ground loss 0.43%
-20 m.
Second tunnel NB SB
excavation (NB)
Stiff clay
ground loss 0.61% Very Stiff clay -20 m.
NB SB
10.5 m.
Sand
(a) Side-by-side Tunnels
Fig. 6 Effect of presence of piled foundation on shape of settle-

ment trough
Soft clay
40
Single tunnel
-15 m. 35 Twin tunnel i = 0.3Z i = 0.4Z i = 0.5Z
First tunnel NB i = 0.6Z
excavation (SB) Stiff clay
30
ground loss 0.69%
Second tunnel Very Stiff clay 25
excavation (NB) -26 m.
Depth, Z (m)
SB
ground loss 0.80% Dense sand
20
15
10
(b) Vertically stacked 5
0
Fig. 5 Surface settlement trough over the MRT twin tunnels 0 5 10 15 20
Trough width parameter, i (m)
Settlement troughs having shapes not to follow the Gaussian
function were normally affected by existence of adjacent build- Fig. 7 Settlement trough width parameters (i) versus depth
ings or flyovers. The existing piled foundations within the trough
zone obstructed the ground movement and caused an unsymmet-
rical trough shape (Fig. 6). 6.2 Subsurface Ground Movement
Maximum surface settlement occurred above the centerline Monitoring of subsurface ground movements was made by means
of first tunnel after the shield passage and normally the Gaussian of borehole extensometers and inclinometers installed at a num-
shaped trough was observed. For the side-by-side twin tunnel ber of sections along the tunnel alignment. The layouts of the in-
case, once the boring of the second tunnel was made at the sec- struments in the two contracts are shown in Fig. 8.
530
cal/empirical approach (e.g. Verruijt & Booker, 1996; Logana-
Extensometer installation in Extensometer installation in than & Poulos, 1998; etc.) and the numerical approach (e.g. FEM
South contract North contract PLAXIS, FDA FLAC, etc.) were investigated (Tavaranum,
Extensometer
2003). For the first approach, it was found that the method of Lo-
Inclinometer Extensometer Inclinometer ganathan & Poulos (1998) method generally can give reasonable
casing casing fits to subsurface settlement pattern with field data for both single
Soft Clay
Soft Clay
and twin tunnels cases (8 out of 12 sections). The gap parameter,
g, from back calculation was in the range of 20-60 mm (Tim-
pong, 2002). Reasonable prediction was also offered by the Ver-
Stiff Clay Stiff Clay
3 - 6 m. ruijt & Booker (1996)’s method. However, they could not suffi-
ciently give reasonable prediction of lateral ground movement for
cases of excavation with high face pressures, where outward
Tunnel Tunnel ground movement occurred (6 out of 12 sections).
Fig. 8 Typical layouts of subsurface instruments for shield tun-

neling
6.2.1 Subsurface settlement

For areas above the tunnel centerline, the magnitude of subsur-
face settlement after the tunnel passage generally increased with
depth toward the tunnel roof (Fig. 9a). For the settlement to the
side of the tunnel wall, it was generally observed that the largest
settlement occurred in the soft clay layer near the interface with
the underlying stiff clay (Fig. 9b). The settlement in the stiff clay
along the instrument tube decreased with depth as the measure-
ment points falls outside the movement zone. At some sections, a Fig. 9(a) Subsurface settlement above tunnel centerline
small heave was measured below the tunnel level.
Long term ground settlement behavior for the EPB shield tun-
nel excavation was examined. The subsurface settlement data
showed that after the shield passage, significant increase in the
settlement in long term was only observed in the soft clay layer.
In some cases, the amount of the increase in the settlement 3-4
months after the shield passage was almost 100 percent of the
short term settlement (Fig. 10). On the other hand, the long term
increase in the settlement in the stiff clay layer near the tunnel
roof was generally not significant.
6.2.2 Lateral displacement

The lateral displacements of ground beside the tunnel were ob-
served at 12 locations along the alignment. The instruments were
placed at distance of 0.20-2.6 m from the tunnel wall. Most of the
measurements showed a similar pattern of the lateral ground
movement which can be divided into two different depth zones, Fig. 9(b) Subsurface settlement to the side of tunnel wall
i.e. above and below the level of the tunnel arch. Above In the
upper zone zone, the ground displaced towards the tunnel with
the maximum displacement Fig. 6 Effect of presence of piled E xte nso me te r Mo nito ring R e sult
foundation on shape of settlement trough occurring in the lower N o. 2 5 -IE -00 3 Sutthisan - Ra tc ha da
area of the soft clay layer. The inward displacement within this Ela pse d time (da ys)
depth zone continued to increase after the passage of the shield -50
0 20 40 60 80 100 120 140
tail.
-40
In the lower zone, two patterns of the lateral ground displace-
ment were observed. In most cases,, inward movement toward -30
Settlement (mm)
the tunnel was observed. The magnitude was smaller than those -20
occurring above the tunnel roof. At few sections where high face -10 25-M E-3/08 -2.02 m.
25-M E-3/07 -5.15 m.
pressure was used in the shield tunnelling, the ground showed 0
Soft clay
25-M E-3/06 -8.32 m.
outward movement (Fig. 11). The movement ceased after the 25-ME-3/05 - 11.45 m.
10 -15 m. 25-ME-3/04 - 14.58 m.
passage of the shield tail. This clearly indicates the effect of the Stiff clay
6 .24 m.
25-ME-3/03 - 17.71 m.
20
applied face pressure in the EPB shield tunneling in Bangkok
-18.5 m.
Shield passing
soils. 30
Studies were made to investigate applicable methods for pre-

diction of the ground in shield tunnelling. Both the analyti- Fig. 10 Changes in subsurface settlement with time
531
of the Bangkok subsoils for the analysis. The results show that
Inclinometer No. IN-T7-04 for the soft/medium clay the modulus, Eu should be around 500
(Phetchaburi - Sukhumvit stations Su, where Su is undrained shear strength. For other soil layers
0 Made ground underlying the soft clay, the values are:
Avg. face pressure = 100 kPa
Avg. grouting pressure = 2.5 bar - 1st Stiff Clay Eu = 700 N
-5 C.L. = 18.8 m - Clayey Sand & Sandy/Silty Clay Eu = 900 N
2.35 m from the inclinometer
- 2nd Hard Clay Eu = 1600 N
Soft clay - 3rd Hard Clay Eu = 2500 N
-10
Zone 1 Zone 2 Zone 3 SOUTH CONTRACT
Maximum Lateral Wall Movement/Excavation Depth, dHmax/H
Depth (m)
0.0 0.5 1.0 1.5 2.0

-15 0
Stiff clay
5
Excavation Depth, H (m)

10
-20
Clayey sand
15
A. 2D before the shield face
B. 1D before the shield face 20
-25 C. at the shield face
D. 0.5D behide the shield face
Very stiff clay E. at the shield tail 25
F. 2D behide the shield face
G. 3D behide the shield face 30
H. 4D behide the shield face
35
-4 -2 0 2 4 6 Cantilever mode (H=1.6m-4.0m) Braced mode-soft soil (H=6.5m-10.8m)
Lateral Deformation (mm) Braced mode-stiff soil (H=12.4m-16.7m) Braced mode-stiff soil (H=20.0m-32.6m)
Fig 11 Lateral displacements beside tunnel (After Suwansawat, Fig. 12 dHmax/H versus depth of excavation (H)
2002)
NORTH CONTRACT
Distance from excavation/Maximum excavation depth, D/H
0 1 2 3 4 5 6 7 8 9 10
7 GROUND MOVEMENTS FROM STATION EXCAVA- 0
Settlement/Maximum excavation depth,
TION
0.1
Ground movements associated with the excavation of the 18 sta- 0.2

A
dVmax/H
tions of the project were closely monitored with systematic in- 0.3 B
strumentation programs. The monitoring data provided useful in-

0.4
formation to capture the typical characteristics of lateral ground C
movement and settlement in the areas adjacent to the controlled 0.5

I II III
deep excavation using diaphragm wall in Bangkok subsoil. The 0.6

station boxes were generally 230 m long by 25 m wide with ex- Cantilever mode (H=1.6m-4.0m) Braced mode-soft soil (H=6.5m-10.8m)
Braced mode-stiff soil (H=12.4m-16.7m) Braced mode-stiff soil (H=20.0m-32.6m)
cavation depths of 20-30 m. Hooi (2002) conducted comprehen- Modifed Category A, B & C Peck's Category I, II & III
sive review and interpretation of the monitoring data. The results

