Leak Detection in Water Distribution Networks Based on Water Demand Analysis

This paper deals with the problem of leak detection in Water Distribution Networks (WDN). A method based on water demand analysis District Metered Areas (DMAs) is proposed. Historical leak-free data flow used to extract minimum, maximum values, and statistical distributions differences (errors) between predicted values at different time hours day. The concept sensor fusion applied reduce measurement uncertainties. For this, a virtual generated that considers each hour day feature and, combined, develops more accurate error capable detecting leaks estimating size magnitude. Furthermore, increase accuracy method, prediction errors are analyzed moving window. Finally, performance proposed assessed by using actual real DMAs Barcelona WDN.