Abstract
•Conjugate vs convective CFD modeling is implemented and compared for DCMD.•Temperature and permeate flux distribution for several module lengths is studied.•The convective approach is less CPU demanding.•The convective approach can be used safely for early DCMD module design.•The conjugate approach need to be used for overall DCMD module operation.
Membrane distillation (MD) process technology is swiftly moving to industrial prototyping where efficient scale-up calculations are crucial to shorten product development cycle. The aim of this numerical investigation is to suggest the best modeling strategy that will enable to innovate; modify or check new direct contact MD (DCMD) module designs or operation in a timely manner. Two main modeling strategies are presented, namely convective and conjugate approaches. For a given flat module, it is shown that replacing the permeate side by a modified convection boundary condition that accounts for every known resistance to heat transfer gives similar results in the feed side as a coupled conjugate approach where the membrane, with a modified thermal conductivity, is part of the computational domain. The two methods are compared for different module lengths in terms of permeate flux and temperature distribution. Simulation time reports show the important gain in CPU time when using the convective approach while retaining desired calculation accuracy during scale-up. Furthermore, investigations were carried out to assess the effect of 3D inlet and outlet effects. Results for a laboratory scale module suggest that the convective approach can be safely used during early design stages and scale-up of single modules in the range of high permeate fluxes, while the conjugate approach has to be used for an accurate prediction of permeate temperatures needed in heat recovery strategy and equipment design.