Geotextiles and Geomembranes 22 (2004) 17–35

A field trial for soft clay consolidation usingelectric vertical drains

S.H. Chewa,*, G.P. Karunaratneb, V.M. Kumaa, L.H. Limc,M.L. Tohd, A.M. Heee

aCivil Engineering Department, National University of Singapore (NUS), Engineering Drive 2,

Singapore 117576, SingaporebGeo5 R&D and Services Pte Ltd, (NUS Spin-off Company), P.O. Box 1048, Singapore 911102, Singapore

cKTP Consultants Pte Ltd, Singapore 159471, Singapored IMRE, National University of Singapore (NUS), Singapore 117576, Singapore

eJTC Corporation, Singapore 609434, Singapore

Received 7 February 2003; received in revised form 1 April 2003

Abstract

Electro-osmosis (EO) has been employed successfully in a number of field cases in the past

using metal electrodes. For large-scale field applications such as soft clay stabilization under

reclaimed land from the sea, metal electrodes are unwieldy and expensive. With the advent of

conducting plastics, it has become possible to use electrically conductive vertical drains in the

same manner as prefabricated vertical drains for accelerated consolidation of soft clay. This

paper presents some laboratory studies and a large field trail conducted with an electric

vertical drain in a reclaimed land in Singapore.

r 2003 Elsevier Ltd. All rights reserved.

Keywords: Electro-osmosis; Soft clay; Conductive plastic; Shear strength; Consolidation; DC voltage

1. Introduction

Deposits of soft clay soil pose construction problems related to settlement andstability. A common soil improvement technique in such deposits is the applicationof prefabricated vertical drains (PVD) with or without surcharge to eliminate a largeproportion of primary consolidation settlement and secondary settlement. To reduce

ARTICLE IN PRESS

*Corresponding author. Tel.: +65-6867-6472; fax: +65-6779-1635.

E-mail address: cvecsh@nus.edu.sg (S.H. Chew).

0266-1144/$ - see front matter r 2003 Elsevier Ltd. All rights reserved.

doi:10.1016/S0266-1144(03)00049-9

this waiting time further and to be less dependent on the need to place and removesurcharge, National University of Singapore initiated work in an old science ofelectro-osmosis (EO) but associated with a relatively new area of technology. Here,the electricity is passed through conducting polymer vertical drains (EVDs) toaccelerate consolidation of clay.

Clay particles carry a net negative charge balanced by net positive charges in thepore water. In EO, when electrodes are placed in soft soils for passing direct currentelectricity the ions in the pore water are attracted to the oppositely chargedelectrodes and dragging with them the surrounding free water molecules. The netwater movement is towards the cathode and if drainage is provided, consolidationoccurs beginning at the anode. The usefulness of EO lies in its suitability for rapidimprovement in clay rich soils. For this type of soils, EO is potentially more effectivein reducing the water content than conventional PVD that uses only hydraulicgradient created by external loading. Extensive research with EO has been carriedout on clay rich soils in Canada, Germany, Norway and the UK (Casagrande, 1952;Casagrande et al., 1981; Chappel et al., 1975; Eggestad and Foyn, 1983; Fetzer, 1967;Gray and Mitchell, 1967; Lo et al., 2000; Shang, 1998; Wade, 1976) and recently inSoutheast Asia countries such as Thailand, Malaysia and Singapore (Bergado et al.,2000; Bo et al., 2000; Toh et al., 2001 and Karunaratne et al., 2002).

Improvements in shear strength followed EO treatment in clay soils.Recently, a proposal to consolidate soft marine clay utilizing the EO principleby means of electric vertical drains (EVDs) was accepted for trial fundingby The Enterprise Challenge Unit in Singapore. The location of the trial site wasat Tuas View Reclamation Project with JTC Corp acting as the Piloting Agency forthe trial.

This paper summarizes the laboratory investigations of EO in Singapore marineclay and presents details of a field trial and some observations. The trial consisted ofinstallation of a patented electrically conducting polymer drain for consolidating a8m thick soft clay layer underlain by a 18m thick recently placed sand deposit.Despite the initial high excess pore pressure in the clay EO process with reversedpolarity did make measurable change in pore pressure and record substantialincrease in vane shear strength.

