Cyclic drawdown of water causing the slope failure of canal and dam

Abstract

Man-made structure such as dam or excavated canal are subjected to change in water level in its lifetime. Safety analysis of sudden drawdown case is common in slope design. However, in general practice the analysis does not consider the accumulation of shear strain that might occur with each drawdown event. This accumulation may eventually reduce the shear strength soil and cause the unbalance of slope. Recently, two different earth structure in Thailand, Pa Bon dam and canal slope in soft Bangkok clay area, failed after several years of service. The slope of those structures experienced the sudden drawdown for many years. Coupled stress and strain analysis reveals that the strain accumulation during each drawdown cycle caused the reduction of shear strength and was the major cause of failure. By knowing this, design code may need to add additional design case in order to cover this mode of failure.

Full text

Cyclic drawdown of water causing the slope failure of canal and dam
Suttisak Soralump1, P.W. Noppadol 2, K. Panthi3 and J. Sanmuang4
1 Associate Professor, Department of Civil Engineering, Kasetsart University, Bangkok, Thailand.
2 Professor, Department of Civil Engineering, Asian Institute of Technology, Pathum Thani, Thailand.
3 Graduate Student, Department of Civil Engineering, Kasetsart University, Bangkok, Thailand.
3 Graduate Student, Department of Civil Engineering, Kasetsart University, Bangkok, Thailand.
ABSTRACT
Man-made structure such as dam or excavated canal are subjected to change in water level in its lifetime. Safety
analysis of sudden drawdown case is common in slope design. However, in general practice the analysis does not
consider the accumulation of shear strain that might occur with each drawdown event. This accumulation may
eventually reduce the shear strength soil and cause the unbalance of slope. Recently, two different earth structure in
Thailand, Pa Bon dam and canal slope in soft Bangkok clay area, failed after several years of service. The slope of
those structures experienced the sudden drawdown for many years. Coupled stress and strain analysis reveals that the
strain accumulation during each drawdown cycle caused the reduction of shear strength and was the major cause of
failure. By knowing this, design code may need to add additional design case in order to cover this mode of failure.
Keywords: Slope failure, Sudden drawdown, Coupled Stress Analysis, Numerical Modelling
1 INTRODUCTION
Pa Bon dam located in the southern part of Thailand.
The operation of dam started in 2004 and has been in
operation till date. However, the first movement of
upstream slope was observed in the year 2014. Repair
work had been carried out after the first movement but
the movement still occurred whenever the reservoir
water level was drawn down. This is not the first case,
Huai Ta Ju dam in the northeastern part of Thailand
also has the same behavior but it was not as lucky as Pa
Bon dam, the upstream slope failed after 20 years of
reservoir water cycles. Fortunately, the failure did not
cause the dam breach. Similarity, the failure behavior
occurred in the excavated canal slope located in the
Bangkok soft clay area. There also a case where the
extreme drought caused those slopes to fail as well. The
detail analysis these two cases have been briefly
described in this paper.
2 PA BON DAM SLOPE MOVEMENT
Pa Bon dam is a zoned dam located at southern part
of Thailand. The typical cross section of the dam is
shown in Figure 2. The operation of the dam started in
2004 and was operating without any hindrance until the
year 2014 when the first movement was observed in the
slope of embankment structure. The dam was operated
after minor rehabilitation but it continued to move each
year with drawdown of water. The actual water level
since the filling of dam is shown in Figure 3. It can be
seen that the dam had experience several sudden
drawdown before the observation of slope movement
but no movement was observed prior to 2014. The
fastest drawdown rate was 10.76 m/day and the largest
drawdown distance is 24.33 m. Figure 1 shows the
photo of the movement on the dam slope and this
movement increased each year with the decrease in
water level.
Figure 1: Photograph showing the movement of Pa Bon Dam
after first movement
Couple stress analysis reveals that the factor of
safety of slope start to reduce significantly when the
reduction of water level was substantial. After that the
slope started to move each year even though the
drawdown rate and distance was smaller than previous
year. Residual soil strength testing was carried out from

Impervious Clay Core
Shell (CL)
Soil Foundation
Rock Foundation (Granite)
Riprap
Filter
23
1
114.00
3
1
31
1
1
1.5
1
Chimney Drain 2.00m
1
0.75
Blanket Drain 2.00m
1
1
1
Rock Toe
1
1
10.0
3.0
8.0
2.5
1
1
0.75
10.0
4.5
4.5
4.5
4.5
4.5
4.5
26.2
13.0
2
1
2
1
2
1
2
1
2
1
2
1
Figure 2: Cross Section of Pa Bon Dam
Figure 3: Variation of Safety Factor with Water Level of Pa Bon Dam
collected dam material (Head, 2011; Mesri &
Huvaj-Sarihan, 2012; Stark & Hussain, 2010). Slope
stability analysis showed that the factor of safety
decreased with decrease in water level but no failure
was obtained analyzing the slope at peak strength
(Figure 3) hence, the analysis was based on the
assumption that the strength of dam material had
reached the residual stage due to cyclic loading of water
leading to the movement of the slope. The shear strain
with passage of time was calculated for the soil after
reaching the residual state using both Mohr Coulomb
and Modified Cam Clay Model as shown in Figure 4.
Strain softening behavior leading to progressive failure
on embankment has been successfully determined by
using MCC model (Chai & Carter, 2009) .
Figure 4: Shear Strain with Passage of Time for Pa Bon Dam
(MC and MCC model)
It can be observed that the shear strain was
accumulating with the passage of time and has been
the cause for the reduction of shear strength. This result
resembled to the numerical analysis and laboratory
test conducted for the samples of three gorges dam
(Chen, Zhang, & Chan, 2018). The accumulation of
shear strain was found to be directly proportional to the
height of drawdown. Following the movement each
year, upstream berm were designed to reduce the slope
movement as shown in Figure 2. In a mean time, the
reservoir water level is controlled and maintained at
higher level until the dry season.
3 CANAL SLOPE FAILURE
Lower Chao Phya basin consist of marine clay
deposit called Bangkok soft clay which thickness of
8-12m. These area are suitable for rice production and
are in need of irrigation canal to supply water all year
round. These canals were built systematically in this
area as shown in Figure 5 and later was modified to be
a foundation of road embankment (Figure 6).
Figure 5 : Systematic Canal located at Lower Chao Phya Basin
Statistically, it is observed that every certain year
there is the failure of canal slope. Even though there are

