You are on page 1of 7


discussions, stats, and author profiles for this publication at:

A design method for soft subsoil improvement

with prefabricated vertical drain

Article · January 2000


6 96

2 authors, including:

Jinchun Chai
Saga University


Some of the authors of this publication are also working on these related projects:

Ariake Sea Research Project View project

All content following this page was uploaded by Jinchun Chai on 01 April 2015.

The user has requested enhancement of the downloaded file.

A design method for soft subsoil improvement with prefabricated vertical drain
J. C. Chai & N. Miura
Saga University, 1 Honjo, Saga, Japan

ABSTRACT: A method for designing the soft subsoil improvement using prefabricated vertical drain (PVD) is presented.
It mainly consists of two parts: (1) the way of determining the design parameters and (2) the method of determining the
spacing and suitable improvement depth. Methods proposed by Chai and Miura (1999) for determining the design
parameters are recommended and discussed. Normally, the natural soft deposit is layered and there is no hand-calculable
close-form solution for determining the spacing and suitable installation depth of PVD improvement. A multi-layer
one-dimensional (1D) finite element program has been developed to aid the design. The application of the proposed design
method to two case histories is presented.

1 INTRODUCTION Chai and Miura (1999). Followings are the brief summary
of the methods.
Installing the prefabricated vertical drain (PVD) into soft
subsoil combined with preloading provides an efficient
2 .1 Horizontal coefficient of consolidation of subsoil
and economic way of improving the soft subsoil. It has
been used for highway, airport and port construction on At present, there is no satisfactory laboratory test method
soft clay deposit. For designing a PVD improvement, the to determine the field coefficient of consolidation in
first thing is to determine the values of parameters horizontal direction of subsoil (C h ). Normally, laboratory
affecting the behavior of PVD. The next thing is to decide test under-estimates the field C h value. Back-analysis from
the spacing between PVDs and improvement depth. The field measurements or field test has been recommended for
parameters affecting the behavior of PVD include (1) evaluating the design C h value. Some reported ratios of
coefficient of consolidation in horizontal direction of field coefficient of consolidation in vertical direction (C v ) f
subsoil, (2) discharge capacity of PVD and (3) smear zone over the laboratory values (C v ) l are listed in Table 1.
parameters (diameter, d s , and horizontal hydraulic
conductivity ratio, k h /k s ). There are still uncertainties on
determining these parameters, which make a precise Table 1. Ratio of field over laboratory coefficient of
design difficult. Chai and Miura (1999) discussed the consolidation
effect of these parameters on the behavior of PVD Site (C v ) f /(C v )l Reference
improved subsoil and the methods for determining the Oxford (1) 4-57 Lewis et al. (1975)
parameters have been proposed. The methods are briefly Oxford (2) 3-36 Lewis et al. (1975)
described and discussed. Donnington 4-7 Lewis et al. (1975)
Determining the spacing and suitable installation Avonmouth 6-47 Lewis et al. (1975)
depth of PVD improvement, unit cell (a single PVD Tickton 7-47 Lewis et al. (1975)
surrounded by a soil cylinder) solutions by Barron (1948) Over Causeway 3-12 Lewis et al. (1975)
or Hansbo (1981) are widely used. The solutions assumed Melbourne 200 Walker and Morgan
a uniform subsoil condition. However, most natural clay (1977)
deposit is not uniform, and normally has a crust at surface Penang 70 Adachi and Todo (1979)
and sometimes sandwiches thin sand layers. For
multi-layer subsoil condition, there is no hand-calculable
solution available. Using the uniform subsoil assumption 2 .2 Discharge Capacity of PVD
does not represent the actual case. An easy-to-use design The most reported data on discharge capacity of PVD were
tool considering multi-layer subsoil condition design tool from the test of confining PVD by rubber membrane.
is desirable. A multi-layer one-dimensional (1D) finite However, in field, the PVD is confined by clay. Based on
element program for PVD improved subsoil (PVD-CON) the laboratory test results of 4 types commercially used
is developed. The features of the program as well as the PVDs, Chai and Miura (1999) and Miura and Chai (2000)
theoretical background are described. Application of the found that confined in-clay discharge capacity (Q C ) was
proposed design method is demonstrated by analyzing two lower than corresponding confined in-rubber membrane
case histories of embankment on PVD improved subsoil, value (Q R ), and Q C reduced significantly with elapsed time.
one in Saga, Japan and one in Zhejiang, China. The main reasons for the reduction are (a) clogging caused
by soil particles entered the drainage channel and some
bio-films formed during the test, and (b) the creep
2 DETERMINING THE DESIGN PARAMETERS deformation of the filter of PVD. It is recommended to use
the long-term confined in-clay value (Q C ) for design.
Based on laboratory test data and back-calculated results Considering the long-term confined in-clay test is not a
from field measurements, the methods for determining the routine test, an empirical equation has been proposed to
design parameters of PVD improvement are suggested by estimate the long-term Q C value from Q R value (Miura and
Chai, 2000). deposit improved by PVDs. The program is developed
under Windows environment and has well-designed pre-
and post- analysis processes. The program can consider
QC = QR (1) followings.
0.01t / t o + i (1) Multi-layer improved subsoil.
(2) The effects of both vertical hydraulic
conductivity of natural subsoil and horizontal drainage of
where i is hydraulic gradient within PVD, t is elapsed time, PVDs.
and t o is a time constant of 1 day to make the unit balance. (3) Both full and partial penetration condition of
It is suggested to use i=0.1 in Eq. (1) for determining the PVDs.
design Q C value. Also, in the case that there are no test (4) Effect of surface vacuum pressure.
data (Q R ) available, as a rough estimation, a design (5) Either one-way or two-way drainage.
discharge capacity of 100 m3 /year is recommended. The program has following main options.
(1) Find spacing (S) for a given time and required
2.3 Smear Effect average degree of consolidation.
(2) Find time (t) for reaching a required average
Installing PVDs into the ground creates a completely degree of consolidation for fixed spacing.
disturbed zone around PVDs, which is called smear zone. (3) Check the effect of penetration depth by repeated
The hydraulic conductivity in the smear zone will be analyses.
reduced significantly. Two parameters are needed to (4) For given (or designed) condition calculate: (a)
characterize the smear effect, namely, the diameter of settlement variation with time at different depth, and (b)
smear zone (d s ) and horizontal hydraulic conductivity ratio undrained shear strength profile and its variation.
(k h /k s ), i.e. the value in undisturbed zone (k h ) over that in
smear zone (k s ). It has been proposed that, d s can be
estimated as follows (Chai and Miura, 1999). 3 .2 Theoretical background
Hansbo’s unit cell solution (Hansbo, 1981) is used to
ds = 3⋅ dm (2) model the consolidation due to PVD. In finite element
formulation, the effect of PVDs is considered by
modifying 1D continuity equation of consolidation as
where d m is the equivalent diameter of the cross-sectional follows.
area of mandrel for installing PVD. Considering the factor
that the laboratory test may properly estimate the k s value kv ∂ 2u 8k h u ∂ε
but under-estimate field k h value, field (k h /k s ) f value can be − + v =0 (4)
calculated as (Chai and Miura, 1999): γ w ∂z 2
γ w D μ ∂t

