Water availability is beginning to impact oil sands development and, as a result, several technologies to increase the percentage of recycled water are being evaluated. One such option being re-evaluated is the use of centrifuges to produce dry tailings that can accommodate overburden and soil replacement. Previous evaluations of centrifuge performance to capture water from the clay and silt tailings (mature fine tailings) components demonstrated some success but, at the time, at unacceptable costs. A better appreciation of the long-term costs of mature fine tailings storage has prompted a re-evaluation of centrifuge technology. The use of additives to improve centrifuge performance has significantly improved the results that can be achieved. Aside from the obvious positive environmental benefit of reclaiming the fluid fine (mature fine) tailings, the increase in the amount of water recycled will reduce the demand for fresh water from the Athabasca River. This paper discusses a laboratory-scale study of the water chemistry and clay/silt feed properties affecting centrifuge performance, as well as the results of a 20 tonne per hour pilot. Introduction The term dry stackable tailings is commonly used in oil sands tailings management to mean a mineral stream left over from the bitumen extraction process which can be stored without the need for dikes or other fluid containment structures. The use of water in the bitumen extraction process results in mineral tailings streams in the form of slurries or suspensions that require containment. In general, surface mined oil sands tailings fall into three categories: sand tailings, fine tailings and froth treatment tailings. Although the froth treatment tailings have important environmental implications due to their contamination with solvents or diluents from the froth treatment process, they are generally combined with the other tailings prior to discharge into the recycle water area (tailings settling pond). It is typically a straightforward process to create dry stackable tailings from the sand tailings, and they are often used to build the containment for the remaining fluid fine tailings. Another common practice in the surface mined oil sands industry is to define sand as mineral particles larger than 44 micrometres (µm). Fines are therefore smaller than 44 µm, and this definition can be useful when discussing the long-term tailings properties during mine planning. It has been demonstrated very clearly, however, that it is the clay content that determines all of the important properties of the fine tailings or fluid fine tailings (or mature fine tailings when the solids content exceeds about 25 wt%)(1–3). Over the entire mine, of course, the average clay-to-fines ratio is fixed and the fines content can be a useful approximation tool for large-scale planning purposes and the prediction of sand volumes and the fine tailings volume that may require containment. The clay content, often expressed as a clay-to-fines ratio, can vary across an oil sands lease by as much as a factor of four. It is therefore essential that the clay content, clay-to-water ratio or clay-to-fines ratio be understood in order to predict the properties of the fluid fine tailings on a daily or monthly basis.

The geology and subsequent mineral composition of oil sands deposits have important consequences for their processing behaviour. Differences in oil sands processability and extraction yields can be dependent upon many factors including the composition of the mineral components and the organic complexes that are associated with certain minerals. These mineral-organic associations help provide the bridge which leads to carry over of bitumen with the tailings as well as carry over of water and mineral matter with the product. Characterization of the minerals via various laboratory techniques and the relationship of these measurements to processing behaviour is discussed. Portions of this work were presented at the 1988 annual meetings of The Canadian Institute of Mining and Metallurgy and the Chemical Institute of Canada. Introduction Bitumen extraction from oil sands is carried out commercially via the Clark hot water process and involves mixing the raw oil sand with steam and a caustic solution. The bitumen separates from the sand and floats to the surface of the suspension. The amount of water and mineral matter that is carried over with the bitumen product and the amount of bitumen which is lost with the tailings are important parameters in determining the process efficiency. The recovery of bitumen from oil sands is a complex process With many steps involving the handling of water/oil systems which are often in the form of emulsions. The first stages of the process concentrate on separating the bitumen from the water and mineral matter, whereas later on the main problem is to remove residual water from the bitumen product. In order to facilitate this removal a combination of chemical (demulsifiers) and mechanical (centrifuges) means is often used. Losses in process efficiency can be due to many factors, several of which are related to the mineralogy of the oil sands feed. Previous studies(1–3) have correlated the amount of fine clay with processing problems; the clays were found to stabilize the water-in- oil emulsions formed in the bitumen product. In addition, clay floes can trap bitumen, carrying it into the tailings stream. Understanding the correlations between the mineralogy of the raw oil sands and the ultimate processability will help develop methods to deal with problem feeds as they reach the extraction plant rather than after they reach the refinery or tailings pond. The amount and size distribution of the clays has been shown to be important in determining the processability of the oil sands, but equally important is the nature of the mineral matter. Chemical characterization via electron microprobe analysis is a useful tool in determining both the quantity and size distribution of the critical mineral components. Preliminary results have shown the importance of the iron compounds; pyrites appear to be relatively innocuous although iron carbonates and hydroxides seem to provide a bridge for organic-mineral interactions which lead to processing problems, i.e. carry over of water, clays and minerals with the bitumen and loss of bitumen to the tailings. Direct observation of frozen hydrated samples in the electron microscope has been utilized for evaluation of both the size distribution of the dispersed phase and the chemical composition(4–5). The size distribution of the dispersed phase is important because centrifugal separation of the emulsions will be more difficult with smaller droplets; as will chemical treatments because surface area is higher per unit volume. The nature of the clays and other minerals which are carried over with the bitumen product also directly affects the bitu

