Rock Slope Stability and Water tightness Assessment of Arjo Didesa Dam site, Western Ethiopia


 Arjo Didesa dam is an earth and rock fill dam which is under construction on the Didesa River in western Ethiopia. However, the dam encountered engineering geological problems that affected the water tightness and slope stability of the dam abutments and reservoir area. To solve these problems, different methods such as discontinuity survey, packer test, sampling and laboratory tests were applied to evaluate the engineering geological conditions of materials along the dam abutments and reservoir area on slope stability and water tightness. Slope stability analyses were performed by kinematic, limit equilibrium and finite element methods using laboratory and in-situ tests. Kinematic analyses revealed that both the left abutment slope 1(LAS1) and the right abutment slope 1(RAS1) sections were unstable for planar mode of failures. Further stability analyses using deterministic methods indicated that both sections were unstable only during saturated conditions. Similarly, stability analysis using limit equilibrium and finite element method revealed that both the left abutment slope 2(LAS2) and the right abutment slope 2(RAS2) sections were unstable under saturated conditions. This study also analyzed the permeability of soils and rock masses and the result showed that the permeability values range from 2.693*10− 7cm/s − 6.687*10− 5cm/s and 3.19*10− 5cm/s − 1.3*10+ 1cm/s for soils and rock masses respectively. Integration of surface hydraulic conductivity and subsurface packer permeability tests showed the presence of potential leakage through the dam abutments. Remedial measures such as rock bolts, anchors and shotcrete for slope stabilization and grouting and clay blanketing are recommended to control this leakage.


Introduction
In a country like Ethiopia having with a great water resource, the construction of dams is very important for development.
Both the federal and regional governments planned to construct large and micro dams in different part of the country including the Arjo Didesa Dam in order to improve food security. However, most of the dam constructed on volcanic and sedimentary terrains in different parts of the country have faced engineering geological problems such as leakage, reservoir siltation, slope instability, damages of spillway and dam body without giving their desired purposes (Haregeweyn et al. 2006;Abdulkadir 2009;and Berhane et al. 2013). Nowadays, leakage of water is a common problem almost in all the dam sites in the world (Mozafari et al. 2011). Leakage problem may lead to water losses from the reservoir and slope instability problem of the dam body and dam abutments. These water losses could occur through the dam body, dam foundation and dam abutments. Moreover, geological structures such as joints and faults are causing excessive leakage of water from the reservoir which leads to failure of the dam (Fell et al. 2005). This scenario was, even more, bigger for the dam constructed in highly weathered and fractured volcanic rocks as in the current study area.
Similarly, slope failure is one of the hazards that lead to the damage of different engineering structures such as dams, tunnels and roads. Slope instability may occur due to the factors such as geological defects, adverse slope geometries, high degree of weathering of slope materials, human activity, heavy precipitation and seismicity which can signi cantly destabilize slopes (Basahel and Mitri 2018). From this perspective, considering the stability of the abutments of a dam is very essential as numerous dams all over the world encountered abutment slope stability problems. Vaiont dam failure, which occurred in Italy in 1963, was a typical example of dam abutment failures due to sliding of rock abutments. As a result of this a huge mass of rock blocks slided into the reservoir which generated a wave overtopping the dam and took about 2500 lives (Genevois and Ghirotti 2005). Sliding of a 5 million m 3 mass into the reservoir of the Pontesei Dam reduced the basin capacity by about 50% (Panizzo et al. 2005). After the rst impoundment of the Three Gorges Dam, a reservoir landslide occurred due to the movement on bedding-plane shears (Wang et al. 2004). In a routinely monitored and inspected Clyde Dam, several large landslides have also occurred along the reservoir margins prior to impoundment (Macfarlane 2009). Slope stability analysis has a long history in geotechnical engineering and engineering geology and recently it is one of the most active areas in both practice and research to prevent this natural disaster (Hussain et al. 2019). Currently, a number of methods are being used for slope stability assessment such as kinematic, limit equilibrium (LE) and nite element (FE) methods that are widely used by many researchers (Raghuvanshi 2017 andShaz et al. 2019).
Kinematic analysis is useful to investigate the possible mode of failure of rock masses without determining the factor of safety (Gischig et al. 2011), while the LE and FE methods analyze the stability of the slopes in terms of factor of safety(FOS) (Kanungo et al. 2013). Besides, in the limit equilibrium method, the factor of safety were calculated for predetermined failure surfaces which divided into a number of slices based on the force and moment equilibrium (Kanungo et al. 2013). However, for FE method doesn't require the shape and location of the failure slip surface (Gri ths and Lane 1999). The FE method preserves global equilibrium until failure is reached and can monitor progressive failure up to and including the overall shear failure (Vinod et al. 2017). Moreover, FEM is the most suitable methods for blocky to very blocky rock mass as its results represents the actual eld conditions (Pain et al. 2014).
The present study is aimed to investigate the engineering geological problems of the Arjo Didesa dam site which is under construction on Didesa River in the volcanic terrain of Southwestern Ethiopian plateau (Fig. 1). However, the dam has encountered engineering geological problems like slope instability and water tightness of the dam abutments and reservoir area respectively. The presence of open joints in dam abutments that is lled by secondary ne materials can possibly trigger the lubrication of materials when impounded water moves along the fracture beneath the reservoir. This will lead to a potential leakage of water from the reservoir as the orientation of joints may favorable this phenomenon.
Thus, in order to understand and solve these water tightness and slope instability problems different methods were followed such as geological and engineering geological mapping, discontinuity surveying, packer permeability test, sampling and laboratory testing which helped to evaluate the conditions of materials along the dam abutments and reservoir area. For slope stability analyses of the dam abutments, kinematic, deterministic and numerical modeling methods were applied.

