Geotechnical Failure Investigation of a Reinforced Soil Wall and Remedial

Work Design

Tan, Yean-Chin; Kowng, Yan-Wen & Tan, Su-Kwong

G&P Geotechnics Sdn Bhd, Wisma G&P, 39-5 Jalan 3/146, Bandar Tasik Selatan, Kuala Lumpur, Malaysia.
[email protected]

This paper presents a case study of a displaced reinforced soil (RS) wall which is located on a fill slope with a bulging wall and open-
ing of gaps within the wall. Subsurface Investigation (S.I.) had been carried out using exploratory boreholes, mackintosh probes and
laboratory tests to determine the relevant engineering properties of the subsoil for subsequent slope stability analyses and investigation.
The investigation revealed that the failure is mainly attributed to inappropriate foundation design. This paper describes the site condi-
tions, including foundation design and the details of the geotechnical investigation of the distressed wall. Following the geotechnical
investigation and analyses, remedial works have been proposed and successfully constructed which include reconstructing part of the
wall to be supported by slab with spun piles and the use of lightweight materials of Expanded Polystyrene (EPS).

1 INTRODUCTION The 7.5m high RS wall design consists of well compacted

granular backfill and galvanised steel reinforcing strips of
The platform of a petrol station was constructed with a rein- highly adherent type (5mm thick x 45mm width) with length
forced soil (RS) wall of up to a height of 7.5m. The site is lo- of 5.8m for the top half height of the wall and length of 5.0m
cated in Kuala Lumpur. The RS wall is located on top of a fill for the lower half height of the wall. Typical cross section of
slope and is supported by reinforced concrete (RC) square RS wall design is shown in Figure 2.
piles. The movements of foundation and progressive wall
bulging had led to the opening of gaps within the wall and ag-
gravated the stability of the wall. This paper presents a geo-
technical investigation carried out by the Authors to find out
the causes of the failure. Appropriate remedial measures were
subsequently proposed based on the findings of the geotechni-
cal investigation works.

2 TOPOGRAPHICAL AND GEOLOGICAL

CONDITIONS

Generally the original ground surface at site varies from about

RL 87m to RL 109m before the construction of the platform at
RL 102.83m. The geological map of Selangor Darul Ehsan,
compiled by the Geological Survey Department, indicates that Figure 1 Original Piling Layout of RS Wall
the project site is underlain by Kuala Lumpur Granite. The
texture and composition of granitic rock generally ranges from
coarse to very coarse-grained. The subsoils based on grain
size generally consists of clayey sandy SILT with some layers
of silty sandy CLAY and silty sandy GRAVEL which is typi-
cal of granitic residual soils.

3 ORIGINAL REINFORCED SOIL WALL AND

FOUNDATION DESIGN THAT FAILED

The original RS wall foundation consists of 150mm x 150mm

square Reinforced Concrete (RC) piles at 1m c/c spacing
driven to a penetration length of 9m. A 350mm thick RC slab
is subsequently cast on the piles as the foundation of the RS
wall. The piling layout of the original RS wall design is shown Figure 2 Typical Section of Reinforced Soil Wall
in Figure 1.

1
4 SITE CONDITIONS AND OBSERVATIONS Mackintosh Probes (MP). The locations of field tests for both
stages of S.I are shown in Figure 5.
The RS wall panel showed visible signs of displacement,
bulging and opening of gaps between the wall panels at the
south-western corner of the wall as shown in Figure 3. This
had caused backfilled material to be washed away when rain,
inducing voids behind the RS wall as shown in Figure 4 that
further aggravated the internal stability of the wall.

Bulging and
Gap Opening

Figure 5 First Stage and Second Stage of S.I Works

Figure 6 shows the interpreted subsoil profiles for both stages

of S.I. works. Generally, borelogs of ABH-1 and ABH-2 indi-
cate that the original subsoil consists of sandy SILT and silty
GRAVELS. Based on the subsurface investigation carried out,
Figure 3 Side View of Wall Bulging the overburden soils in borehole BH-1 generally consist of
clayey SAND at the top 5 m, which are the fill materials and
underlain by the original residual subsoil. The depth of fill
materials is consistent with the original topography plan,
which shows that the original ground level is about 6m below
the existing profile. The clayey SAND and sandy CLAY are
generally very loose to loose and soft to medium stiff respec-
tively with SPT ‘N’ values ranging between 3 and 11 up to a
depth of approximately 8 m followed by slightly weathered
granitic bedrock.

