You are on page 1of 19

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/358656437

Correlating groundwater level and shear strength: Kotkai Pakistan landslide as


case study

Article  in  Proceedings of the Institution of Civil Engineers - Forensic Engineering · February 2022


DOI: 10.1680/jfoen.21.00035

CITATIONS READS
5 34

2 authors:

Muhammad Israr Khan Shuhong Wang


Northeastern University (Shenyang, China) Northeastern University (Shenyang, China)
22 PUBLICATIONS   67 CITATIONS    140 PUBLICATIONS   857 CITATIONS   

SEE PROFILE SEE PROFILE

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

C-PEC project in Pakistan View project

All content following this page was uploaded by Muhammad Israr Khan on 20 February 2023.

The user has requested enhancement of the downloaded file.


1 Article
2 Correlating ground water level and shear strength in
3 seismic conditions – Kotkai landslide as case study
4 Muhammad Israr Khan 1, Shuhong Wang 2

5 1, 2
School of Resources and Civil Engineering, Northeastern University, 110819, China
6

7 Abstract: Earthfill dams failure and landsliding during heavy rainfall and flooding is a

8 common problem observed all over the world. This paper examines and discusses the

9 reasons of landsliding due to rainfall and flooding. Climatic rainfall data is inter-

10 connected with the rainfall precipitation and its effect on the increase of pore water

11 pressure inside soil slopes. Various soil slopes having different geometry and material

12 properties are analyzed with the variation of pore water pressure and the slope factor of

13 safeties are calculated. Correlations between the Ground Water Level (GWL) and slope

14 Factor of Safety (FS) are developed both in seismic and non-seismic situations.

15 Correlations between shear strength (), shear stress () and FS are also developed. Kotkai

16 landslide in Pakistan is considered as case study to cross check the correctness of the

17 developed correlations. These correlations can be applied to design a safe soil slope in

18 intense flood and rainfall areas.

19 Keywords: Slopes stabilization; Safety & hazards; Numerical modelling

20
21 1. Introduction

22 Earthfill dams failure or Landslides caused by rainfall pose a serious threat to human and

23 infrastructure. For this reason, there have been many studies conducted to understand

24 the mechanism of landslides. Most of them are based on numerical modelling, analysis

25 and laboratory tests. Soil erosion usually occurs during heavy rainfall and occurs

26 annually in many parts of the world, especially in areas that provide long-lasting and

27 heavy rainfall, large slopes, less or no vegetation and a rich source of fine clayey soil,

28 including colluvium and residual soil. In past many landslide disasters were reported

29 (Xu, 2012, Springman, 2013, Yenes, 2015, Zhang, 2018). One of the main reason of

30 landsliding is the increase of pore water pressure in soil. The increase in pore water

31 pressure is normally due to the rain water infiltration in soil which leads to the increase

32 in pore water pressure along with the decrease in shear strength of soil (Iverson, 2000,

33 Godt, 2009, Zhang, 2015). The stability of a soil slope is also greatly affected by the

34 increase of ground water table (Corominas, 2005, Ledesma, 2009, Conte, 2018b). The

35 presence of cracks in the ground on the slope can cause the infiltration of rainfall even

36 further into the deep soil. This will create more important hydrologic responses to deeper

37 soils than those of shallow soils within the earth profile. This hydrological response of

38 deep soil can lead to the rise of the water table. Another researcher examined the

39 landscape of Tessina in northeastern Italy and concluded that landslides were associated

40 with an increase in groundwater levels (Mantovani, 2000). Similarly there are many other
41 research work conducted which investigated and monitored rain water runoff, and

42 concluded that groundwater levels are directly linked to slope movement (Zhi, 2016,

43 Conte, 2018a, Bogaard, 2018, Cotecchia, 2016, Hong, 2011). The rapid response of deep

44 soil within the soil profile indicates that the flow rate within the deep soil profile is

45 characterized not only by the vertical infiltration but also by the increased flow and lateral

46 flow. Past experience proofs that the one of the major reason of landsliding during rainfall

47 is the increase of water level and hence it increases pore water pressure of the soil slope.

48 There is a direct relation between rain, ground water level and increase in pore water

49 pressure. It depends on the soil type and slope geometry that how much it will affect the

50 slope? On this topic, a very similar work is done by many researchers, see for example

51 (Conte, 2017, Huang, 2017, Martha, 2015, Matsuura, 2008, Piccinini, 2014, Khan, 2021b,

52 Khan, 2021d, Khan, 2021c, Khan, 2020, Khan, 2019). The stability of various landslides is

53 associated to the flow and moving of groundwater due to rainfall or variation of reservoir

54 water level. The mechanical parameters and hydrodynamic pressure of landslides varies

55 significantly, which may speed up the deformation of slope or induce possible landslide

56 instability.