Geology and Tectonic Setting
In the southwestern Ethiopian plateau, the exposed geological units dominantly consist of the Precambrian basement rocks and associated intrusive rock, Paleozoic sediments, Mesozoic sediments, Tertiary volcanic and quaternary deposits (Tefera et al. 1996 andAlen andTaddesse 2003). The study area is geologically characterized by the presences of Precambrian rocks/basement complex which covers the NW and Central part of Arjo area (Merla et al. 1973) while the Paleozoic-Mesozoic sedimentary rocks and Cenozoic rocks consist of Tertiary volcanic and quaternary super cial deposits that comprised of continental clastic sediments of Getema and Wama Sandstones covering the NE and SE part of Arjo area (Alemu 2014, unpublished research work) and the Tertiary volcanic rocks covered large parts of the area including Dam site that represents Jima Volcanics and Wollega basalt (Merla et al. 1973). The geological structures and tectonic setting of the southwestern Ethiopian Plateau are characterized by the presence of lineaments and faults (Alen and Taddesse 2003). The lineaments of the study area have four general directional trends i.e. (i) N-S (ii) NNE-SSW to NE-SW (iii) NW-SE to WNW-ESE and (iv) ENE-WSW directions while faults of area are represented by high-angle normal faults, low-angle normal slip faults and strike-slip faults (Alemu 2014, unpublished research work).

Methods
In this study, engineering geological assessments were carried out through detail eldwork and laboratory analysis to determine the index and engineering properties of soils and rocks that constitute at the dam site and within reservoir area. The eldwork included discontinuity surveying, mapping, digging test pits, in-situ strength and packer permeability test and sampling. Discontinuity survey includes describing discontinuity orientation, spacing, roughness, persistence, aperture and lling material based on the standard suggested by ISRM (1981). Discontinuity data were used to determine the hydraulic conductivity of the surface rock masses and for slope stability analysis. The laboratory test on the soil samples include natural moisture content, speci c gravity, particle size analysis, liquid limit test and permeability tests while the laboratory tests conducted on rock samples are point load and unit weight tests. Besides, the in-situ strength of rocks was also determined using the Schmidt hammer on selected slope sections according to the standard suggested by ASTM (2001). Accordingly, the point load (Broch and Franklin 1972) and in-situ strength Schmidt hammer test (Aydin 2009) were used to determine the uniaxial compressive strength of intact rocks. Rock Quality Designation (RQD) and Geological Strength Index (GSI) assessments were also carried out in the eldwork to determine the quality of rocks at dam site. GSI was estimated based on basic GSI charts proposed by Marinos and Hoek (2000) whereas the RQD has been estimated from joint volumetric count (JV) as follows using eq. 1 as suggested by Palmstrom (1982). RQD = 115 -3.3Jv ……………………………………………………………………….(1) Furthermore, slope stability analyses of the dam abutments were conducted using the Kinematic, limit equilibrium and nite element methods. The kinematic analyses were performed using Dips 6 software  for the structural controlled slope sections i.e. left abutment slope 1(LAS1) and right abutment slope 1(RAS1) sections which are lying on the left and right abutments respectively (Fig. 2) and further stability analysis for these particular slope sections was carried out in terms of factor of safety (FOS) by deterministic method using RocPlane software . Similarly, stability analysis of highly weathered and fractured (non-structurally controlled) slope sections i.e. left abutment slope 2(LAS2) and right abutment 2(RAS2) slope sections (Fig. 2 ) was done by limit equilibrium method (LEM) and nite element method (FEM) using Slide v6.0  and Phase 2 v8.0 software respectively in terms of FOS.
Moreover, water tightness analysis was performed through laboratory permeability tests on soils, packer permeability tests and hydraulic conductivity estimation from the aperture and spacing of fractures from the surface of rock mass using eq. 2 of Snow (1965) and Louis (1974) by assuming that the fractures are parallel and smooth.
Where g is the gravitational acceleration (9.81m/s 2 ), e is the average aperture of fractures (m), μ is the coe cient of kinematic viscosity of the uid (10 -6 m 2 /s for water at 20 c) and λ is the fracture frequency. The general methodological framework of the study is shown in Fig. 2 as follows.