MP tests from the second stage of S.I. works were carried out
along the slope as shown in Figure 5. Figures 7 and 8 show
the MP profiles, which are generally consistent with the fill
thickness.

Figure 4 Voids Behind Wall Due to Backfill Materials Being

Washed Away

5 SUBSURFACE INVESTIGATION

There were two stages of Subsurface Investigation (S.I) works

being carried out at the site. The first stage of S.I. works
(borehole designation with ABH) were carried out prior to
earthwork by the design consultant. The S.I works consist of
two (2) boreholes and ten (10) Mackintosh Probes (AMP).
However, the results obtained from the first stage of S.I works
were limited and insufficient for detailed investigation and
remedial work design. In addition, the boreholes are located
Figure 6 Interpreted Subsoil Profile from Both Stages of
at quite a distance apart from the location of the wall. There-
S.I. Works
fore, some of the important soil parameters could not be estab-
lished due to insufficient data. A second stage of S.I. works
(borehole designation with BH) was subsequently proposed
by the Authors to assist in the investigation. The second stage
S.I. works consist of one (1) borehole and eighteen (18)

2
 Table 1: Summary of Laboratory Test Results
Type of Results
Laboratory Test
Atterberg Limits & Samples collected from residual sub-
Particle Size Distri- soil are mostly clayey sand of low
bution plasticity
Single Stage Direct c = 0 kPa and φ = 36° for granular
Shear Box Test backfill material of the RS wall
C.I.U c = 3 kPa and φ = 33° for residual
subsoil

Liquid Limit wl

0 10 20 30 40 50 60 70 80 90 100 110 120

80 Inorganic Clays of Very High

Plasticity
wl=90
Inorganic Clays of
70 High Plasticity

Inorganic wl=70
60 Clays of Line "A" Ip=0.73 (wl-20)
Medium wl=50
Plasticity

Plasticity Index Ip
50
Figure 7 Mackintosh Probe Profiles for Section 1-1 wl=35

40

30

Inorganic Clays
20 Inorganic Silts of
of Low Plasticity
Very High Plasticity
10
Cohesionless
Soils
0
Inorganic Silts of Medium
Plasticity Inorganic Silts of High Plasticity
Inorganic Silts of
Low Plasticity

Figure 9 Atterberg Limit

Figure 8 Mackintosh Probe Profiles for Section 3-3

5.1 Laboratory Test Results

A series of laboratory tests were carried out on the samples
obtained from the second stage of subsurface investigation
works, they are:

1. Atterberg limits – see Figure 9

2. Particle size distribution Figure 10 Single Stage Direct Shear Box Test Results for
3. Single stage direct shear box test for wall granular backfill Granular Backfill Material of RS Wall
– see Figure 10
4. Isotropically consolidated undrained triaxial test (C.I.U)
for in-situ subsoil – see Figure 11

The results of the above laboratory tests are summarised in

Table 1.

3
 Case 2: RS wall without the piles to simulate the FOS in the
Interpreted Soil Strength Parameters (C.I.U. Tests)
initial construction before installation of piles and if the small
250
piles had displaced
200
Case 3: Local fill slope stability in front of RS wall
T = (σ1' - σ3') / 2

