Multi objective genetic algorithm approach to reduce sanitary sewer overflow

Sanitary Sewer Overflow (SSO) is the discharge of wastewater from the collection network into the environment. The EPA estimates that up to 75,000 SSO events occur in the U.S. each year. SSOs can occur during dry or wet weather and are significant environmental and public health hazards. One of the main causes of SSOs is excessive rainfall derived inflow and infiltration into the network. This study applies a Multi-Objective Genetic Algorithm (MOGA) to identify near optimal solutions to minimize SSO occurrences and costs for sewer rehabilitation strategies. Two approaches are investigated: (1) enhancing the flow capacity of the collection and conveyance system by pipe diameter increase and (2) peak flow reduction using decentralized inline storage tanks. For the flow capacity enhancement strategy, the decision variables are the number of segments to be replaced, their locations and in how many commercial diameters the segments increases. For the peak flow reduction strategy, the decision variables include the number of tanks, their locations and volume. The approach is tested in a 5.9 square miles sewer network, located west of downtown San Antonio, Texas. The MOGA approach characterizes the trade-off between SSO reduction and cost, and provides stakeholders a better understanding of the system and flexibility in the decision making process to eliminate SSOs.