2. Laboratory investigations

2.1. Small specimens

In laboratory studies, one-dimensional consolidation tests were carried out withand without EO. A modified oedometer built with non-conductive acrylic materialwas used as a cell, and the consolidation stress was applied mechanically. Twoinsulated wires, soldered to horizontally placed stainless-steel anode at the bottomand cathode at the top, were then connected to the DC power source.

Remolded Singapore marine clay (WP ¼ 35; WL ¼ 80) mixed to initial watercontent equal to its liquid limit was used to ensure uniform soil properties

ARTICLE IN PRESSS.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3518

throughout the test series. All specimens, 20mm thick initially, were consolidated to100 kPa in stages so as to enable comparison. Two series of specimens, designated as1A and 1B, were then subjected to EO under a voltage difference of 2 and 3V,respectively, while specimen 1C was held as a control specimen without EOapplication. At the end of EO, the treated soil was subjected to further investigation.The loading was extended to further two days to examine the changes in thecoefficient of secondary compression Ca: EO treated soil was tested for water contentvariation with depth and pH.

Fig. 1 shows the void ratio—pressure (e vs. log p0) relationship of samples. EOtreated soil lacked a well-defined sharp kink at the pre-consolidation pressure asobserved by Bo et al. (2000). EO treatment shows an apparent increase in thecoefficient of consolidation Cv; and apparent decrease in compression index Cc andcoefficient of secondary compression Ca determined in a two-day test period. Forloading above 300 kPa, Cv became stabilized, with Test 1A having an averageincrease of about 40% and Test 1B about 200%. A decrease in Ca of EO treated clayamounted to 10–30% in the stress range up to 800 kPa (Toh et al., 2001).

2.2. Large specimens

The feasibility of applying EO to Singapore marine clay was also investigated inmoderately large diameter cylindrical samples, 500mm diameter and 150mm thick.A particular aspect of this study centered on the type and spacing of vertical drainsand other field implications (Lim, 2002). Fig. 2 shows the plan and cross sectionalelevation of the apparatus, where two electrodes were introduced in the clay at liquidlimit. The position of electrode probes for measurement of voltage and current in theclay and the pore pressure transducers (PPT) is also shown. The sample after

ARTICLE IN PRESS

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

1.10

1.20

10 100 1000Load (kPa)

Vo

id R

ati

o

Test 1A

Test 1B

Test 1C

1C (controlled specimen)

1A (2V)

1B (3V)

Fig. 1. Void ratio–pressure relation of clay after EO (Tests 1A and 1B) compared with untreated clay

(Test 1C).

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 19

treatment was subject to geotechnical property investigations using oedometerand laboratory vane shear tests. The properties of Singapore marine clay used areWP ¼ 35; WL ¼ 80; IP ¼ 45 and sand, silt and clay proportions=2%, 46% and44% respectively.

2.2.1. Test scheme

Three EO consolidation tests, designated as Tests 1, 2 and 3, were conducted in thelarge tank. The clay for the three tests was consolidated to 31, 52 and 30 kParespectively, simulating the range of conditions of in situ soft clay normallyencountered in Singapore. Upon completion of 90% average degree of consolida-tion, as observed by settlement and pore pressure readings, the vertical consolidationpressure was removed. Undrained shear strength profiles via laboratory vane andwater content profile were determined across the electrode positions. Afterestablishing the initial parameters, the clay was reconsolidated to the originalvertical pressure.

ARTICLE IN PRESS

Fig. 2. Large diameter apparatus for EO application in soft clay: (a) plan view and (b) elevation.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3520

During the subsequent EO consolidation, a constant DC voltage of 20V wasapplied between the electrodes made up of perforated metal plates. Tests werestopped when settlement was observed to have petered out and pore pressurereadings have stabilized.

2.2.2. Results

The profiles of shear strength after electro-osmotic consolidation, shown in Fig. 3for Test 1 at four different decreasing depths, A; B; C and D; across the electrodes,are similar to those reported by Abiera et al. (1999) and Bergado et al. (2000). Thesoil near the anode had the largest increase in undrained shear strength followed bythe soil at the cathode. The undrained shear strength in the mid-section had minimalstrength increase. Fig. 3 also shows the initial shear strength of the remoulded clay.