150

The acceptability of a slope’s stability is based on its ability to

100
achieve an adequate Factor of Safety (FOS) against slope fail-
c = 3 kPa
50 ure. An appropriate FOS against the failure of a slope is de-
φ = 33°
pendent upon the extent to which that failure could potentially
0
0 50 100 150 200 250 300 350
result in the loss of life. It is also dependent upon the degree
S = (σ1' + σ3') / 2
of economic loss that would result if a failure does occur. The
Lower Upper Adopted assignment of an appropriate FOS for various types of slopes
BH1/UD1/3-3.9 BH1/UD1/3-3.9 BH1/UD1/3-3.9
BH1/UD2/6-6.9 BH1/UD2/6-6.9 BH1/UD2/6-6.9 would be done in accordance to the Hong Kong Geotechnical
Engineering Office’s (GEO) “Geotechnical Manual for
Figure 11 C.I.U Test Results for Residual Subsoil Slopes”. Factor of Safety of 1.4 is recommended in the stabil-
ity analyses of the slopes. In addition, the slopes should
achieve a factor of safety of 1.1 for the condition of “worst
6 GEOTECHNICAL INVESTIGATION INTO THE predicted groundwater level”.
FAILURE OF RS WALL
Based on the results of the analyses as shown in Table 3, the
In order to determine the causes of RS wall bulging and dis- RS wall displacement is not likely to be induced by local in-
placement, the following analyses as shown in Table 2 were stability of fill slope in front of the wall. Based on the slip
carried out based on the interpreted subsoil parameters and failure analyses results, it can be deduced that the pile dis-
backfill material properties obtained from laboratory tests:- placement and lateral resistance are possible critical factors in-
fluencing the global stability. As deformation could not be
Table 2: Type of Analyses Carried Out analysed in the slip failure analyses using the limit equilib-
Type of Analysis Details rium method, finite element method (FEM) analyses are re-
quired to assess the pile displacement and lateral resistance.
Global Stability of the Wall • Slip Failure Analysis
Table 3: Stability Analyses Results
and Slope • Overturning and Bearing
Capacity Check
• Sliding Failure Check Case Long Term Factor of Safety
Internal Stability of RS Wall • Rupture
• Adherence Modified Spencer’s
Section Bishop Method
Structural and Geotechnical • Pile Axial Capacity Check
Method (Non-
Capacity Check of the Piled • Pile Lateral Resistance
(Circular Circular
Foundation Check
Failure) Failure)
• RC Base Slab Check
Degree of Compaction of • Check on Adequacy of De-
With Piles 1.58 (>1.4) 1.69 (>1.4)
Fill Slope Materials gree of Compaction
The following subsection describes in detail the analyses car- Case 1 With Piles
ried out to investigate the causes of distress. (worst case of 1.26 (>1.1) 1.37 (≥1.1)
WL)

6.1 Global Stability Analyses

Without Piles 1.25 (<1.4) 1.36 (<1.4)
Not acceptable Not acceptable
6.1.1 Slip Failure Analysis Case 2
Without Piles
Independent slope stability analyses were carried out to check (worst case of 1.20 (> 1.1) 1.25 (>1.1)
for different modes of failures namely circular and irregular WL)
shaped failure surfaces using the Modified Bishop’s and
Spencer’s Methods respectively. Therefore, all possible slip Local Stability 1.37 (<1.4)
failure surfaces have been checked and the most critical Fac- of 1V:1.5H Fill 1.44(>1.4)
tor of Safety (FOS) for the slopes is computed. Slope Not acceptable

In the slope stability analyses, the following conditions were Local Stability
Case 3
analysed: of 1V:1.5H Fill
Slope
Case 1: RS wall with the presence of 150mm x 150mm rein- 1.37 (>1.1) 1.44 (>1.1)
forced concrete (RC) piles as per the original design by C&S (worst case of
consultant (assuming the piles were not displaced) WL)

4
As shown in Table 3, the FOS for Case 1 is adequate in both 6.2 Internal Stability Analyses
abovementioned groundwater conditions. It indicates the
Internal stability analyses were carried out to check the inter-
global stability of the slope is adequate if the RC piles were
nal stability of the RS wall in terms of rupture, adherence and
not displaced significantly and could provide adequate lateral
wedge stability in accordance to BS 8006:1995. Coherent
resistance against slope movement. However, if the piles have
Gravity Method is adopted as recommended in Clause 6.3 of
displaced or damaged due to slope movement, the global sta-
the BS 8006:1995. As required, three load cases of load com-
bility would be inadequate, as demonstrated by Case 2. Al-
bination with different partial load factors are considered in
though the local stability of the slopes is slightly inadequate
the design.
(FOS=1.37 < 1.4) as shown in Case 3, it is only marginal. In
view of this, the RS wall displacement is not likely to be in-
Table 4 shows the load combinations used for the checking of
duced by local instability of fill slope in front of the wall, but
the design of the reinforced soil wall. In addition, the effect of
rather due to global slope instability with loading from RS
groundwater level on the internal stability is also investigated
wall when the small piles had been displaced, as shown in
using the method proposed by Tan & Khoo (2006).
Case 2.
Combination A considers the maximum values of all loads
The original configuration of RS wall supported by RC piles
and therefore normally generates the maximum reinforcement
has been analysed using a two-dimensional finite element
tension and foundation bearing pressure. It may also deter-
method programme. Figure 12 shows the graphical printouts
mine the reinforcement requirements to satisfy pull-out resis-
of the FEM analyses results (the deformation is plotted in ex-
tance although pull-out resistance is usually governed by
aggerated scale). As shown in Figure 12, the RC piles beneath
combination B.
the RS wall have bent and displaced significantly, and there-
fore part of the RS wall supported by piles also displaced to-
Combination B considers the maximum overturning loads to-
gether. The estimated pile displacement ranges from about
gether with minimum self mass of structure and superimposed
150mm to 170mm with induced maximum bending moment
traffic load. This combination normally dictates the rein-
and shear force to the pile of 36 kNm and 64 kN respectively.
forcement requirements for pull-out resistance and is normally
The maximum mobilised bending moment and shear force in-
the worst case for sliding along the base.
duced onto the 350 mm thick slab is 146 kNm and 192 kN re-
spectively. The comparison of the induced and ultimate bend-
Combination C considers dead loads only without partial load
ing moment and shear force of piles and slab are elaborated in
factors. This combination is used to determine foundation set-
Section 6.3.2 and 6.3.3 respectively.
tlements as well as generating reinforcement tensions for
checking the serviceability limit state.

