Evaluation of Carbonate and Heterogenous Rock Masses for the Dam Foundation: A Case Study at Kanarwe River Basin, Sulaimaniyah, NE Iraq

Abstract


Introduction
Over the past few decades, climate change and population in a semi-arid region, especially in the middle east, have caused water shortages and an increasing number of constructing dams (Mohammed et al., 2021). Recently the Kurdistan regional government decided to build forty dams in the North of Iraq. The dam site selection construction of forty dams is not easy, and it needs a detailed study of geology, hydrology, metrological, satellite image and accurate geomechanical classification of a rock layer (Ajibade et al., 2020;Hoek et al., 2005;Hoek and Brown, 2019;Khaleghi Esfahani et al., 2018). Additionally, the geomechanical classification of some sites in heterogeneous rocks is problematic because it is composed of two or more lithologies with different mechanical properties (Cai et al., 2004;Marinos and Hoek, 2000;Marinos, 2019;Marinos et al., 2006). The dam's construction on heterogeneous rocks, like this study (Fig. 1), is rugged because it combines weak and strong rock layers (Behnia et al., 2018;Marinos, 2019). Calculation of mechanical properties for weak and strong rock layer need expertise, accuracy, and good fieldwork to select an appropriate value for uniaxial compressive strength (UCS), material constant(mi), and GSI (Laubscher, 1977;Marinos and Hoek, 2001;Marinos, 2019;Pepe et al., 2015). The study of rock masses for dam construction in north Iraq is very limited because of domestic problems and economic crises, so it mainly focused on assessing slope failure (Al-Jawadi, 2021;Hostani and Hamasur, 2022). The purpose of this study is to evaluate and classify rock masses in Goma-Qazan proposed site for dam construction according to different geomechanical classification systems, including Rock Mass Rating (RMR), GSI, and DMR. In addition, a new procedure was suggested for the calculation of stress relaxation and damage level (disturbance factor) in rock masses. Finally, suggest convenient dam types such as earth-fill, hardfill, gravity, and arch dams for the proposed site.

Location
The Kanarwe river basin (KRB) watershed is about 1541km 2 within Little Zab (LZ) river basin (Fig. 2). The LZ is covered more than 20000 Km 2 , where 15000km 2 (80%) of it located with the international border of Iraq and of 5000km 2 (20%) of it located with the international border of Iran (Abbas et al., 2017;Saeedrashed and Guven, 2013). The Kanarwe river passes through mountainous terrain, carrying a maximum water 42m 3 /sec flow during March and 0.029m 3 /sec flow in September. The basin is formed from two sub-basin, the Shiler sub-basin and the Qzilja Sub-basin. The river attribute in both subbasins merged near Suraban Village to form the kanarwe river, then flowed and zigzagging to the proposed dam site in the west.   (English et al., 2015)showing the location of the study area

Rock Mass Classification
The rock mass was evaluated at the proposed dam site based on four approaches: a geomechanical, hydrological, structural perspective, and reservoir filling by soil erosion (Fig. 5). The geomechanical evaluation is the most critical part of this article, including a discontinuity survey, Drone survey, laboratory test, and applying different geomechanical classification systems (Table 1).

Discontinuity Survey
First, the detailed discontinuity survey and accurate field observation, including a quantitative description of joint orientation, persistence, and filling aperture, were done at the site to determine block volume and the mechanical properties of rock. A total of 100 discontinuity measurement was taken from the left and right banks of hard carbonate and heterogeneous clastic rock units. However, the weathered surface in some heterogeneous rock cause restriction or difficult measurements of discontinuity attitude on such lithology. Secondly, the discontinuity measurements are processed and analyzed with computer software based on equal-area stereographic projection, called DIPS V6.00. Finally, the volumetric joint count(JV)was calculated based on discontinuity measurements using the (Palmstrom, 2005) method for calculation volumetric joint count(Vb), average spacing of all discontinuity and rock quality designation(RQD) ( Table 2).    ,4,6,9,11,13,15,17,19, and 21 are heterogeneous clastic; the mentioned characteristic cannot be calculated for volumetric joint counting.