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.

The belief that the Utah tar sands deposits are oil-wet has led to a focus on solvent-based bitumen extraction processes, or some form of solvent assisted water-based extraction process for these types of materials. However, under certain conditions, this ore is in fact amenable to a conventional water-based extraction process. The thermal, mechanical and chemical environments necessary to make the Asphalt Ridge ore behave like an Alberta Athabasca oil sand are outlined, along with the typical criteria which must be satisfied for a novel extraction process to be viable. Laboratory-scale demonstrations of the efficacy of a Clark-style hot water extraction process for the Asphalt Ridge tar sands were subsequently confirmed on a twenty tonne per hour pilot scale. In addition, the scarcity of water at the mining and extraction operation in Utah led to the development of an aggressive tailings treatment process, which also offers lessons for tailings handling in the surface-mined oil sands in Alberta. Introduction The CANMET Energy Technology Centre in Devon, Alberta, became involved in the Asphalt Ridge tar sands project when it was a solvent-based extraction operation hampered by a significant emulsion buildup in the recycle water. In working to develop a solution to the emulsion buildup, it became apparent that, using the solvent-based extraction process, solvent losses associated with clay mineral-solvent interactions would be unacceptably high. As a result, a series of standard tests(1, 2) were applied to Asphalt Ridge tar sand samples in order to assess the potential for a solvent-free, water-based extraction process(3, 4). Surprisingly, some of these nominally oil-wet tar sands performed very well, indicating that the Asphalt Ridge tar sand bitumen could be extracted using commercially proven technology developed over the last 40 years in Alberta(5–10). In order to achieve bitumen recoveries similar to those for Athabasca oil sands, significantly higher mechanical energy levels were required, along with high temperatures. Since the early 1990s, the operating temperature used in commercial processing of Athabasca oil sands has been reduced from about 80 °C to less than 50 °C while increasing the mechanical energy input(1, 2, 11, 12). By maintaining both mechanical and thermal energy inputs at high levels, the ‘difficult to process’ Asphalt Ridge tar sand showed bitumen recoveries of approximately 90%; similar to the Athabasca commercial operations. The Asphalt Ridge tar sand samples that did not perform well in bench-top laboratory assessments were found to be weathered or oxidized; conditions that also inhibit extractability in the Athabasca oil sands in Alberta(13–17). The difficulties encountered by earlier researchers in using Canadian technology, or a modified water-based extraction process for the Asphalt Ridge tar sand (without pre-treatment with an organic diluent before ore conditioning), may have been due to improper handling of cores or bulk samples resulting in bitumen oxidation or weathering(18–24). In referring to the differences between the Utah tar sands and those in the Athabasca deposit, "These differences preclude the direct application of the Canadian mining and recovery technology to Utah's tar sands and to other United States tar sands(18)."

Extended abstract of a paper presented at Microscopy and Microanalysis 2010 in Portland, Oregon, USA, August 1 – August 5, 2010.