Solar-driven hybrid HDH–TVC–RO–ERT desalination: ANN-assisted multi-objective optimization, techno-economic assessment and a case study of Siwa Oasis, Egypt

This study addresses the need for high-efficiency desalination under variable solar conditions, as conventional systems are energy-intensive and electricity-dependent for remote regions. To overcome these constraints, a solar-driven hybrid desalination configuration integrating thermal vapor compression (TVC), humidification–dehumidification (HDH), reverse osmosis (RO), and an energy-recovery turbine (ERT) is proposed. Parabolic trough collectors and thermal storage are used to stabilize operation. A thermo–exergoeconomic model, accelerated by an artificial neural network (ANN) surrogate, is integrated with NSGA-II for multi-objective optimization. To explicitly capture solar variability, Pareto fronts for irradiance levels (Gb = 1, 0.75, 0.5 kW/m2) maximize the energy utilization factor (EUF) and minimize the exergy-based levelized cost of product (LCOP). Results indicate marked improvements at the optimized operating point compared with baseline conditions: freshwater production increases from 42.57 to 68.86 m3/h, GOR rises from 14.74 to 23.84, and EUF improves from 8.563 to 15.88, while LCOP and LCOW decrease to as low as 0.2166 $/kWh and 2.06 $/m3, respectively, across the Pareto solutions. A case study for Siwa Oasis, Egypt, demonstrates the practical applicability of the proposed system for decentralized water supply in solar-rich remote areas. Overall, the ANN-assisted framework provides critical design trade-offs, advancing robust solar-assisted hybrid desalination.