Deep Excavation in Hong Kong –

Design and Construction Control

Jack Pappin, John Endicott & John Clark
(James Sze)

Content of the Presentation

• Part I – Design Considerations

• Part II – Numerical Modelling
• Part III – Observational Method

1
 Part I – DESIGN CONSIDERATIONS

Design Codes and Guides

• CIRIA 104
• BS 8002
• DD ENV
• BD 42/00
• Piling handbook
• CIRIA R185
• CIRIA C517
• CIRIA SP95
• GEOGUIDE 1 2nd Ed.
• GCO Publication No. 1/90

2
 Conventional Retaining System Design Approach

Figure 29, GCO 1/90 (Navfac, 1982b)

CIRIA Report C580 Design

• Aim - holistic, consistently reliable &

economic ELS design
• Limit states design
• Maintain simplicity over different
uncertainties – adopt partial factor of
safety on soil strength only

3
Use of C580 in Hong Kong

• Review Group set up by Geotechnical

Engineering Office in 2004
• Recommendation concluded in “Notes on
Design of Excavation and Lateral Support
Works Using the Limited State Partial
Factor Method in the CIRIA Report No.
C580”
• Promulgated by BD/GEO in early 2005
• alternative to global factor of safety
approach in GCO Publication No. 1/90
• review in 2 years before long-term
implementation

Simplified Flowchart of C580 Design Approach (SSI)

ULS Analysis SLS Analysis

• Moderately conservative parameters • Most probable parameters

• Apply partial factors • Unfactored
• Unplanned excavation(0.5m or 0.1H) • Unfactored surcharge load
• Unfactored surcharge load • Most unfavorable groundwater
• Worst credible groundwater level level under normal circumstances

• BM, SF & prop loads (ULS) • BM, SF & prop loads (SLS) × LF
• Toe-in requirement • Lateral wall deflection

BM, SF & prop loads envelope (structural design)

4
 Conventional vs C580 Design Approach - Example

-6mPD

FSP-VL steel
sheetpile wall

Existing
-16mPD Diaphragm wall

Sze & Lo (2005)

Sensitivity analyses – Wall Bending Moment

Max. bending moment (kNm/m) Ultimate Bending moment (kNm/m)

-2000 -1000 0 1000 2000 -600 -400 -200 0 200 400 600
5
0
Soil side Excavation side
Soil side Excavation side
1
Mu = 788kNm/m
2 0
Depth of embedment (m)

4 -5 -6.0mPD Dig Level

Level (mPD)

6 C580 toe level

-10
7

8
-15 Conventional toe level
9

10
-20
ULS (+ve) ULS (-ve)
Conventional (Max.) Conventional (Min.)
SLS x 1.4 (+ve) SLS x 1.4 (-ve)
C580 (Max.) C580 (Min.)

5
 Sensitivity analyses – Wall Shear Force
Max. shear force (kN/m) Ultimate Shear force (kN/m)

-600 -400 -200 0 200 400

0 500 1000 1500 2000
5
0 Soil side Excavation
side
1
Vu = 669kN/m
2 0
Depth of embedment (m)

4 -5 -6.0mPD Dig Level

Level (mPD)
5

6 C580 toe level

-10

8
-15 Conventional toe level
9

10
-20

Conventional (Max.) Conventional (Min.)

ULS SLS x 1.4 C580 (Max.) C580 (Min.)

Sensitivity analyses – Strut Force & Wall Deflection

Max. strut force (kN/m) Horizontal Wall deflection (mm)

0 500 1000 1500 2000 0 20 40 60 80 100

0 5
S1

2 0 S2
Depth of embedment (m)

4 -5
Level (mPD)

-6.0mPD Dig Level

5
C580 toe level
6 -10

7
Conventional toe level
8 -15

10
-20
ULS - S1 ULS - S2
SLS - S1 x 1.4 SLS - S2 x 1.4 Conventional C580 - SLS

6
Control of Groundwater at Passive Zone

Fill
Pump well
Alluvium

CDV

Rock

E.g. KCRC Spurline Kwu Tung Station

Buildability

• Difficult in constructing
• Unrealistic movement criteria

7
Buildable Design

• Easy and fast

construction
• Less risks
• Safer working
environment
• Could be more
cost effective