Fig. 4 shows the percentage increase in undrained strength between the electrodesin Tests 1, 2 and 3. Fig. 5 shows the profiles of percentage decrease in water content.The decrease was lowest in the mid-section and highest near the electrodes, similar toreported laboratory tests by Abiera et al. (1999) and Bergado et al. (2000). Thesewater content profiles correlated well with the undrained strength profiles.

There was a large undrained shear strength increase in Test 1, which had higherinitial water content, than in Test 2, especially near the cathode. It is observed thatthere is a greater percentage improvement in softer clay (Test 1) by EO than in stifferclay (Test 2) as also observed in Lo et al. (1991). This clearly showed that EO is moreefficient in softer clay than stiffer clay.

Fig. 6 shows the variation in the pore pressure within the clay between theelectrodes, the larger variation being associated with the proximity to the anode. Alarge positive pore pressure in the electronic transducer P5 in Fig. 6 could be due to

ARTICLE IN PRESS

0

5

10

15

20

25

30

35

40

45

50

0 50 100 150 200 250 300

Location from Cathode (mm)

Un

dra

ined

Sh

ear

stre

ng

th (

kPa)

Row A

Row B

Row C

Row D

Before EO

-v e +ve

Fig. 3. Undrained shear strength of clay after EO in Test 1.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 21

an experimental error in the electric field. Shang (1998) also observed reduction ofpore pressure at the anodes. Other phenomena observed are (1) the current throughelectrodes initially rose, became stable and then dropped, (2) gases evolved atelectrodes, (3) settlement continued to occur following nearly the trend shown by thecurrent variation (4) all samples strained vertically between 3.3% and 5.6%, (5) pH

ARTICLE IN PRESS

-50

0

50

100

150

200

250

300

350

400

450

0 50 100 150 200 250 300

Location from Cathode (mm)

Ave

rag

e P

erce

nta

ge

Incr

ease

in U

nd

rain

edS

hea

r S

tren

gth

(%

)

Test 1

Test 2

Test 3

Fig. 4. Percentage Undrained shear strength increase at mid-depth of clay in Tests 1, 2 and 3.

0

2

4

6

8

10

12

14

0 50 100 150 200 250 300

Location from Cathode (mm)

Dec

reas

e in

Wat

er C

on

ten

t (%

)

Test 1

Test 2

Test 3

(?)

(?)

(?)

-ve +ve

Fig. 5. Percentage decrease in water content in Tests 1, 2 and 3.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3522

rose to about 14 initially at the cathode and decreased to about 12, at discharge and(6) electrical energy used up varied from about 14 to 28 kWh/m3.

2.2.3. Compressibility of soil

Figs. 7 and 8 show the void ratio—consolidation pressure (e vs. log p0) of themarine clay before and after EO treatment, respectively, for Test 1. There aresignificant changes in compressibility characteristics, such as Cc decreasing from

ARTICLE IN PRESS

-100.00

-50.00

0.00

50.00

100.00

150.00

1 10 100 1000 10000 100000 1000000

Time (s)

Ch

ang

e in

Po

re P

ress

ure

(kP

a) P1P2P3P4P5

P1 P2 P3 P4 P5

-ve +ve

Fig. 6. Variation of pore pressure in clay between electrodes.

0.5

1

1.5

2

2.5

1 10 100 1000 10000

p (kPa)

Vo

id R

atio

, e

Sample 1

Sample 2

Fig. 7. Void ratio–pressure relation before EO application in Test 1.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 23

about 0.67 to about 0.43, 0.45 and 0.57, and Cr from 0.17 to 0.13, 0.14 and 0.15 forthe Tests 1, 2 and 3, respectively. The percentage increase in pre-consolidationpressure based on the assumption of a virgin compression curve with Cc ¼ 0:67ranged from 21% to 173% for the same three Tests. There is, therefore, a tendencyfor the EO treated clay to have reduced compressibility, which would reflectin the field as observed settlement of smaller magnitude under the same appliedpressure.