Rock Mass Strength
The rock description was done systematically along the left and right banks based on lithologic change. The rock mass and discontinuity characterizations were based on RMR78 and RMR89 (Bienawski, 1976;Bieniawski, 1989). The geological strength index values for carbonate units were determined based on Hamasur (2009) GSI chart (Fig. 6); for flysch units, Marinos (2019) chart was used (Fig. 7). The dam foundation stability and deformability were evaluated based on DMR system invented by Romana in (2004).

Point load test
Uniaxial compressive strength (UCS)was determined indirectly from the point load test, according to the procedure of ISRM (1985). Twenty-two units (three specimens per unit), including the carbonate unit and flysch unit (Table 3 and Table 4).

Petrographic study
The petrographic classification of rocks is beneficial for determining depositional facies and rock compositions. Also, it is used to calculate the material constant (mi) and elements of Hoek and Brown for rocks (He et al., 2020). The studied units are composed of an interfingering sequence of rocks characterized by a high rate of sedimentation composed of Rudist-rich limestone, detrital limestone of reefal facies of Aqra Formation and Sandstone, Siltstone layers of deep-flysch facies of Tanjero Formation (Al-Dulaimi and Sa'ad, 2015;Lawa et al., 2017).

Rock Mass Rating
The geomechanical classification for massive carbonate rock was performed according to (Bienawski, 1976;Bieniawski 1989). This geomechanical classification depends on five parameters: uniaxial compressive strength of intact rocks (UCS), rock quality designation (RQD), discontinuity surface condition, groundwater condition, and discontinuity orientation ( Table 2, Table 3), and (Table  4); then these parameters were rated to determine the value of RMR as in (Table 5). Later the RMR values were used for the calculation of Dam mass rating DMR, DMRSAT related to the dam stability against sliding, deformation modulus of rock mass (Ec), and DMRDEF Dam mass rating related to the deformability of foundation. RMR calculated based on (Table 3 and Table 4)were used, and the RMR value ranged from 68-79 based on (Bienawski, 1976), while based on (Bieniawski, 1989), the RMR show value ranged from 66-74 (Table 5).  (1976 or 1989)

Geological Strength Index (GSI)
Hoek first invented the geological strength index in (1994), then updated by (Hoek, 1994;Hoek et al., 1995;Hoek and Brown, 1997). This geomechanical classification of rock mass is based on the structural condition of the rock mass (joint set relation), the surface condition of the joint, and visual interpretation. This classification was later updated and modified to become a quantitative chart instead of a visual plotting (Cai et al., 2004;Hamasur, 2009;Hoek et al., 2013;Hoek and Brown, 2019;Renani et al., 2019;Sonmez and Ulusay, 2002). This study prefers the quantitative GSI chart classification invented by Hamasur (2009) for the carbonate unit because its the most updated GSI chart related to rock mass classification in Iraq. The heterogeneous units were classified based on the most updated chart invented by Marinos ( 2019). The GSI value for carbonate rocks ranges from 74 -76 (Table 6), while heterogeneous rock units range from 35-55 (Table 3). Finally, based on the GSI values of the rock mass units, a clear contrast in rock strength parameters between carbonate and flysch rocks at the site was found. These results will affect the suitability and stability of the foundation for dam construction.