showed the ratio of maximum lateral wall displacement to the ex- Fig. 13 dVmax/H versus depth of excavation (H)
cavation depth at various stages of excavation as shown in Fig.
12. For the cantilever mode occurring in the first stage of excava- 8. BUILDING RESPONSES
tion (H = 1.6 m – 4.0 m), the dHmax/H = 0.20 - 1.60 %. For the
braced wall mode, the ratio dHmax/H = 0.05 – 0.60% while ex-
Along the alignment, there were 339 buildings and structures lo-
cavating in soft clay layer (H = 6.5 m – 10.8 m). The wall deflec-
cated in the zone of influence. These structures were subjected to
tion in the braced wall mode while excavating in the stiff clay or
a detailed building condition survey. Instruments were installed
sand layers (H = 12.4 m – 32.6 m) showed dHmax/H = 0.05 –
to monitor movements and cracks that may be induced during
0.40%. The observed ground settlement adjacent to the excava-
and after the passages of the shields. With the existence of the
tion walls is summarized in Fig. 13. The ratio of the maximum
soft clay layer, buildings constructed in Bangkok need to be
settlement to excavation depth is high in the first stage cantilever
founded on piled foundations. Different length and size of piles
excavation mode. After the excavation advanced down to the
have been used depending on the size and age of the buildings.
braced mode in either the soft clay layer or the underlying stiff
Generally, the depth of pile tips can be categorized as shown in
clay layer, the ratio of the maximum settlement to excavation
Fig. 14.
depth (dVmax/H) was in the range of 0-0.3%. The observed
The monitoring data of all the instrumented buildings (mostly
width of settlement zone (D) is large, up to 7 D/H for braced
reinforced concrete frame types) were evaluated and the broad
mode with 4 D/H for braced mode with excavation base soft clay
picture of the building response to the tunnel excavation could be
and reducing to 4 D/H in deeper stiff clay or sand. The influence
depicted by the summary plots of building settlements versus
zone of the diaphragm wall excavation is smaller than that sug-
number of storey of the building and distance from the tunnel ex-
gested by Peck (1969) for braced excavation using soldier
cavation (Figs. 15 and 16).
piles/sheet piles. Since the FEM design analyses used for all sta-
It clearly shows that buildings on deep piles were less influ-
tion excavations of the MRT followed linear elasto-plastic soil
enced by the tunnel excavation (Types 4 Buildings: settlements <
property, back-analyses based upon the monitoring data were
10 mm). For smaller buildings (Types 2 & 3), the settlement
made to determine the appropriate corresponding modulus values
could be larger (up to 30 mm). For buildings on short piles (Type
532
1), settlement could be excessive. Fig 15 shows that significant ential settlement/distance between the two reference points)., The
building settlement (> 20 mm) only recorded for buildings lo- observed tilts of these buildings were mostly smaller than 1:1000
cated within distance of 30 m from the tunnel centerline. (0.001) which was the alert level set forth for the control of work
in the Bangkok MRT project. According to Burland (1995) a
> 10 storey (masonry) building experiencing a maximum tilt of 1:500 (0.002)
building
and a settlement of less than 10 mm has negligible risk of any
4 to 10 storey
building damage. In fact, very few cases of damages of the existing build-
2 to 3 storey
building &
ings due to shield tunneling occurred in the project. These were
shophouse limited to old 1-3 storey buildings founded on short piles (proba-
Old building
and house
bly with tips within soft clay layers). All of the tall buildings
were not damaged by the tunnel excavation even though the tun-
0.00 m.
nel was excavated right next to the piles. A study is currently
Made ground -1.5 m.
made to evaluate the influence of the shield tunneling on the load
<6m Soft clay
carrying capacity and integrity of the piles.
6 to 14 m -15 m.
14 to 30 m. 1st Stiff clay 9 CONCLUSIONS
-22 m.
15 to 22 m
1. EPB shield tunneling for the first Bangkok MRT was suc-
1st Sand deposit
cessfully completed with the average advance rate of each
Pile length > 22 m -40 m. shield of 60-70 m/week. Similarly, the deep excavation us-
ing diaphragm wall of the 18 stations were completed with-
Fig.14. Categories of building and its foundation in Bangkok out major difficulties.
2. Maximum ground surface settlement above the tunnels
Building settlements vs distance from tunnel ranged from 5 mm to 120 mm, mostly 20-60 mm. The corre-
sponding volume loss in the tunnel excavation was 0.5-
90
2.50%. Tunneling at in water bearing sand layer tended to
Buildings on foundation
80
Type 1
give the largest ground movement.
building settlement (mm)
70 Type 2
Type 3
3. Most of the surface settlement trough could be fitted by
60
Type 4 Gaussian function. The trough width parameter (i) is ap-
50 proximately between 0.4Z and 0.6Z for tunnels in clays and
40 less than 0.4Z for tunnels in sands.
30
4. The long-term settlements in soft clay layer above the tunnel
20 could be significant in some sections but not for the stiff clay
10 layer.
0
0 10 20 30 40 50 60 70 80
5. The observed subsurface settlements and lateral displace-
ments could generally be reasonably fitted with the prediction
distance from tunnel (m.)
by the analytical methods of Verruijt & Booker (1996) and
Loganathan & Poulos (1998). However, the prediction was
Fig. 15 Building settlement versus distance from tunnel poor for lateral ground movements in case where lateral out-
ward ground movement occurred.
Building settlement vs No. of storey
6. Both maximum lateral wall deflection and maximum ground
90 surface settlement near the station excavation using braced
diaphragm walls showed similar ranges of movement. The
80
upper bound magnitude is around 0.3-0.4% of the excavation
building settlement (mm)
70
depth.
60
7. Settlement of adjacent buildings induced by ground move-
50
ment from the MRT excavation showed clear correlation with
40
building heights which reflects depth of piled foundations.
30
20
10 REFERENCES
0
0 10 20 30 40 50 60 Boscardin, M.D. & Cording, E.J. (1989), Building Response to
No. of storey Excavation-Induced Settlement, Journal of Geotechnical En-
gineering, ASCE, 115;1;1-21.
Fig. 16 Building settlement versus number of storey Burland, J.B., 1995. Assessment of risk of damage to buildings
due to tunnelling and excavation, Proc Int. Conf. Earthquake
The measured differential settlements of the buildings were Geotech. Engrg., IS-Tokyo’ 95.
calculated and expressed in term of settlement ratio or tilt (differ-
533
Hooi Kah Yeng 2002. Ground Movements Associated with Sta- Phienwej, N. 1998. Ground movements associated with soft
tion Excavations of the First Bangkok MRT Subway. M.S. ground tunnelling in Bangkok. Proc. World Tunnel Con-
Thesis, AIT, Bangkok. gress’ 98: Tunnels and Metropolises, Sao Paulo, Brazil.
Loganathan N. & Poulos H.G., 1998. Analytical prediction for Suwansawat, S., 2002. Earth Pressure Balance Shield Tunneling
tunneling-induced ground movement in clays. J. Geotech. in Bangkok: Ground Response and Prediction of Surface Set-
Engrg ASCE, 124, no. 9, pp.846-856. tlements using Artificial Neural Networks. Doctoral Disserta-
Mair R.J., Gunn M.J. & O’Reilly M.P., 1981, Ground movement tion, MIT, USA.
around Shallow Tunnels in Soft Clay, Proceedings of The Timpong, S., 2002. Analysis of Ground Movement in Bangkok
10th Conference on Soil Mechanics and Foundation Engi- MRT Blue Line Project. M.S. Thesis, AIT, Bangkok.
neering, Stockholm, Vol.1, pp. 323-328 Tavaranum, S., 2003. Analysis of Subsurface Ground Movement
O’Reilly, M.P. & New B.M., 1982. Settlements above Tunnels in Induced by EPB Shield Tunnelling in Bangkok MRT Blue
United Kingdom-Their Magnitude and Prediction, Tunnel- Line Project. M.S. Thesis, AIT, Bangkok.
ling’82, pp 173-181. Verruijt, A. & Booker, J.R., 1996, Surface settlements due to de-
Peck, R.B., 1969. Deep Excavation in Soft Ground, Proc. 7th In- formation tunnel in an Elastic Half Plane, Geotechnique, Vol.
ternational Conference on Soil Mechanics and Foundation 46, No. 4, pp. 753-756.
Engineering, State of the Art Volume, pp 225-290.
534
Proceedings of International Conference and Exhibition on
Tunnelling and Trenchless Technology
1 – 3 March 2011, Selangor, Malaysia
International Conference and

Exhibition on Tunnelling and
Trenchless Technology
Geotechnical Risk Assessment and

Management for Tunnelling Project in
Congestive Urban Area – Bangkok Experience
Zaw Zaw Aye and Phongpat Kitpayuck

Noppadol Phienwej
535
Geotechnical Risk Assessment and Management for Tunnelling Project
in Congestive Urban Area – Bangkok Experience
Zaw Zaw Aye & Phongpat Kitpayuck

zaw@seafco.co.th
Dr. Noppadol Phienwej

Asian Institute of Technology, Bangkok, Thailand
noppadol@ait.ac.th
Abstract: The M.R.T Chaloem Ratchamongkhon Line is the first underground mass transit system project in Bangkok developed by
the Mass Rapid Transit Authority (MRTA) of Thailand. This paper presents the methods adopted in prediction of tunneling induced
ground movement and associated building damage risk-assessment carried out for the Contract No. 1 of the M.R.T Chaloem Ratcha-
mongkhon Line constructed in congestive urban area. Application of observational method in managing risk of damage to the existing
buildings and structures are also reported.
1 INTRODUCTION • Risk assessment : where the likelihood and consequence of

