Design and optimization of a waste rock cover system for arid environments
Engineered waste rock cover systems are designed to prevent the generation of acid mine drainage in sulphide bearing waste rock dumps left from the metal mining process. Typically, the cover systems are designed to prevent the ingress of water and/or oxygen into the waste rock pile so that oxidization and hence acid generation is minimized. In arid environments the covers are generally designed as water infiltration barriers only. A saturated cover layer is required to prevent oxygen ingress and this is difficult to achieve in an arid climate. Ultimately, the success of a waste rock cover system is dependent on understanding the relationship between climate, cover and waste material properties, and surface vegetation. One method to determine how successful a cover system will be over the long term is to use predictive numerical modelling software, calibrated from an instrumented test cover site. This gives mine operators and regulatory agencies the ability to evaluate how a cover will perform under typical and extreme climatic conditions. The purpose of the research program described here was to design a zero net infiltrative flux cover system that would eliminate acid drainage at the Kidston Gold Mines in North Queensland, Australia. The mine has three main waste rock dumps, which extend over an area of about 337 ha. The climate is semi-arid with well-defined wet and dry seasons. Mean annual precipitation is approximately 712 mm with more than 80 % of the rainfall occurring between November and April. Total annual potential evaporation was estimated to be about 2300 mm. Historically acid mine drainage was collected in evaporation ponds. The first phase of the research involved characterizing the waste rock and potential cover materials, designing and building two field test covers, and installing instrumentation for monitoring of performance. The majority of this work was carried out by the Unsaturated Soils Group, at the University of Saskatchewan. The design of the cover system required the optimization of water storage and evapotranspiration. Intense rainfall events and associated erosion which occurs in the wet season, required that runoff be eliminated. This requirement demanded that the cover have the ability to store all infiltrating rainfall such that it could subsequently be released to the atmosphere through evapotranspiration. Several cover systems using layers of loose and compacted native soils and waste rock were evaluated using the numerical model SoilCover. Two cover profiles were selected for field-testing based on the net infiltrative flux approaching zero. The first was constructed on a mineralized waste rock dump using 50 cm of compacted, oxidized waste rock overlain by 150 cm of loose oxidized waste rock. This cover was designated the optimum design. The second was constructed on a barren waste rock dump using a single 250 cm layer of loose oxidized waste rock. This cover was designated the alternate design. Both covers had a hummocky surface to promote infiltration and prevent runoff. Construction was completed in December 1996. Seeding of the barren dump cover with pasture grasses took place shortly after construction while seeding of the mineralized dump cover was completed at the end of the wet season. Instrumentation was installed in both covers to measure infiltration and cover moisture conditions. A weather station was installed to monitor climatic conditions. This thesis presents the results for the second phase of the research program which was carried out by the author. It includes the evaluation of the performance of the two test covers and calibration of the numerical model SoilCover, such that field response and predictive modelling could be carried out. A final cover design is then presented on the basis of these results. The research involved field sampling and laboratory testing, instrumentation, field data reduction and analysis, and field response and predictive numerical modelling. Field results over the period 1996 to 1999 indicated that both covers responded reasonably well. The average precipitation was about 541 mm or 76 % of normal. Instrumentation showed that the water content profile within the covers increased considerably under intense rainstorm events and then quickly decreased during subsequent evapotranspiration. Infiltration from the base of the covers into the underlying waste rock was less than 1 % of average annual precipitation during the three wet seasons. Calibration and field response modelling with the numerical model SoilCover confirmed infiltration values into the waste rock approaching zero. Predictive modelling was then carried out using extreme wet year climate data from 1990/1991. Based on 1349 mm of rainfall the calibrated SoilCover model predicted over 50 % infiltration through both cover systems. Further predictive modelling was then carried out assuming improved saturated hydraulic conductivities could be achieved in the compacted layer of the optimum cover design. The results showed that infiltration rates through the cover coul? likely be reduced to between 5 % to 25 % of precipitation during extreme wet year conditions. Final closure cover design is described on the basis of these results.