Mineralogical, bulk and field leachate compositions are used to identify important processes governing the evolution of discharges from a coal spoil heap in County Durham. These processes are incorporated into a numerical one-dimensional advective-kinetic reactive transport model which reproduces field results, including gas compositions, to within an order of magnitude. Variation of input parameters allows the effects of incorrect initial assumptions on elemental profiles and discharge chemistry to be assessed. Analytical expressions for widths and speeds of kinetic reaction fronts are developed and used to predict long-term development of mineralogical distribution within the heap. Results are consistent with observations from the field site. Pyrite oxidation is expected to dominate O2 consumption in spoil heaps on the decadal timescale, although C oxidation may stabilize contaminants in effluents on the centennial scale.

Aqueous chemistry methods assessed the kinetic reactivity and reduction capacity of fracture-filling minerals from a granitic groundwater environment at the Aspo Hard Rock Laboratory, a field station for research and development into the geological disposal of spent nuclear fuel. Naturally occurring fracture filling reacted with oxygenated test solutions of known composition in recirculating batch reactors.

The loss of O2(aq) with time was consistent with second-order reaction kinetics where O2(aq) is consumed through reduction by reaction with structural Fe(II) at the surface of the fracture minerals. Values of the second-order rate constant (k, l mole—1 h—1) varied between experiments within the range 1.7 < log k < 2.9 and values for the total concentration of reducing sites as a measure of the reduction capacity of the mineral (St, mole g—1) varied within the range 8.5 x 10—5 < St < 4.3 x 10—4. Values for the rate constant are somewhat less than those published previously for reaction between O2(aq) and Fe(II) surface complexes; i.e. structural Fe(II) appears to be less reactive than adsorbed Fe(II)(aq) but is significantly more reactive than Fe2+.

Values of the rate constant did not depend on release of network ions from the reacting minerals, and did not vary significantly with pH, mineral mass, specific surface area or mineral Fe(II) content. Oxidative weathering of sulphide minerals does not occur to any measurable extent. When corrected to rock/water ratios for granite aquifers, the rate constants correspond to a half-life for O2(aq) on the order of seconds indicating an essentially instantaneous reaction on repository time scales. Comparison of reducing capacity for the fracture fillings and oxidizing capacity for groundwater saturated with O2(aq) indicates that oxidizing fronts within the geological barrier would travel ~4000 times more slowly than the velocity of groundwater flow.