Integrated Sensor Placement and Leak Localization Using Geospatial Genetic Algorithms

Abstract

There is an urgent need to reduce water loss from drinking water distribution systems. A novel framework that integrates the placement of multiple pressure sensors and localization using geospatial techniques is developed and validated to find leaks/bursts as they occur within district meter areas (DMAs). A data-driven leak/burst localization technique, featuring a novel spatially constrained inverse-distance weighted interpolation technique, was developed that quantifies the change in pressure due to a new leak/burst event using pressure sensors deployed in a DMA. The integrated framework uses the same modeling results and geospatial search techniques in both the optimal sensor placement and leak/burst localization steps. It can be adapted for any data-driven or model-based leak/burst localization technique and is not dependent on high hydraulic model calibration requirements such as high density smart meter deployment. Validation is presented using data from 16 engineered events (field work flushing) conducted in an operational DMA. Results show good agreement between the leak/burst localization performance for real and modeled engineered events, demonstrating that the sensor placement technique can accurately predict the expected performance of an operational DMA. This is particularly the case as the number of optimal sensors increases. Engineered events as small as 3.5% of the peak daily flow (6% of the average daily flow) were correctly localized with search areas containing as few as 14% of the pipes in the DMA (using only four pressure sensors).