each hazards are evaluated and combined to estimate the risk cor-
Geotechnical risk is one of the key elements of the technical and responding to each hazard. A risk register is produced at this
financial risk in most of the major construction projects. Though stage for systematic assessment of the risk.
engineers use adequate factor of safety and taking account of • Risk reduction : where the hazards are eliminated if possible
other risk factors in design and construction process, systematic and the risks are reduced by a combination of design changes,
approach of risk assessment and management is not commonly procedural changes and additional monitoring
applied in most of the projects. • Risk control : where the risks are monitored, controlled and
Tunneling works and deep excavation works pose a high de- managed throughout the project. The risks normally diminish as
gree of geotechnical risk that needs to be identified, assessed, the construction progresses.
minimized and managed. Deep excavation and tunneling works In general, risk management is the process of managing re-
generally interact with natural ground and incorporating their sponse and policies to reduce and control risks. It is an integral
characteristics as structural elements of their own stability. This part of the observational method, which concentrates on technical
interaction is more significant in tunneling projects than in other risk (CIRIA Report 185). The merits of using the OM to manage
civil engineering projects particularly in congestive urban area. ground-related risks have been discussed by Whitman (1984),
In this paper, process of geotechnical risk assessment in Bangkok Blockely (1994), Chowdhury (1994) and Godfrey (1996). Roles
first underground mass transit project is presented. of different parties involved in construction project namely the
client, the designer and constructor are described in details by
Clayton (2001).
2. GEOTECHNICAL RISK MANAGEMENT
Geotechnical risk is the risk to construction project originated 3. CONTRACT 1 OF MRTA CHALOEM RATCHA-
from the ground condition. In general case, once risk have been MONGKHON LINE PROJECT
identified and assessed, to manage the risk fall into one or more
of the following major categories. The M.R.T Chaloem Ratchamongkhon Line is the first un-
• Eliminate or avoid derground mass transit system project in Bangkok developed by
• Transfer the Mass Rapid Transit Authority (MRTA) of Thailand. Total
• Mitigate length of the underground structure is 21.5km, comprises of 16
• Tolerate km of twin single-track bored-tunnel, a total of 4 km long18 cut-
Unfortunately, ground-related risks are not often practical to and-cover stations, a twin 1.5km long cut-and-cover approach
completely avoid or transfer. However, systematic approach, tunnel to Depot and other associated structures. The project was
control and management can minimize the risk if not entirely commenced in late 1996 and opened for the public in 2004. The
eliminate. CIRIA Report 185 outlined the four fundamental Contract No. 1, Underground Structure South - southern portion
stages of risk management as follows ; of this initial system comprising 9 km of twin 6m outside diame-
• Hazard identification : where hazards are identified and docu- ter bored tunnels, nine underground stations and cut-and-cover
mented in a register via experience of similar projects, brains- depot approach tunnel, was awarded to the Joint Venture BCKT
torming, structured interviews etc.
537
.
Fig. 1. Tunnel alignment of the Contract 1 (South Contract) of Bangkok MRT

.
One of the major challenges of the Contract No. 1, was the
need of stacked-alignment for a major portion of twin bored-
tunnel underneath Rama IV road, one of the busiest roads of
Bangkok, which called for a stacked configuration at 3 stations,
Lumpini, Si Lom and Sam Yan as shown in figure 2. This re-
quirement led the contractor to design and construct the deepest
underground structures of Bangkok at Si Lom Station - four-floor
stacked-station of 32m deep right under the existing flyover and
adjacent to a number of sensitive structures. Moreover, as
alignment of the lower tunnel had to place in the deeper depth
within first sand layers for the last four sections of the Contract 1,
the process involved in tunnelling works were more complicated.
Regional settlement due to active consolidation of top soft
clay layer and piezomatric drawdown caused by deep well pump-
ing was another key factor for which design and construction of
both tunnels and station boxes had have taken into consideration.
Deep excavation and bored tunneling works in congested ur-
ban environment of Metropolitan Bangkok requires a systematic
process of ground movement prediction, building risk damage
Fig. 2. Configuration of deeper stations with Stack Platform assessment and called for the stringent management of the risks.
In this paper focus is limited to prediction of tunnelling induced
consisting of Bilfinger + Berger Bauaktiengesellschaft, Ch Karn- settlement and assessment of potential damage to existing adja-
chang Public Company Limited, Kumagai Gumi Company Li cent building and structures though rigorous assessment was also
mited and Tokyu Construction Company Limited. The project made for ground movement caused by deep excavations by di-
was commenced in late 1996, successfully completed and opened aphragm wall support cut-and-cover construction method.
for the public in 2004.
538
.
Fig. 3. Generalized soil profile along the tunnel alignment and stations of Contract No. 1 (Phienwej et. al. 2006)
4. SUBSOIL CONDITION ALONG CONTRACT 1 ALIGH-

MENT
Similar to other localities in Bangkok, a typical subsoil profile

along the tunnel alignment is characterized by the alternating
layers of clay and sand deposits. Figure 3 shows the generalized
soil profile along the tunnel alignment of the Contract No.1.
Weathered crust of 2 m thick is commonly found as the top layer.
In urban areas of Bangkok, this layer is covered by fill material.
Soft to very soft, highly compressible dark gray marine clay lies
beneath weathered crust or fill. Depending on the location, this
layer extends up to 12-16 m. About 2 m thick medium clay layer
can be observed between soft clay and underlying stiff clay.
Generally stiff Clay layer occurs directly underneath medium
clay and its depth goes up to 22 m. Below stiff clay layer, first
sand layer of 5-8 m thickness can be found. This first sand layer,
however, is absent in some areas. Stiff to hard clay layer under-
lies first sand and it is found to be about 5 m thick. Second sand Fig. 4 View of diaphragm wall, stanchions and skeleton slabs
layer generally occurs at depths between 45 to 65 m. As can be
seen in Figure 3, tunnel alignment is located mainly in stiff clay Earth-pressure-balanced (EPB) tunneling method using shield
layer. Tunnels were bored in the first sand layer before reaching tunnel boring machines (TBM) was applied for tunnel excavation
to Queen Sirikit Center station and deeper sections between Bon except at the cut-and-cover section of tunnel approach to the de-
Kai (Khlong Toey station) and Hua Lamphone stations. pot. Japan made two sets of Kawasaki TBM were used from Ra-
ma 9 station to Queen Sirikit Convention Center (QSCC) station.
5. STATION BOX CONSTRUCTION AND TUNNELLING Tunnel boring from QSCC station to Hulamphong station was
WORKS carried out by two sets of Herrenknecht TBM. Kawasaki TBMs
were designed to drill through the concrete diaphragm wall –
Cast-in-place rigid diaphragm walls were used as permanent re- having specially manufactured Fiber Reinforcement (FRP) in-
taining structures for all stations and some entrance structures stalled at tunnel eyes of diaphragm walls. Pre-opening of diaph-
and ventilation buildings. Most of the entrance and ventilation ragm wall concrete and removal of steel reinforcement were ne-
buildings were excavated by sheet pile walls. Tip levels of diaph- cessary at tunnel eyes for Herrenknecht TBM. Tunnels were
ragm wall ranges from 19m (at Hua Lamphong station) to 39m bored with overcut by cutter-head at front (total cut diameter dur-
(at Sam Yan and Silom stations) for the final excavation depths ing boring was 6.48m) with backfill grouting. At permanent
of 15m to 31m. Excavated down to 31 below existing ground stage, tunnel was supported by 300mm thick precast segmental
level at Silom station marked the deepest excavation supported concrete rings having 6.30m and 5.70m outer and inner diameter
by diaphragm wall in Bangkok. Top-down construction method respectively. The loads from existing and future structures within
was used for all stations having stanchions for vertical support influence zone were taken into account in tunnel lining design.
and skeleton slabs as bracing during bulk excavation of station Movement joints (Omega Joint) were provided at the connection
boxes (Figure 4). at station diaphragm walls and tunnels to accommodate the diffe-
539
rential settlement between two structures. Figure 5 shows the
tunnel configuration at permanent stage.
Tunnel alignment generally follows the existing main Bang-
kok roads, having single-bored twin tunnels for north and south
bound tunnels at operation stage. From Rama 9 station to initial
section of Khlong Toey station, twin tunnels were bored at same
level (about 21m below existing ground level). Due to the pres-
ence of existing flyovers located in the middle, two tunnels were
required to change the alignment having stacked one above the
other.
Fig. 7. View of Bangkok major road where twin tunnels were

bored underneath at approximately 21m below existing ground
level
6. OVERVIEW OF GROUND MOVEMENT PREDICTION

AND BUILDING DAMAGE ASSESSMENT PROCESS AP-
PLIED IN THE PROJECT
Well accepted published literatures were used as basic reference

and method in ground movement prediction and building damage
assessment. For the station boxes where cut-and-cover methods
were used for excavation, the work of Peck (1969) and Clough
Fig. 5. Typical sectional configuration of tunnel at permanent and O’Rourke (1990) were used whereas published papers of
stage Burland et. al. (1977) and Boscardin & Cording (1989) were ap-
plied for bored tunnels. Building damage assessment was carried
out in 3 stages; Stage 1 Assessment, Stage 2 Assessment, and
Stage 3 Assessment. “Greenfield Site” assumption was adopted
in Stage 1 and 2 Assessment ignoring any influence of the stiff-
ness of the structure and type of supporting foundation. Stage 2
Assessment was further extended by Additional Stage 2 Assess-
ment. In Additional Stage 2 assessment, the variation of settle-
ment trough shape and magnitude with depth were considered in
identifying the risk of building damage. In Stage 3 assessment, a
detailed qualitative assessment of structure and vulnerability of
each building or structure to damage were taken into account,
coupled with detailed structural analysis in assessing the build-
ings falling in risk category 3 classified by Mair et. al. (1996).
7. SUMMARY OF INSTRUMENTATION PROGRAM
Instrumentation played a major role in the construction of this

