Two Case Studies on Soil Nailed Slope Failures

Liew, Shaw-Shong1 & Liong, Chee-How2

1
Director, Gue & Partners Sdn Bhd, Kuala Lumpur, Malaysia
2
Senior Engineer, Gue & Partners Sdn Bhd, Kuala Lumpur, Malaysia

39-5 Jalan 3/146, The Metro Centre, Bandar Tasik Selatan, 57000 Kuala Lumpur, Malaysia
e-mail: gnp@gueandpartners.com.my

Abstract

The use of soil nails for slope strengthening works has been gaining popularity in Malaysia since its first
application in 1980s in view of the attractive benefits of simple and fast installation method to reinforce steep
cut slope. Prior to design, it is of utmost importance that subsurface investigation be carried out to determine
the subsoil profile and the relevant engineering parameters. Soil nailed slope design in accordance to
established procedures could then be performed with better confidence after the investigation. This paper
presents the investigation results of two soil nailed slope failures in Malaysia and discussion on the lessons
learned from these two failures.

The first failure site (Site A) is underlain by completely weathered Shale facies, with the existence of
mudstone and siltstone. The failure consisted of seven upper berms of 1V:1H cut slope (a total of 42m in
height) and five lower berms of 4V:1 H soil nailed slope (total of 30m in height) reinforced with 12m length
soil nails. When the slope failure occurred, all the soil nails except for the soil nails at the lowest berm had
been installed.

The geology of the second failure area (Site B) consists primarily of weathered metamorphic rock with
massive granit ic intrusion. The failure involved a steep soil nailed slope (4V:1H) up to a total slope height of
seven and a half berms, with the maximum height of about 45m. The top slope was reinforced with 6m length
soil nails, and the lower slope were reinforced with 12m length soil nails.

During the investigation, additional subsurface investigation works were also carried out to confirm the
subsoil profile and shear strength parameters. The investigation results reveal that the first failure was due to
inadequate Factor of Safety (FOS) for global stability and the mechanism of progressive failure. The dominant
cause for the second failure is attributed to facing failure at the soil nail head.

Keywords: Facing failure, nail head, progressive failure, soil nail

1. CASE HISTORY A – FAILURE OF FIVE BERMS SOIL NAILED SLOPE + SEVEN BERMS
CUT SLOPE
The original slope consisted of the upper cut slope and lower soil nailed slope with the following
configuration as shown in Figure 1:
a) Seven upper berms of 1V:1H cut slope with a 2m width berm provided at every 6m height interval. The
total slope height was 42m.
b) Five lower berms of 4V:1H soil nailed slope with a 2m width berm provided at every 6m height interval.
The slope was reinforced with 12m length soil nails and the total slope height was 30m.
These cut slope and soil nailed slope were constructed to facilitate the formation of a new road.

-1-
1.1 Topography and Geological Condition
The site is located on high ground with reduced level ranging from RL210m to RL330m. Based on the draft
geological map interpreted by Mineral and Geoscience Department, Malaysia (1974), the site is underlain by
Shale facies. This facies consist of mudstone and siltstone.

Figure 1: Designed Slope Profile and Remedial Slope Figure 2: Front View of Failed Slope
Profile

1.2 Site Conditions and Observations

During the geological mapping and subsurface investigation works after the failure, it was observed that the
cut slope face varie d from a relatively smooth surface to irregular rough surface. In general, the site was dry
and no water seepage was observed. The geological mapping reveal that the joint sets mapped at the slope
surface were daylighting towards the main road as shown in Figure 3. These unfavourable joint sets were
likely contribut ing to the slope failure. Joints with in-filling material like iron oxide and silt were also
observed as captured in Figure 4. Most of the exposed materials on the slope surface were Grade III to V.

Figure 3: Daylighting Joint Sets towards Main Road Figure 4: Joints with Infilling Material

-2-
1.3 Subsurface Investigation and Laboratory Tests
A subsurface investigation consisting of two boreholes and relevant laboratory tests were planned and
implemented to establish the subsoil profile and obtain necessary soil strength parameters. The locations of
the borehole s and the interpreted subsoil profile are shown in Figures 5 and 6 respectively.

