Missouri University of Science and Technology

Scholars' Mine
International Conference on Case Histories in (1993) - Third International Conference on Case
Geotechnical Engineering Histories in Geotechnical Engineering

Jun 1st

Performance of Prefabricated Vertical Drains in

Improvement of Malaysian Soft Marine Clay
F. Haji Ali
Universiti Malaya, Kuala Lumpur, Malaysia

B. Kim Huat
Universiti Pertanian Malaysia, Malaysia

Follow this and additional works at: http://scholarsmine.mst.edu/icchge

Part of the Geotechnical Engineering Commons

Recommended Citation
Ali, F. Haji and Huat, B. Kim, "Performance of Prefabricated Vertical Drains in Improvement of Malaysian Soft Marine Clay" (1993).
International Conference on Case Histories in Geotechnical Engineering. 24.
http://scholarsmine.mst.edu/icchge/3icchge/3icchge-session07/24

This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International
Conference on Case Histories in Geotechnical Engineering by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright
Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact
scholarsmine@mst.edu.
1!1!1!1!!1!1 Proceedings: Third International Conference on Case Histories in Geotechnical Engineering, St. Louis, Missouri,
~ June 1-4, 1993, Paper No. 7.13
'-:..=

Performance of Prefabricated Vertical Drains in Improvement of Malaysian

Soft Marine Clay
F. Haji Ali B. Kim Huat
Department of Civil Engineering, Universiti Malaya, 59100 Kuala Department of Civil Enginering, Universiti Pertanian Malaysia,
Lumpur, Malaysia Serdang, Selangor, Malaysia

SYNOPSIS In Malaysia, an increasing need has developed for various types of construction on sites
underlain by soft cohesive soil. The use of vertical drains in conjunction with preloading is one
of widely used methods to improve the geotechnical properties of the soft soil. This is due to its
relatively cheap cost and availability of drains as well as the practical ease of application. This
paper presents a number of high quality case studies which had been carried out to study the
effectiveness of the vertical drains. The results of the studies are presented and discussed. The
performance as well as the effectiveness of the soil improvement method are evaluated.

embankments were constructed i.e. four 3m high

INTRODUCTION
The objective of vertical drainage is to bring
about more rapid rate of settlement and in-
crease in shear strength in highly compres-
sible subsoil with low permeability. Since the
seventies, prefabricated vertical drains have
largely replaced sand drains. Most of the
prefabricated drains are modelled after the
original Kjellman 'wick '(Kjellman, 1948) ; that
is, they are band-shaped and have a rectangular
cross section of typically about 100 mm x 4 mm.
Majority of them are composite drains consisting
of a plastic core with grooves, studs, or chan-
nels, wrapped with a geotextile filter sleeve.
Until recently, the geotechnical engineer had no
choice but to accept the filter and core as sup-
plied by the manufacturer. Today, however, suffi-
cient research and experience exists to make it
possible to custom design the drain system and
specify the desired material properties for each
component.
Many millions meters of prefabricated verti-
cal drains have been installed in Malaysia for
the treatment of areas such as soft marine clay
and soft silty clay in abandoned tin mining
ponds. Some degree of uncertainty arose in the
effectiven~ss of the method because of unexpected
observations such as little dissipation of pore
pressure and slow rate of settlement in some of
the soil improvement works. Then a number of high
quality case studies had been carried out in
Malaysia to study the applicability of this
method in local soft clay.
CASE HISTORIES
Muar Flats Trial Embankments
In the study on trial embankments on soft marine
clay carried out by the Malaysian Highway
Authority in 1986, various consultants,
specialist contractors and suppliers of
proprietary systems were invited to show the ef-
fectiveness of their methods of design and con-
!3truction of embankments. Altogether thirteen
embankments and nine 6m high embankments
(Malaysian Highway Authority, 1989). Vertical
drains were installed in one of the 3m embank-
ments (known as scheme 3/4) and in four of the 6m
embankments (known as schemes 6/5, 6/7, 6/8,
6/9). For comparison, two untreated control em-
bankments (one for each height) were also con-
structed, known as schemes 3/2 and 6/6.
The soft clay layer varies between 15 to 20m
in thickness. The upper soft clay layer is an
olive green very soft marine clay, approximately
6m thick. Very thin and discontinuous fine sand
partings, decaying leaves, near vertical roots
and bioturbations are found within this layer.
The second soft clay layer consists of an olive
green very soft marine clay layer approximately
11 m thick. Many shell fragments are found within
this layer. No sand lenses are however observed
in the entire lower sequence.
A standard instrumentation scheme was
adopted for all the embankments. The types of in-
strument installed includes surface settlement
markers, settlement gauges, inclinometers and
piezometers.
For the scheme 3/4, a fabricated vertical
drain type A was used (see Table 1). The drains
were installed up to the depth of 20 m. The con-
figuration of the scheme,is illustrated in Fig.
1 For scheme 6/5 the use of 20 m long sand
columns to improve the ground beneath a 6m high
embankment was proposed as shown in Fig. 2. A to-
tal fill thickness of 9. 8 m inclusive of 1. 0 m
sand blanket was placed. An estimated 2. 6 m
settlement was expected leaving 1 2 m to be
removed at the end of 15 months. In scheme 6/7 it
was proposed that a vacuum preload equal to a 3.5
m thick fill be applied over most of the treat-
ment period to induce a settlement of 3.6m to
take place leaving 6m height of fill above the
original ground level. Vacuum preload was to be
carried out with prefabricated vertical drain in-
stalled on a 1.32 m rectangular grid. The scheme
is shown in Fig. 3. Three different makes of
prefabricated drains were used i.e. type B, type