first underground mass transit system project in congestive urban
area with difficult soil condition. The primary objective of the in-
strumentation program was to monitor the performance of the
deep excavation and tunneling operation to ensure safe execution
of the construction works and that adjacent structures were not
adversely affected. Furthermore, the instrumentation program
Fig. 6. View of EPB Tunnel Boring Machine after cutting was established to provide feedback in application of the obser-
through diaphragm wall vational method. Extensive sets of instrumentation were installed
around the predefined influence zones of station box excavation,
tunneling and associated building, structures and utilities. In ad-
dition to other geotechnical instruments, inclinometers, extenso-
meters, various types of settlement points, tile-meter, crack meter
and crack gauge were installed to monitor the ground and asso-
ciated building and structures movement.
540
8. BUILDING CONDITION SURVEY 9. BUILDING DAMAGE ASSESSMENT AND APPLICATIO
OF OBSERVATIONAL METHOD
Prior to commencement of the major works, all the existing
buildings and structures within the predefined influence zone The main reasons of applying observational method in risk as-
were systematically surveyed. The primary objectives of the sur- sessment were ;
veys were;  Uncertainty in Ground Movement Prediction and Staged
 To record pre-construction status of all buildings and struc- Assessment of Buildings
tures within the influence zone. The records included were  Due to the different types of buildings and structures along
age of building, building type, storey height, expected foun- MRTA alignment it was recognized that results obtained
dation types, visible defects supported by sketches and pho- from Building Damage Risk Assessment might be con-
tographs as appropriate servative for some buildings / structures and could be un-
 To utilize the information collected from the building / derestimated for others
structure condition survey in the assessment of the buildings  To minimize unnecessary implementation of building pro-
and evaluate sensitive structures with respect to construction tection measures without compromising safety
aspects and monitoring
 To establish a benchmark to monitor the possible effect of Figure 8 shows the link between observational method and
construction and feedback information in applying observa- geotechnical risk management, particularly risk of damage to ex-
tional method and if necessary to implement the building isting buildings and structures. As mentioned in earlier sections
protection measure of the paper, controlling and managing the risk is one of the most
Expected foundation types of different buildings were classi- important factors in successful completion of the tunnelling
fied based on the height, number of floors and age of the build- works in urban environment and hence all parties involved had to
ings. Building survey information was also used to identify the pay special attention on the issue.
types of instrumentation required.
Geotechnical
OM
Risk Management
Identify Risk
Risk of damage to building
Design & Plan
Obtained data from fully instrumented

first few locations, reanalyzed and back-calculated
Analysed Degree
Construction Control
of Risk
(Building Damage Category)
Monitoring & Review
Trigger Criteria
Exceeds No
Control
Implement
Yes Planned Action
Fig. 8. Flowchart showing the OM process involved in Building Damage Risk Assessment
Stage 1 No further
Prediction and preparation of surface settlement < 10mm
assessment
(Green Field Site Condition)
>10mm
Stage 2 No further
Determination of building risk category < slight
assessment
(limiting tensile strain) based on surface settlement
> slight
Additional Stage 2 No further

Determination of building risk category (limiting tensile strain) < slight assessment
based on sub-surface settlement
> slight
Stage 3
Detailed Assessment
(building structural check, consider building stiffness, Soil-structure interaction)
Fig. 9. Diagram showing process of Staged-Assessment
541
Process of Stage-Assessment is shown in figure 9. After car- total settlement induced by the twin tunnels. 10mm and 25mm
rying out the initial assessment of potential building damage, it settlement contours were plotted in the topographic plan with
was recognized that without applying the observational method, buildings and other structures including flyover and khlong
the result obtained from assessment process could be conserva- bridges in the tunnel alignment. Buildings falling outside 10mm
tive for some buildings and could be underestimated for others. settlement contours were eliminated from further assessment.
For example, the settlement estimated in Second Stage Assess- Remaining buildings with settlement greater than 10mm were
ment shown in above diagram was based on surface movement. short-listed for stage 2 assessment.
For the buildings located in close proximity to the excavation
founded on piles, it was considered that subsurface settlement 10.3 Stage 2 Assessment
may be more critical. Stage 2 Assessment was then further ex-
tended by Additional Stage 2 Assessment in which the variation The buildings and structures short-listed in stage 1 were assessed
of settlement trough shape and magnitude with depth were consi- in this stage. Stage 2 assessment involves calculation of maxi-
dered in identifying the risk of building damage. Settlement of mum tensile strain induced in the building and structures due to
the buildings due to the excavation works were predicted assum- tunnel boring and classification of risk category. Risk category
ing the “green filed” condition. Neglecting restraints from foun- of the buildings was determined applying the method proposed
dation and structure, it was assumed that buildings follow the by Mair et. al. (1996),
ground settlement trough at foundation level (estimated pile tip
level). However, if the building is supported by deep founda- 10.4 Additional Stage 2 Assessment in tunnel zones
tions, assessment result could be too conservative.
As most of the buildings and structures in Bangkok along tunnel
10. PREDICTION OF GROUND MOVERMENT INDUCED alignment are supported by piled foundations, it was considered
BY TUNNELLING WORK AND RISK ASSESSMENT that Stage 2 Assessment using surface settlement profile might be
insufficient. Hence, Additional Stage 2 Assessment considering
10.1 Prediction of surface settlement the variation of settlement trough shape and magnitude with
depth (subsurface settlement) was carried out for the critical
The Gaussian normal distribution curve developed by Peck buildings.
(1969), O’Reilly and New (1982) was used to estimate the In general, subsurface settlement trough will be narrower and
ground settlement induced by a single tunnel.
steeper than those of the surface as shown in the sketch below. A
simple approach was used in predicting the subsurface settlement
profile with following assumptions.
First stage assessment of the risk of damage to buildings and (1) The shape of the subsurface settlement profile caused
structures in tunneling zones involved preparation of settlement by the tunnel construction is characterized by a Gaus-
profile induced by tunnel excavation. Settlement was computed sian distribution similar to surface settlement profile.
at every 100m interval along tunnel alignment using the method (2) The ground settlement influence line for tunnel in clay
based on Gaussian normal distribution curve presented above. is the same for subsurface as shown in the figure 10 be-
Settlement induced by each tunnel (north and south bound) was low:
predicted independently and then superimposed to determine the .
Fig. 10. Distribution of surface and subsurface settlement induced by tunnel boring.
542
Therefore, substituting the distance above the tunnel axis Z0-Z  The risk of damage to the building in terms of magni-
for Z0, equations derived from Gaussian normal distribution tude of building slope with reference to the limiting val-
curve subsurface settlement profile was determined. In addition ues specified by various researchers.
to this method following equation can also be used to predict the The following assumptions were made in the conservative
subsurface settlement. approach of Stage 3 assessment.
 Buidling movement is influence by excavation induced
K’ = [0.14+0.26(1-z/z0)] / (1-z/z0)
settlement of both main station box and associated
underground structures (e.g. Entrance Building)
By substituting K’ values for K in equation derived from
 Building settlement was casued by subsurface
Gaussian normal distribution curve the subsurface settlement at
settlement at foundation level as buildings are supported
required depth can be calculated. by piled foundation.
 No allowance was made for structural stiffness of the
building.
0
 No allowance was made for possible effects of soil-
-10 structure interaction.
-20 Surface_Settlement_NB Tunnel
A summary of risk category determination in Stage 3 Assess-
-30
Surface_Settlement_SB Tunnel ment based on the predicted and monitored settlement of the
Surface_Settlement_Total
building adjacent to station excavation work is shown in the Ta-
-40 Settlement @ 12m _ NB Tunnel
Settlement @ 12m_SB Tunnel ble 1 below.
-50 Total Settlement @12m
SB-TUNNEL CENTER
-60
NB_TUNNEL CENTER Table 1. Summary of risk category determination based on Bos-
-70 carding and Cording (1989)
0 10 20 30 40 50 60 70 80
Distance (m)
Angular Distortion Risk catego-
Data Taken
ry / degree
From Sv1 Sv2 L
S
 of damage
Fig. 11. Surface and sub-surface troughs (mm) (mm) (mm) (m)
Transverse Settlement Trough due to Tunnel Boring
Settlement
48 23 25 18.6 0.0013 2 / slight
Settlement Array, CS-7D(SS-7-53A-0 )
Prediction
20
CL_NB
15 CL_SB
12/11/99 12:00
Actual
10
04/01/00 12:00
5 19/01/00 12:00
25/01/00 21:00
0
-5
26/01/00 06:00
26/01/00 09:00
Monitoring 53 14 39 18.6 0.0020 2 / slight
Data
26/01/00 12:00
-10 26/01/00 15:00
26/01/00 18:00
*Predicted horizontal movement was used in computing

-15
27/01/00 00:00
Settlement, m
-20 27/01/00 03:00
horizontal strain for actual miniotring data.

27/01/00 06:00
-25
27/01/00 09:00
-30 27/01/00 12:00
27/01/00 15:00
-35 27/01/00 21:00
-40 28/01/00 03:00
28/01/00 12:00
From the above table, following conclusions can be made.
-45 29/01/00 12:00
31/01/00 12:00
 Predicted angular distortion broadly agrees with that of

-50
02/02/00 12:00
-55 09/02/00 12:00
actual monitoring data. Hence, settlement prediction and