kh k kh 2l 2 k h
( ) f = C f ⋅ ( h )l (3) μ = ln +
n 3
ln s − + π (5)
ks ks s ks 4 3q w
where (k h /k s ) l is the ratio determined by laboratory test, where: γ w is the unit weight of water, z is depth, t is time,
and C f is the ratio of field horizontal hydraulic ε v is volumetric strain, u is excess pore pressure, k v is
conductivity over corresponding laboratory value. The C f hydraulic conductivities in vertical directions, l is drainage
values for few natural clay deposits are given in Table 2. length, D is the diameter of unit cell, q w is discharge
capacity of PVD, n=D/d w (d w is the equivalent diameter of
Table 2. C f values for few clay deposits PVD), and s=d s /d w . k h , k s and d s are defined previously.
Site C f Method for Reference Two ways have been incorporated in the program for
evaluating (k)f calculating the undrained shear strength (S u ) of subsoil.
Bangkok clay at 25 Back-analysis Chai et al. One is using Modified Cam clay theory (Roscoe and
AIT campus (1995) Burland, 1968), in which S u can be expressed as:
Bangkok clay at 4 Back-analysis Chai et al.
Nong Ngu Hao (1996)
p' M 2 +η2 Λ
Malaysia Muar 2 Back-analysis Chai and Su = M( ) (6)
clay deposit Bergado (1993) 21+ Λ M2
Ariake clay 4 Back-analysis Chai and Miura
(Japan) (1999) where p’ is mean effective stress, M is the slope of critical
Louiseville 1 Self-boring Tavenas et al. state line in p’-q plot (q is deviator stress), η =q/p’, and
(Canada) permeameter (1986) Λ =1- κ/ λ ( κ and λ are slopes of unloading-reloading and
St-Alban 3 Self-boring Tavenas et al. virgin loading curves in void ratio, e, versus ln(p’) plot).
(Canada) permeameter (1986) In the program, κ is fixed as λ/10.
Soft mucky clay 6 Back-analysis Shen et al. Another way is to use an empirical equation (Ladd,
(eastern China) (2000) 1991), which relates the S u with effective vertical stress
σ’ v and overconsolidation ratio (OCR).
S u = S ⋅ σ ' v ⋅(OCR ) m (7)
3.1 Features of the program

