North-Central Section (44th Annual) and South-Central Section (44th Annual) Joint Meeting (11–13 April 2010)

Paper No. 9
Presentation Time: 3:45 PM

MODELING THE HYDROLOGIC EFFECTS OF LONGWALL MINING ON SHALLOW BEDROCK AQUIFERS USING MODFLOW WITH TMR


BOOTH, Colin, Department of Geology and Environmental Geosciences, Northern Illinois University, Davis Hall 312, DeKalb, IL 60115 and GREER, Christopher B., Geology and Environmental Geosciences, Northern Illinois University, 312 Davis Hall, DeKalb, IL 60115, cbooth@niu.edu

High-extraction (longwall) underground coal mining creates extensive drainage, complex fracturing and subsidence of the overlying strata that makes it difficult to apply commonly available groundwater models to examine and predict groundwater impacts. The variably saturated, heavily fractured roof zone immediately above the extraction is particularly difficult to model; however, standard groundwater models may be more easily applied to the shallower aquifer system (of most interest for groundwater resources) that is typically separated hydraulically from the mine and immediate roof zone by a confining zone at intermediate depth in the overburden. Nevertheless, several problems must be resolved, particularly the changing hydraulic properties, advancing subsidence zone, steep hydraulic gradients and sharp spatial changes over the longwall panels. We are applying MODFLOW (Groundwater Vistas® version) with Telescopic Mesh Refinement (TMR) to simulate the hydrologic responses in the shallow system, using a well-documented case study (Jefferson County, Illinois, 1988-1995) as the conceptual model and data base for subsidence impact and recovery. The mine is about 220 m deep, overlain by a shale-dominated overburden that includes a 23-m-thick sandstone aquifer at a depth of around 22 m, and a thin cover of glacial deposits. The uppermost 60-70 m of the overburden is modeled. In TMR, which has been extensively used in other hydrogeological problems but not for longwall modeling, a finely discretized local model (LM, here about 7 km2) is embedded in a coarser regional model (RM, about 20 km2). Currently, the base RM is calibrated and we are experimenting with LM approaches to the longwall hydrologic problems. The primary driver for head changes during subsidence is rapid increase in fracture porosity during the tensional phase; this is simulated by short-term well sinks defined in MODFLOW stress steps. There is no MODFLOW structure to accommodate changes in hydraulic properties, so these are modeled as stress zones in discrete steps associated with the advancing subsidence front. We expect that this development of MODFLOW with TMR will demonstrate an application that can be readily used by hydrogeologists to better evaluate and predict the hydrological impact of longwall coal mining.