Multi-effect desalination: Dynamics, validation and optimization through diverse algorithmic frameworks

doi:10.1016/j.seppur.2026.136919

2026-01-15

PRE

S. Karishma

Amiya K. Jana *

DOI：10.1016/j.seppur.2026.136919

原文链接

原文求助

摘要

En 中文

This work presents a comprehensive multi-objective optimization study on a multi-effect desalination system with thermal vapor compression (MED–TVC), focusing on the parallel cross-feed (PCF) configuration. A dynamic model is developed to capture transient behavior under realistic operating conditions and is validated against data from four large-scale desalination plants, showing excellent agreement with the gain output ratio. To identify the operating regimes, which maximize performance ratio while minimizing freshwater production cost and environmental impact in terms of carbon dioxide emissions, and to explore trade-offs between them, a diverse set of evolutionary algorithms is strategized. All approaches identify feasible pareto-optimal solutions but exhibit distinct biases. For the MED–TVC system, genetic algorithm-based methods (non-dominated sorting genetic algorithm II (NSGA-II), NSGA-III, and strength pareto evolutionary algorithm II) yield well-balanced trade-offs; indicator-based method (indicator-based evolutionary algorithm (IBEA)) favors cost and efficiency-focused solutions; generalized differential evolution III (GDE-III) and swarm-intelligence-based methods (speed-constrained multiobjective particle swarm optimization (SMPSO) and optimized multiobjective particle swarm optimization (OMOPSO) methods) generate robust mid-range solutions. These findings underscore that algorithm selection directly influences engineering recommendations for MED–TVC systems. This work establishes a validated computational framework for sustainable optimization and highlights the importance of aligning algorithm choice with performance, cost, or environmental priorities.