QUANTIFYING AND PREDICTING THE LONG-TERM EVOLUTION OF COAL MINE DRAINAGE IN A PITTSBURGH COAL BASIN

Coal mine discharges (CMD) in Appalachia are important sources of metals, SO₄, and acidity that can degrade aquatic habitats and water supplies for decades. The Irwin Coal Basin (ICB), in the bituminous field of Pennsylvania, contains a series of partly to completely flooded abandoned underground mines separated by leaky barriers within the Pittsburgh Coal seam. Historical and recent data for eight CMD sites across the ICB were used to evaluate spatial and temporal water-quality trends over a decadal timeframe. Since the 1970s, these CMD sites have become less acidic, with exponential decreases in acidity, SO₄, and Fe concentrations that now approach steady state; two have transitioned from net-acidic to net-alkaline character. Along a ~25-km flow path along the ICB from northeast to southwest, corresponding with increasing overburden thickness and residence time, CMD becomes more alkaline with increasing Na concentrations. We hypothesize that 1) cation-exchange reactions contribute to the evolution of Na/SO₄+HCO₃ water-type, and 2) siderite equilibrium acts as a source and sink of Fe. First-order exponential decay models were adapted to include a steady-state asymptote consistent with siderite equilibrium and background groundwater chemistry. With a specified intercept, a second-order model better fit the rapid decay immediately after the “first flush.” Although projections are uncertain, the models suggest that siderite equilibrium could maintain dissolved Fe >20 mg/L over the next 40 years. A first-flush chemical evolution model developed with PHREEQC indicates that the spatial and temporal trends in pH, net-acidity, SO₄, Fe, and major cations could be explained by the progressive dilution of first-flush water by ambient groundwater plus sustained water-mineral reactions involving pyrite and carbonates (calcite, dolomite, siderite) plus cation-exchange by clays (illite, chlorite, mixed layer illite/smectite). This model considers physical and chemical processes to explain observations and estimate trends rather than fitting observed data to exponential functions. The processes are modeled to occur throughout time, considering realistic constraints such as mixing ratio of groundwater (residence time), eventual decay of reactant phases, and anticipated mineral equilibrium conditions.