Excavation and Disturbance Factor
Calculating rock excavatability significantly controls the rock mass's strength properties after excavation. To determine which types of excavability methods can be applied at the site for removing rocks at the dam foundations and embankments, Pettifer and Fookes's (1994) chart was applied. These methods are constructive for determining the value disturbance factor(D), which is used for the estimation of blasting damage level and stress relaxation during excavation at a dam site (Hoek et al., 2002;Hoek and Diederichs, 2006) E m (GPa)=(1-D/2)x(σc/ 100) 1/2 x10 (GSI-10)/40 (Hoek et al., 2002) (1) The new factor D "allows for the effects of blast damage and stress relaxation". The D factor can be estimated according to guidelines for tunnels, slopes, and pit quarries but not clarified for dam foundations, which is limited to the application of GSI classification and its program(RocLab program) for dam foundations. Excavations at dam foundations must be done carefully, D should be very low, but it cannot be zero because of decompression. Tentative guidelines for determining D value were invented in 2003 by Romana (Table 7), but these guidelines contain some ambiguities about types of rock, such as the term "good rock…..etc.".
This study uses Pettifer and Fookes (1994) and (Franklin et al., 1971) classification for determining the disturbance factor for rock mass based on the excavatability method for rock types and overcomes previous ambiguities in tentative guideline Romana (2003), RocLab software 1.031, and GSI chart Hoek et al., (2002) for determining D value. These methods represent the 3D dimensional relation between discontinuity spacing and point load strength index Is50, providing a more realistic assessment to determine the D value.
The spacing of discontinuities and point load strength index 1s50 results were plotted on the Pettifer and Fookes chart (Fig. 10) to determine the excavation types of each rock mass unit. The plotted results show that extra hard ripping to blasting is required for excavation in massive carbonate rocks. In contrast, easy ripping excavation is required to excavate in flysch heterogeneous rock. The excavation will be led to the di-stabilization rock slope and dam foundation.  Table 7. Rock disturbance calculation based on excavation types (Romana et al., 2003) Rock mass description (Romana et al., 2003) Excavability types based on Pettifer and Fookes's (1994)  For this study, the value of 0.4 is applied in the RocLab software for massive limestone of the Aqra Formation, while the value of 0.2 is applied for poor rock mass or flysch rock of Tanjero Formation based on excavabilty condition (spacing vs Is50)by Pettifer and Fookes (1994) chart (Fig.10)   Fig.9. Assessment of excavatability of rock at the dam site for this study using the (Pettifer and Fookes, 1994)chart

Dam Mass Ratings
The dam mass rating is a new geomechanical classification used to evaluate dam foundations. It is an event by Romana (2003) as an adaptation for RMR classification. The RMR classification is inapplicable and challenging for the evaluation of dam foundations because it is limited only to rock mass descriptions in tunnels, slopes, and foundations (Romana, 2004). The DMR is applicable for "overall stability against sliding, needed depth of excavation for the foundations, consolidation grouting treatment, possible consequences of excessive relationship-following Rocha concepts-between the deformation modulus of the dam and the rock mass foundation" (Romana et al., 2003). Another problem in RMR is water rating which is unique and not varied in the calculation of RMRBD. However, in DMR, the water rating(WR) varies depending on water pressure(ru) which is controlled by valley geometry, the water level, and the efficiency of the grouting curtains (if exist). The WR can be found by following the equation: WR = 10log (1/ru) -1.5 (valid for 0.02 < ru < 0,7) (2) Were ru is the water pressure and found by: ru = u/σv. (3) where u is the water pressure and σv is the total vertical stress. Finally, The WR rating value (Table 8)is applied in RMRBD for determining water rating (fifth parameter). Thus the DMR geomechanical classification becomes very crucial in the evaluation of dam foundation and its work on two concepts; the first one is the relative deformability of rock mass DMRDEF related to the strength of rock, and the second one is the stability of the dam foundation DMRSTA During the filling of the reservoir. Table 8. Relation between Water rating(WR)and water pressure (ru) (Romana et al., 2003) WR 15 10 7 4 0 ru(bineniawski) 0 0-0.1 0.1-0.2 0.2-0.5 >0.5 ru(formula 2) 0.002 0.07 0.14 0.28 0.7

