Difference analysis and recognition of hydraulic oscillation by two types of sudden faults on long-distance district heating pipeline

Abstract

The implementation of long-distance district heating (LDH) systems, which rely on renewable energy sources, waste heat recovery, and waste heat reuse, offers a scientifically supported and environmentally friendly approach to achieving low-carbon emission reduction goals. As the common sudden failure on the long-distance district heating pipeline (LDHP), the leakage and valve failure will lead to strong hydraulic oscillation, which may cause heating failure if they are not recognized and handled in time. In this paper, the difference in hydraulic oscillation induced by leakage and valve failure is investigated with the hydraulic transient model. Based on the above analysis, a recognition method combining the support vector machine (SVM) and particle swarm optimization (PSO) is proposed to identify these two types of sudden faults. Taking a 20 km LDH system as an example, the effectiveness and robustness of the recognition method are investigated. The results show that the types, locations, and status of these two types of sudden faults can be accurately recognized and the robustness of the recognition method is strong when the sample time is less than 10 s. Our research deepens the understanding of the characteristics of the hydraulic oscillation induced by the leakage and valve failure on the LDHP and gives a new idea for the intelligent recognition of sudden faults online.