57 Keeping all the previous research work in consideration, this paper examines the slope

58 stability analysis of a pre-defined soil slope with varying water level at bottom of the

59 slope and its FS. The same site was considered for investigation as already considered in

60 previous paper of same authors (Khan, 2021a).


61 2. Testing and Methodology:

62 Many laboratory experimentations were performed to conclude the mechanical

63 properties of soil required for analysis. These tests include Atterburg limit test, sieve

64 analysis, moisture content testing, porosity test, triaxial test, direct shear test, compaction

65 test, dry density test and specific gravity test. The soil samples were collected from site

66 by conducting four number of boreholes at different location. Four number of boreholes,

67 fifteen triaxial tests, eighteen number of Atterburg’s limit test, twenty two direct shear

68 test and almost fourty sieve analysis tests were performed to compute the mechanical

69 properties of soil in detail. Similarly all other tests such as porosity, moisture content test,

70 compaction and specific gravity tests were also conducted. All these tests were conducted

71 repetitively to make sure the material properties are precisely computed. Analysis is

72 performed such that the value of Ground Water Level (GWL) at the slope bottom face is

73 changed with a constant rate of 3 inches till 5 feet. This analysis is performed by

74 considering both seismic and non-seismic conditions. A limit equilibrium two

75 dimensional slope stability analysis software is used for the analysis by considering

76 Mohr-Coulomb criteria. The slope FS is computed and correlation is develped between

77 the GWL and FS. Figure 1 presents the slope considered in this analysis and mechanical

78 properties of soil after all the required testings is mentioned in table 1. While table 2

79 presents the FS with varying GWL –Nonseismic.


80 Figure 1. Slope model
81
82 Table 1. Mechanical properties of soil
S. Borehole Borehole Borehole Borehole
Description Average
No. 1 2 3 4
Unit weight
1 22.22 22.22 21.43 23.79 22.42
(kN/m3)
Cohesion
2 13.00 10.12 8.86 7.88 9.97
(kN/m2)
3 Friction 32.00 34.00 42.00 42.00 37.50
Specific
4 3.76 3.80 3.71 3.68 3.74
gravity
Moisture
5 30.00 28.00 23.00 8.00 22.25
content (%)
Strength Mohr- Mohr- Mohr- Mohr- Mohr-
6
type Coulomb Coulomb Coulomb Coulomb Coulomb
83 3. Results and Discussions

84 In this analysis, both Seismic Factor of Safety (SFS) and Non-Seismic Factor of Safety

85 (NSFS) are computed. The horizontal seismic coefficient is considered as 0.3 and vertical

86 seismic coefficient is ignored as it is normally very low. The initial GWL is considered on

87 top of the slope with increasing downward at 1 feet interval.

88 Table 2. FS with varying GWL -Nonseismic


Shear Strength Shear Stress
Description GWL NSFS
(kN/m2) (kN/m2)
0 152.931 102.242 1.496
1 ft. 152.931 102.242 1.496
2 ft. 20.105 14.516 1.385
3 ft. 25.3379 24.407 1.038
4 ft. 29.5036 31.2691 0.944
Original slope 5 ft. 35.2768 40.6444 0.866
6 ft. 39.0218 48.0115 0.813
7 ft. 41.8529 52.7819 0.793
8 ft. 48.2241 63.0494 0.765
9 ft. 51.0875 68.1292 0.750
10 ft. 55.6853 75.6637 0.736
89