3. Field trial

A field trial was conducted in 2001 in Singapore to examine the feasibility of EO inconsolidation of soft clay using vertical drains. The trial consisted of installing andelectrifying two plots of land with a patented electrically conductive plastic verticaldrain known as EVD. The site under the jurisdiction of the JTC Corp was at TuasView Reclamation Phase IV, Singapore, in the Western part of Singapore Island.

The site consists of 18.7m of recently placed sand fill on 8m thick in situ softmarine clay in the seabed underlain by stiff clay and sedimentary rock of JurongFormation. The excess pore pressure in the soft clay from the reclamation fill wasabout 100–180 kPa at the time of initiation of this trial.

In a 50m� 50m area anode and cathode drains, all with filter sleeves, wereinstalled on a 1.2m2 grid using a standard PVD installation rig with no specialequipment or modification. Fig. 9 is an illustration of the installed EVD with theinstallers in the background.

There were four sub-plots to investigate various EVD configurations, as shown inFig. 10.

1. Sub-plot X contained the normal EVD with two stainless-steel wires at 1.2mspacing. Additionally it contained another set of copper EVDs with 6mm2 copper

ARTICLE IN PRESS

0.50

1.00

1.50

2.00

2.50

1 10 100 1000 10000

p (kPa)

Vo

id r

atio

, eSample 1

Sample 2

Sample 3

Before EO

Fig. 8. Void ratio–pressure relation of soft clay after EO treatment in Test 1.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3524

wires attached to the exterior of the drain filter, the copper wires being fullyinsulated in the upper sand layer but fully exposed within the underlying softmarine clay layer. These were also spaced at 1.2m so the effective drain spacingbecame 0.6m. The inclusion of the copper wires was designed to indicate themaximum power input capable of each drain despite the fact that the copper wireat the anodes might not last more than a day due to electro-chemical corrosion.

2. Sub-plot Y had the same configuration as Sub-plot X; except that the wire with anarea of cross section of 4mm2 in area was a bare copper wire, throughout thewhole depth of sand and soft marine clay. This again would not last but betweenX and Y; the loss of energy in the sand layers could be established.

3. For Sub-plot 2A EVD drains were produced with 2 copper wires and 3 copperwires in place of 2 steel wires within the core. These were installed in the sectionmarked in Fig. 10 as ‘Sub-plot 2A’.

4. Sub-plot 2B was a representative of the remaining area where two steel-wireddrains were used but with 1.2m spacing.

All EVDs were electrically insulated from the metallic anchor plates so as toprevent any possible short-circuiting via more conductive stiff clay underneath thesoft clay.

Each sub-plot had a piezometer and a deep settlement gauge locatedapproximately at the center. After EO treatment, in situ vane shear tests werecarried out in the soft clay.

Voltage probes and voltage lines were installed to take voltage readings at depthsat designated points in each subplot.

ARTICLE IN PRESS

Fig. 9. Installed EVD with drain installation machines in the background.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 25

3.1. Site operation and observations

1. Sub-plot Y was electrified first as it was meant more to provide information onthe maximum electric power required. Also during the intended polarity reversal,both anode and cathode exterior copper wires would be lost under electro-chemical corrosion. An input voltage of only 5V was applied at the start becausea fairly high current would be reached due to the high conductivity of copperwires and the huge losses in saline saturated sand layers. Electro-chemicalcorrosion, shown in Fig. 11, ended the treatment with polarity reversal in 72 h.Since the bare copper wire is exposed to both sand and clay of about equalthickness, about half the voltage would be lost in the sand layers. This had a veryslight effect on the pore pressure, as shown in Fig. 12.

ARTICLE IN PRESS

Fig. 10. Plan of trial site with different sub-plots and instrumentation (bare symbols are EVD locations).

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3526

2. In Sub-plot X the exterior copper wires were insulated within the sand layer, somore voltage could be applied into the marine clay and the treatment could last alittle longer due to the larger wire cross-section area (6mm2 instead of 4mm2).For the first treatment of this sub-plot, only drains with exterior copper wireswere connected to the power source. An input start-up voltage of 14V was

ARTICLE IN PRESS

Fig. 11. Electro-chemical corrosion.