E.g. KCRC Tsuen Wan West Station

Over-excavation

Unworkable sequence vs Opportunistic contractor

8
 Optimization of ELS Design at Kowloon Bay
vert. 2m c/c

Fill

MD

ALL

CDG/MDG

Original Design Optimized

Development at Sheung Shing Street, KLN

9
Bottom-up with Raking Strut Option

Unrealistic Movement Criteria - Diaphragm Wall

Settlement measured at Chater Station in mid 70s

10
 How about Sheetpile Wall Installation ?

5 2005-03-08 2005-03-14

10

15

Settlement (mm)
20

25

30 6mm

35

40

45 40

44

48

52
54

58

62

66

70
72

76

80
42

46

50

56

60

64

68

74

78

82
Chainage (m)

Installation may attribute a substantial part of induced movements

Sheet piling – Giken (Silent) piling

E.g. East Rail TSTE Mody Road Subway, CLP Cable Tunnel TWS access shaft

11
Conclusions of Part I

• C580 design approach would result in consistently

reliable design.
• Shortest Toe-in might not be most economic.
• Buildability - one of key factors for successful
execution of ELS works.
• Get Contractor’s involvement at earlier stage if
possible.

Part II – NUMERICAL MODELLING

12
Numerical Modelling

• Earth pressure Coulomb 1776Sub-title

Numerical Modelling

• Earth pressure Coulomb 1776

• Applied Pressure Diagram Terzaghi & Peck 1948

13
Numerical Modelling

• Earth pressure Coulomb 1776

• Applied Pressure Diagram Terzaghi & Peck 1948
• Beam on Springs 1976

Numerical Modelling

• Earth pressure Coulomb 1776

• Applied Pressure Diagram Terzaghi & Peck 1948
• Beam on Springs 1976
• Finite Elements/Differences 1986

14
Numerical Modelling

• Earth pressure Coulomb 1776

• Applied Pressure Diagram Terzaghi & Peck 1948
• Beam on Springs 1976
• Finite Elements/Differences 1986
• Bricks on Strings Simpson 1992

Numerical Modelling

• Earth pressure Coulomb 1776

• Applied Pressure Diagram Terzaghi & Peck 1948
• Beam on Springs 1976
• Finite Elements/Differences 1986
• Bricks on Strings Simpson 1992
• 3D Applications in practice 2002 - 2005

15
Layout Plan of Extension to the Tsim Sha Tsui
Station Concourse

Typical Cross Section showing Ground

Conditions and the Proposed Works

16
View of the 3D FLAC Model Developed
showing the Lateral Wall Deflections along
the length of the Excavation

Plan view from FLAC 3D Model Showing the

Displacement of the Tunnel Linings along the
length of the Excavation

17
Wall Deflections and Tunnel Deformations
from FLAC 3D at a Section about 40 m from
the southern end

18
 Monitoring of Tunnel Convergence – Downtrack Tunnel
4
3 5
2 6
1 7
Convergence in mm
20
16
12
8
4
0
-4
-8
-12
-16
-20
10 Dec 02 09 J an 03 08 Feb 03 10 Mar 03 09 Apr 03 09 May 03 08 J un 03 08 J ul 03 07 Aug 03 06 Sep 03 06 Oct 03 05 No v 03 05 Dec 03 04 J an 04

1- 4 1- 5 1- 6 1- 7 7 -2 7 -3 7 -4

1- 4 1- 5 1- 6 1- 7 2- 7 3- 7 4- 7

Measured Deformation of Excavation Lateral

Support Wall
Displacement (mm)
-30 -25 -20 -15 -10 -5 0 5 10 15 20 25 30
0.00

5.00

10.00
Depth (m)

15.00

20.00

INC01
INC02
25.00
INC03

30.00

19
• 3-D modelling can be useful for complex geometries

• 3-D modelling can be useful for complex geometries

• Often variations in ground conditions or
construction procedure render sophisticated
constitutive models redundant

20
• 3-D modelling can be useful for complex geometries
• Often variations in ground conditions or
construction procedure render sophisticated
constitutive models redundant
• KIS = Keep It Simple

• 3-D modelling can be useful for complex geometries

• Often variations in ground conditions or
construction procedure render sophisticated
constitutive models redundant
• KIS = Keep It Simple
• Less likely to get a plausible result based on wrong reasons
• Easy to check for approval
• Easy to compare with monitoring results
• Easy to back analyse

21
 Part III - USE OF OBSERVATIONS MADE
DURING EXCAVATION

What Constitutes a Well Performing

Excavation?

• Common Objectives of Client and Contractor

22
What Constitutes a Well Performing
Excavation?

• Common Objectives of Client and Contractor

• totally safe against collapse

What Constitutes a Well Performing

Excavation?

• Common Objectives of Client and Contractor

• totally safe against collapse
• adequate protection to nearby structures and roads etc.