Geology of the Study Area
The lithological units in the study area consist of rhyolite, basalt, alluvial soil, residual soil and colluvium deposit (Fig. 3). The rhyolite rock is exposed at dam abutments and in the reservoir area which is characterized by high degree of weathering and fracturing and porphyritic in texture. In this rock unit, the joints were partially lled with ne soils and soft gouge. The basaltic rock is the dominant rocks in the study area which is characterized by different degree of weathering from highly to slightly weathered and aphanitic in textures, dark to grey in color. The colluvial deposit is deposited under gravity force along the dam abutment slopes. It is a mixture of an extremely wide range of soil types which ranges from ne to coarse-grained soils. These materials are likely to be highly permeable. Furthermore, the reservoir area is dominantly covered by alluvial and residual soils. Alluvial and residual soils are covering a large part of Didesa plain which is part of the reservoir area and is characterized by mixtures of major soil types from silty clay to coarse sands and gravels. As observed from in-situ test pits and drilled boreholes along the dam axis, the thickness of this soil ranges from 2 to 5m.

Discontinuity Surveying
The discontinuity survey analysis revealed that majority of the joint sets were having larger continuity, vertical to subvertical orientation and the aperture and spacing of discontinuity varied from 0.1cm to 0.7cm and 1-120cm respectively (Table 1). They are characterized by slightly rough to smooth discontinuities and moderately to highly weathered ones as per ISRM (1981) ( Table 1). Majority of the joints have favorable orientation for leakage as these discontinuities crossed the dam axis nearly parallel to the direction of river ow. As shown in Fig. 4 in the left abutment, JS1 has a very favorable orientation for leakage whereas in the right abutment, JS3 has a very favorable orientation for leakage as these are oriented parallel to the direction of river ow. Therefore, these conditions will lead to a potential leakage of water from the reservoir area. Hence, special treatment is required at the dam abutments during the construction stage in order to control this leakage.

Rock mass Classi cation
In the present study the two most widely used rock mass classi cation methods, namely rock mass rating(RMR) (Bieniawski, 1989) and geological strength index(GSI) (Marinos and Hoek, 2000) have been used to characterize the rock mass and estimate the rock mass quality at the left and right abutments of the dam. The RMR system consists of ve basic parameters that represent different conditions of rock mass and discontinuities (Table 1). Based on RMR, the rock mass quality falls in the range of poor to good quality rocks with RMR values ranging from 33 to 63 (Table 1) while in the case of GSI, the rock mass quality was classi ed in the range of disintegrated to very blocky with GSI values ranging from 15 to 65 as estimated in the eld (Table 2).

