Experimentally validated predictive control of feedwater flow in solar thermal desalination using forecasted irradiance

This study experimentally validates a forecast-based predictive control strategy for regulating feedwater flow in solar thermal desalination systems. While previous works have largely relied on simulation-only approaches or complex multi-component architectures, real-world deployment has remained limited due to challenges related to system complexity, computational demand, and scalability. To address these barriers, the present work integrates a simplified single-tank thermal energy storage (TES) unit with a low-complexity control algorithm driven by forecasted solar irradiance. Unlike conventional fixed-flow or reactive systems, the proposed approach determines the daily feedwater quantity in advance based on the next-day solar irradiance forecast, and applies hourly flow control adjustments in response to real-time thermal conditions. The system was deployed in Taif, Saudi Arabia, under dynamic solar conditions. It successfully maintained TES outlet temperatures within ±5% of the 75 °C setpoint, while adjusting feedwater flow by up to 27% to match thermal availability. Compared to baseline fixed-flow operation, the system improved solar energy utilization by 8.1% and reduced thermal losses by 14.6%. These findings confirm the technical feasibility of predictive control in field conditions and establish this method as a scalable, low-complexity solution suitable for off-grid desalination in resource-constrained environments. The demonstrated resilience and operational simplicity also offer a promising foundation for integration into decentralized, climate-adaptive freshwater systems.