230.00

250.00

270.00

290.00

310.00

330.00

350.00

11-Jun 16-Jun 21-Jun 26-Jun 1-Jul 6-Jul 11-Jul 16-Jul 21-Jul

Date and Time

Po

re P

ress

ure

Rea

din

gs

(kP

a)

P1 at Sub-plot Y

P2 at Sub-plot X

P3 at untreated area

Electrical Treatment ofSub-plot X(at 14.0V)

2nd Electrical Treatment ofSub-plot X(at 33V)

Electrical Treatment ofSub-plot Y(at 6.0V)

Fig. 12. Pore pressure variation in Sub-plots X, Y and untreated area.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 27

feasible. Within 6 h, pore pressure dropped by 2.6 kPa. Water bubbles wereobserved at the cathode wire ends, between the wire and its plastic insulation, asshown in Fig. 13. The measured drop in pore pressure is a point measurement andcan be estimated as being half the maximum (at anode) since the piezometer wasplaced approximately mid-point between a pair of anode and cathode drains.

Treatment of Sub-plot X continued for 22 h with no polarity reversal. Themaximum pore pressure drop measured was 10 kPa (refer to Fig. 12). There weretwo reasons for not reversing the polarity during treatment of Sub-plot X: Onewas to leave the cathode copper wires intact so that a further treatment of thisplot is possible where the current is ‘‘forced’’ to get into the clay layers. The otherwas to have a comparison with another sub-plot where polarity reversal was done.

Sub-plot X was given a second treatment by connecting all the remaining drains(comprising EVD with two embedded steel wires with or without the externalcopper wire) as anode drains and leaving the cathode drains unchanged. Thus thecurrent was ‘‘forced’’ to move to the marine clay layer since access to the cathodedrains is predominantly via the exposed copper wire within the soft clay. In ordernot to loose the cathode wires, polarity reversal was not applied. Again a cleardrop in the pore-pressure was registered as shown in Fig. 12. The build up ofresistance in this sub-plot is shown in Fig. 14.

3. Sub-plot 2B was a representative of the remaining area where two steel-wireddrains were used but with 1.2m spacing and insulated from the anchor plate. Itwas powered up but again it had little or no effect on the pore-pressure (Fig. 15).

4. In Sub-plot 2A, only copper wired drains, a section of which had 3 wires and thebalance 2 wires encapsulated within the conductive plastic core, were used. Sincethe wires were all encapsulated, the treatment could continue for a longer period,

ARTICLE IN PRESS

Fig. 13. Salt deposits at cathode wire.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3528

free from the fear of electro-chemical corrosion. About an hour after thegenerators were switched on, tiny bubbles of water were noticed surfacing at thelittle gap between the wire and its encapsulation (by the conductive plastic) at thecathode drains, as shown in Fig. 16. These occurred on the cathode (previouslyanode) drains when polarity was reversed. Treatment was carried out for 314.8 h

ARTICLE IN PRESS

0.000

0.200

0.400

0.600

0.800

1.000

1.200

1.400

Jul-01 Jul-01 Jul-01 Jul-01 Jul-01 Jul-01 Jul-01 Jul-01 Jul-01Date

Res

ista

nce

(o

hm

)

0

50

100

150

200

250

300

Cu

rren

t (A

)

Resistance

Current

Fig. 14. Resistance of Sub-plot X during 2nd E-O treatment.

300.00

310.00

320.00

330.00

340.00

350.00

360.00

370.00

380.00

19-Jun 21-Jun 23-Jun 25-Jun 27-Jun 29-Jun 1-Jul 3-Jul 5-Jul 7-Jul 9-Jul 11-Jul 13-Jul 15-Jul 17-Jul 19-Jul

Date

Po

re P

ress

ure

Rea

din

gs

(kP

a)

P5 at Sub-plot 2A

P3 at untreated area

P4 at Sub-plot 2B

Fig. 15. Pore pressure variation in Sub-plots 2A, 2B and untreated area.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 29

(or 13.12 days) in total, with polarity reversal commencing after the first 25 h. Inorder to provide the generators a cooling period, polarity reversal was applied at a7–10 h-interval. Treatment was ended when it was observed that the resistance ofthis plot had increased significantly and exponentially as shown in Fig. 17.

