Batch extraction method to estimate total dissolved solids (TDS) release from coal refuse and overburden

Abstract

A rapid batch extraction method was evaluated to estimate potential for total dissolved solids (TDS) release by 65 samples of rock from coal and gas-bearing strata of the Appalachian Basin in eastern USA. Three different extractant solutions were considered: deionized water (DI), DI equilibrated with 10% CO 2 atmosphere (DI + CO 2 ), or 30% H 2 O 2 under 10% CO 2 (H 2 O 2 +CO 2 ). In all extractions, 10 g of pulverized rock (<0.5-mm) were mixed with 20 mL of extractant solution and shaken for 4 h at 50 rpm and 20–22 °C. The 65 rock samples were classified as coal (n=3), overburden (n = 17), coal refuse that had weathered in the field (n = 14), unleached coal refuse that had oxidized during indoor storage (n = 20), gas-bearing shale (n = 10), and pyrite (n = 1). Extracts were analyzed for specific conductance (SC), TDS, pH, and major and trace elements, and subsequently speciated to determine ionic contributions to SC. The pH of extractant blanks decreased in the order DI (6.0), DI + CO 2 (5.1), and H 2 O 2 +CO 2 (2.6). The DI extractant was effective for mobilizing soluble SO 4 and Cl salts. The DI + CO 2 extractant increased weathering of carbonates and resulted in equivalent or greater TDS than the DI leach of the same material. The H 2 O 2 +CO 2 extractant increased weathering of sulfides (and carbonates) and resulted in the greatest TDS production and lowest pH values. Of the 65 samples, 19 had leachate chemistry data from previous column experiments and 35 were paired to 10 field sites with leachate chemistry data. When accounting for the water-to-rock ratio, TDS from DI and DI + CO 2 extractions were correlated to TDS from column experiments while TDS from H 2 O 2 +CO 2 extractions was not. In contrast to column experiments, field SC was better correlated to SC measured from H 2 O 2 +CO 2 extractions than DI extractions. The field SC and SC from H 2 O 2 +CO 2 extractions were statistically indistinguishable for 7 of 9 paired data sets while SC from DI extractions underestimated field SC in 5 of 9 cases. Upscaling comparisons suggest that (1) weathering reactions in the field are more aggressive than DI water or synthetic rainwater extractants used in batch or column tests, and (2) a batch extraction method utilizing 30% H 2 O 2 (which is mildly acidic without CO 2 enrichment) could be effective for identifying rocks that will release high amounts of TDS.