An Evaluation of Accelerated Drying of Reclamation Soil Covers by Convective Airflow

Abstract

The coke beach instrumented watershed is a reclamation cover test site constructed in 2005 on top of petroleum coke at the Mildred Lake mine operated by Syncrude Canada Ltd in Fort McMurray, Alberta. Petroleum coke is a by-product of the oil sands extraction process. The cover system monitoring has shown an annual water loss which cannot be readily accounted for based on a standard water balance analyses. The water loss is most pronounced in late spring/early summer. It is hypothesized that the cause of the enhanced water loss is a result of convective drying of the cover as a result of airflow from the atmosphere into the underlying coke. This airflow could be the result of density gradients across the cover system as a result of temperature contrasts between the atmosphere and the coke or could be the result of oxygen consumption processes associated with the oxidation of methane released from the underlying fine tailings. The primary purpose of this study was to design and implement a field monitoring system to determine whether convective airflow was occurring across a series of trial closure cover systems overlying petroleum coke, and to utilize the resulting data in a water-loss calculation and determine the effect on the overall water balance for the site. The coke beach instrumented watershed contains two cover system trials: a shallow cover system and a deep cover system, with nominal cover depths of 0.40 m and 1.0 m respectively. Each cover system was instrumented with soil monitoring instrumentation which has been continuously monitored since cover construction. The shallow cover system included a meteorological monitoring system to complete the water balance monitoring. Additional studies have been carried out on each cover system including regular vegetation monitoring, and hydraulic conductivity testing. The major field research associated with this thesis was the installation and monitoring of differential pressure between the subsurface soil air and the ambient conditions. Three clusters at variable depths (0.4 m, 1.1 m, and 2.0 m) were installed on each cover system. In addition, air permeability testing of the cover system and underlying coke was performed to collect more data points for which to assess airflow rates. The results of the field monitoring program showed that differential pressure gradients existed across the cover systems relative to ambient conditions, and each cover system showed enhanced drying during the field monitoring years. The pressure gradients measured at each cover system were sufficient to induce substantial airflows, with measured differential pressures exceeding 40 Pa during peak periods. However, the estimated airflow rates did not appear to be sufficient to account for all of the enhanced drying observed in the water balance. This lack of airflow is likely a result of low permeability of the cover system material where the differential pressure systems were installed. However, it should be noted that across both cover systems, substantial cracking has occurred as a result of the dry conditions of the soil material. These cracks create macro-pores through which increased flow rates are possible due to larger void space. Given the differential pressure gradients measured across the cover systems, airflow through these cracks is possible and may have an effect on the soil moisture in material in the area surrounding the cracks. Further refinement in the research may be able to determine the effect the cracks have on airflow and cover drying. In addition to allowing for potential increased airflow rates, the macro-pores and cracks may also give rise to increased net percolation/bypass flow during snow melt infiltration and/or heavy rainfall events. The cover systems have approximately a 1% slope with substantial surface roughness and consequently runoff from the covers is unlikely. However, very little to no ponded water was observed on top of the cover systems even following large rainfall events. This indicates that there is a high potential for bypass flow or increased net percolation which may not have been represented by point measurements of permeability. No monitoring of net-percolation or infiltration was included as a component of the water balance; lysimeters were installed as part of the initial meteorological system installed in 2005, however, they were not regularly maintained or monitored since installation and as such were not usable. Hypothetical analyses were carried out in order to determine the water removal via airflow under higher permeability conditions. This analysis found that an increase in permeability of three orders of magnitude would result in sufficient airflow to cause enhanced drying of each cover system such that the additional moisture loss not currently accounted for by the water balance was completely accounted for by airflow alone. It is possible that the water balance of the covers is affected by both enhanced drying as a result of airflow as well as some form of preferential or bypass flow during infiltration events. Further research is required to more fully characterize the latter mechanism.