This study represents another contribution in the MEND series on the prediction and control of acidic drainage from waste-rock dumps. The Samatosum Minesite in southeastern British Columbia implemented full-scale layering of acid-generating and acid-neutralizing rock in its waste-rock dump, based on a series of column tests.
This study has reviewed existing information, reinterpreted old data, and obtained new data for the layered waste-rock dump at Samatosum and the column tests which simulated it. Most of the rock units at Samatosum had significant proportions of net-acid-generating rock, except for mafic pyroclastics (MAF). As a result, all rock was divided into MAF and Potentially Acid Generating (PAG) rock. The PAG was encapsulated by MAF layers within the dump, and the overall ratio of Neutralization Potential (NP) to Total Acid Potential (TAP) within the dump was 3:1.
Column tests containing 10-50 kg of rock simulated various layered and unlayered sequences of PAG and MAF rock. These columns were operated up to 5.5 years to determine if the dump design was appropriate. However, the two columns containing layers of PAG and MAF to simulate the dump had NP:TAP ratios close to 1:1, instead of 3:1.
Because effluent concentrations of metals like calcium and magnesium were not measured during column testing, rates of NP consumption and depletion could not be determined. However, simple geochemical relationships and data from other mines indicated that the ratio of NP consumption to sulphide depletion in the layered columns was likely greater than 1:1. Additionally, not all measured NP at Samatosum is available for neutralization: up to 10 t CaCO3/1000 t is unavailable. Therefore, the columns were predicted to eventually release net acidity if they had continued. This contradicted previous predictions, primarily because previous work assumed that the rate of NP consumption was lower and equivalent to the rate of alkalinity production. In reality, alkalinity production typically represents only a portion of total NP consumption.
Comparisons of pre-test and post-test analyses on various grain-size ranges from the columns indicated layering as small as 0.2 m did not alter the reaction rates and the geochemical behaviour of the PAG and MAF material. In other words, layering with MAF did not slow reaction rates in the adjacent PAG layers. However, the effluent from the layered columns did represent a composite from the two materials.
Significant accumulation of sulphate from sulphide oxidation was noted in PAG, where it originated, and in basal MAF layers within the columns. Also, NP was significantly depleted in MAF layers, particularly in the finest grain-size range. In fact, NP depletion in the finest MAF particles caused some MAF to become net acid generating and suggested that NP in the coarser particles may not be readily available. This also supported the prediction of net acidity from the columns.
Monitoring of drainage from the dump at Samatosum has revealed signs of impending net acidity. In fact, pH at one station has fluctuated between neutral and acidic values, with the acidic values becoming lower and more persistent with time. However, unlike the columns, the dump has an overall ratio of NP:TAP of 3:1, and thus is not predicted to generate net acidity.
The reconciliation of predictions with monitoring data is based on (1) the observations in the columns that layering does not suppress reaction rates in the PAG and (2) coarser rock can preferentially channel water through a layer. Consequently, the appearance of net acidity in some dump drainage is simply the result of physical conditions rather than a failure of geochemical principles. This highlights the importance of physical design and physical hydrogeology in any future design and construction of a layered dump. If drainage does not pass through all available neutralizing layers, acidic drainage may appear even in the presence of excess neutralizing potential.
A final note focusses on metal leaching. Although acidic pH may be prevented with carefully designed and implemented layering, leaching of metals is not so easily controlled. Because reaction rates in even small layers of net-acid-generating Samatosum rock, on the order of 0.2 m, could proceed unattenuated by adjacent net-acid-neutralizing rock, metal leaching can probably occur at accelerated rates in layered dumps. If site-specific solubilities of secondary minerals are relatively high, aqueous metal concentrations may then exceed water-quality requirements even in near-neutral drainage. Therefore, layering is not necessarily a control technique for metal leaching.
