Leakage Localization through modelling - Creating a method for leak localization in plastic pipes

Abstract

Maintaining and managing the drinking water supply is one of the biggest challenges in the years to come globally. Dwindling supply in some parts of the world leading to water shortages and too much untreated water in other parts puts the water supply under constant and increasing strain. When drinking water is becoming harder and harder to acquire, the management of the clean water we produce become one of the top priorities. Today, the water loss in Sweden is on average 20 %, which is higher than average compared to other countries in Europe. To reduce the water loss, several methods may be applied to locate the leakage. One such method is based on a calibrated MIKEURBAN-model and pressure measurements in the area. The company DHI created the Leak Localization software in 2016. Based on the commonly available MIKEURBAN modelling software, Leak Localization takes pressure measurements in the area and compares them to the calculated pressure from the MIKEURBAN-model. Based on this difference, the software calculates the probability of the leakage location and then generates a GIS-map to display the results. In this project, a standardized method will be created to facilitate the usage of the software. The method will also be tested and evaluated based on the efficiency of the entire process. All parts of the method will be actively used, investigated and explained, and in most cases compared to other methods of data gathering or leakage localization on the market. A full comparison between all available leakage localization techniques are impossible due to the limited scope of this project. The focus of the comparison will therefore be on the data gathering aspect of the method created. The results reveal several flaws but also the potential of the software. The possibility of a passive system that activates when the pressure drops and automatically locates leakage regardless of pipe material or dimension may change the approach of leakage localization. To fundamentally change established methods previously requiring a significant number of man-hours to an automated process is a significant advance. The main flaws are the difficulty of placing a significant number of pressure sensors on the drinking water network, the unrefined user interface and the lack of an automated calibration method for MIKEURBAN. Nearly all the flaws are currently being worked on, making this method of leak localization one of the most viable in the future.