Within 7 h of electricity application through EVD, piezometer reading in P5surged approximately 3 kPa. When polarity was reversed 24 h later, pore pressurereading surged 14 kPa within 3 h. Subsequent polarity reversals caused similar surgein pore pressure readings but with decreasing magnitude. There was a significantdrop in excess pore pressure between 28th June 2001 and 30th June 2001. Continued

ARTICLE IN PRESS

Fig. 16. Salt deposits near copper wire of EVD.

0.000

0.500

1.000

1.500

2.000

2.500

3.000

3.500

27-Jun 29-Jun 1-Jul 3-Jul 5-Jul 7-Jul 9-Jul 11-Jul 13-Jul

Date

Res

ista

nce

(o

hm

)

Fig. 17. Resistance of Sub-plot 2A during EO treatment.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3530

electricity application and reversed polarity did not create further changes to porepressure readings. Fig. 15 shows pore pressure readings in (P5) Sub-plot 2A and in(P3) nearby untreated section (Sub-plot 2B) in Plot 2.

The readings indicated that significant pore water pressure changes had occurredwhen electricity was applied. However, direction of pore water movement could notbe inferred from the single piezometer. In addition, pore pressure was not lowered asexpected in EO, but increased instead. Increase in pore pressure was expected in theinitial stages of EO.

3.1.1. Field vane shear test

All treatment of the various sub-plots as described above ended by tremendousincrease in electrical resistance and no further changes to pore pressures duringprolong treatment. The vane shear test, however, would show the strengthimprovement of the clay in the variously treated areas.

Based on the pore pressure response to the electrical supply, Sub-plots Y and 2Bwere unlikely to show any additional improvement. Therefore vane shear tests weredone in Sub-plots X and 2A. The primary difference between these two plots must benoted. Sub-plot X had a mix of steel wired and copper wired drains plus drains withexternal copper wires exposed in the soft marine clay (insulated in sand). It also hadno polarity reversal. In contrast, all copper wired-drains in Sub-plot 2A had polarityreversed. For purpose of comparison, a vane shear test was done in the untreatedarea as well.

Fig. 18 shows the field vane shear results, VS-C-1, in the untreated area arisingfrom the improvement that comes from PVD only, and VS-C-3 in Sub-plot X. Theresult here seems to indicate that electrical treatment in Sub-plot X had no effect atall. Sub-plot X did register, during both treatments, a clear drop in the pore pressure.Since polarity reversal was never applied in this plot, EVD improvement wasexpected to be maximum at the anode and gradually reducing to nil at the cathode asobserved in laboratory trials. The result, VS-C-4 shown in Fig. 18, of the test near

ARTICLE IN PRESS

20

21

22

23

24

25

26

27

28

29

30

0 20 40 60 80 100Undisturbed Shear Strength (kPa)

Dep

th (

m)

120

VS-B-1, Before EO, 11 Jun, Near Sub-plot X

VS-C-1, 13 Jul, Non-EO treated Area

VS-C-3, 16 Jul

VS-C-4, 23 Jul

Fig. 18. Field vane shear in Sub-plot X.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 31

the anode displayed a significant improvement, clearly due to EVD. Although inboth vane shear tests in this sub-plot started at the ground surface, mid-pointbetween anode and cathode, after boring 18–28m below, it is possible, with just asmall inclination, to end up on either side, nearer the anode or cathode drain. Themeasurements from the 2nd test (Fig. 18) are likely to have been taken closer to theanode. In the untreated area and in Sub-plot 2A, where polarity reversal wasapplied, this did not quite matter. Fig. 10 shows the locations where the Field VaneShear tests are conducted.

