LAUSR

Abstract Long-distance district heating pipelines (LDHPs) are important infrastructures in large scale district heating (DH) system. Their safety and stability are crucial for stable and efficient operation of the whole DH system. Pipeline leakage is a common accident in the operation of DH system, accurate and fast leak detection technique is an indispensable measure for ensuring efficient repair and avoiding large-area heating failure. In this paper, a fast and accurate leak detection technique for LDHPs is proposed. The technique utilizes the numerical hydraulic transient model of the LDHP and the particle swarm optimization (PSO) algorithm to find the leak point. A pressure wave velocity identification procedure is also established for the leak detection. Simulation tests are conducted on a 20 km LDHP to investigate the influences of instrumental error and sampling frequency of measured data on the accuracy of the leak detection. Results show that when the sampling frequency is higher than 0.1 Hz, the relative errors of the detected leak point areas and leak point positions can be less than 2.73 % and 2.40 %, respectively. Effective leak detection of LDHPs can be therefore conducted through the proposed leak detection technique and measuring instruments with high precision and proper sampling frequency.