A REASSESSMENT OF ANNUAL WATER BALANCES FOR OIL SANDS RECLAMATION SOIL COVERS PLACED OVER COKE

Abstract

There were initial concerns that the soil covers constructed by Syncrude Canada on the Coke Beach Instrumented Watershed were experiencing elevated water losses which was limiting revegetation plant growth. Previous studies conducted on the site aimed at unravelling the cause of the accelerated water loss but failed in achieving this objective. However, these studies found that the enhanced drying of the covers may be due to other processes than the water balance components but not convective air flow. A bypass flow was hypothesised to be the likely cause of the accelerated drying of the covers. From this background, this study was designed to; Construct a daily water balance for long-term (2005-2017) monitoring data set based on field meteorological and soil monitoring data; a) develop a calibrated physics based model of the hydrologic performance of the two reclamation covers using a soil-vegetation-atmosphere transfer (SVAT) model; b) use this calibrated SVAT model to identify the key processes controlling the hydrological performance of the reclamation covers with a focus on identifying which processes (e.g. preferential flow, convective air flow) or hydraulic properties (e.g. dual porosity water flow storage) have the greatest influence over performance. In achieving these objectives, this study adopted two approaches (system dynamics and physics based model) in constructing the daily water balance and to identify which processes control the hydrological performance of the covers. From the results of the water balance components using the system dynamics model, a mean AET of 38% of the annual mean PPT was found to be lost to the atmosphere on the shallow cover while on the deep cover, a mean AET of 44% of mean annual PPT was found to be lost. The mean net percolation for the shallow cover was 63% of annual PPT compared to the mean net percolation of 58% of annual PPT lost through the deep cover. Comparing the change in water ability of the covers, the shallow cover was found to be losing -2% of annual PPT while the deep cover lost -3% of the annual PPT. From the detail evaluation of the water storage ability of the covers, it appears the SD approach underestimates the AET and overestimates net percolation in this scenario and hence the results by the SD approach may not be the reality. With the simulated water balance approach (physics based model) a mean AET of 76% of annual PPT was observed to be lost to the atmosphere on the shallow cover compared to a mean AET of 86% of annual PPT found to be lost on the deep cover. Similarly, the mean percentage of precipitation released as NP was 24% on the shallow cover and 21% on the deep cover. The predominant volume of NP was associated with snowmelt and rainfall while the cover was frozen. The higher AET on the deep cover than the shallow cover is an indication of the additional water stored within the thicker cover and available for use by vegetation. From the water balance components volumes estimated by the physics based model, evapotranspiration was identified by this study as the main component greatly influencing the performance of the covers and causing the elevated drying in the growing season, followed by the net percolation.