Dynamic analysis and optimization design of parallel/cross-feed MED-TVC systems considering fouling characteristics

• Published in: International journal of heat and mass transfer Vol. 238; p. 126490
• Main Authors: Zhang, Hao, Zhang, Yuanmin, Song, Xuewu, Zhao, Hongxia, Sun, Wenxu, Jia, Lei
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.03.2025
• Subjects: ANN surrogate model; Dynamic analysis; Input disturbance; MED-TVC desalination systems; Non-fouling operational optimization; NSGA-II; Non-fouling operational optimization; NSGA-II; MED-TVC desalination systems; ANN surrogate model; Input disturbance; Dynamic analysis
• ISSN: 0017-9310
• DOI: 10.1016/j.ijheatmasstransfer.2024.126490

•A new approach for the optimization design of MED-TVC considering fouling characteristics.•Dynamic analysis of input parameters on fouling crystallization is conducted.•ANN surrogate models for input parameters and performance metrics are developed.•Quantitatively compared optimization results with and without fouling consideration.•Achieved a maximum PR of 12.96 and a minimum FWPC of 1.157 $ m−3 considering fouling. Multi-effect distillation with thermal vapor compression (MED-TVC) desalination systems are regarded as effective solutions to address freshwater scarcity. However, fouling within the evaporator significantly reduces freshwater production and restricts long-term operational efficiency. To this end, this study proposes a non-fouling operational optimization approach for MED-TVC systems utilizing artificial neural networks (ANN) and non-dominated sorting genetic algorithm II (NSGA-II). A dynamic model is developed to reveal the sensitivity of various input parameters concerning fouling formation, performance ratio (PR), and freshwater production cost (FWPC), whose accuracy is validated against field data. The ANN surrogate model and NSGA-II are employed to determine the optimal input parameters under Case-1 (random fouling) and Case-2 (non-fouling) scenarios, focusing on maximizing PR and minimizing FWPC. Additionally, a dynamic response analysis of four input disturbances is conducted under optimized non-fouling conditions to prevent fouling caused by parameter fluctuations. Results indicate that PR and FWPC are most sensitive to the motive steam flow rate (mms) and the number of effects (N), whereas fouling formation is more subject to the seawater flow rate (msw) and seawater temperature (Tsw). The ANN surrogate model demonstrates reliable predictive performance, achieving a maximum PR of 12.96 and a minimum FWPC of 1.157 $ m−3 following non-fouling optimization. This paper offers valuable insights into achieving non-fouling and efficient operation of the MED-TVC systems.