989
C and typeD (see Table 1). The drains were 18.5
m long and spaced at 1.32 m centres. The scheme
6/8 involved the placing of a fill of thickness
equal to the final height of 6m above original
ground level plus the anticipated consolidation - APPROX. Z.Om
THICK BERM
settlement so that at the end of consolidation, a OGL..
height of 6 m would remain above the original +2.4m
ground level. Consolidation settlement was to be ::I I: If I: II: II II: I Ill
accelerated by use of prefabricated vertical
drains spaced at 2m centers on a square grid. ::::::I: l;:: I: ll::: ll
I I I I I I 11 I 1 I I I I I I I I I I I
0.6 m THICK SAND
MAT OVER 0.2m
LEVEI.LtiG FILL
Vertical drain type A was used and the lengths I i:: i:: IIi I i::::: l: i i
I I I 1 I I I I I I I I I I II I I I II
were 20 m throughout the entire width of the em- I: IIIII: I l I: l: ll:: II:
bankment and berms as shown in Fig. 4. The con- ...~ 21 NOS. SCP/SD AT 2.2 m CENTRES ,1-
figuration of scheme 6/9 is shown in Fig. 5 , was NOTE: OGL. o ORIGINAL GIIOUND 0 10m 20m
arrived at after several stages of embankment LEVEL
construction. Embankment construction was ter-
minated after placement of a fill thickness of
about 4. 7 m and the scheme discontinued because Fig. 2. Scheme 6/5
the embankment approached f ai 1 ure. However,
monitoring of movements and pore pressures were
still continued. Vertical drains type A was used
and the lengths were 18m. They were installed on
a triangular grid at the spacing illustrated in
Fig.5.

Table I. Description of Vertical Drain Cores.

I Properties NOTE : OGL o ORIGINAL

GROUND 31 NOS. VERTICAL DRAINS
Drain Description LEVEL
...L AT 1 32 m CINTRES
L
SECTION
0
...........
e tOm

Fig. 7. Volume Definitions

B caaposite Polyester
(core)
100 X 5 11D1 Core is 3-di.mensional,
open mat-structure.
Non-woven filter
8 4m FINAL TOTAL FILL -
THICKNESS
OVE%
J. I
9.7 m DESIGN TOTAL
FILL. THICKNESS 4.0m FINAL TOTAL FILL
-44.4m