Table 4: Partial Load Factors for Load Combinations associ-

ated with Walls (from BS8006:1995)

Load Combination

Effects A B C

Mass of reinforced soil body 1.5 1.0 1.0

Mass of backfill on top of RS
1.5 1.0 1.0
Wall
Figure 12 FEM Analysis – Displaced RC Piles Earth Pressure behind struc-
1.5 1.5 1.0
ture
Traffic Load:
6.1.2 Overturning and Bearing Capacity Check
(a) on reinforced soil block 1.5 0.0 0.0
The long term bearing capacity of the wall is governed by the
pile foundation supporting it. As for overturning, the inde- (b) behind reinforced soil
1.5 1.5 0.0
pendent analyses carried out indicate that the wall is stable block
against overturning with factor of safety against failure of
greater than 2.0. 6.2.1 Rupture
Rupture is the tensile capacity of the reinforcement to resist
6.1.3 Sliding Failure Check the tensile forces generated from the lateral earth pressure and
The sliding resistance is provided by the soil-to-soil contact the external loads. This is specified under BS 8006:1995
between the reinforced soil wall fill and the slab supporting it. clause 6.6.5.2.5. Independent analyses carried out indicate
The analysis indicates that the sliding resistance is adequate that the reinforcement is adequate against rupture failure.
against the lateral forces imposed on the RS wall. However, it
is important to note that the lateral forces have been trans-
ferred to the piles which should have been designed to cater 6.2.2 Adherence
for it.
Adherence is the resistance of the reinforcing fill against local
sliding of the reinforcement from the tensile forces generated.

5
The adherence check is specified under BS8006:1995 clause
6.6.5.2.4. Based on the design calculations provided, the 0 10 20 30 40 50 60 70
original design methodology of RS wall is in accordance to 0 0
Adherence (FMT)
BS8006:1995. However, the values for the coefficient of fric- Adherence (BS8006)
tion (μ) used are following the French Ministry of Transport’s 1 Tensile Force - Case C 1
Tensile Force - Case B
(FMT) recommendations for Reinforced Earth Structures Tensile Force - Case A
2 2
(FMTRRES) instead of BS8006:1995.
3 3

Depth (m)
The FMT’s recommendations use the parameter f* which is
similar to BS8006:1995’s μ parameter. The value used varies 4 4
with depth of the reinforced soil wall as follows (Cl. 2.3.3.1,
FMTRRES): 5 5

