Abstract: In order to resolve the problems existing in the application of the water networks integration optimization method in the thermal power plant, the transfer process of mass impurities from water regeneration units was expanded in this study. The complete mass transfer process was used to describe the process of water utilization of impurity load units, and the maximum mass transfer capacity and rate of regeneration are used to describe the processes of regeneration. In addition, a directed graph was used to describe the connections between units of water networks. Through analyzing the relationship between the maximum concentration, impurity load, and water supply process of water units, the mathematical mechanism of the optimization process was obtained so as to build a regeneration recycling water networks optimization model. A 3300 MW thermal power plant was chosen to conduct model application research. The water networks of the case power plant was divided into 9 water subsystems, and equivalently divided into 11 water units, then a concentration correction model had been constructed for the regeneration unit to achieve the maximum mass removal process. The chloride ion, entire solids of suspended matter, and sulfate ion were picked to perform a water-saving optimization design, and the model was simplified by solving the KKT conditions of the optimization problem. After optimization, the use of fresh water of the water networks was reduced by 6.7%, total emissions of sewage were reduced by 62.9%. The case study showed that the constructed regeneration recycling water networks model based on the "supply-utilization-consumption-discharge" process of water networks was successful, and it could realize the water-saving optimization design of the water networks in thermal power plants considering multiple impurities. Furthermore, it resolved the issues of inaccuracy in the description of the regeneration process in the use of the integrated optimization method of water networks in thermal power plants, as well as the inappropriate description of the water regeneration unit for water-saving reconstruction of water networks.

Key words: water system integration, optimal design of water networks, process systems, nonlinear optimization, industrial water conservation

摘要： 为解决水网络集成优化方法在火电厂水网络优化应用中存在的问题，该研究对杂质的传递过程进行了扩展，使用移质推动描述杂质负荷单元用水过程，以及用最大移质能力及移除率描述再生过程，并通过邻接矩阵描述网络单元间的连接关系，从而构建再生循环水网络模型；对3300 MW火电厂进行模型应用研究，将案例水网络划分为9个用水子系统，等效拆分为11个用水单元并对移质单元构建浓度修正模型从而实现最大移质过程，选取Cl^-、总悬浮固体(SS)、SO_4^(2-)三个杂质指标进行节水优化应用，通过对该优化问题KKT条件进行求解从而实现模型的简化，优化后水网络新鲜水取用量减少6.7%、排污量减少62.9%。通过案例研究表明，构建的基于“供-用-耗-排”过程的再生循环水网络模型是成功的，可实现火电厂多杂质指标、复杂系统考虑再生循环的水网络节水优化设计，解决了水网络集成优化方法在火电厂应用中单元间回用过程描述不够准确以及再生单元描述不适用于节水改造情景的问题。

关键词: 水系统集成, 水网络优化设计, 过程系统, 非线性优化, 工业节水