Using GIS and Spatial Optimization Techniques for Improvements in RDII Studies

Urban sanitation networks are difficult problems to analyze due to the non-Euclidean nature of pipelines that requires global solutions to optimization problems. The placement of flow monitors within such a network requires a graph theoretic approach that simultaneously considers all possible solution sets. Currently, it is common practice to adopt an arbitrary, local and sequential method for placing a flow monitor. In an effort to improve the results of a sanitary sewer inflow and infiltration study, the spatial optimization of 53 flow monitoring devices was studied. To achieve optimization, the principals of graph theory, specifically the facility location problem, was implemented with allowances for flow directions. Using a directed graph vector line drawing of the Lawrence Indiana sewer system, a network was created, and a facility location model was applied for 53 locations using a weighted length of pipe as the connection cost. The results showed a 45 percent increase in the amount of sewer segments that could have been monitored using this methodology instead of the ad hoc placement of monitors that were based on sequential and hence sub-optimal choices. This translated to a 50 percent increase in relative efficiency.