23
What Constitutes a Well Performing
Excavation?

• Common Objectives of Client and Contractor

• totally safe against collapse
• adequate protection to nearby structures and roads etc.
• makes provisions for a safe working environment

What Constitutes a Well Performing

Excavation?

• Common Objectives of Client and Contractor

Protective Factors
• totally safe against collapse
• adequate protection to nearby structures and roads etc.
• makes provisions for a safe working environment

24
What Constitutes a Well Performing
Excavation?

• Common Objectives of Client and Contractor

Protective Factors
• totally safe against collapse
• adequate protection to nearby structures and roads etc.
• makes provisions for a safe working environment

• permits reasonably rapid progress of work

What Constitutes a Well Performing

Excavation?

• Common Objectives of Client and Contractor

Protective Factors
• totally safe against collapse
• adequate protection to nearby structures and roads etc.
• makes provisions for a safe working environment

• permits reasonably rapid progress of work

• cost as low as possible after considering the above

25
What Constitutes a Well Performing
Excavation?

• Common Objectives of Client and Contractor

Protective Factors
• totally safe against collapse
• adequate protection to nearby structures and roads etc.
• makes provisions for a safe working environment
Commercial Factors
• permits reasonably rapid progress of work
• cost as low as possible after considering the above

Dominating Factor...
Ground Conditions Uncertainty

• The state-of-the-art in modern geotechnics gives

us...
• increasingly powerful design tools
• easy access to information on the outcome of many past
projects

26
 • But still views diverge on the correct balance
between protective and commercial aspects.
• variance between analytical tools
• significance of disturbance to field samples
• use of average values or more conservative
• whether soil properties may change during the excavation
process

The Only Sure Way is the Observational Way

The Observational Method in ground engineering is

the continuous review and refinement of a design
based on observations of field behaviour

27
Examples

• Tseung Kwan O Station

• observational method was used to justify omission of a layer of
props and a buried prop at later stages of the project

• otherwise it seems that use of observational design is the

exception in Hong Kong

Example of Observational Method

Typical Section of Excavation

Diaphragm Wall
Ex. Ground Level Ex. Ground Level
Strutting
S1
S2
Sandy Fill
Sandy Fill S3
Stiff Silty Sand
S4
S5
Stiff Silty Sand
Marine Clay
Formation Level

Marine Clay
Jet Grout Raft Clayed Sand

Hard Boundary Clay

MD Silty Sand
Strut S5 deleted by
Observation Approach

Hard Boundary Clay

28
 Why Use Observational Design?
Direct Alteration of Soil Properties

Heating / Dessication due to Jet Grout Hydration

Why isn't OM used for every excavation?

• time lag between measurement and reporting and

interpretation
• reliability of reported measurements
• authority approval for change in sequence of work

29
Web Based Real Time Instruments
(eg.strain
Reporting
gauges on
struts)

dataloggers
Users
with solar panel
connected
over the
Internet

Server

Instrument Reliability

30
Authority Approvals Strategy

• Submit two designs for approval

• Design A – Adopts "moderately conservative" parameters
agreeable to the authority (i.e. normal practice)
• Design B – Targets more favourable parameters
• Proponent (probably the contractor) decides how much
better performance is expected to be... could be "most
probable" or even more optimistic

• Identify a Test Stage in the Excavation

• Up to the test stage, A and B follow the same path, but diverge
thereafter
• Identify test conditions to verify that more favourable parameters
are safe
• At design approval stage, the authority approves only the test
conditions, not the target parameters

S1 a S1 a S1 a

S2 b S2 b S2 b

S3 c S3 c Ø2, C2 S3 c
Ø1, C1
Most
Moderately S4 d d S4 d
Probable
Conservative S5 e
Parameters
Parameters f f

Design A Test Stage Design B

Test Condition –
Strut load, deflection

31
Issues to Address in adopting the
Observational Method

• Most critical is to avoid a delay of say one month

waiting for consent after submission of a report on
reaching the test stage, therefore...
• make consistently up-to-date archive of observations available
to the authority and the supervising engineer
• build a simple control framework into the web based real time
reporting system so that compliance with test stage conditions
can be instantly verified
• adopt a "self-regulation" process whereby the Supervising
Engineer (RSE) certifies the compliance with the test stage
conditions

Summary

• Limit State Design

• for consistency and economy
• Buildability
• most problems are the result of buildability not being properly
considered in the design, the leanest solution may not be the
best
• Sophisticated 3D modeling
• invaluable for unusual, complex analysis, esp. for effects on
existing sensitive structures, but requires specialist expertise
in its application
• Observational Method
• viable tool for reducing cost and improving safety that can and
should be integrated into existing procedures for authority
control

32
Thank you

33