Implementation of a modeling concept to predict hydraulic and geochemical conditions during flooding of a deep mine

Abstract

More than four decades of intensive uranium mining in Germany resulted in a legacy of surface and underground mining objects requiring remediation in a densely populated area. The remediation concepts have to take into account the predictions of future environmental impacts for various remediation activities. Controlled flooding of deep uranium mines to a natural flooding or overflow level is the general remediation solution selected. Thereby the local and regional hydraulic and geochemical conditions of lager areas at and around the former mining site are influenced over a long period. For the optimization of the operations as well as the application for the necessary permissions model predictions are regularly used. In the presented case of the Königstein deep mine flooding the modeling concept followed combines commercially available groundwater modeling software to account for the relevant hydrogeological and geochemical processes in the mine and the influenced surroundings. It was found that there are no off-the-shelf solutions available allowing to model both the hydraulic and geochemical conditions on a local and regional scale. Furthermore a flexible modeling concept is required to include additional modeling tools if necessary and to allow better traceable model predictions. Based on monitoring of the ongoing flooding process the concept is further developed and the models are continuously checked, adopted and extended. Thereby the system understanding is improved allowing to further narrow the prediction uncertainties. In addition to a description of the applied modeling tools and their interaction the paper presents the experiences gained during this process and their implications for the further improvement of the model predictions.