02/03/00 12:00
-60
05/04/00 12:00
-65 Predicted
-70
-25 -20 -15 -10 -5 0 5 10 15 20 25
assumed effective depth of piled foundation is fairly
Distance, m
reasonable.
Fig. 12. Actual settlement profile monitored at fully instru-  The settlement of actual monitroing data is slightly
mented location in comparison with prediction higher than that of prediction. This may be due to the
additional settlement caused by diaphragm wall panel
10.5 Stage 3 Assessment construction which influence zone might have extended
to 15m from the diaphragm wall (equivalent distance to
The method used in Stage 3 Assessment of the buildings in tun- bottom of soft clay level ).
Limiting values of the building slopes at two critical condi-
nel zone was based on the superimposed predicted subsurface
tions specified by various researchers as shown in the table 2 be-
settlement trough induced by twin tunnels.
low were also used in determining the potential building damage
Based on the pre-construction state building condition survey
of stage 3 assessment.
and detailed structural survery, the method for damage
assessment was selected. Two main approaches were applied. Table 2. Summary of Limiting Building Slope Specified by Var-
ious Researchers
 Conservative approach, mainly applied in additional
stage 2 assessment with some refinements Limiting Values by Type of Damage
 Frame analysis using FEM structural software Various Researchers Cracking in walls Structural Damage
and partitions
1. Skempton and 1/300 to 1/500 1/150
10.6 Stage 3 Assessment by conservative approach
Macdonald
Conservative approach applied in Stage 3 Assessment involved 2. Meyerhof 1/500 1/250
consideration of two separate criteria, namely ;
3. Polshin and Tokar 1/500 1/200
 The risk of damage to the building specified by angular 4. Bjerrum 1/500 1/150
distortion and horizontal strain.
543
10.7 Stage 3 Assessment by structural frame analysis  To compare the result obtained from above item with al-
lowable values.
Some buildings of critical condition were assessed by structural
 To establish the final trigger values based on above com-
analysis using detailed structural survey information as basic da-
parison.
ta. By simulating the differential settlement of the building pre-
dicted from Additional Stage 2 Assessment, development of  To confirm the settlement profile from actual monitoring of
moment and shear in the main structural members were deter- last array is within the alert level of final trigger value.
mined with the use of structural analysis software. Computed
moment and shear were assessed in comparison with the capacity
of the structural members estimated from collected building con-
dition survey and detailed structural survey data. The process in-
volved in structural assessment can be summarized below.
 Factor of safety of each structural members were de-
termined base on predicted differential settlement of
the buildings.
 A series of assessment were performed to determine
the critical differential settlement causing structural in-
stability of the members or differential settlement at
limiting factor of safety.
 Trigger values were established based on above results.
 Based on the trigger values and the actual building
condition and response in conjunction with monitoring
data, the most sensitive structural members were pro-
Fig. 13. Position of MWA water tunnel in relation to MRT tun-
tected, mainly by strengthening approach.
nels (clear distance are 3.38m and 3.18m for North and South
bound respectively) at the first crossing.
10.8 Damage Assessment on Sensitive Buried Structures
Apart from the flyovers (bridges), the most critical structure re-
quired for Stage 3 Assessment were MWA main water tunnel as
tunnel alignment crosses the water tunnels at two critical loca-
tions – the first near Rama IV and Rachadaphisek road intersec-
tion and the second at Rama IV and Siphraya road intersection
near Sam Yan station. The position of water tunnels in relation to
MRT tunnels drawn in the subsoil profile at relevant locations
are shown in Figure 13 and 14.
According to the available information, MWA tunnel at the
first crossing is of 2.5m internal diameter steel lining with
150mm segmental ring and at the second crossing MWA tunnel
is of 2.5m diameter segmental ring with lightly reinforced con-
crete lining.
In preliminary stage 3 assessment, based on the predicted
subsurface profile using the method described in published paper
Fig. 14. Position of MWA water tunnel in relation to MRT tun-
“Ground Movement Prediction and Building Damage Risk-
nels (clear distance are 1.5m and 2m for North and South bound
Assessment for the Deep Excavations and Tunneling Works in
respectively) at the second crossing
Bangkok Subsoil”, Zaw et. al (2006) e, structural assessment of
water main was carried out and results were compared with struc-
For the first location of MRT-MWA tunnel crossing, target
tural capacity of water tunnel calculated from the material prop-
ground loss from tunnel operation was set at 0.5%, whereas
erties indicated in the construction drawings provided by MWA.
ground loss based on allowable movement curvature of water
As bending moment induced in the water tunnel was considered
tunnel derived from predicted pattern of subsurface settlement
critical parameter in assessing the damage which is governed by
was 0.65%. Back-calculated ground loss derived from the moni-
both maximum settlement and curvature of the water tunnel fur-
toring data of rod-extensometer placed at the crown of MWA wa-
ther assessment was proposed coupled with observational ap-
ter tunnel was within target value. In addition to the preliminary
proach.
assessment based on prediction of subsurface settlement using
 To carry out back-calculation using monitoring data ob-
simple empirical formulas, a detailed soil-structure interaction
tained from the settlement arrays (settlement points, exten-
was analyzed using 2-dimensional FEM considering behavior of
someter data) at first 3 locations before reaching to
tunnel boring by earth pressure balance (EPB) method. Stress re-
MWWA water tunnel near Rama IV – Rachadapisek inter-
section, to determine the volume loss and relationship be- lief from the effect of cutter-head, tail void and over-cut from
tween surface and subsurface settlement as well as best es- EPB tunnel boring was considered in the analyses. Adding
timation of subsurface K value and subsurface settlement ground movement due to consolidation caused by disturbance
profile at water tunnel level. from tunnel operation surface and subsurface settlement were
 To perform beam-on-elastic foundation analysis using the predicted by finite element analysis. Detailed parameters and as-
subsurface settlement profile obtained from predicted sub- sumption made in analysis was presented in the published paper
surface settlement bending moment likely to occur in of Sakai & Sugden (2000).
MWWA water tunnel.
544
After the successful passing of both tunnels at the first MWA
crossing, another trial area tunneling was planned to further de- REFERENCES
termine the ground movement associated with various tunneling
parameters well before reaching to more critical crossing after Boscarding, M.D. and Cording, E.G. (1989), Building Response
Sam Yan station. Tunnel section between Silom and Samyan sta- to Excavation-Induced Settlement. Journal of Geotechnical
tion, in which soil profile similar to that of second crossing was Engineering, ASCE, 115;1;1-21
selected and comprehensive instrumentation program was estab- Burland, J.B., and Wroth, C.P. (1974), Settlement of Buildings
lished to monitor surface and subsurface settlement as well as ho- and Associated Damage. SOA Review, Conference on Set-
rizontal movement of the ground at depth. tlement of Structures, Cambridge, Pentech Press. London,
pp. 611-654
Burland J.B., Broms J.B. and de Mello VFB (1977), Behavior of
Distance (m)
Foundations and Structures, SOA Report Session 2, Proceed-
0 10 20 30 40 50 60 70 80 90
ings of 9th International Conference, SMFE, Tokyo, 2:495-
546
-10 Clayton C. R. I (2001), Managing Geotechnical Risk, Improving

Productivity in UK Building and Construction, Institute of
Civil Engineers and Thosmas Telford Ltd. UK
-30 Clough, G.W., and O’Rourke, T.D., (1990), Construction In-
duced Movements of In-situ Walls, ASCE Geotechnical
Total Settlement Special Publication No. 25, pp. 439-470
-50
Clough C.W. and T.D. O’Rourke (1990), Deep Excavation and
Tunneling, 7th ICSME, Mexico City
Graftio H. (1936), Some Features in Connection with the Foun-
Fig. 15. Predicted subsurface settlement trough of assumed dation of Svir 3 Hydro-electric Power Development,” Pro-
ground loss 1% calculated from empirical formula and actual ceedings of the first international conference in Soil Mechan-
subsurface settlement (indicated by dark-circle) observed from ics, Cambridge, Mass., 1, pp. 284-290
rod extensometer – first MWA tunnel crossing. JV BCKT, (1997) Geotechnical Interpretation Report (GIR) for
MRTA Initial System - UGS, Bangkok
The observed ground movement induced by tunneling works Mair, R.J., Taylor, R.N., and Burland, J.B., (1996), Prediction of
were within predicted values. Well organized plan, systematic Ground Movements and Assessment of Risk of Building
approach of observational method to achieve control values Damage. Geotechnical Aspects of Underground Construction
backup by extensive instrumentation and close co-operation in Soft Ground, pp. 712-718, Balkema, Rotterdam
among relevant personnel were the key factors contributed to O’Reilly M.P and New B.M (1982), Settlements above tunnels in
successful completion of EPB tunneling particularly in the most United Kingdom – their magnitude and prediction, Tunneling
sensitive points at MWA water tunnel crossings. The monitored ’82, pp. 173-181, London, IMM
subsurface settlement data from extensometer and predicted set- Peck R.B. , (1969) Deep Excavation and Tunneling in Soft
tlement trough of assumed ground loss 1% ground settlement Ground, 7th ICSME, Mexico City, State of the Art 3, 225-290
profile of the first MWA tunnel crossing is presented in figure Peck, R. B. [1969]. “Advantages and limitations of the observa-
15. tional method in applied soil mechanics”, Geotechniques 19,
No.2, pp. 171-187
Phienewej N., Photayanuvat C., Suwansawat S., Mcconochie D.,
10. KEY FACTORS IN SUCESSFUL COMPLETION OF THE (2006) “Geotechnical Aspects in Design and Construction of
PROJECT Chaloemratchamongkhon Underground MRTA, Bangkok”,
Proceedings of the International Symposium on Underground
The M.R.T Chaloem Ratchamongkhon Line, one of the most Excavation and Tunneling, Urban Tunnel Construction and
modern underground mass transit systems in the region was suc- Protection of Environment, EIT, Bangkok, Thailand, , pp.
cessfully completed and has been in operation for over 6 years. 271-279
All deep excavation works and tunneling were successfully com- Powderham, A. J. [1996]. “An overview of the observational me-
pleted without any significant damages to the adjacent buildings thod : development in cut and cover and bored tunneling
and structures. Well organized plan, systematic approach of ob- projects”, The Observational Method in Geotechnical Engi-
servational method to achieve control values backup by extensive neering, The institute of civil engineers, Thomas Telford Pub-
instrumentation and close co-operation among relevant personnel lishing, London, pp. 7-24
were the key factors contributed to successful completion of the Zaw Zaw Aye , Dhiraj Karki and Christian Schulz [2006].
project. The purpose of this paper is to provide a set of technical “Ground Movement Prediction and Building Damage Risk-
guidelines for ground movement prediction and building damage Assessment for the Deep Excavations and Tunneling Works
risk assessment of deep excavation and tunneling works in Bang- in Bangkok Subsoil, Proceedings of the International Sympo-
kok subsoil and application of observational method which may sium on Underground Excavation and Tunneling, Urban Tun-
partly serve as a reference for the future underground mass tran- nel Construction and Protection of Environment, EIT, Bang-
sit system projects in urban area of Bangkok. Successful comple- kok, Thailand, , pp. 281-297
tion of the first MRT project in Bangkok marked the practicality
of well-executed deep excavation works and tunneling by EPB
method coupled with observational method in the sensitive urban
environment.
545
Geotechnical Special Publication no. 225
25-29 March 2012, Oakland, California, USA
Geocongress 2012 - State of

the Art and Practice in
Tunneling effects on pile group response in

Bangkok
Thayanan Boonyarak and Charles W. W. Ng
The Hong Kong University of Science and Technology
547
GeoCongress 2012 © ASCE 2012 3119
Tunneling effects on pile group response in Bangkok
Thayanan Boonyarak1 and Charles W. W. Ng2

