Long-Term Time-Series Analysis to Understand Groundwater Flow in Abandoned Subsurface Mines with Application to a Coalfield in Liège, Belgium

Complex underground flow processes can occur in flooded mine workings. As the groundwater rebounds, outbreaks, flooding, and slope stability problems can occur where hydraulic pressures build up in less drained areas. A time-series statistical analysis was conducted to understand how exploited areas in an abandoned coalfield were connected and to calculate groundwater response times to rain events by spatially and temporally correlating piezometric levels and discharge rates. Ten years of flow rate and water level data were statistically analyzed for an abandoned coalfield in Liege (Belgium). Then, the results were compared to results from physically-based simulations (a 3D groundwater flow model) based on data from the first 2 years of monitoring. The statistical approach gives qualitative indications on the interconnections between the different areas of the coalfield, as well as on the storage capacity/transmissivity of the aquifer. Improved understanding of this hydrogeological behavior can be used to prevent post-mining accidents and assess the associated risks.