2018-02-05_PERMIT FILE - C1981019

Home Browse Search

October 19, 2017 Page 6 development of regional groundwater conditions and lithological distributions relative to the Collom Pit. The dewatering evaluation was carried out in two distinct stages. The first stage involved calibrating hydraulic parameters of the Collom area strata, which were based primarily on the ground response to the pilot dewatering test as well as other in-situ test results. Subsequently, the calibrated strata properties were used in the predictive analyses to assess and optimize the effects of variable dewatering well locations, for variable well depths and screen depths, and variable dewatering time periods. 2.4 Calibration of Strata Hydraulic Parameters At the start of the calibration study, an 8,000 -ft x 8,000 -ft zone was defined to encompass the Collom Pit. The overall model included a buffer on either side of the Collom Pit so that boundary effects on the pit's hydrologic response were limited during simulation and regional flow divides were accounted for. The overall horizontal model dimensions were 24,000 ft x 24,000 ft (Figure 2). The model north axis was rotated to N23E in order to orient the Collom Pit perimeter roughly parallel to the model vertical boundaries. AAI identified the coal seams (131-132, F6, FA -FB, G7 -G9, GA -GB) associated with the thickest/weakest carbonaceous mudstone (CMS) layers, which are considered critical for stability of the walls of the box cut. The overburden and interburden non -coal layers associated with the above coal layers were simulated in the calibration and predictive models as single composite layers of the constituent sandstone, mudstone, and siltstone rock types, given their similarities and for modeling efficiency. The Km volcanic ash bed, a known low -permeability layer, was simulated as the penultimate bottom layer in the developed models, as layers underlying the Km bed are unlikely to have any impact on Collom dewatering. However, a more permeable layer below the Km bed formed the model bottom, to avoid model numerical convergence issues. In total, 17 layers were included in all calibration and predictive models, which are listed in Table 1. All layers were simulated as confined aquifers for computational efficiency. The convergence criterion for all models was 0.1 ft of head. The weak CMS layers were incorporated into the composite non -coal layers. They were considered unlikely to depressurize or yield water during the anticipated dewatering timeframe and will be simulated as fully saturated and pressurized (thus, weakened) in the highwall stability analyses. The layer elevations, based on up-to-date local geologic models, were imported into the calibration models after rotating the world coordinates to model coordinates. In order to perform the calibration, the pilot well (location screen depths) and the observation well information were imported into the model. The locations, stand pipe and piezometers depths, and observed heads during the pilot dewatering test for the observation wells (C -05-16A, C -05-16K, C -05-16E, C -05-16H, C -05-16I, C -04-16C, C -04-16G, C -04-16J, C -04-16L, and C -04-16M) were imported into the calibration model. The overall grid size was coarser away from the dewatering test cluster (500 ft x 500 ft) and finer around the dewatering and observation wells (50 ft x 50 ft), for improved modeling accuracy. The dewatering well itself was placed at the center of a 10 -ft x 10 -ft grid. An annual groundwater recharge of 2 inches per year from precipitation was assumed in all the models, for conservatism, which is higher than the 0.7 inches per year estimated in the regional water Agapito Associates, Inc.