Design, construction and performance of passive systems for the treatment of mine and spoil heap drainage

Abstract

Passive treatment systems for mine water pollution remediation require no chemical or energy inputs, and should only incur limited maintenance costs if properly constructed. This thesis investigates two such passive treatment systems that have been installed at sites in County Durham, UK. The overall objectives of the investigation were to improve understanding of the operation of passive systems in order to produce more accurate design guidelines, and to develop a new passive treatment option to complement existing technologies. To put the research into context the causes, impacts and extent of mine water pollution in the UK and overseas are outlined. In addition a detailed and up-to-date review of passive treatment research and development is provided, which highlights areas in which further research is required. At Quaking Houses, County Durham, a full-scale compost wetland has been constructed to treat an acidic and metal-rich discharge from the spoil heap of an abandoned coal mine. Over the 27 month study period the discharge had pH >_ 4, and mean iron, manganese and aluminium concentrations of 5.4 mg/L, 3.6 mg/L and 6.2 mg/L respectively. The unique design and construction of the 440 m2 wetland is described. An examination of the performance of the wetland illustrates that it has yielded significant improvements in water quality since its construction. Mean reductions in acidity, iron and aluminium concentrations are all around 50%. Bacterial sulphate reduction may be an important contaminant removal mechanism, but its proportional significance is unclear. Comparison of the system at Quaking Houses with other wetlands reveals that the mean reduction in acidity in terms of mass removal per unit area (5.01 g/m2/d) is comparatively low. However, it is shown that comparing wetland performance on this basis may be misleading where influent contaminant concentrations are relatively low, because contaminant removal is demonstrably influent concentrationdependent. A first-order removal assessment method of performance may be more appropriate, and on this basis the Quaking Houses wetland performs well when compared to systems in the USA. An algorithm for predicting wetland effluent iron concentrations is derived using a linear multiple regression technique. Time series analysis strongly suggests that some effluent water quality variables (and therefore wetland performance measures) are significantly affected by seasonal climate changes. In particular aluminium and acidity removal rates rise with increasing air temperature. This implies that to meet the same effluent water quality targets passive treatment systems in cold climate locations may need to be larger than equivalent systems in warmer climates. At Kimblesworth, also in County Durham, a pilot-scale passive treatment system has been operated for 4 months. The Kimblesworth discharge is a net-alkaline pumped mine water containing up to 2 mg/L iron and up to 1 mg/L manganese. The novel system at this site was designed to rapidly remove iron by oxidation and accretion of iron to high surface area media. The monitoring programme has revealed the system to be very efficient. Iron concentrations are consistently reduced to < 0.5 mg/L, and removal rates are an order of magnitude greater than in wetland systems. Furthermore, lithium tracer tests reveal that the residence time of water in the reactors is just 5- 10 minutes. The results of this research suggest that the technology could be applied at full-scale. The compost wetland at Quaking Houses and the new reactor at Kimblesworth may well complement each other. The Quaking Houses system generates alkalinity and removes metal contaminants, but effluent water still contains marginally elevated concentrations of iron (mean 2.4 mg/L). The Kimblesworth system is shown to be very effective at rapidly removing low concentrations of iron, and thus could conceivably be used as a polishing unit following a compost wetland such as that at Quaking Houses. Alternatively high surface area media reactors such as those at Kimblesworth could be used to treat discharges in topographically difficult locations, where the installation of a wetland is not feasible.