Figure 5: Layout Plan showing Borehole Locations

Figure 6: Interpreted Subsoil Weathering Profile

-3-
A series of laboratory tests was carried out on the samples recovered from the subsurface investigation works.
The tests included Atterberg limits, particle size distribution, multiple reversal shear box test, Consolidated
Isotropically Undrained (CIU) triaxial test with pore pressure measurements and petrographic analysis.
Selected test results are presented in the following section.
0 250 500 750 1000 1250 1500 1750 2000 2250 2500 0 50 100 150 200 250 300 350 400 450 500 550 600

2500 2500 600 600

Legend 550 550

2250 Shear Box (BH2-MZ2) 2250 Legend
Shear Box (BH1-MZ1)
CIU (BH2/MZ2-S1)
500 CIU (BH1/MZ1-S1) 500
2000 CIU (BH2/MZ2-S2) 2000 CIU (BH1/MZ1-S2)
CIU (BH2/MZ2-S3)
450 CIU (BH1/MZ1-S3) 450
CIU (BH2/MZ3-S1)
1750 CIU (BH2/MZ3-S2) 1750
CIU (BH2/MZ3-S3) 9
o ) 400 400
=3 )
o
φ' p 33

Shear Stres s, k Pa
Shear Stress, kPa

1500 , 1500 350 350

30 'p=
= o ) , φ
3)
o
(c ' p 33 30
th = = =3
1250
ng φ' r 1250 300
h (c ' p , φ' r 300
,
tre =0 gt 0
kS c' r ren ' =
ea t h( 250 St th
(c r 250
1000 P ng 1000 ak ng
tre Pe t r e
lS lS
ua 200 ua 200
sid s id
750 Re 750 Re
150 150

500 500
100 100

250 250 50 50

0 0
0 0
0 50 100 150 200 250 300 350 400 450 500 550 600
0 250 500 750 1000 1250 1500 1750 2000 2250 2500
Normal Stre ss, kPa
Normal Str ess, kPa
Figure 7: Effective Shear Strength for Grade III Figure 8: Effective Shear Strength for Grade IV
Material Material
Based on the British Soil Classification System, most of the tested samples are silt of intermediate to high
plasticity. Three sets of CIU tests were also carried out on samples of Grade III and IV material. In addition,
two multiple reversal direct shear box test were also performed on the reconstituted samples from the CIU
specimens. The results of both the CIU and shear box tests are shown in Figures 7 and Figure 8. Table 1
summarises the peak and residual strength parameters from the tests.

Table 1: Peak and Residual Strength of Grade III and IV Materials

Weathering Grade Effective cohesion, c’ Effective friction angle, φ’
Grade IV Peak 30 kPa Peak 33o
Residual 0 kPa Residual 33o
Grade III Peak 30 kPa Peak 39o
Residual 0 kPa Residual 33o

1.4 Rainfall Record

Rainfall record was obtained for the failure area and it was found that there was no record of abnormal high
rainfall before the slope failure event.

1.5 Potential Failure Modes of a Soil Nailed Slope

The Manual for Design and Construction Monitoring of Soil Nail Walls (1998) recommends that design of
soil nailed slope essentially requires the following potential failure modes to be assessed:
a. Nail tendon failure – The structural capacity of the nail tendon should be adequately provided in the design.

-4-
b. Nail pullout failure – The ultimate pull-out resistance between the nail and soil/rock should be checked to
ensure that the Factor of Safety (FOS) against slope failure is adequate. The soil nail pull-out resistance is
usually verified by a pull out test.
c. Face failure – The facing structural design requires provisions of adequate concrete thickness and
reinforceme nt for moment and shear capacity to resist earth pressures and punching shear force.
d. Overall failure – The overall failure is usually checked using slope stability analysis software by means of
limit equilibrium method. Adequate FOS against overall failure should be provided according to the authority
requirement or relevant codes.

1.6 Causes of the Failure

As this particular soil nailed slope failure resembles an overall failure as discussed in Section 1.5, slope
stability analyses were carried out using limit equilibrium and finite element methods to investigate the causes
of the failure.

a) Limit Equilibrium Method

Slope stability analysis using Bishop method is performed using the established subsoil profile and shear
strength parameters from the subsurface investigation. The results of the analyses indicate that the global
Factor of Safety (FOS) is marginally higher than 1.0 even when all the soil nails were completely installed at
the lower 5 berms. In addition, the FOS for local stability for the 1V:1H upper cut slope is also marginally
more than 1.0. Figures 9 and 10 show the analyses results for both global and local stability.

b) Finite Element Method

In addition to the limit equilibrium method, finite element analyses using the Mohr-Coulomb strength criteria
and elasto-plastic model were also carried out to simulate the slope cutting at various stages (Figures 11 to 14)
and to reveal the likely failure mechanism. From the modeling of each excavation stage, it is apparent that the
development of plastic points within the soil body indicates mobilisation of peak strength in these soil
elements. When the cutting is in progress, the plastic points gradually develop and propagate to the lower
areas. Eventually, a well-defined shear surface is formed when the excavation reached the lower berm. This
demonstrates the case of a progressive failure.