90 Table 3. FS with varying GWL -Seismic


Shear Strength Shear Stress
Description GWL SFS
(kN/m2) (kN/m2)
0 131.861 152.715 0.863
1 ft. 114.821 133.068 0.863
2 ft. 17.0184 21.6352 0.787
3 ft. 20.0125 36.2968 0.551
4 ft. 22.8012 46.6241 0.489
Original slope 5 ft. 26.6262 60.8787 0.437
6 ft. 28.9822 71.7342 0.404
7 ft. 30.7978 78.8274 0.391
8 ft. 37.1869 102.021 0.365
9 ft. 39.9793 112.064 0.357
10 ft. 55.6853 75.6637 0.357
91 Table 4 and 5 presents the summary of SPSS linear regression analysis.
92 Table 4. Regression analysis summary in case of NSFS
Variables
Variables Variables
Model Method
entered removed
1 FS and GWL N/A Linear
Model Summary
Adjusted R Standard error
Model R R square
square of the estimate
1 0.917 0.841 0.824 0.12783
Anova
Model 1 Sum of squares df Mean square F Sig.
Regression 0.780 1 0.780 47.760 0
Residual 0.147 9 0.016
Total 0.927 10
Coefficients
Standardized
Model Unstandardized coefficients t Sig.
coefficients
B Standard error
Constant 1.429 0.072 19.813 0
GWL -0.084 0.012 -0.917 -6.911 0
93
94 Table 5. Regression analysis summary in case of SFS
Variables
Variables Variables
Model Method
entered removed
1 FS and GWL N/A Linear
Model Summary
Adjusted R Standard error
Model R R square
square of the estimate
1 0.915 0.837 0.819 0.08724
Anova
Model 1 Sum of squares df Mean square F Sig.
Regression 0.352 1 0.352 46.288 0
Residual 0.068 9 0.008
Total 0.421 10
Coefficients
Standardized
Model Unstandardized coefficients t Sig.
coefficients
B Standard error
Constant 0.816 0.049 16.583 0
GWL -0.057 0.008 -0.915 -6.804 0
95 From table 4 and 5, by using SPSS linear regression analysis, the correlation between FS

96 and GWL in case of NSFS from table 2 comes out to be:

97 NSFS = 1.429 - 0.084 GWL (1)

98 Applicability value such as R2 of equation 1 is 84.1%.

99 Using SPSS linear regression analysis, the correlation between FS and GWL in case of

100 SFS from table 3 comes out to be:

101 SFS = 0.816 - 0.057 GWL (2)

102 Applicability value such as R2 of equation 2 is 83.7%.

103 The applicability value of both these correlations is very high. For any slope stability

104 design project, these correlations can be used to know about the value of FS in case of

105 varying GWL.

106 Figure 2, 3, 4 and 5 presents the shear strength versus distance graphs.

Bishop simplified FS = 1.496


Radius (109.295) Center (106.986, 113.484)
Shear Stress & Shear Strength kPa

180
160
140
120
100
80
60
40
20
0
29.410
0.000
5.882
11.764
17.646
23.528

35.292
41.174
47.055
52.937
58.819
64.701
70.583

Distance m

Shear Strength kPa Shear Stress kPa

107 Figure 2. Shear strength graphs with variation of GWL


Bishop simplified FS = 0.736
Radius (103.466) Center (100.508, 113.484)

80

70

60

Shear Stress & Shear Strength kPa


50

40

30

20

10

0
14.927
19.903
24.878
29.854
34.830
39.805
44.781
49.757
54.732
59.708
0.000
4.976
9.951

Distance m

Shear Strength kPa Shear Stress kPa

108 Figure 3. Shear strength graphs with variation of GWL

Bishop simplified FS = 0.863


Radius (109.295) Center (106.986, 113.484)
180

160
Shear Stress & Shear Strength kPa

140

120

100

80

60

40

20

Distance m

Shear Strength kPa Shear Stress kPa

109 Figure 4. Shear strength graphs with variation of GWL


Bishop simplified FS = 0.357

Shear Stress & Shear Strength kPa


Radius (106.818) Center (106.986, 113.484)

150
100
50
0

Distance m

Shear Strength kPa Shear Stress kPa

110 Figure 5. Shear strength graphs with variation of GWL

111 Figure 2 to 5 proofs that shear strength of soil has a close relation with GWL. With increase

112 of GWL, the shear strength of soil also increased up to some extent and then declined after

113 the peak point is reached. In both seismic and non-seismic analysis, FS value is less than

114 1.5 in all the cases, which can be seen in table 2 and 3. Less than 1.5 FS means an unstable

115 or risky slope. So this slope needs stability solutions i.e. either to apply stepping

116 technique or insert nails. Another solution is to improve the soil strength by replacing the

117 soil or through compaction etc. Constructing a retaining wall is also an alternate solution

118 to make the unstable slope as stable.

119 During this study, it is observed that FS,  and  also have very close relation. From table

120 2, a correlation between these three parameters with varying GWL is established in

121 nonseismic conditions as follows by using SPSS analysis.

122 FS = 1.159 + 0.013  - 0.017  (R2 is 91.9%) (3)

123 From table 3, the correlation between SFS,  and  with varying GWL comes out to be:

124 FS = 0.581 + 0.008  - 0.005  (R2 is 71.9%) (4)


125 In case of seismic shear strength correlations, the equation applicability is 71.9%, which

126 can be increased with more data availability on site.

