Development of a two-fluid model for predicting phase-changing flows inside thermal vapor compressors used in thermal desalination systems

[Display omitted] •A new simulation method for a two-phase flow in TVCs was developed.•The effects of condensing steam on the performance of a TVC were studied.•The method considers different velocities between liquid and vapor phases.•The method is able to predict the flow properties in wide ranges of TVC application.•The model predicts the performance of a TVC more accurate related to other models. The aim of this study is to investigate the performance of “thermal vapor compressors (TVCs)” as one of the main components of “multi-effect distillation (MED)” systems. A single-phase vapor flow is normally required for continuous operating of a TVC, though a mixed liquid-vapor steam flow is often formed in TVCs due to a steam condensation in the supersonic flow through nozzles. A two-phase flow in a TVC undesirably reduces its performance, and lowers fresh water production rates of a desalination system. More accurate prediction of the “entrainment ratio” as a main performance parameter of a TVC completely influences on the gain output ratio (GOR) of a desalination system. For this purpose, a mathematical model was developed with respect to the phase-changing flow, and the model was later validated with experimental data. An iterative multiphase flow methodology based on a non-equilibrium condensation theory was developed to explore the difference between “single-fluid” and “two-fluid” models. Interactions between the liquid and the vapor phases were thoroughly evaluated through comparing variations in nucleation rates, droplets radii, number of droplets, and so on. The advantage of this method over other numerical methods is that this method is capable of considering different velocities for the liquid and the vapor streams based on the Eulerian-Eulerian approach, where formation and collapse of droplets can be precisely predicted. Results revealed that the performance of a TVC can better be predicted in a two-fluid model when compared with a single-fluid model.