FOS=1.08 FOS=1.01

Figure 9: Limit Equilibrium Analysis Results for Figure 10: Limit Equilibrium Analysis Results for
Local Slope Stability (1V:1H upper cut slope) Global Stability

Based on both the analyses results, it was evident that the cause of the failure was due to inadequate FOS
against overall failure. Nail tendon failure, nail pull-out failure and facing failure were unlikely to govern the
failure mechanism as no signs of these failures were observed.

-5-
Figure 11: Development of Plastic Points in Finite Figure 12: Development of Plastic Points in Finite
Element model (After cutting of two upper berms) Element model (After cutting of seven upper berms)

1.7 Summary
From the investigation and analyses results, the findings are summarised as below:
a. Limit equilibrium analyses revealed that the FOS against global and local stability are just marginally
higher than 1.0.
b. The unfavourable daylighting geological structures were likely contributing to the slope failure.
c. The finite element analysis shows that the shear failure surface gradually developed as the excavation
progresses to the lower berms of the slope, indicating the mechanism of progressive failure.

Figure 13: Development of Plastic Points in Finite Figure 14: Development of Plastic Points in Finite
Element model (After cutting of eleven upper berms) Element model (After cutting of twelve upper berms)
2. CASE HISTORY B – FAILURE OF 7.5 BERMS SOIL NAILED SLOPE
A soil nailed slope (4V:1H) with maximum of 7.5 berms was formed to facilitate the construction of a new
road at a mountainous terrain. The maximum height of the soil nailed slope was about 45m. The soil nailed

-6-
slope was strengthened with 12m length soil nails except the highest berm where 6m length soil nails were
installed. Every berm (6m height) was reinforced with three rows of soil nails. The horizontal spacing of the
soil nail was 1.5m c/c.

Figure 15: Layout Plan showing the Failure Area Figure 16: Soil Nails Remained Intact on the Slope after
Failure

2.1 Topography and Geological Condition

The site was hilly with valley terrain with levels ranging from about RL1090m to RL1135m as shown in
Figure 15. The slope face was facing a valley with a river running through the valley. Based on the Geological
Map of Peninsular Malaysia ; 8th edition, 1985, the site comprises of granite and metamorphic rock types.

2.2 Site Observations

Site visit was carried out after the failure of the soil nailed slope. Figure 16 shows most of the soil nails
remained intact at the failed slope surface. The site was found to be dry and no water seepage was observed on
the failed slope surface. Geological mapping was also carried out at the failure area to map the discontinuities
and weathering profile of the slope.

2.3 Subsurface Investigation and Laboratory Tests

A subsurface investigation was planned to establish the subsoil profile and to obtain the necessary soil
strength parameters for the investigation purpose. Two boreholes, two observation wells, Mackintosh Probes,
hand augers and collection of block samples have been performed. Monitoring of groundwater was also
carried out during and after the subsurface investigation works. Groundwater level was found to be about 20m
and 35m below ground level for boreholes BH-1 and BH-2 respectively.

The interpreted subsoil profile is shown in Figure 17. The subsoil profile was interpreted based on the
subsurface investigation results, drilling records of the soil nails and site observations after the failure event.
Consolidated Isotropically Undrained (CIU) triaxial test with pore pressure measurement were also carried out
for soil samples recovered from the boreholes.

-7-
 Figure 17: Interpreted Subsoil Profile

Most of the tested samples were classified as very silty gravel/sand. The samples show high composition of
coarse material (sand and gravel) content, ranging from about 50% to 90%.
As shown in the interpreted subsoil profile, varying degrees of subsoil weathering was encountered within the
soil nailed slope. Due to the difficulties in recovering and testing of relatively harder material (Grade IV and
III) to obtain realistic shear strength parameters, the rock mass strength method proposed by Hoek et al (2002)
was adopted to derive the equivalent Mohr-Coulomb strength parameters. The Hoek-Brown failure criterion
for rock mass is an empirical method which provides rock mass strength parameters by incorporating
geological parameters and observations in the failure model. A software “RocLab” was used to input the
required geological parameters and generate the Mohr-Coulomb failure envelope. The strength parameters
from Hoek-Brown failure criterion was then adopted in the slope stability analysis.
2.4 Rainfall Record
Rainfall record was also obtained from three rain gauge stations nearby to the failure area. No exceptionally
heavy rainfall was recorded in two months prior to the failure event.
2.5 Cause of Failure
Based on the site observation, nail tendon failure and pull-out failure were unlikely to be the cause of the
failure. The conditions of the soil nails (Figure 16) after failure event indicated that the failure might be highly
related to face failure, which was discussed in Section 1.5 of this paper. The Manual for Design and
Construction of Soil Nail Walls (1998) states that, “The facing structural design requires provision of
adequate concrete thickness, reinforcement and moment capacity to resist the earth pressures applied to the
facing span between adjacent nail heads, and provision of adequately sized bearing plates to provide adequate
punching shear capacity”. Assessment was then carried out on the facing of the soil nailed slope, which was

-8-
designed to be 100mm thickness (G25) with 2 layers of A6 mesh reinforcement as shown in Figure 18. It was
found that the flexural strength and punching shear strength of the facing was grossly inadequate. Therefore,
facing failure was deemed imminent.