1
Polyester jacket. O.Sm SAND BLANKET THICKNESS IN BERM
0.6m CEVELLING FII.L. WITH -- - -
(fil. jacket) HOR.DRAINS AT 2.0m ~ --.. t.Om DEEP ~DE

c Polypropylene 100 x 3 IIDI Core consists of 38

(core)

Polypropylece
parallel rows of
rect:angul.ar castle-
shaped groove. Filter
CENTRES

::P. . .e. . / .
~ilifRI0~~ 1'jW~E~a~~

u~~"".
TRENCH

.-

~
(fil.jacket) jacket is non-woven E
oontinuous filaments 2
D caaposite Polyethylene 100 x 4 11111 Core consists of sharp -..,j<.
(core) stllds.
-.1<-------------- ------,!<
Polyethylece/ NOTE: OGL. .f~t~AL GROUND SECTION oe,"""',/!to&,;....,~om
Polypropylene
(fil.jacket)
1--------- ---------- ----------- ---------- ---------------------- Fig. 4 Scheme 6/8

rO.Sm THICK SAND BLANKET OVR 0.8m

~RL:::E:,gNSAII - l 4..
THICK L.IVELLING FILL
v 7m ., 115m ., ttm ., tOrn f 43m

- I [ 8.0 m TOTAL FILL THICKNESS lEFOR E

111!110\IAL OF SURCHARGE !TOTAL DESIGN I
FILL THICKNESS WAS 11.1 mJ 'I 1 1 1
[FLL10m DESIGN T01ll
----- - --
Z.Sm SANDII.ANKET 4.7m FINAL. TOlliU.o:~. THICKNESS
~3.!1m FINAL TOTAL. FlU.
~4.8m
OVER O.em LEVELLING l'lLL THICICNESS

~
l'lLL WITH HOR. - _ _ _ _ _ _-.. THICKNESS IN BERM
DGL.
+2.4111 .., ''e FINAL. TOTAL
ss DRAINS ICf' Z.Om t..,.. ..- ', 1m OEEP SIDE
\oW'''""'"'' ~;+ t+'.io+l H-illr'!\ f'ILL THICKNE

~;:~ . .. . . ~~.~ .. .. .. .,. . .,. ' , .. ~EN~~

--,-
J
NOTE: OGL. OIIGIHAL GROUND LEVEL
~Om NOTE: OSL ORIGINAL IIIOUND
LEVEL.

Fig. 1. Scheme 3/4 Fig. 5. Scheme 6/9

990
Pore Pressure Time Behaviour
The rates of pore pressure dissipation for
various embankment schemes are plotted against
depth in Fig. 6. As illustrated pore pressure
dissipation occurred only at schemes 6/5 and 6/7.
The excess pore pressure at Schemes 3/4 , 6/8 and
6/9 , all with type A drains, exhibited low rates
of dissipation, only marginally faster than the
schemes without vertical drainage (control
embankments).

Settlement
The measured settlements are due to volume I
changes plus lateral movements. The total
measured settlement (St) may be regarded as the I
~
sum of two components:
-ps
i) Sc, the component of settlement due to
volume change, Vc; 12L---L-~~~--~--~--~~
-40 -!0 -20 -tO 0 10 20 30
ii) Sh, the component of settlement due to DECREASE INCREASE
lateral movement. Rate of Pore Pressure Change (mm/day)
Fig . 6.
For the purpose of separating the measured
settlement to obtain these two components, an as-
sumption has to be made of the shape of the base
of the fill which is due solely to volumetric
changes. A trapezoidal shape shown in Fig. 7 is
assumed.

