A Leakage Detection System with an Efficient Prioritization at a District Meter Area Level

The leak repair process is one of the costliest operations for the water supply companies due to the amount of water lost, the customer inconvenience, and the further damages made to the infrastructure. To tackle this problem, a new approach is presented for detecting and prioritizing leaks at a district metered area (DMA) level. The performance of the algorithm is based on the extraction of the most meaningful features of the water flow daily, and the solution predicts the minimum night flow using neighbor DMA data to analyze and filter the noise for detecting and prioritizing leakages. The purposed solution has been validated with real historical data using a subset of 171 similar DMAs with data coming from January 2016 to July 2020. The algorithm was able to detect leakages successfully in nearly 72% of the cases, making around 18% of false positives. In conclusion, the presented approach consists of a full-fledged methodology able to predict the minimum night flow, detect leakages reliably, and prioritize them at a DMA level efficiently and reliably.