f * = fo* (1 – z/zo) + tan (φ’). z/zo for Z ≤ Zo = 6m. 6 6

7 7
f * = tan (φ’) for Z > Zo
8 8
* 0 10 20 30 40 50 60 70
Where the minimum fo value is taken as 1.50 in absence of
Tensile Force / Adherence (kN)
accurate measurements as recommended by FMT.
Figure 13 Tensile Force Induced and Adherence Profile
British Standard for Reinforced Soil, BS8006:1995 states that With Depth
the coefficient of friction is derived from μ = tan δ’. However,
BS8006:1995 does not state the value of δ’ and in consistent
with the standards, the value of δ’ is obtained from British 6.2.3 Effect of Groundwater Level on Internal Stability
Code of Practice for Earth Retaining Structures
(BS8002:1994). Clause 3.2.6 of BS8002:1994 states that for In order to investigate the effect of groundwater tables on the
design values, the values of parameter δ’ can be adopted as internal stability of RS wall, groundwater table of 1/3 and 2/3
approximately two-thirds of the peak angle of friction of fill of RS wall height are assumed in the analyses according to
materials in the reinforced soil (2/3*φ’). BS8006:1995. The tensile force does not significantly increase
with the presence of groundwater table (GWL) as can be seen
Using BS8002:1994’s design values for δ’ results in coeffi- from Figure 14.
cient of friction value of 0.45 (for a soil peak angle of friction
*
of 36°. However, using FMT’s recommendations, the f pa- The effect of groundwater table on the adherence is also in-
rameter values varies from maximum 1.5 at the top of the RS vestigated. The adherence (in accordance to BS8006) with the
wall to tan φ’ (which is 0.73) from the depth below 6 m. presence of the groundwater table (at 1/3H and 2/3H) is ex-
Therefore, the FMT’s recommendations of μ values are higher pectedly lower than that without the presence of the ground-
than the British Standards by 62%. The design of RS wall us- water table as illustrated in Figure 15.
ing FMT recommendations is more optimistic than design us-
ing British Standards.
0 10 20 30 40 50 60 70 80
Figure 13 presents the results of the tensile force and adher- 0 0
ence (calculated in accordance to BS8006:1995 and FMT) for No water effect - Case A
all cases. As can be seen from Figure 13, the tensile force in- 1 No water effect - Case B 1
No water effect - Case C
duced is higher than adherence. The non-compliance on ad- Water@1/3H - Case A
Water@1/3H - Case B
herence failure according to BS8006:1995, is due to insuffi- 2 Water@1/3H - Case C
2
cient reinforcement strip length provided there is no change to Water@2/3H - Case A
Water@2/3H - Case B
the strip width. The reinforcement for adherence is marginally 3 Water@2/3H - Case C 3
Depth (m)

adequate at the top 3m based on FMT’s recommendations. It GWL @ 2/3H

4 4
can be seen that the adherence reduces at the depth of ap-
proximately 3.5m for both BS8006 and FMT. This is attrib- 5 GWL @ 1/3H 5
uted to the reduction in the length of steel reinforcing strips as
explained in Section 3. 6 6

In addition, the level of compaction of the reinforced fill ma- 7 7

terials during construction (if inadequate) will also affect the
adherence. 8 8
0 10 20 30 40 50 60 70 80
Tensile Force (kN)
Figure 14 Acting Tensile Force Profile with Depth
(Case A, B, C)

6
 ing of the wall. Therefore, the geotechnical axial compression
0 10 20 30 40 50 capacity of the piles is adequate.
0 0
No water effect - Case A,B,C
Water@1/3H - Case A,B,C 6.3.2 Pile Lateral and Bending Resistance Check
1 Water@2/3H - Case A,B,C 1
Shear force and bending moment mobilised in the RC piles
2 2 are extracted from the FEM analyses. As mentioned in Section
GWL @ 2/3H 5.1.1, the maximum bending moment and shear force induced
3 3
on the RC piles are 36 kNm and 64 kN respectively. The steel
Depth (m)

reinforcement details of the 150mm x 150mm RC pile that are

4 4
commonly available in the local market using 4Y10 with link
5 GWL @ 1/3H 5 of φ5.5@69mm are adopted for computation of moment and
shear resistances. The ultimate bending moment and shear
6 6 force capacity of 150mm x 150mm pile are 10 kNm and 31
kN respectively and are found to be grossly inadequate. As a
7 7 result of inadequate lateral resistance, the RC piles have struc-
turally failed under bending and shear.
8 8
0 10 20 30 40 50 Since the piles are embedded within the 4m thick soil layer,
Adherence (kN)
these piles maybe subjected to axial and lateral loads which
Figure 15 Adherence Profile With Depth (BS8006) are induced by these soil movements. Tschebotarioff (1973)
presented design method to predict maximum induced bend-
0 10 20 30 40 50 60 70 80 ing moment for such scenario. As such, additional analysis us-
0 0 ing Tschebotarioff’s method was also carried out to check the
Adherence (FMT)
adequacy of the piles to resist such loads. The estimation of
1 Adherence (BS8006)
1 the lateral force and bending moment is based on the assump-
Tensile Force - Case C
Tensile Force - Case B tion of fixed pile head condition. The results show that the es-
Tensile Force - Case A
2 2 timated lateral force and maximum bending moment are 28
kN and 40 kNm respectively which are relatively consistent
3 3 with the findings from FEM analyses. However, it shall be
Depth (m)

noted that this method only provides an estimate of the pile

4 4 bending moment and does not consider either the lateral de-
GWL @ 1/3H
flection or the axial response.
5 5