Vv is the vertical volume obtained from intergra-

tion of the settlement profile of the fill ground
interface ( see Fig. 8 ) and ;

Vh is twice the lateral movement volume obtained Fig. 7. Assumed Settlememnt Profile Due to
from intergration of the inclinometer profile to Consolidation Settlemement
account for movement at both edges (see
Fig. 8 )
Sh is computed from:

sh = st - sc
Figs. 9(a) to (c) show the three different com-
ponents of settlement for various schemes. It can
be seen that the variation of compression settle-
ment with time, due to consolidation process, for
scheme 3/4 (with vertical drains) does not differ
much from that for scheme 3/2 (control
embankment). Similarly, for schemes 6/8 and 6/9
with the same type of prefabricated vertical
drains, the rate of compression settlement is SOFT CLA.Y
more or less the same as for the control embank-
ment (scheme 6/6). Howe\l'er, for the other two
schemes, particularly scheme 6/5, the rate of
setlement is high. Fig. 8. Volume Definitions

Degree of Consolidation.
in Fig. 10. It can be seen that a significant
The degree of consolidation achieved may be com- degree of consolidation had been achieved for
puted from: scheme 6/7. This high degree of consolidation is
consistent with the degree of pore pressure dis-
u = sc I s sipation achieved. Unfortunately the inclinometer
at scheme 6/5 stopped functianong at an early
Where: stage but the early trend and high rate of pore
pressure dissipation indicates that this scheme
s = (Cc/1+e 0 ) h log (1+ 4P/P 0 ') would also have achieved a significant degree of
consolidation. All other schemes, which showed
The values of u for various schemes are plotteo minimal pore pressure dissipation, exhibited a

991
 100~---------------------------------------,
!
~ 90

I
W".t
~
::;, 80

I 100 200 300 400

TIME -DAYS
600 700
0
100 200 300

!~~
TIME DAYS
-1
Fig. 10. Degree of Consolidation for Various
Schemes
-2

-3 0

-3.80 -z
100 200 300 400 !500 600 700

I
TIME- DAYS -4
0

! -
-0.1
~ -
~~
-tO

Iii -12
~-tz
f'! i
~ ...
-t4 a: -t4

-t&

I -t
-t8
Scheme 6/7
-20 l..,_._L_..___.__..___,_--.1
o to 20 .110 40 so ao
100 200 300 400 !500 700 UNDISTURBED VAHE SHEAR STRENGTH (kPal
TIME- DAYS

Fig. 9. Comparisons of Components of Settlement Fig. 11. Changes of Strength

for Various Schemes
low degree of consolidation.
Change In Strength
Fig. 11 ( a) and (b) show the change in strength
for schemes 6/5 and 6/7 respectively, where sig-
nificantly large dissipation of pore pressure
took place. It can be seen that appreciable
strength gains throughout the full depth of the
soft clays were observed at these schemes which
is consistent with the pore pressure dissipa-
tion.
Trial Embankments at Juru
The trial involved the construction of two em-
bankments to an elevation of 3.6 m above sur-
rounding ground, one being the control(untreated)
and the other the "drained" embankment (treated
with the prefabricated vertical drains). Both em-
bankments had been instrumented with identical
layouts and the instruments are basically
required to measure excess pore pressures,
horizontal displacements and settlements under
the embankment loading. The objective was to as-
certain whether vertical drains do accelerate
consolidation of the Juru marine soft
clay, (Mohamad, et al, 1991 ). Based on the in-
terpretation of piezocone soundings and borehole
logs the soft clay layer is about 14 m in
thickness.
Type B drains were used for the ground
treatment. The drains were installed on a 1.2 m
square grid configuration and were installed to
an average depth of 14m (estimated depth of sand
layer).
Figs. 12 and 13 show the variation of
settlement with time for treated and control em-
bankments respectively. Also plotted in the
figures are the variation of fill thickness with
time. The maximum fill thickness was achieved in
about five months. Comparing the settlements of
the original ground surfaces below the centre of
the embankments, it can be seen that the settle-
ment of the treated embankment is almost twice
that of the control embankment.
Figs 14 and 15 show the development and dis-
sipation of pore pressure below the treated and
control embankments respectively. Significant
different in pore pressure dissipation can be
clearly seen between the two embankments. As ex-
pected the pore pressure dissipation below the
treated embankment is faster due to the presence
of the vertical drains.
Both settlement and pore pressure observa-
tions in these trial embankments indicate the ef-
fectiveness of the soil improvement method.
DISCUSSION
From the above observations it has been found

