Optimal Synthesis of Batch Water Networks with a Flexible Scheduling Framework Using Dynamic Programming

Abstract

Abstract Batch processes have received considerable attention from both industry and academia because of their flexibility to produce low volume, high value-added products. Sustainable water management and strict environmental regulation drive the development of systematic methods for water integration in batch plants. In batch processes, scheduling plays a critical role towards resource conservation. In some cases, production rescheduling can further reduce freshwater consumption and wastewater generation. Although many papers about simultaneous optimization of batch water network integration and product scheduling have been published, the insight-based method is scarce. This work proposes a dynamic programming (DP) method for the optimal design of batch water networks with a flexible product scheduling. The proposed methodology is explained as follows. Firstly, water-using units are ranked based on the ascending of inlet limiting concentration. The whole process is divided into N stages and each concentration locates at a specific stage. Secondly, the target of freshwater consumption of the process can be determined using the recursive function of DP model. Thirdly, once the target is determined, the optimal match between water sources and sinks are identified. The timing of each operation could be further determined. If the batch water network is feasible, the final batch water network is determined. If the obtained water network is infeasible, it will return to the second step to relax the target or chose another alternative with the same target until obtaining feasible batch water network. Two illustrative examples from literature were considered to demonstrate the feasibility of the proposed method. In the first case study a regenerator with a fixed outlet concentration is incorporated in batch water network. In the second case study is presented a batch plant with multiple contaminants in which a fixed removal ratio model is employed.