6
6.3.3 RC Base Slab Check
6
Independent analysis of the adequacy of the RC base slab was
7 7 carried out based on the information from drawings prepared
by the original consultant. Based on the drawings, BRC mesh
8 8
A10 was provided at the top and bottom of the base slab of
0 10 20 30 40 50 60 70 80 350mm thick. As mentioned in section 6.1.1, from the FEM
Tensile Force / Adherence (kN)
analyses, the maximum bending moment and the shear force
Figure 16 Adherence and Tensile Force Profile with induced onto the RC slab are about 146 kNm/m and 192 kN/m
Depth (With Water Effect) respectively. Based on BS8110:1985, the ultimate moment
and shear resistances of the slab are 55 kNm/m and 102 kN/m
When comparing the adherence and tensile force induced with respectively. Therefore, the moment and shear resistances of
the presence of groundwater table (Figure 16), it was found the slab are grossly inadequate.
that the adherence is still inadequate. The adherence calcu-
lated using FMT is also included for comparison. In both
methods of calculation of adherence, the tensile force exceeds 6.4 Compaction of Fill Slope Material
the adherence with and without the presence of groundwater
Compaction tests were carried out on the fill materials during
table indicating inadequacy of design length.
filling of the platform and slopes and prior to the construction
of the RS wall. The information on the locations of the tested
6.3 Structural and Capacity Check of Piled Foundation samples was not made available at the time of investigation.
Figure 17 shows the compaction test results which indicate
that the degree of compaction of tested samples ranges from
6.3.1 Pile Axial Capacity Check 82% to 99%. In other words, 56% of the tested samples have
The geotechnical capacity of the piles is estimated by using a degree of compaction less than 95%, indicating some local-
the borehole information. The calculations show that the geo- ised areas with inadequate compaction. This can be observed
technical axial capacity of the piles is adequate for the pro- from some of the Mackintosh Probes results (Figures 7 to 8),
posed pile working load of 200kN. High strain dynamic load in which soil consistency increases and decreases alternately
tests were carried out on two piles and the results indicate that with depth at some MP locations.
the tested piles had achieved an axial static resistance of more
than twice the working load at the time of testing before plac-

7
 7 PROPOSED REMEDIAL WORK
Bulk Density vs Degree of Compaction

2.5 The proposed remedial work includes the following construc-

tion sequences:-
Bulk Density(Mg/m3)

2
a) Installation of sheet pile
1.5 b) Excavation of backfill material to the base of the rein-
forced soil wall
1
c) Hacking the base slab and drive φ400mm spun piles
0.5 (Grade 80, Class B) filled with reinforced concrete plug
d) Reinstatement of Reinforced Soil (RS) wall and partial re-
0
80 82 84 86 88 90 92 94 96 98 100 placement of backfill material with Expanded Polystyrene
Degree of Compaction(%)
(EPS)
Test A Test B

Since the convenient store is located near the slope, temporary

Figure 17 Compaction Test Results
sheet piles were necessary to minimise ground movement at
the convenient store during the adjacent excavation.
6.5 Findings
The spun piles were closely spaced and filled with reinforced
Based on the above findings, the causes of RS wall displace- concrete plug to provide adequate lateral resistance and the
ment and bulging are as follows: piles were driven into the competent hard layer. As such, it is
not necessary to adopt raking piles to provide the lateral resis-
a) Main factors: Foundation instability mainly due to inade- tance.
quate pile lateral resistance against lateral movement and in-
adequate shear and moment resistances of the RC piles and Longer wall reinforcement length was also used to achieve the
slab required factor of safety of the wall against adherence failure
while Expanded Polystyrene (EPS) was adopted to reduce the
b) Inadequate FOS for RS wall reinforcement strip adherence vertical overburden pressure at the wall section where re-
and inadequate fill compaction also contributed to some bulg- placement of wall reinforcements could not be done due to
ing of the wall site constraints. Figures 19 to 24 illustrate the details and se-
quence of the proposed remedial works.
The inadequacy of the abovementioned main factors is con-
firmed via visible cracks observed on the slab upon exposing
the base of the RS wall (as shown in Figure 18) for the reme-
dial work.