992
 2
Fig. 12. Settlement and Displacement Vs. Time
(Treated Embankment)
4
2
Fig. 13. Settlement and Displacement Vs. Time
(Control Embankment)
that there are cases where the soil improvement
using vertical drains is not effective even
though the types of soft clay are more or less
the same. Possible reasons for the ineffective-
ness are 1) soil disturbance during drain place-
ment, and 2) poor performance of drains.
The view has been expressed that soil dis-
turbance during drain placement may cause a dras-
tic decrease in permeability and strength and in-
crease in pore pressure resulting in a prolonged
period of stabilization ( Casagrande and Poulos,
1969). The soil is disturbed during drain place-
ment mainly in two ways: 1) by being compressed
and sheared especially when a mandrel is pushed
into the ground and 2) by being smeared, which
causes the sand-silt veins to be covered with
clay. Both actions reduce the soil permeability
around the drains. The disturbed and smeared zone
slows down water flow and consequently consolida-
tion as well
993
Fig. 14. Excess Pore Pressure Vs. Time
(Treated Embankment)
Fig. 15. Excess Pore Pressure Vs. Time
(Control Embankment)
The performance of a band-shaped vertical
drains system is affected by the quality of the
drains. One of the most important hydraulic
properties of the band-shaped drains is the
longitudinal permeability or the discharge
capacity.The discharge capacity is a measure of
the vertical drain's well resistance. The
vertical drains must provide low entry resis-
tance to the flow of water and to discharge it
with little well resistance.
In order to ensure that the vertical drains
perform satisfactorily, the flow capacity should
satisfy certain minimum requirement which can be
determined from the results of discharge capacity
tests in which the effects of lateral pressure,
creep and drain deformation are incorporated.
Drain type A which has been used in schemes
3/4, 6/8 and 6/9 for the Muar Flats trial embank-
ment does not have the filter jacket. Discharge
capacity tests had been carried out on various
types of drain including drain type A by the main
author ( Faisal, 1991 ). Significantly large drop
in discharge capacity was observed after subject-
ing drain type A to lateral pressure and con-
solidation settlement. The poor performance was
caused by clogging of the drain due to soil par-
ticles entering the cores through small holes
around the drain. The drain deformed by sharp
kinks as the surrounding soil settled a.nd this
resulted in large reduction of discharge
capacity.

CONCLUSIONS
There is little doubt that vertical drains have
sometimes been used in vain due to reasons such
as disturbance effect and poor drain performance.
Still, it is certain that in many instances they
have shortened the time required for soil stabi-
lization. But where they have been used and con-
solidation proceeded as planned, it is difficult
to assert that the timely completion of stabi-
lization was due to the drains and would not have
happened anyway. For this reason the best way to
assess the effectiveness of vertical drains is to
compare the time rate settlement of the same or
nearby embankments built on ground with and
without drains. Also the best way of comparing
the effectiveness of various types of drains is
to compare the time rate settlement of nearby 'em-
bankment built over ground with the different
drain types.
REFERENCES
Casagrande, L. and Poulos, S., ( 1 969), " On the
Effectiveness of Sand Drains", Canadian
Geot. Journal, pp. 287.
Faisal Haj i Ali, ( 1 991 ) , " The Flow Behavior of
Defromed Prefabricated Drain", Geotextiles
and Geomembranes, Vol. 10, No. 3, pp. 235-
248.
Kjellman, W.,(1948), Discussion to "Consolidation
of Fine-Grained Soils by Drain Wells ",
Transactions, ASCE, 113, pp.748-51.
Mohamad, R., Nicholls, R. A., Albakri, W. H.,
Dobie, M. J. and Othman, K. ( 1991)," Verti-
cal Drains Trial at Juru, Malaysia:
Perfromance During Construction", 9th Asian
Regional Conference On Soil Mechanics and
Found. Eng.,Bangkok, Thailand, pp.S09-514.
The Malaysian Highway Authority,(1989), Proc. In-
ternational Symposium On Trial Embankments
On Malaysian Marine Clays,Kuala Lumpur.

994