1
Research student: thayanan@ust.hk
2
Chair Professor, Ph.D, CEng, FICE, FASCE, FHKIE, FHKEng: cecwwng@ust.hk
The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon,
HKSAR
ABSTRACT
Tunneling adjacent to an existing pile group may induce additional force,

bending moment and settlement of a pile. Moreover, it may cause load redistribution
and differential settlement in a pile group. In this paper, the construction of a water
diversion tunnel with an outside diameter (D) of 5.55 m constructed in Bangkok is
reported. The tunnel was constructed in very stiff clay using an earth pressure
balance shield with a cover-to-diameter ratio of 4.5 and very close to a pile group.
The pile group was located within a distance of 1D from the tunnel springline.
Tunneling induced ground responses were measured by using surface settlement
points, extensometers and inclinometers. To investigate the influence of tunnelling
on the pile group, a three-dimensional, elasto-plastic, coupled-consolidation
numerical analysis was also conducted to simulate the pile group responses.
Observed ground responses and computed changes of horizontal effective stress and
excess pore water pressure around piles are reported. Redistribution of pile shaft
resistance, load transfer in the pile group and pile head settlements are investigated.
INTRODUCTION
Tunneling adjacent to an existing pile group may induce additional bending moment,
cause redistribution of axial force in each pile and load transfer in a pile group. Any
stress change in the ground due to tunnelling is the main concern to engineers in
evaluating capacity and additional settlement of a pile group.
Chen et al. (1999) and Lee & Ng (2005) carried out numerical analyses to
study the influence of tunneling on adjacent pile. They found that negative skin
friction was induced along the pile shaft as a result of stress relief. Tunneling
adjacent to a pile group was also studied by Pang et al. (2005) and Cheng et al.
(2007). It is reported that the maximum bending moment induced in a pile occurs at
the tunnel axis in the transverse direction of tunnel advancement while no significant
bending moment occurs in the longitudinal direction.
In this paper, monitored soil displacements due to a 5.5m diameter shield
tunnelling adjacent to a pile group are reported and compared with computed finite
element results. The capacity and settlement of the pile group due to the tunneling
are also investigated.
549
THE SITE
Figure 1a shows plan of tunnel alignment and adjacent pile groups of elevated
railways. The layout of instrumentation is also shown in the Figure. Various
instruments were installed to observe tunneling induced ground displacement and
structural response. Three surface settlement points (S1, S2 and S3), two
extensometers (Ex1 and Ex2) and one inclinometer tube (I1) were installed to
measure ground response. Two pier settlement points (C1 and C2) were mounted on
pier to measure pier settlement. The monitored data is interpreted from “BMA flood
diversion tunnel project: Sansaeb Canal and Ladprao Canal Section” (Lim et. al,
2009). The total length of tunnel in this project is about 5 km.
Figure 1b shows a typical soil profile of the study area. The first upper layer
is 12 m thick soft clay. Below it is the succession of medium stiff clay of 3 m, stiff
clay of 5 m, very stiff clay of 14 m and stiff clay of 14 m thick layers. A 12 m very
dense sand layer underlies the clay layer. A water diversion tunnel with an outer
diameter (D) of 5.55 m was constructed in very stiff clay at cover-to-diameter ratio
(C/D) of 4.5 using an earth pressure balance shield machine. The tunnel alignment is
very close to a (2×2) pile group with center to center of pile spacing 3 m. The outer
diameter and length of the shield are 5.70 m and 8.05 m, respectively. The diameter
of each concrete bored pile is 1.0 m. The length of piles is 60 m and pile toes are
embedded in very dense sand layer. All four piles are rigidly connected with 1.8 m
thick pile cap. The clear distance between the nearest pile and the tunnel springline is
3.65 m. The tunnel axis depth is approximate at the middle of pile length.
3.00
(a) Instrumentation (b) S2
0.00 m C1
Surface settlement point -3.0 m
Soft
Sukh
Pier settlement point clay Ex1&Ex2

Extensometer -12.0 m Medium
umvi
-12.0 m
Inclinometer -15.0 m stiff clay -15.0 m
Ex1 Stiff
t 71 R
-20.0 m clay -20.0 m
Very 3.65
oad
stiff -27.55 m
clay
3 S1 -34.0 m 5.55 m
.6 Sukhumvit Road
18 I1
C2 C1 Stiff -40.50 m
4.35 I1
clay
8
.3
13 S2 3.65 -48.0 m
4- 1.0 m
Ex2 Concrete pile
S3 Very
dense
Scale: sand
0 10 20 Shield advancing -60.0 m
SECTION A
Figure 1. (a) Plan of study area and lay out of instrumentation (b) sectional view
A-A and soil profile
NUMERICAL ANALYSIS
To investigate the influence of tunnelling on the pile group, a three-dimensional,

elasto-plastic, coupled-consolidation numerical analysis was carried out using Finite
Element program ABAQUS (Hibbitt et al., 2008).
550
Figure 2a shows the geometry of a 3D finite element mesh. A plane of

symmetry is identified at the tunnel axis, so only half of the domain is modeled. The
mesh is 40 m wide, 50 m long and 70 m high. Eight-node brick elements, four-node
shell elements and two-node beam elements are used to model the soil, pile cap and
piles, respectively. Roller and pin supports are assigned on the four vertical sides and
bottom of the mesh as boundary conditions, respectively.
Figure 2b illustrates tunnel excavation simulation technique adopted in the
finite element analysis (FEA). This technique is known as Displacement Controlled
Method (DCM), modified from Cheng et al. (2007). During shield advancement
over-excavation due to pitching angle is often encountered at the site. Therefore, it is
assumed that soil displacement at the upper part of the shield is larger than the lower
part of the shield and the shield’s invert is in contact with soil element. To simulate
each tunnel advancing step, displacement boundary is specified to each node around
the excavated perimeter. Consequently, soil element around the shield is pulled to a
final tunnel diameter of 5.55 m. Also, soil inside the excavated section is removed
simultaneously. A volume loss of 2.5% was imposed on tunnel boundary. By using
this technique, tunnel lining is not required. In order to simulate an earth pressure
balance shield, a roller support is assigned to the shield face to restrain horizontal
soil movement in front of the shield. Shield advancing speed of 5 m per day is
modeled in the analysis. Tunnel alignment is simplified from curve in actual
condition to be straight in FEA. The presence of pile foundations of surrounding
buildings are ignored in the analysis.
In Figure 2c piles and pile cap are arranged in rotated square. Total load on
the pile group is assumed to be 18,000 kN. Excessive pore water pressure generated
in response of the applied load on pile group is allowed to dissipate before tunnel
advancement in the numerical analysis.
(a) 50 m (b) (c) 18000 kN
Y=0 40 m Excavated
diameter,
VL = 2.5% P2
P4
P3
Tunnel Tunnel axis
70 m 3m
axis 3m P1
Final tunnel
diameter
5.55 m φ 1.0 m
Beam
Note: element
Direction of shield advancement
Plane of symmetry
Figure 2. (a) Geometry of finite element mesh (b) tunnel excavation simulation
(c) detail of piles and pile cap
551
An elasto-plastic soil model with Mohr-Coulomb failure criterion is used to

simulate soil behaviour. Soil parameters are selected based on a soil investigation
report and adopted from previous studies [Tanaka et. al (2001), Seah et. al (2000)].
Permeability of clay layers and sand layer is assumed to be 1x10-9 and 1x10-3 m/s,
respectively. Soil parameters are summarized in Table 1. The concrete piles and pile
cap are modeled as linear elastic with Young’s modulus and Poisson’s ratio as 30
GPa and 0.3, respectively. The unit weight of concrete is assumed as 24 kN/m3. The
water table is assumed to be located at the ground surface with a hydrostatic initial
pore water pressure profile.
Table 1. Soil parameters in finite element analysis