several possible causes of failure, on the basis of
information available, it was hypothesized that the
failure had occurred due to sudden drawdown of water
due to prolong drought. The fluctuation of water level
in canal is presented in Figure 9.
Figure 6: Road Embankment Foundation
Figure 7 : Failure of embankment along Canal Bank
Couple stress-pore water pressure analysis was
carried out and it was determined that the factor of
safety of slope greater than unity at the time of failure
as well. The slope stability analysis in this case too
showed the fluctuation of safety factor with water level
but no failure was obtained analyzing the slope in peak
strength (Figure 9). The shear strain was calculated for
the slope using both MC and MCC model and it
showed that the shear strain was accumulating with
passage of time. The accumulation observed in MCC
model was very significant as compared to MC model.
The movement of the slope was also observed at the
different section of the canal. The behavior of the soil
was tested in centrifuge model (Figure 11) with actual
variance of the water level (Sasingha & Soralump,
2017) . The failure of the slope is shown in Figure 10 .
After the slope movement, test was conducted by using
the pile and without the use of pile. It was found out
that the pile can be used for the stabilizing the slide
slope
Figure 8: Shear Strain with Passage of Time for Pa Bon Dam
(MC and MCC model)
Figure 9 : Figure showing the variation of safety factor with water level with passage of time

Figure 10: Failure of Road Along the canal (Sasingha &
Soralump, 2017)
Figure 11: Photo of centrifuge testing to simulate the effect of
repeated drawdown
4 MITIGATION MEASURES OF STRAIN
ACCUMULATION DUE TO SUDDEN
DRAWDOWN
During the design of slope, cyclic accumulation of
shear strain is not considered as a major concern but
various cases were encountered where movement of the
slope might have occurred due to cyclic drawdown of
water. To prevent the failure of slope due to
accumulated strain from cyclic sudden drawdown, the
following mitigation measures have been are proposed:
1. Increase the required FS during the design for the
case of sudden drawdown. This methods helps in
minimizing the strain movement in each
drawdown cycle.
2. Strengthening the slope can be carried out by the
use of piles (He, Hazarika, Yasufuku, & Han,
2015). This has been successfully done in many
road along the irrigation canal in soft Bangkok
clay area. Short wooden piles were used to
reinforce the slope and reduce accumulation.
Figure 12: Reinforcement of Slope using Short Wooden Piles
5. CONCLUSION
From the above case studies of embankment failure
of canal slope and slope movement of 2 embankment
dam is Thailand, following conclusion can be drawn:
1. Failure can be mitigated by designing the dam
considering the cyclic impact of water during
the sudden drawdown each year
2. Analysis regarding drawdown depth, height of
dam and slope of dam structure requires further
investigation
3. The mitigation measure proposed in Section 4
should be used for preliminary increasing the
shear strength of the slope structure
6. REFERENCES
Chai, J., & Carter, J. P. (2009). Simulation of the progressive
failure of an embankment on soft soil. Computers and
Geotechnics, 36(6), 1024-1038.
doi:10.1016/j.compgeo.2009.03.010
Chen, Y., Zhang, G., & Chan, D. (2018). Cyclic response and
modeling of saturated silty clay due to fluctuations in
reservoir water level of the Three Gorges Dam, China.
Soils and Foundations, 58(3), 702-715.
doi:10.1016/j.sandf.2018.02.023
He, Y., Hazarika, H., Yasufuku, N., & Han, Z. (2015). Evaluating
the effect of slope angle on the distribution of the
soil–pile pressure acting on stabilizing piles in sandy
slopes. Computers and Geotechnics, 69, 153-165.
doi:10.1016/j.compgeo.2015.05.006
Head, K. H. (2011). Manual of Soil Laboratory Testing Volume
2: Permeability, Shear Strength and Compressibility
tests (Third ed.): Whittles Publishing.
Mesri, G., & Huvaj-Sarihan, N. (2012). Residual Shear Strength
Measured by Laboratory Tests and Mobilized in
Landslides. Journal of Geotechnical and
Geoenvironmental Engineering, 138(5), 585-593.
doi:10.1061/(asce)gt.1943-5606.0000624
Sasingha, M., & Soralump, S. (2017). Use of short piles for
stabilizing the side slope of the road embankment
along the canal. International Journal of Geotechnical
and Geological Engineering, 11(2).

Stark, T. D., & Hussain, M. (2010). Shear Strength in Preexisting
Landslides. Journal of Geotechnical and
Geoenvironmental Engineering, 136(7), 957-962.
doi:10.1061/(Asce)Gt.1943-5606.0000308