INVESTIGATIONS AND MODELING TO UNDERSTAND AND MITIGATE IMPACTS OF ABANDONED MINE DRAINAGE (Invited Presentation)

Abandoned mine drainage (AMD) from legacy coal mines degrades more than 5,000 miles of streams across Appalachia. The AMD can have a range of pH values (2 to 8) along with elevated concentrations of SO₄, Fe, Al, Mn, and other constituents. The pH of AMD is unstable; CO₂ outgassing causes pH to increase, while Fe oxidation and hydrolysis cause pH to decrease. Net-alkaline AMD may have sufficient alkalinity to maintain pH 6; however, even net alkaline AMD may be a long-term source of contamination because of elevated Fe, Mn, and SO₄. Treatment of AMD can accelerate pH change and decrease contaminant loadings to streams, potentially mitigating aquatic impacts. A specific treatment strategy may be appropriate depending on the AMD flow rate and chemistry, site characteristics, funding, and operational logistics plus the chemical and biological characteristics of the receiving water body. Nevertheless, AMD sources typically become less acidic with decreasing concentrations of contaminants over a decadal timeframe and, eventually, may not warrant treatment. As the AMD evolves, management strategies may be adjusted; however, the rate and extent of evolution can vary. Because of varying conditions by site, accurate predictions and effective solutions to AMD problems may warrant study and experiments to understand hydrochemical and biological interactions that influence the chemistry of untreated and treated AMD. Such understanding is needed to develop models that explain spatial and temporal variations in AMD quality. For example, the “TreatTrainMix2” model has been developed to simulate changes in water quality during passive or active treatment. This tool, which may be applied to optimize long-term management strategies, has been incorporated with the AMDTreat 6.0 cost analysis model. By adjusting kinetic variables or chemical dosing, effects of independent or sequential treatment steps that have different retention time, aeration rate, or quantities of reactive solids can be simulated for a specified influent. Given the system size, based on reaction time estimates, the AMDTreat model may be used to estimate costs for treatment installation and annual operations and maintenance. Thus, various passive and/or active treatment strategies and associated liability can be identified for current or future influent.