The results of the Field Vane Shear test and percentage change in undrained shearstrength, as shown in Fig. 19 indicate a large shear strength increase especially in theupper half of clay layer in Sub-plot 2A. The increase in Sub-plot 2A below 27.0m isaffected by the closer proximity to stiff clayey silt, which has lower water content. Inaddition, the massive amount of overburden pressure from 18.7m of reclamation sandfill on the soft clay may have a slight ‘‘depth effect’’ as suggested by Lo et al. (1991).

The substantial difference between EO treated and untreated clay indicates thatEO caused greater improvement in the shear strength than conventional treatmentby PVD action.

The increase in undrained shear strength of soft marine clay with consolidation forSingapore Marine clay follows a Cu/Po0 of 0.26–0.28. If EVD were to perform as ahydraulic PVD installed at the same 1.2m spacing, the undrained shear strengthwould increase with time as shown with the broken line in Fig. 20. The vane shearstrength before and after EO treatment yielded a Cu variation as shown by the solidcurve, which was achieved in 13 days. The time taken by the PVD for the equivalentgain in strength would have been about 130 days illustrating a ten-fold reduction intime to achieve the same strength in soft marine clay. It should be noted thatcombined EO and PVD effect have been taken following Carillo’s concept and a

ARTICLE IN PRESS

15.0

17.5

20.0

22.5

25.0

27.5

30.00.0 20.0 40.0 60.0 80.0 100.0 120.0

Undisturbed Shear Strength (kPa)

Dep

th f

rom

Gro

un

d S

urf

ace

(m)

Before EO, in untreated area

After EO, in untreated area

Before EO, Sub-plot 2A

After EO, Sub-plot 2A

Fig. 19. Field vane shear strength in Sub-plot 2A.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3532

conservative parabolic variation was assumed between the cathode and the vanepoint, which was at mid-point between the anode and cathode.

3.1.2. Settlement

Although deep settlement gauges were installed in Sub-plots X and Y, theyregistered virtually no change. This could be attributed to the fact that the sub-plotswere too small to register a settlement independent of its surrounding (Tan et al.,1987). The main cluster of settlement plates for the entire plot was located in thecentral area where it was untreated electrically, so it would only show PVD settlement.

The settlement of the center of the PVD area subsequent to EVD treatment wasnot influenced by the EVD treatment that lasted about 14 days.

3.1.3. Power requirement

Sub-plot 2A consumed a total energy of 6223 kWh during 315 h and roughly halfwent to the soft marine clay. Taking the volume of clay to be 1728m3 the energyconsumed was 1.8 kWh/m3.

4. Conclusions

The field trial has shown that the soft clay, beneath the 18m of sand fill, can bereached effectively by EVD under the following conditions:

(a) Sufficiently high conducting EVD being installed in the clay;(b) Effective insulation of the metallic shoe, and(c) Provision for reversal of polarity.

ARTICLE IN PRESS

0 50 100 150 200 250 300Time (Day)

20

30

40

50U

ndra

ined

She

ar S

tren

gth

(kP

a)Actual Path

Conventional PVD Path

EVD Treatment

Fig. 20. Vane shear strength–time variation in EO application in comparison with that in PVD

application.

S.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 33

Shear strength improvement is clearly visible in the EO treated clay, which is alsoindicated by the pore pressure response observed. This aspect was observed both inthe laboratory small and large specimens as well as in the field trial. Site porepressure arising from 18m of sand fill was extremely high for EVD. Part of theelectrical energy would have spent in overcoming the hydraulic gradient towardsthe drain filter. Pore pressure response to EVD would be much more apparent hadthe pore pressure been smaller. Laboratory tests have revealed the importance ofvoltage gradient between electrodes, which has been observed in the field as well.

Since the EO treated plan area was small in extent compared with the depth, andthe positive friction from the sides of the untreated land partly overcomes the weightof the treated block, the settlement was not observed in the treated area within thetime of treatment, apart from the change in pore pressure. Based on the vane shearstrength values, converted to effective stress or otherwise, it is estimated that EOimprovement is about 10 times faster than the PVD treatment.

