Enhanced mathematical modeling for optimizing solar stills with AI exploitation

Abstract

As global water demand rises, 2.2 billion people face clean water scarcity , exacerbated by climate change . Addressing this, the United Nations' SDG 6 targets sustainable water management. Current technologies for providing water resource vary from natural reserve assessments to advanced treatment methods, including desalination and recycling. The present study focuses on solar still desalination, particularly suitable for off-grid applications due to its integration with renewable energy. This paper introduces two innovative open-source Python model designated in order to optimize solar stills designs, accounting for various parameters and materials, balancing efficiency and cost. The first one was validated by experimental data, the model accurately predicts performance with a 4 % error margin and the second software adopts an approach allowing machine learning (ML) for data analysis, focusing on clustering and prediction tasks. This algorithm has RMSE of 0.027. This research underscores solar stills' potential in delivering a sustainable, cost-effective solution for areas lacking water infrastructure, thus contributing to achieving SDG 6. The findings advocate for a nuanced approach to material and design choices, and other adjustment such as baffle inclusion or reducing water film thickness, markedly influence output considering both economic and environmental feasibility. • Open-source model to optimize solar still designs and predict performance • Applied IAPWS equations for water, saline water, and moist air • Included head losses and condensation correlations • Validated model with experiments; achieving RMSE = 0.027 for ML model • Demonstrated design impact, water film thickness, and wind effects