The desalting process of an electrostatic desalting unit was studied using the collision time of two droplets in a water-in-oil (W/O) emulsion based on force balance. Initially, the model was solved numerically to perform a process analysis and to indicate the effect of the main process parameters, such as electric field strength, water content, temperature (through oil viscosity) and droplet size on the collision time or frequency of collision between a pair of droplets. In decreasing order of importance on the reduction of collision time and consequently on the efficiency of desalting separation, the following variables can be classified such as moisture content, electrostatic field strength, oil viscosity and droplet size. After this analysis, a computational fluid dynamics (CFD) model of a biphasic water–oil flow was developed in steady state using a Eulerian multiphase framework, in which collision frequency and probability of coalescence of droplets were assumed. This study provides some insights into the heterogeneity of a desalination plant which highlights aspects of design performance. This study further emphasizes the importance of two variables as moisture content and intensity of electrostatic field for dehydrated desalination by comparing the simulation with the electrostatic field against the same simulation without its presence. The overall objective of this study is therefore to show the necessity of including complex phenomena such as the frequency of collisions and coalescence in a CFD model for better understanding and optimization of the desalting process from both process safety and improvement.