Oil sands, also known as “tar sands“ or “bituminous sands“, can be either loose sands or partially consolidated sandstone saturated with a highly viscous form of petroleum. Compared with conventional techniques, extraction of oil from tar sands is expensive and hence depends on oil prices and the availability of efficient and sustainable extraction techniques.
Oil extraction from oil sands requires a large amount of water for different processes. Hot water used to reduce the viscosity of the oil makes up the largest share. After oil/water separation, the water is sent to the tailings ponds and can be reused as “recycle water“ without any further treatment. River water is primarily used as boiler feed water, but it also finds use as cooling water in the summer. Due to their resistance to organic matter and oil residues, desalination behavior, and thermal stability, ceramic nanofiltration membranes (NF membranes) can contribute to the development of new and more efficient recycling processes, including partial heat recovery.
In a current project started in 2013 together with partners Shell Global Solutions International B.V., Shell Canada Ltd., and Andreas Junghans - Anlagenbau und Edelstahlbearbeitung GmbH & Co. KG, 19-channel elements with ceramic NF membranes are being tested in an oil field in Canada.
The goal of this project is to make the recycle water usable for other purposes besides the current one (boiler feed water). From an environmental point of view, use of recycle water instead of river water would be beneficial, but this is currently not possible due to the high residual bitumen and solids contents of the tailings. Ceramic NF membranes completely remove suspended solids and residual bitumen and at the same time reject most of the multivalent ions, thereby enabling a much higher yield in the subsequent reverse osmosis process.
The 19-channel NF membranes showed rejection of alkaline earth metals (Ca and Mg) of up to 80 % and of alkali metals (Na and K) of up to 55 %. The permeate was free of organic matter. Long-term tests performed over several months confirmed the stability of the membranes.
Economically feasible preparation of the large amount of boiler feed water and cooling tower water required is not possible with the 19-channel NF membranes due to the prohibitively high membrane costs. For this reason, Fraunhofer IKTS is looking for ways to reduce the membrane fabrication costs. One approach is to increase the membrane surface area per membrane element and thus reduce handling requirements. The ultimate goal is to use honeycomb substrates with a surface area of up to 20 m2 per element instead of 19-channel tubes. In the first scale-up phase, ceramic NF membranes were pre-pared on 163-channel substrates with a membrane surface area of 1.3 m2 per element, five times as high as it originally was. The sol-gel process, on which membrane preparation is based, was adapted to the smaller channel diameter and the reduced suction of the 163-channel substrate. In lab measurements, the membranes showed the same flux and retention behavior as that of 19-channel NF membranes.
The next step is to increase the membrane surface area to approx. 5 m2 per element. This requires the use of new handling technologies due to the size and weight of the elements. This development work started in 2015.
We thank Shell Global Solutions International B.V., Shell Canada Ltd., and Alberta Innovates – Energy and Environment Solutions for financial support. In addition, we thank Andreas Junghans - Anlagenbau und Edelstahlbearbeitung GmbH & Co. KG and the Rauschert GmbH as well as its subsidiary
inopor GmbH for a fruitful partnership.