Geotechnical Study of Unstable Slopes: A Case Study at Sunkoshi Power-

house Site, Central Nepal

By: - G.S Pokharel
Despite having enormous hydropower potential in our country, various
geological problems frequently arise, therefore, a case study of Sunkoshi Power
station owned by NEA has been discussed in the paper which deals with a specific
problem of slope instability. The Sunkoshi Power station was prone to natural
hazards before its construction; for instance, an outburst of Zhanzangho glacial lake
in 1964 in Tibet and after utility completion in 1971, it was affected six major natural
disasters-(a) In 1976, by a debris flow at the slope no. 2 of Sunkoshi Powerhouse,
(b) In 1981, by a high intensity flood in Bhotekoshi due to outburst of Zhanzangho
glacial lake, (c) In 1982, by damming of Balephi Khola near Phalamesangu thereby
causing LDOF in Bhotekoshi (d) In 1986, by a GLOF in Sunkoshi, (e) In 1995, by
a debris flow in slope no. 2 of Sunkoshi Powerhouse, (f) In 1996, by an intensive
debris flow in Larcha Khola.
The paper deals with the event of July 5 1995. A large failure in the form of
mass wasting process occurred in Slope no. 2 of the Sunkoshi Powerhouse area early
in the afternoon of July 1995, similar to that in 1976. The slope is west facing slope
of about avg. 25.5 degrees and stretches from elevation of 790m to about 1,100 m
and a length of about 630m. The slope angle is gentler in the middle and steepest at
the top. The ground water table depth varied from 29.5 m in the dry season to 10m
in the wet season around August & September. A series of check dams were
constructed across the toe of the slope after the debris torrent breaching the power
canal in 1976, and stone masonry walls along this flow channel, which were
displaced by up to 25 cm. After the first study (1977-78), 4 piezometers were
installed in the lower & middle parts of the slope drilled up to colluvium in 1979 to
monitor the groundwater table. Piezometer P2 remained clogged at the very

Prepared By: Pareekshit Poudel Page 1

beginning, P3 remained dry for the whole monitoring period, P1 & P4 remained
mostly dry and water table rose above the bottom only during the wet season, & P1
located on the bottom most part of the slope recorded the highest water table around
the beginning of September. Fifteen reference monuments marked as A1, A2, etc.
were installed to monitor movement at different parts of the slope, which showed
total displacements varying from 0.09 m at A1 to 1.32 m at B1 over a period of 14
years; the horizontal displacement of the slopes occurred during the time of rising of
water level in the piezometers.
NEA’s SRC Lab, (1990, assisted by JICA, HMGN’s Water Induced DPTC)
investigated Slope no. 2 with the aim of trying to formulate a permanent solution
and devise measures to stabilize the slope by addressing the following: (a) The nature
of instability blocks (single or multiple), (b) The causes of natural mass wasting, (c)
Immediate and long term measures for stabilizing the slope. The study of geological
map of the area showed that the unstable part of the slope was composed of
colluvium which was made of fragments of phyllites of various sizes & shapes in
different states of weathering. The finer material was represented by silt & sand size
particles derived from the weathering of phyllites. This soil is greatly affected by the
seasonal rise of the groundwater table in the monsoon. The rise in the water table
not only lubricated the slip surface but also caused the shear strength of the soil to
decrease resulting in a decrease in slope stability.
Based on distinct physical characteristics like presence of gullies, topography,
prominence of a type of failure mechanism etc., slope was divided into separate
segments and unstable segments were investigated by exploratory drilling,
piezometers were installed in the drilled holes to monitor the water levels &
reference piles were installed all over the slopes to monitor and record slope
movement. The areas with the most relative movement were investigated in detail
to try and arrive at engineering solution.

Prepared By: Pareekshit Poudel Page 2

 The slope was divided into three blocks A (oldest landslide at the lowermost
end of slope, subdivided into A1 & A minus A1; movements were observed in Block
A1 since 1972 as indicated by horizontal displacement (about 25 cm) of the masonry
drainage walls and in block A minus A1 since 1992 in the form of progressing cracks
along the drainage channel which lies along the toe of the slope between the slope
and the power canal. ), B (central part of the slope; no movement) and C (uppermost
and the steepest part of the slope; at present, it contains two increasingly unstable
areas).
A seismic profile was made in 1991 & first stability analysis was carried out.
Two holes were drilled at the depth of 36 m at the top of blocks A1 and A minus A1
and piezometers were installed in each of these holes. RCC piles were also installed
in rows to monitor the slope movement along with the water levels.
Short term countermeasures were recommended to cope with the immediate
threat like, in Block A to:-(a) Stop cultivation and especially paddy cultivation at the
top of block A1, (b) Treat upper part of the slope with red clayey soil so as to reduce
infiltration of water, (c) Extend and improve the drainage system to prevent water
logging, (d) Install five horizontal drainage holes up to the lowest water level to
control the rise in the groundwater table during the monsoon season. Similarly,
measures recommended in unstable areas of Block C, are to: (a) Construct a drainage
channel along the upper part of instability to prevent surface flow in the exposed
soil, (b) Design a system of RCC drains to minimize the flow of water through the
exposed ground in order to minimize soil loss, (d) Implement bio-engineering
measures in areas between the drainage channels, (e) Cover the exposed ground
surfaces with suitable geo-textiles or gunny bags.
The preliminary and final stability analysis and overall investigations
concluded that movement of Block A1 and A-A1 are separate phenomena occurring
at different times with different failure mechanisms. Block B was found stable

Prepared By: Pareekshit Poudel Page 3

because no movement was recorded from the segment. Block C was found to have
tension cracks in upper surface that could develop into a potential slip surface.
There were no evidence that suggested the instability of the slope as a single
phenomenon affecting the slope, rather it seemed that the slope composed of a
number of local instabilities in Blocks A and C.
As a treatment measure of the above instable blocks apart from the temporary
measures given above, following were also proposed:
For Block A1:
 Apply bio-engineering to prevent of control soil loss.
 Continue to monitor the movement of each block as stabilization of Block A1
may lead to stabilization of block A-A1.
For Block C:
 Carry out bioengineering in the areas between the drainage channels.
 Cover and fill the tension cracks in the vicinity of the instabilities with red
clay to prevent infiltration of water into the cracks.
 Construct a retaining wall with horizontal drains along the upper part of the
landslide to control slumping of soil mass weakened by moisture during the
wet season.

Prepared By: Pareekshit Poudel Page 4