Acknowledgements

The authors wish to acknowledge the TEC Unit, Public Service Division,Singapore, for their support in funding the field trial reported in this paper. Theyalso acknowledge JTC Corp, Singapore, for its financial contribution, provision ofthe trial site and collaborative field support. The third and fifth authors also wish toacknowledge National University of Singapore for the provision of researchscholarships and the facilities at the university laboratories during the investigationsreported in this paper. The supply of EVD and the field operation by RasWillRepresentative Pte Ltd, Singapore, are also gratefully acknowledged.

References

Abiera, H.Q., Miura, N., Bergado, D.T., Nomura, T., 1999. Effects of using electro-conductive PVD in

the consolidation of reconstituted Ariake Clay. Geotechnical Engineering Journal 30 (2), 67–83.

Bergado, D.T., Balasubramaniam, A.S., Patawaran, M.A.B., Kwuenpreuk, W., 2000. Electro-osmotic

consolidation of soft Bangkok Clay with prefabricated vertical drains. Ground Improvement 4,

153–163.

Bo, M.W., Choa, V., Zeng, X.Q., 2000. Electro-osmosis properties of Singapore marine clay. Coastal

Geotechnical Engineering in Practice, International Symposium. Vol. 1, Yokohama, Balkema,

pp. 211–216.

Casagrande, L., 1952. Electro-osmosis stabilization of soils. Journal of Boston Society of Civil

Engineering 39, 51–83.

Casagrande, L., Wade, N., Wakely, M., Loughney, R., 1981. Electro-osmosis Project British Columbia,

Canada. Proceedings of the 10th ICSMFE, Vol. 3. Stockholm, pp. 610–697.

Chappel, B.A., Burton, P.L., 1975. Electro-osmosis applied to unstable embankment. Journal of

Geotechnical Engineering, ASCE 10 (18), 739–773.

Eggestad, A., Foyn, T., 1983. Electro-osmotic improvement of a soft sensitive clay. Proceedings of the 8th

ECSMFE, Vol. 2. Helsinki, pp. 597–603.

Fetzer, C.A., 1967. Electro osmotic stabilization of west branch dam. Journal of Soil Mechanics and

Foundation Division, ASCE 93 (4), 85–106.

ARTICLE IN PRESSS.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–3534

Gray, D.H., Mitchell, J.K., 1967. Fundamental aspects of electro osmosis in soils. Journal of Soil

Mechanics and Foundation Division, ASCE 93 (SM 3), 875–879.

Karunaratne, G.P., Chew, S.H., Lim, L.H., Toh, M.L., Poh, W.G., Hee, A.M., 2002. Electro-osmotic

consolidation of soft clay with conductive polymeric vertical drain. Proceedings of the 7th

International Geosynthetics Conference Nice, Vol. 3. France, pp. 350–355.

Lim, L.H., 2002. Electro-osmotic consolidation of soft clay. Masters Thesis. National University of

Singapore.

Lo, K.Y., Ho, H.S., Inculet, I.I., 1991. Field test of elctroosmotic strengthening of soft sensitive clay.

Canadian Geotechnical Journal 28, 74–83.

Lo, K.Y., Micic, S., Shang, J.Q., Lee, Y.N., Lee, S.W., 2000. Electrokinetic strengthening of a soft marine

sediment. International Journal of Offshore and Polar Engineering 10 (2), 133–144.

Shang, J.Q., 1998. Electro-osmosis-enhanced preloading consolidation via vertical drains. Canadian

Geotechnical Journal 35, 491–499.

Tan, S.A., Lee, S.L., Karunaratne, G.P., Choa, V., 1987. Design of drains and surcharge in reclamations.

Soils & Foundations 27 (4), 89–98.

Toh, M.L., Lim, L.H., Karunaratne, G.P., Chew, S.H., Ong, G.S., Hee, A.M., 2001. Electrical Osmotic

Consolidation of Clay. Proceedings of the 14th KKNN Symposium in Civil Engineering. Kyoto,

Japan, pp. 605–610.

Wade, M.H., 1976. Slope stability by electroosmosis. Proceeding of the 29th Canadian Geotechnical

Conference. Vol. 10. Vancouver, pp. 44–66.

ARTICLE IN PRESSS.H. Chew et al. / Geotextiles and Geomembranes 22 (2004) 17–35 35