An integrated approach for non-revenue water reduction in water distribution networks based on field activities, optimisation, and GIS applications

Abstract

This paper describes the development of an integrated approach for water pipe network calibration and lea quantification. The proposed approach provides a practical procedure that merges field measurements and mathematical modeling for leak quantification and network calibration. The use of Evolutionary Algorithms (EA) using both Genetic Algorithms (GA) and Shuffled Complex Evolution – University of Arizona (SCE-UA) to calculate the pinpointed leak outflows well as to identify any faulty meters. Also, by treating pinpointed physical leaks as junction points, the model, via EPAnet connection, calibrates the network and calculates their demands (which is equivalent to physical losses) and estimating pipes roughness factor (CHW) for better network calibration. Besides, throughout the process, the developed model uses GA and SCE-UA but departs from traditional use in using floating-point representations and introducing a new self-adaptive constraint handling function, which yields optimal results in fewer generations (model runs). Two pilot networks are selected, a well-known hypothetical network (Hanoi network) to present the general mathematical model, while a real network (Faisal city network) is used to test the integrated approach. The model integrates EPAnet for the required hydraulic modeling during the simulation and GIS for input data integration and output representation during the whole procedure. The results prove the approach's accuracy and efficiency.

Keywords

• Leak detection
• Network calibration
• Optimization
• Water losses