A 1D finite element program named PVD-CON has been where S and m are constants. In the program, m is fixed as
developed for calculating the consolidation of soft clay 1.0.
T.P 71.0 (U nit:m)
2.0 2.0
5.0 21.0 25.0 21.0
Sand M at
1 :7 .0

Em bankm ent
± 0.0 Top crust
A riake clay (A c1)
-5.0 A s1


A riake clay (A c2)
Piezom eter
A s2 Inclinom eter casing
A riake clay (A c3)
Length of PV D :L=25.5m
D ense sand £ Sub-surface settlem ent system

Fig. 1 Cross section of test embankment in Saga airport

Table 3 Soil parameters for Saga Airport test embankment

Soil Layer Thickness Ch Cv γt Cc eo M OCR
(m) (m 2 /day) (m 2 /day) (kN/m 3 )
1(clay) 1.0 0.1 0.067 15.0 0.58 2.0 1.3 5
2(clay) 3.0 0.08 0.053 14.5 1.0 2.0 1.2 2
3(sand) 1.5 54 54.0 15.5 0.1 1.8 1.5 1.2
4(clay) 5.0 0.045 0.03 14.5 2.0 2.5 1.2 1.2
5(clay) 5.0 0.065 0.043 14.5 2.0 2.5 1.2 1.2
6(clay) 5.5 0.087 0.058 14.5 2.0 2.5 1.2 1.2
7(sand) 2.5 178 178.0 16.0 0.1 1.7 1.5 1.2
8(clay) 1.5 0.26 0.173 16.0 0.7 1.75 1.3 1.2

4 ANALYZING TWO CASE HISTORIES Table 4 Parameters related to PVD improvement

Item Symbol Unit Values
4 .1 Test Embankment in Saga Airport (Case 1) Case 1 Case 2
Saga Airport is located 13 km south of Saga city on a Drain diameter dw mm 48.3 53.0
reclamation land close to Ariake Sea. The deposit mainly Unit cell diameter De m 1.7 1.58
D e /d w n - 35.2 29.7
consists of soft and highly compressible Ariake clay. For
Smear zone ds mm 300 355
verifying the effect of vertical drain improvement, 3 test
embankments were constructed on natural, PVD improved d s /d w S - 6.2 6.7
and sand drain (SD) improved subsoil (Bergado et al., Hydraulic k h /k s - 10 15
1996). At the test site, the soft layer is about 25 m deep conductivity ratio
consisting of 3 clay layers (Ac1, Ac2, and Ac3) and 2 thin Discharge capacity qw m3 /yr 85 100
sand layers (As1 and As2) underlying a thick dense sand
layer. The embankments had the same geometry with a fill
thickness of 3.5 m, base width of 71 m by 71 m, and top The calculated surface settlements on embankment
width of 25 m by 25 m. The PVDs were installed to centerline with different PVD penetration depth are
around 25 m deep with an improved area of 45 m by 45 m. compared in Fig. 2. The measured data are included in the
figure also. It can be seen that reducing the PVD length
PVDs were installed in a square pattern with a spacing
from 25 m to 21 m (just penetrate through the second clay
S=1.5 m. Fig. 1 shows the geometry of the embankments
layer, Ac2) not influences the consolidation rate much.
and main instrumentation points. However, if reducing the PVD length to 15.5 m, it reduces
The soil parameters needed for performing the consolidation rate of the deposit considerably.
PVD-CON analysis are listed in Table 3 and the drain The further calculations were conducted on finding
properties are given in Table 4. These parameters are the spacing and time for given condition. At 1 year after
same as those reported by Chai and Miura (1999) for plane starting construction, the corresponding spacings for
strain finite element analysis. The coefficient of average degree of consolidation of 90%, 95% and 99% are
consolidation was calculated from the values of hydraulic 2.0 m, 1.6 m, and 1.22 m, respectively. With actual
conductivity, stress state and corresponding deformation designed condition (S=1.5 m), the calculation showed that
parameters. Other necessary data are the groundwater the average degree of consolidation at the end of
level of 1.0 m below ground surface and the embankment construction was about 80%. The time for reaching 90%
construction time of about 200 days. Surcharge load was degree of consolidation was about 265 days and 340 days
70 kPa (3.5 m fill thickness). for 95% degree of consolidation.
Fill height (m) 0