Rock Mass Strength and Deformation parameters
In this study RocLab software (Rocscience 2002) was used to determine the geomechanical properties of the rock mass for non-structurally controlled slope sections. The input parameters used in RocLab software were given in Table 3. The GSI values determined in the eld (Table 2) and disturbance factor(D) values taken from (Torres and Brent 2002) were used in the slope stability analysis. Similarly, the material constant of the intact rock (mi) was also taken from (Rocsience 2002) and for rhyolite rock, the value of mi is considered to be 25. The Young's modulus(Erm) was determined using eq. 4 (Hoek and Diederichs 2006). For isotropic rocks, in some rock engineering applications with limited eld data, a value between 0.2 and 0.3 is a common estimate for Poisson's ratio (Gercek 2007). Thus, for this study its value was assumed to be 0.3. The intact modulus (E i ) was determined using eq. 3 and the modulus ratio (MR) value of the rhyolite rock was assumed to be 400 as proposed by Deere (1968). This relationship is useful when measurement of Ei is di cult for completely undisturbed rock samples. The estimated geo-mechanical properties of the rock masses were presented in Table 4. Similarly, for structurally controlled slope sections, the shear strength parameters of the rock mass along the discontinuity failure planes were determined from RocData software   (Table 5). The joint wall compressive strength (JCS) corresponding to 25% of uniaxial compressive strength of intact rock was determined from Schmidt hammer tests whereas the joint roughness pro le (JRC) value was determined visually in the eldwork by comparing the appearance of the discontinuity surface with the standard developed by Barton and Choubey (1977). The values of the basic friction angle ( b ) for most of the smooth and unweathered rock surfaces range between 25 0 and 35 0 (Barton, 1973). Hence in this study, the basic friction angle of 30 0 was considered.

Kinematic Analysis
Discontinuity and slope orientation that were used in the Dips software for kinematic analysis were measured in the eld (Table 6) whereas the friction angle along the failure plane was estimated in RocData software ( Table 6). The kinematic analyses revealed that planar mode of failure occurred due to the intersection of joint set 2 (JS2) and Joint set 3 (JS3) at LAS1 (Fig. 5a) while at RAS1, planar mode of failure occurred due to joint set 3 (JS3) (Fig. 5b).

Stability Analysis using Deterministic Method
The stability analysis for the planar mode of failure was performed by deterministic method using RocPlane software (Rocscience 2002) in terms of FOS (Park and West 2001). The input parameters used in RocPlane software were given in Table 7. The horizontal seismic acceleration of the study area from Johnson and Degraff (1991) was estimated to be 0.01g. This was used in this analysis. Where γ is the unit weight of the rock, w is waviness, is friction angle, c is cohesion, u w is water pressure, α is the seismic coe cient FOS was calculated under different anticipated conditions i.e. static condition(without earthquake loading) and dynamic condition(with earthquake loading) by varying water pressure level, percent lled in the joints i.e. dry condition when 0% is lled, a moderate condition when 50% is lled and fully saturated condition when 100% is lled (Table 8). The analysis result revealed that at both LAS1 and RAS1 slope sections, the FOS is less than one only during saturated conditions (Table 8) and the slope was found to be unstable (Wyllie and Norrish 1996) (Fig. 6). This indicated that pore water pressure signi cantly contributed to destabilize these particular slope sections as compared to seismic activity as the study area falls in low seismic risk zone (Asfaw 1986). In addition, the eld the geometry of the slopes and orientation of discontinuity also played a greater role in destabilizing these particular slope sections.

Stability Analysis by LEM
LE and FE method is the most suitable for the analysis of highly weathered and fractured slope materials (Mahboubi et al. 2008). Hence, the slope stability analysis of highly weathered and fractured slope sections of the left abutment slope 2(LAS2) and right abutment slope 2(RAS2) were carried out using LE and FE methods in terms of Factor of Safety. The LE methods are based on force and moment equilibrium while the FE methods use the stress-strain relationships to determine the behavior of the model (Memon 2018). Slope stability analysis using LEM was performed using GLE/Morgenstern-Price/ method using Slide v6.0 software (Fig. 7). Hence, FOS was calculated under different anticipated conditions i.e. before impoundment of water in the reservoir (dry condition) and after impoundment of water in the reservoir (saturated condition) under both static and dynamic conditions (Table 10). The input parameters used in slide software were given in Table 9. The shear strength parameters (cohesion and friction angle) of rocks that were used in the analysis were given in Table 5 whereas the geometry of the slope was measured in the eld. Where C is Cohesion, is Friction angle, γ is the unit weight of the rock, α is Horizontal seismic acceleration and HWR, MWR and SWR represent highly weathered rhyolite, moderately weathered rhyolite and slightly weathered rhyolite respectively.
The analysis results revealed that at both LAS2 and RAS2 slope sections, the FOS is less than one and the slope was found to be unstable during saturated conditions and stable under dry conditions (Table 10) (Wyllie and Norrish 1996). This indicated that the pore water pressure signi cantly contributed to the instability of these particular slope sections as compared to horizontal seismic acceleration. The other factor affecting the stability of these slope sections was the geometry of the slopes that needs to reduce the angle of these slopes for a better slope stability.