127 4. Kotkai Landslide Pakistan

128 Various events of landslides at Kotkai have been reported to occur. During recent

129 landslide, about 120 m stretch of road has completely displaced. However, during the

130 site visit, the road has been realigned after cutting the upside slopes. Secondly,

131 consultants also tried to realign the road through the left ridge shown in Figure 6, which

132 however was not practical due to steep gradient. During this activity, top layer of

133 alluvium, which was relatively impermeable, has been removed and underlying shale

134 has been exposed. These exposed portions of shale created a threat for source of seepage

135 intake and consequently increases the potential of further landslides in this area.

136 Figure 6. Alignment of Road Through Left Ridge


137 It is suggested that realigned road should remain intact instead of adopting the old

138 alignment by constructing the heavy fill foundation for road over the loose sliding

139 material. However, retaining wall is proposed along the outer edge of the road, which

140 will be placed over the compacted foundation. Beneath the foundation of retaining wall,

141 dowels bars have been proposed to increase the shear strength of foundation material. In

142 addition, it is suggested that realigned road should be placed over the 0.5 meter

143 compacted granular fill in order to drain the seepage water. Breast wall is proposed to be

144 constructed with a toe drain. Downside and upside slopes are proposed to be protected

145 with dry stone pitching. Table 6 presents the reliability percentage of all the developed

146 four correlations.

147 Table 6. Results comparison using analysis methods and developed correlations

Result by
SSA Results by the
S.
Description Correlation methods developed
No
(Bishop correlations
Simplified)
NSFS = 1.429 - 0.084
1 Factor of Safety 0.92 0.94
GWL
2 Factor of Safety SFS = 0.816 - 0.057 GWL 0.79 0.82

FS = 1.159 + 0.013  -
3 Factor of Safety 0.97 0.99
0.017 

FS = 0.581 + 0.008  -
4 Factor of Safety 0.96 0.99
0.005 
148
149 Table 6 proves that all the four developed correlations in this work have applicability

150 value higher than 95%.

151
152 5. Conclusions

153 A soil slope embankment is analyzed in detail with considering the local soil material

154 and its proposed model in two phases, i.e. seismic and nonseismic conditions.

155 Correlations between FS, GWL,  and  are developed both in seismic and nonseismic

156 conditions mentioned in equation 1, 2, 3 and 4. It is found that all these diffent parameters

157 have very close relation as the applicability value R2 of the correlations, such as from

158 equation 1 to 3 is 84.1%, 83.7% and 91.9% respectively. This high R2 is a clear sign that all

159 these parameters are interdependent. These correlations can be used in design projects

160 provided that the material properties are in range of table 1. Further work can be done

161 with considering other soil parameters and other soil material types, such as silty clay,

162 clayey sand, sand and clayey silt etc.

163 Abbreviations:

164 GWL = Ground Water Level

165 SFS = Seismic Factor of Safety

166 NSFS = Non Seismic Factor of Safety

167 FS = Factor of Safety

168 c = Cohesion of soil

169  = Friction angle of soil

170  = Unit Weight of soil

171  = Shear strength of soil


172  = Shear stress of soil

173 kPa = kiloPascal

174 SPSS = Statistical Package for the Social Sciences (Software)

175 R2 = Equation percentage of variation (Applicability value)

176 Availability of data and materials

177 The analysis data such as the SPSS regression analysis, excel graphs and software analysis

178 files are available as supplementary files.

179 Competing interests

180 The authors decalres no conflict of interest.

181 Author’s information

182 Muhammad Israr Khan is PhD scholars at Northeastern University China under the

183 supervision of Prof. Dr. Shuhong Wang.

184 References
185 Bogaard T, Greco R (2018) Invited perspectives: hydrological perspectives on

186 precipitation intensity-duration thresholds for landslide initiation: proposing

187 hydrometeorological thresholds. Nat Hazards Earth Syst Sci. 18(1):31-39.

188 Conte E, Donato A, Pugliese L, Troncone A (2018a) Analysis of the Maierato landslide

189 (Calabria, Southern Italy). Landslides. 15:1935–1950.

190 Conte E, Donato A, Troncone A (2017) A simplified method for predicting

191 rainfallinduced mobility of active landslides. Landslides. 14:35–45.


192 Conte E, Troncone A (2018b) A performance-based method for the design of drainage

193 trenches used to stabilize slopes. Eng. Geol. 239:158–166.

194 Corominas J, Moya J, Ledesma A, Lloret A, Gili JA (2005) Prediction of ground

195 displacements and velocities from groundwater level changes at the Vallcebre

196 landslide (Eastern Pyrenees, Spain). Lanslides. 2(2):83–96.

197 Cotecchia F, Lollino P, Petti R (2016) Efficacy of drainage trenches to stabilise deep slow

198 landslides in clay slopes. Géotech. Lett. 6:1-6.