Depth (m)
End of construction
1 Initial
PVD length 25 m (actual) 20
PVD length 21 m
Settlement (m)

PVD length 15.5 m

1 Measured
0 20 40 60 80
Undrained shear strength (kPa)
Fig. 3 Undrained shear strength variation (case 1)

3 4.2 Test embankment at Hongzhou-Ningbo expressway

0 200 400 600 (case 2)
Elapsed time (days) Hongzhou-Ningbo expressway passed the eastern coastal
region of China, where soft clay (mucky clay in China) is
widely deposited. To find an efficient subsoil improvement
Fig. 2 Comparing the surface settlement (case 1)
method, serious test embankments were constructed with a
total length of 3.15 km. One of the test embankments was
For actual designed case, the predicted undrained constructed on PVD improved subsoil. At the test site, the
shear strength variation of subsoil is given in Fig. 3, which thickness of soft layer is about 23 m. The soil profile
can be used for checking the stability of the embankment consists a thin top crust underlying soft to very soft clay
and in return to adjust the construction speed of the layers. Below the soft layers, there is a thin clay sand
embankment. layer followed by bedrock. The embankment cross-section
These calculations demonstrated that with the aid of together with subsoil profile is shown in Fig. 4. The soil
PVD-CON program, an efficient and economical design of parameters for PVD-CON analysis are listed in Table 5
PVD improvement could be easily achieved. (Shen et al., 2000).

26m Surface settlement
.5 gauges
1:1 Embankment

TC Sand mat t=0.5m
SC1 PVD Drain Inclinometers
-10 MC Piezometers
MSC settlement gauges
-20 SC2
Cross section
PVD spacing
TC:top crust,SC:silty clay

PVD MSC:mucky silty clay

MC:mucky clay

Fig. 4 Embankment cross-section of Case 2

Table 5 Soil parameters for test embankment at Hongzhou-Ningbo expressway, China

Soil Layer Thickness C h Cv γt Cc eo M OCR
(m) (m 2 /day) (m 2 /day) (kN/m 3 )
1(clay) 1.2 0.079 0.079 19.3 0.19 0.81 1.0 5
2(silty clay) 3.6 0.028 0.011 18.5 0.37 1.07 1.0 1.5
3(mucky clay) 4.5 0.14 0.093 14.3 0.69 1.36 0.8 1.0
4(mucky clay) 5.0 0.227 0.152 14.3 0.64 1.36 0.8 1.0
5(soft silty clay) 4.0 0.324 0.169 17.9 0.42 1.1 0.8 1.0
6(silty clay) 4.7 0.132 0.061 19.3 0.23 0.81 1.0 1.0
Parameters related to PVD behavior are given in Above two cases demonstrated that PVD-CON is a
Table 4 also. The groundwater level is about 1.5 m below useful and powerful design tool for soft subsoil
ground surface. The total construction time was about 260 improvement with PVDs. The figures of settlement curves
days. The surcharge load was 117 kPa (5.88 m fill and undrained shear strength distributions can be
thickness). automatically plotted by PVD-CON. Good agreement
The calculated surface settlements on embankment between calculated and measured data shows that the
centerline are compared with measured data in Fig. 5. The recommended methods for determining the design
1D analysis predicted the surface settlement reasonably parameters are adequate.
well. The analysis results by assuming PVD length of 10
m and 14.3 m are also indicated in Fig. 5. For the subsoil
condition considered, it seems that using an improved 5 CONCLUSIONS
depth of 14.3 m (penetrate the soft mucky clay layer only)
is an economic option. The analyses were also conducted (1) Based on the laboratory and field evidences, the
to find spacing and time for given condition. At 1 year methods for determining the design parameters of soft
after starting the construction, the spacings for resulting in subsoil improvement with prefabricated vertical drain
average degree of consolidations of 90%, 95%, and 99% (PVD) are recommended and discussed.
are 2.0 m, 1.58 m, and 1.29 m respectively. Under actual (2) The theoretical background and the features of a 1D
designed condition (S=1.5 m), the average degree of finite element program (PVD-CON) for consolidation
consolidation at the end of construction was about 83%. analysis of PVD improved subsoil are described. The
The time for 90% average degree of consolidation was 290 program can consider the multi-layer subsoil
days and 338 days for 95% degree of consolidation. condition and it has well defined pre- and post-
The calculated undrained shear strength distributions analysis processes.
corresponding to initial and at the end of construction (3) Two case histories of embankment on PVD improved
conditions, respectively, are plotted in Fig. 6. Within the subsoil were analyzed by the procedures described in
very soft mucky clay layer (around 5 to 15 m depth), the this paper. It has been demonstrated that the proposed
undrained shear strength increase is more than 50% during method is a powerful tool for designing soft subsoil
the construction period. improvement with PVD.

