RESUMO
When dispersed in aqueous solution, droplets of bitumen (extra heavy oil) are known to acquire negative surface charges. The resulting electrostatic repulsion, according to traditional DLVO theory, is far too strong for any droplet coalescence to occur. However, from experience it is known that bitumen droplets do coalesce in aqueous suspensions. Furthermore, the process appears to be random, with the probability of coalescence increasing sharply with the drop size. To explain these facts, we modeled the bitumen-water interface as a heterogeneous surface comprising charged "patches"; the zeta potentials of the patches constitute a random variable that is assumed to be Gaussian. The traditional DLVO theory, according to this model, remains sound on the local scale (i.e., for patches interacting across an intervening water layer). Such a theory can predict the probabilities of coalescence in remarkable detail. A parameter central to this theory is the lateral extent of the charged patches, which was estimated to be in the neighborhood of 0.6 µm.