199 Godt JW, Baum RL, Lu N (2009) Landsliding in partially saturated materials. Geophys.

200 Res. Lett. 36(2):206–218.

201 Hong YM, Wan S (2011) Forecasting groundwater level fluctuations for rainfall-induced

202 landslide. Nat. Hazards. 57(2):167–184.

203 Huang FM, Luo XY, Liu WP (2017) Stability analysis of hydrodynamic pressure

204 landslides with different permeability coefficients affected by reservoir water level

205 fluctuations and rainstorms. Water. 9:450.

206 Iverson RM (2000) Landslide triggering by rain infiltration. Water Resour. Res. 36(7):1897–

207 1910.

208 Khan MI, Wang S (2021a) Method for predicting factor of safety and seepage due to the

209 variation in dam width and other parameters. Proceedings of the ICE - Geotechnical

210 Engineering. Ahead of print. DOI: https://doi.org/10.1680/jgeen.21.00042


211 Khan MI, Wang S (2021b) Slope Stability Analysis to Correlate Shear Strength with Slope

212 Angle and Shear Stress by Considering Saturated and Unsaturated Seismic

213 Conditions. Applied Sciences 11(10):4568.

214 Khan MI, Wang S (2020) Comparing the various slope stability methods to find the

215 optimum method for calculating factor of slope safety. Earth and Environmental

216 Science 480:1-13.

217 Khan MI, Wang S (2021c) Developing correlations for advance prediction of slope factor

218 of safety using linear regression analysis – Karachi landslide is a case study. Polish

219 Journal of Environmental Studies.

220 Khan MI, Wang S (2021d) Slope Stability Analysis to Develop Correlations between

221 Different Soil Parameters and Factor of Safety Using Regression Analysis. Polish

222 Journal of Environmental Studies 30(5):4021-4030.

223 Khan MI, Wang S, Zhangze (2019) Analysis of earth fill hydraulic dam with varying crest

224 length and permeability to develop correlations. Earth and Environmental Science

225 304:1-8.

226 Ledesma A, Corominas J, Gonzàles A, Ferrari A (2009) Modelling slow moving landslides

227 controlled by rainfall. Proceeding of the 1st Italian workshop on landslides, Napoli.

228 1:196–205.

229 Mantovani F, Pasuto A, Silvano S, Zannoni A (2000) Collecting data to define future

230 hazard scenarios of the Tessina landslide. Int. J. Appl. Earth Obs. Geoinf. 2(1):33–40.
231 Martha TR, Roy P, Govindharaj KB, Kumar KV, Diwakar PG, Dadhwal VK (2015)

232 Landslides triggered by the june 2013 extreme rainfall event in parts of

233 Uttarakhand state, India. Landslides. 12:135–146.

234 Matsuura S, Asano S, Okamoto T (2008) Relationship between rain and/or meltwater,

235 pore-water pressure and displacement of a reactivated landslide. Eng. Geol.

236 101:49– 59.

237 Piccinini L, Berti M, Simoni A, Bernardi AR, Ghirotti M, Gargini A (2014) Slope stability

238 and groundwater flow system in the area of Lizzano in belvedere (Northern

239 Apennines, Italy). Eng. Geol. 183:276–289.

240 Springman SM, Thielen A, Kienzler P, Friedel S (2013) A long-term field study for the

241 investigation of rainfall-induced landslides. Geotechnique. 63:1177–1193.

242 Xu Q, Zhang S, Li WL, Van Asch TWJ (2012) The 13 august 2010 catastrophic debris flows

243 after the 2008 Wenchuan earthquake, China. Nat. Hazards Earth Syst. Sci. 12:201–

244 216.

245 Yenes M, Monterrubio S, Nespereira J, Santos G, Fernandez-Macarro B (2015) Large

246 landslides induced by fluvial incision in the Cenozoic Duero Basin (Spain).

247 Geomorphology. 246:263–276.

248 Zhang LL, Fredlund M, Fredlund DG, Lu HH, Wilson GW (2015) The influence of the

249 unsaturated soil zone on 2-D and 3-D slope stability analyses. Eng. Geol. 193:374–

250 383.
251 Zhang Y, Meng XM, Jordan C, Novellino A, Dijkstra T, Chen G (2018) Investigating

252 slowmoving landslides in the Zhouqu region of China using Insar time series.

253 Landslides. 15:1299–1315.

254 Zhi MM, Shang YQ, Zhao Y, Lü Q, Sun H (2016) Investigation and monitoring on a

255 rainfall-induced deep-seated landslide. Arab. J. Geosci. 9(3):182.

256

View publication stats

You might also like