Assessment of passive treatment and biogeochemical reactors for ameliorating acid mine drainage at Stockton coal mine

Abstract

Acid mine drainage (AMD) at Stockton Coal Mine, located near Westport, New Zealand, is generated from the oxidation of pyrite within sedimentary overburden exposed during surface mining. The pyrite oxidation releases significant acidity, Fe, and sulphate together with trace metals to the receiving environment. Aluminium is also elevated in drainage waters due to acid leaching from overburden materials. Thirteen AMD seeps emanating from waste rock dumps, and associated sediment ponds were monitored at Stockton Coal Mine to characterise water chemistry, delineate their spatial and temporal variability, and quantify metal loads. Dissolved metal concentrations ranged from 0.05-1430 mg/L Fe, 0.200-627 mg/L Al, 0.0024-0.594 mg/L Cu, 0.0052-4.21 mg/L Ni, 0.019- 18.8 mg/L Zn, <0.00005-0.0232 mg/L Cd, 0.0007-0.0028 mg/L Pb, <0.001-0.154 mg/L As and 0.103- 29.3 mg/L Mn and the pH ranged from 2.04-4.31. Currently this AMD is treated further downstream by a number of water treatment plants employing a combination of ultra fine limestone and calcium hydroxide; however, in the interest of assessing more cost-effective technologies, passive treatment systems were investigated for their treatment and hydraulic efficacy and as potential cost-effective options. Biogeochemical reactors (BGCRs) were selected as the most appropriate passive treatment system for ameliorating AMD at Stockton Coal Mine. Results of mesocosm-scale treatability tests showed that BGCRs incorporating mussel shells, Pinus radiata bark, wood fragments (post peel), and compost increased pH to ≥6.7 and sequestered ≥98.2% of the metal load from the Manchester Seep located within the Mangatini Stream catchment. The following design criteria were recommended for BGCRs incorporating 20-30 vol. % mussel shells as an alkalinity amendment: 1) 0.3 mol sulphate /m3 substrate/day for sulphate removal (mean of 94.1% removal (range of 87.6-98.0%)); 2) 0.4 mol metals/m3/day for metal (mean of 99.0% removal (range of 98.5-99.9%)) and partial sulphate (mean of 46.0% removal (range of 39.6-57.8%)) removal; and 3) 0.8 mol metals/m3/day for metal (mean of 98.4% removal (range of 98.2-98.6%) and minimal sulphate (mean of 16.6% removal (range of 11.9- 19.2%)) removal. At the maximum recommended loading rate of 0.8 mol total metals/m3/day an average of 20.0 kg/day (7.30 tonnes/year) of metals and 85.2 kg acidity as CaCO3/day could be removed from the Manchester Seep AMD by employing BGCRs. The design hydraulic residence time (HRT) would be 3.64 days. On an acidity areal loading basis, a design criterion of 65 g/m2/day was recommended. Tracer studies conducted on the BGCRs indicated ideal flow characteristics for cylindrical drumshaped reactors and non-ideal flow conditions for trapezoidal-shaped reactors indicative of shortcircuiting, channelised flow paths and internal recirculation. Consequently, this resulted in compromised treatment performance in the trapezoidal-shaped reactors. The relaxed tanks in series (TIS) model could be successfully applied to model the treatment performance of drum-shaped reactors; however, the model was unsuccessful for trapezoidal-shaped reactors. Because most pilot and full-scaled vertical flow wetlands (VFWs) have consisted of trapezoidal-prism basins excavated into the ground, the rate-removal methods previously recommended (e.g. mol metals/m3/day) should be applied to BGCR design, evaluation and operation rather than results of hydraulic and reactor modelling. Overall, a staged passive treatment approach is recommended. The first stage should consist of a sedimentation basin to remove sediment, the second stage a BGCR to remove acidity and metals and the third an aerobic wetland to provide oxygenation and tertiary treatment of metals (primarily Fe) from BGCR effluent. Preliminary analysis indicates that BGCRs are potentially a more cost-effective means of treating AMD at Stockton Coal Mine compared with the current active lime-dosing plant by over $125/tonne of acidity ($197/tonne for BGCRs versus $324/tonne for lime dosing (60% efficient)); however, their successful implementation would need to recognise current treatment goals, required areal footprint and inherent maintenance requirements.