Slab

Figure 18 Slab Cracks Observed

In addition, due to the opening of the gap and bulging of the

RS wall caused by the factors described above, the fill materi-
als (granular materials) in the RS wall have dropped and Figure 19 Installation of Sheet Pile
washed out through the openings. The condition deteriorated
more rapidly during rainy season. Due to the loss of materi-
als, the internal stability of the wall was badly affected, induc-
ing further widening of gap and bulging of the wall. The proc-
esses repeated until a large void was formed in the wall (as
shown in Figure 4) and depression at the top of the wall.

8
 Figure 23 Section A-A

Figure 20 Excavation of Backfill Material, Hack Base

Slab and Install Spun Piles

Figure 24 Section B-B

7.1 Finite Element Method (FEM) Analyses

Prior to the actual construction of the proposed remedial
works, finite element method (FEM) analyses on the proposed
remedial solution were carried out to analyse the overall con-
figuration in order to determine the performance of the pro-
posed remedial solution. Figure 25 shows the model of RS
Wall used in the FEM analyses. The Mohr Coulomb soil
model was adopted in the analyses. In FEM modelling, both
the spun pile and RC Raft were modelled as linear elastic ma-
terials with properties of Young’s Modulus and Poisson’s ra-
tio. The spun piles and the RC raft were modelled as beam
Figure 21 Reinstatement of RS Wall and Partial Re- elements in order to determine the shear force and bending
placement of Backfill Material using EPS moment for detailed design thereafter. A summary of the ma-
terial properties is presented in Table 5.

Strip

RS Wall
Original
RC Raft Ground
Profile

Spun Piles
G.W.L
Hard Layer

Figure 25 Reinforced Soil Wall Model

Figure 22 Detail ‘A’

9
Table 5: Material Properties for FEM Analyses
γsat Eref
Name Type ν Slope
(kN/m3) (kN/m2)
Granular Drained 18 0.3 B- B+
50,000
Backfill
Existing Soil Drained 20 0.3
12,500 R.E. Wall

Hard Layer Drained 20 0.3

125,000
RS Wall Elastic - 0.2 -
Panel
RC Raft Elastic - 0.2 -
Pile Elastic - 0.2 -
Strip Elastic - - -

cref φ (°) EA EI
Name
(kN/m2) (kN/m) (kNm2/m)
Granular 1 36 - -
Backfill
Existing Soil 3 33 - -
Hard Layer 3 33 - -
RS Wall - - 3.64 E6 5,945 Figure 26 Location Of Inclinometers
Panel
RC Raft - - 6.5 E6 33,900
Pile - - 1.65 E6 34,976
Strip - - 4.69 E4 -

The checking for punching shear, normal shear, maximum

shear and bending moment on the RC raft indicates that a
minimum thickness of 300mm slab with T16@125mm rein-
forcing steel bars are required. The structural elements were
designed according to BS8110: 1985.

7.2 Reinforced Soil Wall Design

In the remedial design, the RS wall has been designed with
consideration of external and internal stability according to
BS8006:1995. The coherent gravity method was adopted for
internal stability check.

8 PROPOSED MONITORING PROGRAMME

Figure 27 Details of Inclinometers
A monitoring programme was recommended to monitor the
long-term performance of the proposed remedial work. The R.E. Wall Slope
Displacement (mm)
proposed monitoring programme consisted of three (3) incli- 0
nometers. However, one inclinometer was damaged during in- -5 5 15 25 35 45
-2
stallation. Therefore, only two (2) inclinometers were in-
stalled. The locations and typical details of inclinometers are -4
9-Feb-04
shown in Figures 26 and 27. Readings of the instruments -6 24-Feb-04
Depth (m)

12-Mar-04
were recommended to be taken fortnightly for the first two (2) -8 Design Limit
13-Apr-04
17-May-04
months and once a month thereafter for eight (8) months. 15-Jun-04
-10 16-Jul-04
Subsequently, the monitoring was carried out once every four Slope 10-Jan-05 (Final)
Design Limit
(4) months for a year. -12 B- B+

-14
R.E. Wall
Figures 28 to 31 show that the inclinometer monitoring results
-16
for Inclinometers 1 and 2 respectively in comparison with the
FEM analyses. Upon reviewing the inclinometer monitoring
Figure 28 Inclinometer Monitoring Results for Incli-
results, it was found that the lateral displacement of the incli-
nometer 1 (Major Axis)
nometer showed a trend of stabilising readings and are within
the acceptable limit as interpreted from the FEM analyses.
Therefore, it can be concluded that the proposed remedial
works are effective.