Slope Stability Analysis by FEM
Slope stability analysis of LAS2 and RAS2 slope sections were further evaluated using the nite element method (FEM) in terms of the critical strength reduction factor (SRF) to con rm the stability conditions of these slope sections. Hence, the critical SRF was calculated using Pahse2 V8.0 software under different anticipated conditions (Figs. 8 and 9). The shear strength reduction option in Phase2 software was used to perform a nite element slope stability analysis and compute a critical SRF. The input parameters used in Phase2 were given in Table 11. Where C is Cohesion, is Friction angle, γ is the unit weight of the rock, α is Horizontal seismic acceleration and HWR, MWR and SWR represent highly weathered rhyolite, moderately weathered rhyolite, and slightly weathered rhyolite respectively.
The analyses result revealed that when both slope sections are at static and dynamic saturated conditions, the critical SRF is less than one and the slopes were unstable while when the slopes are in dry conditions, the critical SRF was greater than one and the slope was stable according to Wyllie and Norrish (1996) (Figs. 8 and 9). This showed that the effect of water pressure plays a greater role in destabilizing these particular slope sections as compared to horizontal seismic acceleration.
When the results of the FE analyses were compared with the result obtained from the LE method, the FOS results determined from the FEM were in very good agreement with those computed by the LEM (Fig. 10). Furthermore, slip failure surfaces were almost the same for both techniques. Besides, the analyses results revealed that the slope sections of the left and right abutments of the dam are unstable only during saturated conditions. Several researchers have previously compared the results of slope stability analyses using the LE and FE methods (Cheng et al. 2007 andVinod et al. 2017).
They concluded that the results of both methods were generally in a good agreement for homogenous slopes. But, the FE method is more advanced and valuable than the LE method in case of slope modeling and design with heterogeneous materials, complex slope geometry and different loading patterns. In general, the present study concluded that the results of both FE and LE methods were in good agreement under different anticipated conditions when simple slope geometry and heterogeneous material were considered.

Engineering Characterization Of Soils
The grain size analysis result revealed that the dominant soil types found in the study area are identi ed as ne-grained soils (Silty-clay soils). Similarly, the atterberg (liquid limit) test analysis result revealed that the soils in the area contains an average liquid limit of more than 50% and have a plasticity index ranging from 17.18-44% (Table 13) and classi ed as high plasticity soils (Ramana 1993). Moreover, natural moisture content and the speci c gravity of the soils range from 25.18-41.96% and 2.54 to 3.3 respectively. According to USCS, the soil samples from the reservoir area fall in MH/OH and CH/OH region which implies clayey-silty soil and silty clay soil respectively (Table 12).