Soil 1
Description: Soil 1
Model: MohrCoulomb FOS=1.12 58
65 1
Wt: 18 63 66
2 57 64
6
Cohesion: 6 61
55
Phi: 32 59
53 62
Description: Soil 2
4 3
51 56 Soil 2
2
49 54 60
Model: MohrCoulomb 47 52
67
Wt: 18 69 5
50
45 48
Cohesion: 8 43 68
87

Phi: 36 41 46
39 86
44
Description: Soil 3 71 70
37 42
Soil 3
7 40 3
Model: MohrCoulomb 34 38
Wt: 19 32 36
73 72
30 80
Cohesion: 70 35
28 33
Phi: 27 26 31
24 29
Description: Soil 4 75 74
22 27
Model: MohrCoulomb 20 25
18 23
Wt: 20 16 21
Cohesion: 120 78 77 76
14 19
Soil 4
12 17
Phi: 27 10 15
8 13
81
11
9

85

Figure 18: Details of Gunite Facing at Site Figure 19: Stability Analysis Result 83

Apart from the assessment of the facing failure, slope stability analysis using limit equilibrium method was
also carried out to assess the possibility of overall failure. The interpreted subsoil profile, CIU results and
strength parameters from Hoek-Brown failure criterion were collectively adopted in the stability analysis
model. The stability analysis result (Figure 19) revealed that FOS of the soil nailed slope was only 1.12 which
was marginally stable. Such low FOS is not adequate for long-term stability.

2.5 Summary
The investigation has revealed the following findings:
a. The main cause of failure was due to inadequate design of the gunite facing of the soil nailed slope.
b. The FOS against overall failure is inadequate and was marginally stable.
c. Rainfall/groundwater was not a triggering cause to the failure event.

3.0 RECOMMENDATION AND CONCLUSION

The following recommendations and conclusions can be made from the investigation of the two case histories
of soil nailed slope failures:
a. It is necessary to carry out subsurface investigation at high cut area especially if soil nailed slope is to be
constructed. Inspection and examination on the exposed slope material and geological structures should
also be carried out during various stages of construction. If the subsoil profile, geological structures or
groundwater table are found to be different from the design model, then the design shall be reviewed with
the updated information. This design feedbacks and verifications are crucial in order to ensure safety of a
soil nailed slope. Sometimes, further design optimisation is possible if the ground condition is more
favourable.
b. Engineering assessment shall be carried out for all four potential failure modes: nail tendon failure, nail
pull-out failure, facing failure and overall failure for the design of soil nailed slope. The design of the
facing is often neglected by designers thinking that it s sole purpose of the facing is to protect against
surface erosion only and neglect its role as a structural element to resist the earth pressure. It should be

-9-
 noted that the design of facing is especially critical when the soil nailed slope is high and steep. The facing
should be designed to resist the earth pressures, bending moment and punching shear force from the pulling
of soil nail under the earth pressure. Inadequate facing design could lead to failure of soil nailed slope as
depicted in case history B.
c. Representative shear strength parameters should be used for the design of a soil nailed slope. Strength tests
such as CIU and direct shear box tests should be carried out to obtain the strength parameters for design.
For rock mass, the current available approach still relies on empirical method. The reliability of the method
is very much related to the adopted values for the empirical factors with respect to the database of failure
case and visual classification based on rock mass characteristic.

REFERENCES

British Standard Institution, BS8081:1989 British Standard Code for Ground Anchorages.
Bryne, R.J. et al. (1998) Manual for Design and Construction Monitoring of Soil Nail Wall. Federal Highway
Administration of US.

Geological Map (1974), Malaysia Minerals and Geoscience Department.

Geological Map (1985), Malaysia Minerals and Geoscience Department.

Hoek, E., Carranza-Torres, C. and Corkum, Brent. (2002) Hoek-Brown Failure Criterion-2002 Edition.
INGEOTUNELES , Tunnelling Engineering Series, Vol.7, Chapter 1, pp.35-47.

-10-