AmericanChemicalSociety.com

Washing of oil-contaminated soils, extracting bitumen from oil sands, recovery of oil from storage tank sludge, and washing of drill cuttings containing oil-based drilling fluids are processes that depend on the principles of oil-particle-water separation. The following is the final summary of some of the key findings of computer modeling and experimental approaches to understand particle-oil-water separation.

Previously, we reported the development of a computational fluid dynamics (CFD) model to simulate the separation of oil from a sand particle via the external shear of water.  The flow was assumed axisymmetric, allowing the implementation of the Navier-Stokes equations in spherical coordinates. The oil/water interface was tracked using the Volume of Fluid (VOF) method that, for the first time, was implemented in spherical coordinates.  The flow model was validated by predicting the equilibrium positions of a fluid drop that partially wets a solid particle under gravity. That work has produced one manuscript on the VOF implementation in spherical coordinates that's currently under review in the International Journal for Numerical Methods in Fluids, and one paper published in the Canadian Journal of Chemical Engineering that deals with the validation of the model under gravity and its use to predict the flow conditions that favor particle-oil-water separation. 

Due to limitations that will be explained later, flow experiments for single particle systems could not be conducted using the initially proposed experimental setup. Instead, particle-oil-water separation experiments were conducted using a technique that reproduces the type of conditions that are used in actual soil washing conditions. These basic conditions were simulated using the CFD code, leading to a range of Capillary numbers (Ca# = shear/surface tension forces) resulting from a range of interfacial tensions produced using different surfactant concentrations.  For each value of Ca#, the model was run until steady state conditions (or complete oil detachment) was obtained. At this point, the filling angle at steady state was plotted against Ca#.  A filling angle of 180° means that the oil completely coats the particle, and 0° means complete oil detachment.  The features of this curve resemble the shape of capillary curves used to describe oil recovery from reservoirs, that is, the filling angle decreases with increasing Ca# as it approaches a critical Ca# (~ 0.5 in these studies).  This critical Ca# was the same even when changing the volume fraction of oil or the contact angle at the water-oil-particle contact line.  These findings were reported in the Canadian Journal of Chemical Engineering article. Another finding, yet to be published, is that the dynamics of the process (i.e. the time that it takes to reach steady state) is highly dependent on the ratio of the viscosity of the oil to the viscosity of the aqueous solution.  High viscosity ratios slow down the process to the point that even when at a Ca# higher than the critical Ca#, no separation was observed. As will be discussed below, it was determined that an optimal viscosity ratio and a critical Ca# are the key parameters required to define particle-oil-water separation conditions. 

Parallel to the modeling effort, a single particle flow chamber was constructed using a polycarbonate enclosure, as per the initial proposed scope of work. A 5 mm metal particle was tethered to a flow distributor located on the top of the chamber. The oil was coated with a neutrally buoyant mixture of bitumen, chloroform and toluene. Due to design limitations (large pressure drops at moderate flow rates) the Reynolds numbers achievable with that configuration were substantially lower than 1, and as was later confirmed by the model, no separation or substantial changes occur in that region.  A falling particle setup has been introduced using a high speed camera (1000 frames per second) to capture the dynamics of oil detachment. While the predicted separation is observed in the falling particle experiments, particle rotation issues need to be resolved before the data can be published.  

Our batch soil washing experiments were conducted using beach sand contaminated with a mixture of toluene and bitumen, then left in the fume hood to evaporate the toluene and age the coating. This contaminated sand was washed with water containing 3% NaCl and a surfactant, sodium dihexylsulfoccinate (SDHS) in concentrations ranging from 0 to 0.1%. Toluene was used as a solvent to aid in the separation.  1gram of the bitumen-contaminated soil was pre-soaked with a prescribed amount of solvent (toluene) and soon after washed with 4 mL of the surfactant solution (2 minutes of mild agitation).  The vials were then centrifuged and, for the cases when oil was liberated, a clear phase of oil was observed on the top of the vial. The fraction of oil removed was then correlated to the Weber number obtained during the cleaning experiment.  The Weber number is analogous to the Capillary number but accounts for the ratio of inertial to surface tension forces. This Weber number has previously been used to study the break-up of oil droplets in shear. The residual oil content in the particle after washing was plotted against the Weber number and a curve similar to the capillary curve described above was obtained. Once again, as the Weber number increases (lower interfacial tension), the fraction of oil retained by the particle decreases, especially as the Weber number approaches a critical value.  Furthermore, increasing the volume of solvent (toluene) used in the pre-soaking stage facilitated the separation, an effect that is linked to the reduction in viscosity of the bitumen-toluene mixture. These findings have been presented at two international conferences and will be submitted for publication shortly. 

 Using modeling and experimental approaches we have determined that viscosity ratios and interfacial tension are the two properties that dominate (and can be managed through process design) particle-oil-water separation. We have determined that this separation can be accomplished when the viscosity ratio approaches a value of 1 and when using washing conditions that meet the critical transport number (Capillary or Weber number).  Further development of these findings into separation processes for cleaning oil-impacted soils and oil extraction is ongoing.