Height (m)

Adachi, K. and Todo, H. 1979. A case study on settlement

4 of soft clay in Penang. Proc. 6 th Asian Regional Conf.
Analyzed On Soil Mech. and Found. Engrg, 1: 117-120.
Measured Barron, R. A. 1948. Consolidation of fine-grained soils by
drain wells. Trans. ASCE, No. 113:. 718- 742.
0 Bergado, D.T., Anderson, L.R., Miura, N., and
Balasubramaniam, A.S. 1996. Soft ground improvement,
Measured in lowland and other environments. ASCE Press, New
0.5 PVD length 18.3 m (actual)
Settlement (cm)

PVD length 14.3 m York, p. 427.

1 PVD length 10.0 m Chai, J. C. and Bergado, D. T. 1993. Performance of
reinforced embankment on Muar clay deposit. Soils and
1.5 Foundations, 33(4): 1-17.
Chai, J.C., Miura, N., Sakajo, S., and Bergado, D. T. 1995.
2 Behavior of vertical drain improved subsoil under
2.5 embankment loading. Soils and Foundations, 35(4):
0 200 400 600 800 1000 1200 49-61.
Chai, J.C., Bergado, D.T., Miura, N., and Sakajo, S. 1996.
Elapsed time (day) Back calculated field effect of vertical drain. Proc.
Second Int. Conf. Soft Soil Engrg., Nanjing, China, 1:
Fig. 5 Comparing the surface settlement (Case 2) 270-275.
Chai, J. C. and Miura, N. 1999. Investigation of factors
affecting vertical drain behavior. J. of Geotechnical and
Geoenvironmental Engrg., ASCE, 125(3): 216-226.
0 Hansbo, S. 1981. Consolidation of fine-grained soils by
prefabricated drains. Proc. 10th Int. Conf. Soil Mech.
End of construction and Found. Engrg., Stockholm, 3: 677-682.
Ladd, C. C. 1991. Stability evaluation during staged
Depth (m)

10 construction. J. of Geotech. Engrg., ASCE, 117(4):

Initial 541-615.
Lewis, W. A., Murray, R. T. and Symons, I. F. 1975.
Settlement and stability of embankments constructed on
soft alluvial soils. Proc. the Institute of Civil Engineers,
20 59: 571-593.
Miura, N. and Chai, J. C. 2000. Discharge capacity of
prefabricated vertical drain confined in-clay.
0 20 40 60 80 Geosynthetics International, International,
Undrained shear strength (kPa) Geosynthetics Society, 7(3).
Roscoe, K. H. and Burland, J. B. 1968. On the generalized
stress-strain behavior of wet clays. Proc. of Engrg.
Fig. 6 Predicted undrained shear strength (Case 2) Plasticity, Cambridge Univ. Press, 535-609.
Shen, S. L. Yang, C. W., Miura, N. and Chai, J. C. 2000. ASCE Special Conf. on Use of in Situ Tests in Geotech.
Field performance of PVD improved soft clay under Engrg., Blacksburg, 1034-1048.
embankment. Proc. of Inter. Symp. on Coastal Geotech. Walker, L. K. and Morgan, J. R. 1977. Field performance
Engrg. in Practice, IS-Yokohama 2000. of a firm silty clay. 9 th Inter. Conf. on Soil Meh. and
Tavenas, F., Tremblay, M., Larouche, G., and Leroueil, S. Found. Engrg, Tokyo, 1: 341-346.
1986. In situ measurement of permeability in soft clays.

View publication stats