Predicting sizes of toluene-diluted heavy oil emulsions in turbulent flow. Part 1—Application of two adsorption kinetic models for [formula omitted] in two size predictive models

Abstract

Predicting oil-in-water emulsion droplet size and stability in turbulent flow is useful for choosing separation processes in heavy oil extraction. In order to study the process, averaged droplet sizes for toluene, n-heptane and mineral oil were measured during stepped-down turbulent mixing. Droplet sizes were measured for five concentrations of heavy oil in toluene at oil fraction 0.05 in model process water as a function of time during turbulent mixing. Mixing was conducted with a Rushton turbine in a cylindrical tank. Size distributions were measured with the Mastersizer 2000 laser diffraction instrument. Dynamic interfacial tensions σ dyn (by drop volume technique, DVT) and static interfacial tensions σ (by du Nouy ring) as a function of time were measured. Data were fitted with two adsorption kinetic models (diffusion limited Dl, and Langmuir based Lb) to find “equilibrium” interfacial tensions σ E needed to predict Sauter mean diameter d 32 in Hinze–Kolmogorov (H–K) and Wang–Calabrese (W–C) models. Data were also analyzed by using a classical breakage–coalescence kinetic model. Results showed that by using σ dyn with W–C model the predicted d 32 changed proportionately with σ dyn . For any σ E the d 32 predicted using the W–C model was closer to the experimental d 32 than the d 32 predicted by the H–K model. The σ E 's obtained by the Dl and Lb models were similar when fitted to DVT data giving identical W–C predicted d 32 . However, only the Dl model fitted the σ -time data obtained by the du Nouy ring technique and the σ E in W–C model gave d 32 closest to experimental d 32 for droplets of 1–10 wt% heavy oil in toluene. The predicted d 32 of droplets for the 25-wt% heavy oil was larger than the experimental d 32 . This difference was attributed to interfacial film thickness and decreased elasticity discussed in an earlier publication.