Performance of Mesocosm-Scale Sulfate-Reducing Bioreactors for Treating Acid Mine Drainage in New Zealand

Abstract

Water chemistry was monitored monthly for ten months from an acid mine drainage (AMD) seep emanating at Stockton Coal Mine within the Mangatini watershed in New Zealand. Metal concentrations of the seep water were Fe (4.31-146 mg/L), Al (7.43-76.7 mg/L), Cu (0.0201-0.0669 mg/L), Ni (0.0629-0.261 mg/L), Zn (0.380-1.39 mg/L), Cd (0.000540-0.00134 mg/L) and Pb (0.0049-0.0056 mg/L), pH was 2.49-3.34 and total acidity (pH 8.3) was 78.5-626 mg/L as CaCO 3. Water chemistry signature prompted laboratory mesocosm studies measuring the effectiveness of sulfate-reducing bioreactors (SRBRs) for generating alkalinity and sequestering metals. Alkaline materials utilized in the SRBRs included industrial waste products such as mussel shells, nodulated stack dust (NSD) derived from the cement industry, and limestone. Organic substrate materials included post peel, a by-product from fence post manufacture, Pinus radiata bark and compost. Seven SRBRs comprised of varying substrate mixes received aerated AMD for nearly four months. AMD was sourced from the pond that collected the seep water. The SRBR containing NSD successfully removed all metals, but effluent was caustic with pH>9. Bioreactors consisting of 20-30% mussel shells were most successful at immobilizing metals and generating circumneutral effluent. Systems containing mussel shells sequestered more than 0.8 moles of metals/m 3 of substrate/day at stable operating conditions and yielded effluent concentrations (removal efficiencies) of 0.120-3.46 mg/L Fe (96.5-99.8%), 0.0170-0.277 mg/L Al (99.5-99.9%), <0.0005-<0.001 mg/L Cu (>99.7->99.9%), <0.0005-0.0020 mg/L Ni (99.3->99.7%), <0.001-0.005 mg/L Zn (99.7-99.9%), < 0.00005 Cd (>98.3->98.9%) and <0.0001-0.0001 Pb (99.5-<99.7%). The system consisting of limestone as the only alkalinity generating material was less effective (15.4-64.3 mg/L Fe). Results from duplicate systems but different reactor shapes indicated reactor dimensions influence flow characteristics and therefore treatment efficacy.

Variability of Stockton Coal Mine drainage chemistry and its treatment potential with biogeochemical reactors

“…The P2 reactor containing 12% mussel shells and 5% limestone outperformed S4, the larger trapezoidal reactor with the same substrate complement, indicating cylindrical drum reactors (562 mm substrate depth) perform better than trapezoidal reactors (400 mm substrate depth) for this BGCR design. Metal removal in S2 and S3 (both contained 20 vol% mussel shells) was similar throughout this study except at the highest loading rates tested of 1.4 mol metals/m 3 substrate/day (McCauley et al 2008). Metal removal from S2 was less than that of S3 at the highest metal loading rates tested, possibly due to increased hydraulic short circuiting caused by the higher percentage of flat, plate-like bark.…”

“…Wildeman et al (2006) recommended design criteria of 0.3 mol metal removal/m 3 of substrate/day for SRBRs containing a mixture of organic materials and crushed limestone, which is comparable to the removal rates measured from BGCR S1 in this study (containing similar reactor substrate media including limestone as an alkalinity amendment). This study found metal removal 0.8 mol/m 3 substrate/day for BGCRs containing 20Á30 vol% mussel shells (McCauley et al 2008(McCauley et al , 2009). Effluent pH was also typically 6.7.…”

“…Influent and effluent sampling was typically conducted on a weekly or biweekly basis and prior to increasing AMD flow rate. Further details about the experimental set-up and treatment efficiencies of these mesocosm-scale systems, which were operated for nearly four months, has been previously reported (McCauley et al 2008(McCauley et al , 2009.…”

“…Mesocosm biogeochemical reactor treatability tests Biogeochemical reactors containing 20Á30 vol% mussel shells (P1, S2 and S3) showed the best metal removal for each respective reactor shape (McCauley et al 2008(McCauley et al , 2009. Metal removal was also good for P3 (containing NSD), but its effluent pH was 9.0Á10.5, and therefore considered too caustic for discharge to a freshwater ecosystem so its operation was terminated before the other BGCRs.…”