10
 Slope Stability Analyses
B- B+
Displacement (mm) • Local stability of the fill slope in front of the RS
0
-10 -8 -6 -4 -2 0 2 4 6 8 10 12 14
wall was not the causes of the bulging problems.
-2 • Global stability of overturning and sliding of the RS
-4 wall were adequate, except for the global slip failure
-6
through the slope under the loading from the wall.
Depth (m)

Slope stability analyses indicate that the global slip

-8 9-Feb-04
24-Feb-04
failure under the loading from the RS wall was in-
-10 12-Mar-04 adequate if the RC piles had been displaced.
13-Apr-04
-12 17-May-04
15-Jun-04

-14
16-Jul-04 RS Wall Internal Stability Analysis
-16 • According to BS8006:1995, all aspects of RS wall
Figure 29 Inclinometer Monitoring Results for Incli- internal stability were adequate except for adher-
nometer 1 (Minor Axis) ence. The wall strip reinforcements had inadequate
FOS against adherence failure.
• There was insignificant increase in the acting tensile
R.E. Wall
Towards Slope force in the presence of the groundwater table for
Displacement (mm)
0 7.5m RS wall. However, the adherence was com-
-5 0 5 10 15 20 25 30 paratively lower with the presence of the groundwa-
-2 ter table than that in the absence of the groundwater
Design Limit table. As such, it is recommended to consider the ef-
-4
fects of the groundwater table (hydrostatic pressure)
Depth (m)

-6
9-Feb-04
24-Feb-04
on the reinforced soil wall internal stability design.
12-Mar-04
13-Apr-04
17-May-04
-8 15-Jun-04 Structural and Geotechnical Capacity Check of Piled Founda-
16-Jul-04
10-Jan-05 tion
-10 Design Limit

• Finite element method (FEM) analyses show sig-

-12 nificant movement of the 150mm x 150mm RC piles
Figure 30 Inclinometer Monitoring Results for Incli- supporting the wall.
nometer 2 (Major Axis) • The shear and bending moment resistances of
150mm x 150mm RC piles and the base slab are in-
adequate.
B-
B+ Remedial Work Design
Displacement (mm)
0
• The remedial works consist of reconstructed part of
-6 -4 -2 0 2 4 6 8 10 12 14
-2
RS wall to be supported by 450mm diameter spun
piles (filled with reinforced concrete plug) and par-
-4 9-Feb-04
tial replacement of existing backfill materials with
Depth (m)

24-Feb-04
12-Mar-04 lightweight Expanded Polystyrene (EPS).
-6 13-Apr-04
17-May-04 • The monitoring results of the inclinometer show a
15-Jun-04
16-Jul-04 trend of stabilising readings and are within the ac-
-8
ceptable limit as interpreted from the FEM analyses.
-10 The proposed remedial works are effective.

-12
REFERENCES
Figure 31 Inclinometer Monitoring Results for Incli-
nometer 2 (Minor Axis) British Standard Institution. 1995. BS8006:1995 British Stan-
dard Code of Practice for Strengthened / Reinforced Soils and
Other Fills.
9 SUMMARY AND CONCLUSIONS British Standard Institution. 1994. BS8002: 1994 British Code
of Practice for Earth Retaining Structures
The investigation results of RS wall failure at the petrol sta- The French Ministry of Transport. Reinforced Earth Struc-
tion were presented. The wall bulging occurred progressively
tures: Recommendations and Rules of the Art.
had caused the backfilled material to be washed away during
rain and induced voids behind the RS wall. In an attempt to Tschebotarioff G.P. 1973. Foundations, Retaining and Earth
identify possible causes of failure, a comprehensive investiga- Structures, 2nd Edition. McGraw-Hill, New York.
tion was carried out. Tan, Yean-Chin & Khoo, Chee-Min. 2007. Effects of Hydro-
static Pressure on Limit State Design of Reinforced Soil Wall
The conclusions and recommendations can be summarised as Internal Stability, Proc. of 16th SEAGC, Kuala Lumpur (sub-
follows: mitted).

11