Stability of the dam foundation against sliding
Based on DMR geomechanical classification, the stability of the dam foundation against sliding (DMRSTA) depends on three factors: (RSTA), the significative governing joint orientation, and dam axis orientation. It can be calculated from the following equation: DMRSAT= RMRBD + CF* RSTA.
(4) DMRSAT is the dam foundation stability against sliding, RMRBD is the basic dry rock mass rating, CF geometric correction factor is calculated based on (Eq 5), and RSTA is the adjusting rating factor for the dam stability CF=(1-Sin (5) CF=(1-sin |αd -αj|)2, where αd is the direction upstream-downstream of the dam axis, and αj is the dip direction of the significative governing discontinuity (in this study is bedding planes). Based on the above equations, the rock mass units at the dam foundation were evaluated. The values of RMR BD were taken from (Table 5), CF from (Table 11), and R STA from (Table 12), and the effect upstream downstream of the dam axis was analyzed by Dip software (Fig. 11). Table 9. Degree of dam sliding safety and DMRSAT (Romana et al., 2003) DMRSAT Less than 30 30-60 Greater than 60 Degree of safety Serious concern Concern No primer concern Table 10. Provisional procedures for dam foundation excavation and consolidation grouting (Romana, 2004) (+) minimum (desirable) -gravity dams include (CVC, RCC and hard-fill concrete) -rockfill dams included are the ones sensible to settlement (with concrete -CFRDor asphaltic -AFRD -face upstream.  Table 12. Rating of the adjusting factors for the dam stability, RSTA, according to joints orientation (Romana, 2004) DS (dip downstream), US (dip upstream), A (any dip direction), Gravity dams include CVC (Conventional vibrated concrete), RCC (Roller compacted concrete), and hard-fill concrete dams.
For the calculation of DMRSTA, each unit was evaluated separately based on RMR basic dry from (Table 5), The RSTA factors for each kind of dam and CF (Geometric factor) ( Table 13). DMRSAT results for each unit are correlated with dam stability against the stability safety (Table 9) to determine foundation stability problems at the dam site. The carbonate rock mass units do not need any concern about foundation stability problems, while heterogeneous rock mass units need concern (Table  13). Finally, based on the standard guideline for grouting dam foundation (Table 10), all carbonate units (Table 14A) do not need grouting, but heterogeneous units need tentative grouting (Table 14 B).     Fig.10 Stereographic projection of discontinuity, dam axis and direction of the significant governing discontinuity (in this study is bedding plane)

Influence of the Foundation Deformability on Dam Behaviour (DMRDEF)
Two instances are risky for a concrete dam's usual behaviour. First, if Em (deformation modulus) of the rock mass variation along the dam axis, the second one if Ec/Em (deformation modulus concert/deformation modulus of rock mass) crosses the stander allowable limit for each kind of dam (Table 15 & Table 16). For assessing the above problem and risks, the Em value was calculated based on the serafim-Pereira formula if RMRBD < 60 or DMRDEF < 55 Em (GPa ) = 10 (RMR-10)/40 ………………………....Eq 6 but if RMRBD > 60 or DMRDEF > 55 the Em (GPa ) = 2RMR -100 (Romana, 2003a )…….Eq 7 The above equations are accepted for determining Em from RMR Basic dry RMRBD. RMRBD adds the first four parameters (compressive strength of intact rock, RQD of the rock mass, spacing and condition of the significant governing discontinuity) of RMR plus 15.  (Romana, 2004) (Romana et al., 2003) Based on the correlation Em values of rock mass units results with stander guidelines of dam modulus deformation and dam height (Table 17), it appears that some rock mass units have problems with specific dam types and suggested dam height of 80m (Table 18). The dam height of 80m was selected based on hydrological and topographic conditions at the selected proposed dam site Table 17. Deformability problems in concrete dams according to the value of DMRDEF (Romana et al., 2003) Ec/ Em Influence on dam Problems <   Romana et al, 2003) rocks required blasting led to disturbance and stress relaxing in the rock mass by 0.4, while excavation in flysch rocks required easy ripping, which led to disturbance and stress relaxing in the rock mass by 0.2. The results of the GSI value for rock masses in twenty-two units at the dam site reveal that the value of the carbonate rocks is nearly between 74-76, while flysch rock gives a lower value, 35-55. The (DMRSTA) shows a good indication of rock foundation stability and ranges from 82.9-43.7 except for some rock mass flysch units, which can be treated during dam construction. The foundation desirability evaluation was done based on RMRDB, indicating that massive carbonate units are suitable for gravity, hardfill, earth-fill, and Arch dams; meanwhile, the flysch rocks required systematic grouting for Arch dams and spotted grouting for gravity dams. Ec/Em (deformation modulus of the dam/deformation modulus of foundation rocks) values for massive carbonate rock range from 0.18-0.93; they indicate site suitability for arch, gravity dam, and hardfill dams. While the Ec/Em values for flysch rock mass units range from 1-6.4, they indicate that some units have a serious problem situation and need treatment based on the types of dams. The DMR DEF (RMR related to relative deformability) value between 69-80.2 for massive carbonate units shows no deformability problems, but the DMRDEF value for flysch rock mass ranges from 40-50 and shows no deformability problems. However, only the arch dam shows a deformability problem