Permeability of Soils
The result of permeability tests in the laboratory revealed that the coe cient of the permeability of soils in the study area ranges from 2.693*10 − 7 cm/s to 6.687*10 − 5 cm/s (Table 13) which can be classi ed as impervious to semi-pervious permeability respectively (USBR 1987). It can be concluded that the permeability of the soils in the reservoir site is relatively small and watertight. The permeability of rock mass is mainly affected by the discontinuity characteristics such as orientation, spacing, conditions of discontinuity and the stress applied on the fractures (Berhane et al. 2013). The permeability of rock masses was estimated using Eq. 2 and packer test for surface and sub-surface rock mass respectively.
Accordingly, for surface rock mass the results showed that the hydraulic conductivity of rock mass ranges from 2.9*10 − 4 m/s to 1.3*10 − 1 m/s and 1.3*10 − 4 m/s to 1.2*10 − 2 m/s with permeability class ranging from moderate to high permeable classes at dam abutments and in the reservoir area respectively (Bell 2007) and the distribution of permeability of surface rock mass at dam abutments and in the reservoir area were shown in Fig. 11. This implies that the rock mass in the area is affected by the high degree of fracturing and weathering. Besides, in the abutments of the dam, there are widely opened joints lled by secondary ne materials particularly in the left abutment while in the right abutment some fractures are not lled and are interconnected to each other. Therefore, there will be a signi cant amount of water leaking through these openings and serious water losses from the reservoir will occur through the dam abutments after the water is impounded in the reservoir.
Similarly, the subsurface exploration of the Arjo Didesa dam site was conducted by drilling boreholes to determine the geotechnical properties of the subsurface rock mass. This drilling result showed that the subsurface rock mass that is found at the dam site was characterized by different degree of weathering and fracturing. However, the degree of weathering is relatively decreasing as depth increases (Fig. 12). Packer tests were conducted by Oromia Water Works Design and Supervision Enterprise in 11 drilled boreholes on the dam abutments and at the center of the river in order to determine the permeability of subsurface rock mass. The depth of these boreholes varies from 20m to 92m. The test section interval was varied from 2.5 to 5m and it depends on the condition of the rock mass with depth. Furthermore, the permeability of a rock mass along the dam axis was interpreted and classi ed based on Lugeon values (Fell et al. 2005) ( Fig. 15) (Table 14). This suggests that the presence of highly permeable and open joints in the rock mass which is classi ed as permeable rock (Fell et al. 2005).
Many researchers stated that as the overburden increases with depth, the permeability of the rock mass decreases (Lee andFarmer 1993 andBerhane 2013). Similarly, the permeability of the rock mass decreases as depth increases (Figs. 13 and 15).
The ow type can be determined from the packer test and it is used to determine the characteristics of discontinuities and grouting types. Thus, at the left abutment, the laminar ow is the dominant ow type (> 40%) which is followed by the dilation ow type (> 23%) and the less dominant ones include turbulent, void lling and washout ow types (Fig. 14). The dominated laminar ow in the left abutment indicated the presence of a parallel joint opening with low lugeon values (Rozo, 1933). On the other hand, turbulent ow type was observed most frequently at the center of the river (Fig. 14) which shows the presence of either open joints with high discharge or tight joints with low discharge (Ajalloeian 2013). Similarly, in the right abutment void lling is the dominant ow type (> 28%) followed by dilation (> 21%) and washout (> 21%) and less dominant ones such as laminar (14%) and turbulent ow (14%) (Fig. 14). The void lling behavior indicated that the presence of the larger joint apertures and during the test, water is progressively lling the open discontinuities or swelling occurs in the discontinuities (Houlsby 1976). On the basis of this study, the following conclusions and recommendations can be made.
The current study assessed the geological and geotechnical conditions of materials with respect to slope stability and water tightness along the dam abutments and reservoir area. Detailed eldwork and laboratory analysis was conducted on representative samples to determine the index and engineering properties of the soils and rocks.
Slope stability assessments of the dam abutments were carried out through different approaches such as the Kinematic, limit equilibrium and nite element method. Kinematic analysis revealed that at both LAS1 and RAS1 slope sections were potentially unstable due to planar mode of failures. Further stability analysis using the deterministic method revealed that both LAS1 and RAS1 slope sections were found to be unstable in saturated condition which develops pore water pressure that signi cantly contributed to destabilize these slope sections.
Similarly, stability analyses of LAS2 and RAS2 slope sections were performed using FE and LE methods in different conditions. The analyses result showed that both slope sections were unstable only in saturated conditions as pore water pressure can be developed to destabilize these slope sections. The FOS calculated from FE method was slightly smaller than that was calculated by LE method. Thus, it could be concluded that the results of both FE and LE methods were in good agreement at different anticipated conditions when simple slope geometry and heterogeneous materials are considered.
The surface discontinuity survey and packer test results revealed that permeability of the rock masses at the dam abutments range from low to high permeability classes. Thus, excessive leakage will be expected after the reservoir is impounded especially through the dam abutments through fractured rocks.
The soil permeability tests in the study area range from impervious to semi pervious ones. Thus, the leakage path is long and the potential leakage from the reservoir is not expected after the nal lling of the reservoir but during the initial reservoir lling.
Rock bolt and anchors are recommended as remedial measures to prevent planar sliding and stabilize slopes. For LAS2 and RAS2 slope sections, the failed material should be rst removed to reduce the slope angle and then shotcrete should be applied to increase the stability of the slope by providing su cient drain holes to remove pore water pressure that is building underneath the shotcrete surface. Grouting is recommended at both abutments to prevent the leakage problems when larger open joints are present. This study strongly recommends further assessment to be done for the design and installation of different remedial measures and the effect of rapid drawdown of water on slope instability of the dam abutments. Declarations