Accurate predictions of long-term dewatering requirements by groundwater models depend on the quality of calibrated hydraulic parameters and boundary conditions. The latter is very important for large-scale dewatering projects when drawdown propagates a significant distance away from the mine area and predicted mine inflow/dewatering rates predominantly depend on the hydrogeological behavior of the model boundaries. This paper presents the results of a multiyear data collection and groundwater modeling study to define large-scale dewatering requirements for a mine project in a complex hydrogeological setting characterized by large bedrock transmissivity and groundwater storage, high recharge from precipitation, and the presence of leaky lateral and vertical boundaries. Over a period of 5 years, dewatering requirements were evaluated for scoping, pre-feasibility, and feasibility studies using a groundwater model calibrated to various hydraulic stresses, including short-term hydrogeological testing in geotechnical boreholes, 2 to 3 day airlift tests, 5 to 14 day pumping tests in prototype dewatering wells, one year of highly variable flow and water level data obtained during exploration decline (initial mine) dewatering, and observations of groundwater recovery once dewatering ceased.
Based on the size and length of the hydraulic stress applied to the groundwater system and observed responses, a series of conceptual and numerical models were developed for the project. When new data were obtained, an iterative approach was used to evaluate the robustness of the model calibration, assess the validity of the conceptual model, and revise to the predicted dewatering requirements. In total, the groundwater model was recalibrated five times to account for new data. The process required revising the conceptual model with regards to transmissivity, vertical hydraulic conductivity, groundwater storage, recharge from precipitation, and boundary conditions. Dewatering predictions were completed for two mining methods (block cave and open pit) and the high variation in predicted dewatering requirements highlights the importance of understanding vertical recharge to the groundwater system and the hydrogeological role of lateral model boundaries. In complex hydrogeological systems, these two factors cannot be precisely evaluated based on short- to intermediate-term hydraulic testing; model calibration to long-term testing provided superior input to the conceptual models and resulted in more defensible mine dewatering predictions. Predictability of the final version of the model was verified by reproducing groundwater recovery observed during flooding of the exploration decline.
