Pilot-scale membrane distillation modeling: Validation, comparison, and consensus

Pilot-scale air-gap membrane distillation (AGMD) models have shown inconsistent predictive performance due to the lack of modeling consensus and validation across diverse systems. This study presents the first robust, cross-system validation and refinement of four pilot-scale AGMD models using an aggregated experimental dataset of 2716 datapoints that cover a wide range of feed salinities, module sizes, membrane and spacer configurations, and both AGMD and vacuum-assisted AGMD (V-AGMD) systems. Modeling consensus was achieved under real-world operating conditions by (i) using empirical correlations for the Nusselt number, (ii) applying a resistance-in-series approach for the thermal conductivity of the membrane, (iii) incorporating the ePTFE support layer into the air gap, and (iv) accounting for entrapped distillate in the air gap. The adjustments were guided by measurements, imaging, and literature, and were analyzed using root-mean-square error and mean error. The resulting modeling consensus and aggregated dataset provide a reliable foundation for AGMD system design, model calibration, and performance optimization across the broader range of operating conditions for which MD is being applied. This work also provides additional insight into how entrapped distillate influences the performance of AGMD.