Material γt Su SPT c’ φ’ ϕ’ ν' E’
(kN/m3) (kPa) (N/ft) (kPa) () () (kPa)
Soft clay 16 18-20 N/A 1.5 18 - 0.15 1,200
Stiff clay 1 19 90-120 12-15 2.0 19 - 0.15 5,000
Very stiff clay 20 160-200 25-30 3.0 20 - 0.15 7,000
Stiff clay 2 19 100-130 13-18 2.0 19.5 - 0.15 6,000
Very dense sand 20 - 50-60 0.1 35 6 0.15 40,000
MONITORED DATA AND COMPUTED RESULTS
In this paper, only measured ground horizontal displacement is reported. Reference

point of the shield distance used in this paper is taken at the shield face. A
monitoring section is selected at the center of the pile group (i.e., at Y=0) as
reference point for tunnel advancement.
Ground horizontal displacement. Measured horizontal ground displacement in

transverse direction of tunnel advancement is shown in Figure 3. When the shield
was more than 12 m away from the monitoring section, measured horizontal ground
displacement was insignificant. When the shield reached at 10 m behind the
monitoring section (i.e., Y= -10 m), soil moved outward from the shield. It indicates
that stress redistribution of soil due to soil arching may occur. More discussion and
explanation is given in the later section. Maximum outward movement of 3 mm from
the tunnel axis was observed at this stage.
When the shield reached the monitoring section (i.e., Y=0 m), soil abruptly
moved inward towards the shield and caused maximum inward displacement. Large
displacement occurred from depth 25 to 32 m which is considered about 1D (tunnel
diameter) from the tunnel axis. Measured horizontal displacement at the tunnel axis
depth was about 15 mm inward to the shield. When the shield reached at 25 m ahead
of the monitoring section (i.e., Y=25 m), measured horizontal displacement reduced
by 3 to 5 mm. The reduction may be caused by stress redistribution. No significant
horizontal ground displacement was observed when the shield advanced further.
Computed ground horizontal displacements show a similar trend as compared
with the measured data. When the shield reaches at 10 m away from the monitoring
section (i.e., Y=-10 m), computed outward soil movement from the shield occurs
552
from depth 23 m to 40 m. When the shield reaches the monitoring section (i.e., Y=0
m), computed results show an abrupt soil displacement towards the shield. In
contrary to the measured data, when the shield passed at 25 m (i.e., Y=25 m) beyond
the monitoring section, a slight additional horizontal displacement towards the shield
occurs. The differences between measured and computed results may be caused by
tail void grouting is not simulated in the FEA. Therefore, outward movement in FEA
as the shield advances beyond the monitoring section does not occur. It can be
noticed that, the computed ground horizontal movement from FEA can reasonably
capture the measured ground response at the depth of tunnel axis. However, there is a
significant difference between computed and measured ground response near the
ground surface. Large measured horizontal ground displacement may be triggered
by other activities rather than tunneling such as surcharge induced by heavy traffic
load near the inclinometer tube.
Horizontal displacement (mm)

-20 -15 -10 -5 0 5 10 15 20
0
Shield distance, Y
5 Measured (Y=-10m)
Measured (Y=0m)
10
Measured (Y=25m)
15 Computed (Y=-10 m)
Depth (m)
Computed (Y=0m)
20
Computed (Y=25m)
25
Tunnel
30 -
I1
axis depth
35
+
Horizontal displacement sign

40 convention
-Y +Y + : inward to tunnel axis
- : outward from tunnel axis
45
Figure 3 Comparison of measured and computed horizontal ground

displacement of I1 (measured data from Lim et al, 2009)
COMPUTED PILE GROUP RESPONSE
Load transfer in the pile group. Figure 4a shows changes of computed load at pile
head with respect to the shield advancement. The changes of load at the pile head are
noticeable when the shield advances from Y = -5 to -2 m. Load on P2 significantly
increases while load on P1, P3 and P4 decreases. Since P2 is the farthest pile from
the shield, it is the least affected by tunneling. This explanation is verified with
computed mobilized shaft resistance in the following section. Consequently, load at
pile head from P1, P3 and P4 is transferred to P2. Steady state of load transfer in pile
group is achieved when shield passes at Y=10 m ahead of the monitoring section.
553
Maximum reduction and increment in load on pile head of P1 and P2 is 0.8% and
1.2%, respectively.
Pile head settlement and pier tilting. Figure 4b illustrates the computed results of
pile head settlement against shield advancement. In addition to effects of tunneling,
the observed pier settlement was highly influenced by other factors such as vibration
induced by heavy traffic and the elevated train. Furthermore, pier settlement was
barely noticeable by conventional surveying instrument. Therefore, the measured
pier settlement result is not reported in this paper. Maximum computed pile head
settlement occurs at P3 since significant shaft resistance reduction in this pile occurs
when the shield is approaching to the monitoring section. Moreover, load at pile head
on P3 is almost constant (see Figure 4a). Pile head settlement in P1 and P2 is almost
equal although different mechanism is developed. Pile head settlement in P1 would
be mainly caused by reduction of horizontal effective stress and development of
negative shaft resistance which is discussed in following section. Elastic shortening
due to increase of load at pile head would be the major cause of pile head settlement
in P2. The reduction of load at pile head in P4 is higher than in P3 (see Figure 4a).
Therefore, minimum pile head settlement is induced on P4. Tunneling induced pier
settlement is about 1.3 mm. The differential settlement in the pile group causes the
pier to tilt in the longitudinal direction of the tunnel axis toward P3. The maximum
inclination of the pier becomes 1:12,500 when the shield is 5 m away from the
monitoring section (Y = -5 m). In addition, the computed differential settlement
between two piers is 1:27,000 at the end of tunnel driving (the distance between two
piers is 35 m). Those computed results are considered very small for the
serviceability limit state criteria.
(a) Shield distance, Y (m) (b) Shield distance, Y (m)
-20 -10 0 10 20 30 -20 -10 0 10 20 30
1.5 P2
0.0
P4
1.0 P3
P1
Δδp (mm)
0.5
ΔP (%)
0.5
-Y +Y
0.0
1.0
-0.5 P1 P2 Avg. 1.3 mm
-1.0 P3 P4 1.5
Figure 4 (a) Changes of computed load at pile head (b) induce pile head
settlement
Mobilized shaft resistance. Figure 5 shows computed mobilized unit shaft

resistance of P1, P2, P3 and P4, during shield advancement. The significant change
of shaft resistance occurs in P1 as shown in Figure 5a. Mobilized shaft resistance
decreases significantly at depth 13 to 24 m. Negative shaft resistance develops in P1
when the shield reaches at the monitoring section and advances further, indicating
that soil settles more than pile at depth 12 to 25 m. To maintain load equilibrium in a
pile, mobilized shaft resistance in P1 below the tunnel axis increases. The change of
mobilized shaft resistance in P2 is the lowest as shown in Figure 5b. Negative shaft
554
resistance did not develop in P2, which suggests that impact of tunneling on P2 is the
smallest. It is suggested that tunneling influence zone is about 1.5D from the tunnel
axis. P3 and P4 respond to the shield advancement in the same trend as shown in
Figure 5c and 5d, respectively. The only difference that can be noticed is significant
change occurs in P3 when the shield is at 5 m away from the monitoring section. The
significant change in P4 occurs when the shield is at 5 m beyond the monitoring
section. In other words, the maximum reduction of shaft resistance on each pile
occurs when the shield face is slightly beyond the pile.
(a) fs (kPa) (b) fs (kPa) (c) fs (kPa) (d) fs (kPa)
-10 -5 0 5 10 15 -5 0 5 10 15 -5 0 5 10 15 -5 0 5 10 15
0
5
P1 Initial
P2 P3 P4
Y=-5m
Y=0m
10
Y=5m
15 Y=25m
20
Depth (m)
25 Tunnel
30 axis
35
40
45 P2
P4
P3
P1 Sign convention for horizontal axis
+ : positive shaft resistance (piles settle more than soil)
- : negative shaft resistance (soil settle more than pile)
-Y +Y
Figure 5 Computed mobilized unit shaft resistance of (a) P1 (b) P2 (c) P3 (d) P4
Induced axial force and bending moment in pile. Figure 6a illustrates computed
tunneling induced axial force in each pile when shield is at 5m ahead of monitoring
section. The maximum induced axial force occurs in P1 at the depth of tunnel axis.
The trend of induced axial force distribution of P2, P3 and P4 is similar to P1. The
differences of magnitude of induced axial force are due to development of negative
shaft resistance in P1 (see Figure 5a). It can be noticed that minimum induced axial
force occurs in P2 as a result of minimum reduction of unit shaft resistance (see
Figure 5b).
Figure 6b shows the distribution of computed bending moment of each pile in
longitudinal direction of the tunnel axis (Mx). Maximum Mx occurs in P1 when the
shield is at 5 m beyond the pile (Y = 5 m). The bending moment (Mx) causes every
pile to bulge in the same direction as the shield advancement at the depth of tunnel
axis. Bending moment in each pile extends to depth 50 m, which is about 4D below
the tunnel axis.
Figure 6c illustrates the distribution of computed bending moment of each
pile in transverse direction of tunnel axis (My). Maximum My occurs in P1 when the
shield is at 5 m beyond the pile. The bending moment (My) causes the pile to bulge
towards the shield due to stress relief around the shield. By comparing bending
moment of a pile in 2 directions, My is significantly larger than Mx due to higher
555
stress reduction around the shield in transverse direction than in the longitudinal
direction.
The ultimate bending moment capacity of a reinforced concrete bored pile
with diameter 1.0 m is approximate between 1,000 and 1,400 kN.m, given the
reinforcement ratio is 0.5% of the pile section. Therefore, maximum computed
induced bending moment in P1 is only about 20% of its capacity.
(a) ΔFx (kN) (b) Mx (kN.m) (c) My (kN.m)

