Optimal Planning of Regional Water Distribution Systems: A Case Study

A well planned regional water distribution system could provide a reliable and high quality drinking water supply to the residents. A framework for optimal planning of regional water distribution systems is presented. The framework is comprised of four stages: (1) determine the scope of the regional water supply system, (2) partition to subregions and establish water supply relationship, (3) optimal connection layout of transmission mains between the subregions with water supply relationship, and (4) pipe diameter optimization of the regional water distribution system. Most work of the first two stages is left to the decision maker of municipal planning bureau and water utilities. For the last two stages, it is concentrated on setting up mathematical models and finding optimal solution for these models. Dijkstra Algorithm is used at the third stage to find the least cost for layout of transmission mains between adjacent subregions. At the fourth stage, a model for diameter optimization of regional water distribution systems was proposed with the objectives of minimal investment and operation costs and maximum entropy reliability. The model, in which diameter of the pipes need to be optimized are the decision variables, takes into account both economic and reliability issues. Linked to EPANET as the hydraulic engine, the shuffled frog-leap algorithm (SFLA) is adopted to solve the model. Finally, the proposed planning framework is applied and demonstrated on the planning of the real-life regional water distribution system in Tianjin City, north China, which serves more than 10 millions people. The Tianjin case could play an exemplary role to the other similar planning of regional water distribution systems.