Infiltration and Drainage through Coarse Layered Soil: A Study of Natural and Reclaimed Soil Profiles in the Oil Sands Region, Alberta, Canada
Natural coarse textured soils comprise a significant portion (approximately 20%) of the area to be mined at Suncor, Syncrude (aurora mine), Albian/Shell, and CNRL mines in the Alberta’s oil sands (Macyk, 2006). Although similar in soil textural classifications, the undisturbed areas support a range of ecosite types which exhibit different moisture regimes, suggesting that there are natural mechanisms controlling the plant available water sufficient for forest development. The global objective of this study was to evaluate the potential for textural variability to enhance water storage in coarse textured soil. The observations of the infiltration and drainage behaviour of natural and reclaimed coarse-texture soils in this study have demonstrated that this potential exists and can be applied in reclamation design to achieve the ranges of soil water storage needed to establish different ecosites. Field based infiltration and drainage testing, pit excavation and sampling have been completed on 14 sites (7 natural and 7 reclaimed). Bulk saturated hydraulic conductivity and field capacity were estimated for each of the 14 sites based on the field test results. The observed transient water dynamics give an indication of the effect of layering on these material properties. Laboratory analysis of water content (650 samples), particle size (650 samples), water retention (35 samples), organic carbon (100 samples) as well as calibration of field instrumentation were completed on a large number of samples (approximate values shown in brackets above) across all sites. The laboratory analysis was used to characterize textural variability (mean and standard deviation of the particle diameter) for the layered sites and estimate the soil water retention curve (SWRC) relationships for the range of soil textures encountered at the study sites. Pedotransfer functions (PTFs) were used to investigate if there were significant differences in the residual sum of squares between estimated and measured SWRCs. The measured organic carbon was used to aid in estimating permanent wilting point (WP) used in the calculation of the available water holding capacity (AWHC) of all profiles. An investigation into the calibration of the moisture capacitance probe (MCP) was undertaken as part of a comparison of the measured and simulated volumetric water content (VWC) profiles. Water storage at the cessation of drainage was related to the soil texture and textural variability as measured in the laboratory. Sites with more textural variability generally stored more water for plant use. There appeared to be a limit to what can be considered ‘useful’ textural variability. If adjacent soil layers had too extreme a contrast in texture and therefore hydraulic conductivity, unstable/preferential flow (i.e. bypassing of some of the water and nutrients from plant roots) occurred. The total porosity calculated from field samples was often higher than the maximum measured VWC in each layer which may be indicative of one or more factors that resulted in less than full saturation being attained within the targeted 1 m depth of saturation during the test. Some of these factors include: errors in sampling leading to an overestimate of total porosity; lateral flow along textural interfaces; air entrapment within the rapidly advancing wetting front; unstable/preferential flow as a result of the high contrast in hydraulic conductivity (fine over coarse) between adjacent layers (i.e. Ks Ratio >20) or where tests were conducted on slopes (i.e. funnel flow). This latter case was common at the reclaimed sites. A modelling study of one uniform (SV10) and one layered (NLFH1) natural site was conducted. The models were built by incorporating soil properties of the layers in the various soil profiles as estimated from field and/or laboratory testing. This study offers a comparison between various PTFs and their ability to capture the soil-water storage/dynamics during infiltration and drainage testing. The Arya PTF gave a better estimation of the laboratory measured SWRCs. However, when modeling the measured infiltration and drainage testing for the relatively uniform site SV10, the Arya PTF and Modified Kovacs (MK) PTF performed similarly. The Arya PTF performing slightly better for the infiltration phase and the MK PTF performing slightly better for the drainage phase. Both PTFs gave a reasonable estimation of water storage but the MK PTF gave a better estimation of the water storage with time as compared to the Arya PTF. For the highly layered site NLFH1, neither model performed well. The Arya PTF gave a substantially better estimation of the infiltration phase and gave the better estimation of the magnitude of water storage with time, the MK PTF performed marginally better for the drainage phase and gave a better estimation of the shape of the water storage with time. Generally, the study showed that the replication of the profile water storage requirements for the layered natural ecosites (‘b’ and ‘d’ ecosites) has been achieved and can be achieved by layering (or even mixing) available coarse textured reclamation materials. This study has indicated that replicating the highly uniform ecosites (‘a’ ecosites) is where the bigger challenge lies in reclamation. Reclaiming with a diversity of target ecosites is essential to achieving the pre-disturbance land capability standard that the mine operators are bound by. The temptation may exist to simply condone reclamation that has met or exceeded the pre-exisiting land capability. However, problems with ground water recharge and regional water distribution are likely to arise if large areas of lower functioning ecosites are replaced with higher functioning ecosites.
DegreeMaster of Science (M.Sc.)
DepartmentCivil and Geological Engineering
SupervisorBarbour, Sidney L.
CommitteeSi, Bing C.; Elshorbagy, Amin A.
Copyright DateApril 2014
available water holding capacity
soil water retention curve
1-D numerical modelling