-100 0 100 200 300 400 -100 -50 0 50 -100 0 100 200
0
P1
P2
10
P3
P4
20
Tunnel
Depth (m)
30 Sign axis
convention Sign
for Mx + convention
P2 + for My
40 P4
P3
P1
50
-Y +Y
X
+ +
60
Figure 6 Computed (a) induced axial load distribution (b) bending moment in
longitudinal direction, Mx (c) bending moment in transverse direction, My in
each pile due to tunneling
Changes of horizontal effective stress and induced excess pore water pressure.
Figure 7a shows changes of computed normalized horizontal effective stress
(Δσ’h/σ’h0) with respect to shield advancement. In this study, coupled-consolidation
analysis is conducted which allows both volume and pore water pressure to change
during tunnel advancement. Soil elements around piles at the depth of tunnel axis are
selected to be investigated. As the shield reaches between Y = -10 to -5 m, Δσ’h/σ’h0
of P1, P2 and P4 increases. This is caused by horizontal effective stress relief around
the shield. Subsequently, arching effect or stress redistribution occurs on soil element
in front of the shield and causes outward movement from the shield. The measured
horizontal ground displacement confirmed the explanation of stress redistribution.
When the shield advances further and reaches at Y= -5 to 3 m, Δσ’h/σ’h0 suddenly
drops due to stress relief around the shield. Horizontal effective stress around piles
slightly increases when the shield advances beyond the monitoring section (Y = 3 to
25 m). This is due to stress redistribution behind the shield causing a slight outward
movement at the piles. The highest change of Δσ’h/σ’h0 occurs at P1 with the highest
increment Δσ’h/σ’h0 of 1% and the maximum reduction Δσ’h/σ’h0 of -3%. The
tunneling impact in terms of changes of horizontal effective stress on P2 is the lowest
as expected.
556
Figure 7b shows changes of excess pore water pressure at the depth of tunnel
axis which has a similar trend to Δσ’h /σ’h0. When the shield approaches to the
monitoring section (Y = -20 to -10 m), Δu/u0 around every pile increases. It indicates
contractive behavior of soil around piles due to stress redistribution induced by
tunnel driving. When the shield reaches at Y= 3 m, Δu/u0 abruptly decreases to
minimum then it gradually rebounds when the shield advances beyond the
monitoring section. The dilative behaviour suggests that the resultant horizontal
displacement of soil around each pile is towards the shield. The maximum change of
Δu/u0 occurs on P1 whereas the minimum occurs on P2. The difference of Δσ’h/σ’h0
and Δu/u0 among P2, P3 and P4 is not significant. It indicates that the influence zone
of tunneling extends to P1 only.
(a) Shield distance, Y (m) (b) Shield distance, Y (m)

-20 -10 0 10 20 30 -20 -10 0 10 20 30
2 2
1 1
Δσ'h/σ'h0 (%)
Δu/u0 (%)
0 0
P2
-1 -1 P4
P1 P3
-2 P2 -2 P1
-3
P3 -3
P4 -Y +Y
-4 -4
Figure 7. Changes of computed (a) horizontal effective stress and (b) excess pore
water pressure of soil element at depth of tunnel axis
CONCLUSIONS
Based on field observations and computed results, the following conclusions may be
drawn. Within the pile group, P1, P2, P3 and P4 denote the closest, the farthest, the
front and the rear pile to the tunnel, respectively.
1. The maximum measured transverse horizontal ground displacement was 0.2%D
(i.e., 15 mm) at depth of the tunnel axis when the shield was at the monitoring
section. When the shield was approaching to the monitoring section, outward soil
movement away from the shield was observed due to stress redistribution. As the
shield reached the monitoring section, an abrupt change of inward horizontal
ground displacement toward the shield was observed as a result of stress relief
around the shield.
2. The influence zone of load transfer and pile head settlement are computed when
the shield advances from 1D (tunnel diameter) away from the center of the pile
group to 1D beyond the center of pile group. As the shield advances towards the
pier, computed pile load at P2 increases because it is the least affected one by
shaft resistance reduction due to tunneling. On the other hand, loads at pile head
of P1, P2 and P3 decrease because of load re-distribution within the pile group.
The maximum pile head settlement occurs on P3 causing the pier to tilt in the
longitudinal direction of tunnel towards P3.
557
3. Based on computed results, tunneling causes a significant reduction of shaft

resistance in P1 at depth above the tunnel axis. To maintain equilibrium, shaft
resistance below the tunnel axis increases. The influence zone of shaft resistance
reduction is estimated to be 1.5D from the tunnel axis.
4. The maximum computed induced axial force, longitudinal and transverse bending
moments occur in P1 at the depth of tunnel axis when the shield advances at 1D
beyond P1. The largest induced axial force occurs due to the development of
negative shaft resistance. The bending moment of the pile in the longitudinal
direction of tunnel advancement (Mx) causes the pile to bulge in the same
direction as the shield advancement. The bending moment of the pile in the
transverse direction of tunnel advancement (My) causes the pile to bulge towards
the shield due to stress relief around the shield.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the financial support provided by the
General Research Fund 617608 from the Research Grants Council of the HKSAR.
REFERENCES
ABAQUS, (2008). Theory Manual, v 6.8. Hibbitt, Karlson and Sorensen Inc. RI
Chen LT, Poulos HG, and Loganathan N. (1999). “Pile responses caused by
tunneling”. J. Geotech. Geoenviron. Eng. 125(3), 207-215.
Cheng CY., Dasari GR., Chow YK.,and Leung CF. (2007). “Finite element analysis
of tunnel–soil–pile interaction using displacement controlled model”.
Tunnelling and Underground Space Technology 22, 450–466
Lee GTK, Ng CWW. (2005). “Effects of Advancing Open face tunneling on an
existing loaded pile”. J. Geotech. Geoenviron. Eng., 131(2), 193-201.
Lim S., Teparaksa W. and Shibuya S. (2009). “Effects of shield tunneling through
the obstructions in Bangkok”. J. of the SEA Geot. Soc., 40 (3): 145-159
Loganathan N. and Poulos HG. (1998). Analytical prediction for tunneling-induced
ground movements in clays. J Geotech Geoenviron Eng, 124(9):846–856
Pang CH., Yong KY. and Chow YK. (2005). “Three-dimensional numerical
simulation of tunnel advancement on adjacent pile foundation.” Underground
Space Use: Analysis of the Past and Lessons for the Future , 1141-1147
Peck RB (1969). “Deep excavation and tunneling in soft ground.” Organizing
committee (eds) Proc. of 7th Int. Conf. in Soil Mechanics and Found. Eng.,
Mexico City. Sociedad Mexicana de Mecanica de Suelos, 225–290
Seah, TH. and Lai, KC. (2000). “Strength and deformation behavior of soft Bangkok
clay”. Geotechnical Testing Journal, Vol. 26, No. 4
Tanaka H., Locat J., Shibuya S., Soon TT, and Shiwakoti DR. (2001).
“Characterization of Singapore, Bangkok, and Ariake clays”. Can. Geotech.
J. 38: 378–400
558

SEAFCO - Special Foundations Practice in Bangkok

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SEAFCO - Special Foundations Practice in Bangkok

Uploaded by

Copyright:

Available Formats

Current Practices in Deep Foundations and Diaphragm Wall Construction

in Thailand is published to mark the auspicious occasion of SEAFCO

Current Practices in Deep Foundations and Diaphragm Wall Construction in Thailand

Seafco Public Company Limited

Current Practices in Deep Foundations

Zaw Zaw Aye

Prof. Dr. Suchatvee Suwansawat

Quality Control Tests on Deep Foundations

Sonic Integrity Test on Piles Founded in Bangkok Subsoil-Signal

Geotechnical Engineering : Research and Practice

Ground Movement Prediction and Building Damage Risk-Assessment

N on-G route d v s. B ase -G route d P ile s (S ilom)

ameter for base grouted and non- 10 Silty Clay

base grouted piles respectively.

base grouting involves;

• Replacement of sediments with 50

grout under the pile base. Dense Sand

• Squeezing water out of the 60

sediments. B a s e -G ro u te d P ile A - D ia .1 .5 x5 5 .1 6 m (3 S ta g e s - to ta l 2 0 0 0 ltr)

• Improving the sand stratum

from 25 bars to 60 bars depending on the depth of

the project area. Sand

Non-Grouted vs. Base-Grouted (V-Rangsit) -CNW

-30 Age: 61 days -30 10

Exposed contiguous bored pile wall.

Layout of the project.

Rebar Cage Installation for Diaphragm Wall.

Concrete casting of 3.0m thick foundation mat in progress.

Architects & Engineers

Substructure works in progress.

Designer & Engineers

Diaphragm Walls: 2,460 sq.m (1.5m thick x

Soil mixing: 1,734 cu.m

Diaphragm wall and bored pile construction in progress.

Owner & Designer

Bird’s eye view of the completed underground car park.

Designer Layout of pile foundation and diaphragm walls.

Exposed diaphragm wall and contiguous piles. Bored piling in progress.

Diaphragm Walls: 13,672sqm

Structural Concrete: 42,230cu.m

Construction of Diaphragm Wall under the Existing Highway.

Layout of the Project.

tural Center) Station (S12) is one

Long. The perimeter walls of the

barrettes (1.2mx3.0m) embedded

44.5-55.0m deep in conjunction TOP SLAB +98.900

with pre-placed steel stanchions

top down construction and to

ORANGE LINE SLAB

Glass Fibre Polymer Reinforce- Stiff Clay DIAPHRAGM WALL

ment was used in diaphragm wall BASE SLAB +80.450

Cross section of station box (Schematic).

TYPE OF WORK: Foundation Piles, Diaphragm Wall and Barrettes

DESIGNER: Ove Arup and Partners International Limited

Layout of station, diaphragm wall and barrettes.

Subsoil conditions at the site.

Panel excavation with grab.

Left – Lowering rebar cage at-

Centre – Setting up SEAFCO

Right – Placing a steel stanchion

Concreting a barrette with pre-placed steel stan-

All rights reserved. No part of this publica-

Monitoring and Checking Bentonite A jack-out pressure cell to be installed

Checking concrete quality of barrette