Sensor placement optimization for leak detection in a gas system installation
Please login to view abstract download link
Gas pipeline networks constitute critical infrastructure for energy transport, yet they are vulnerable to external or internal leaks that may lead to severe safety, environmental, and economic consequences. Early and reliable leak detection is therefore a major objective in pipeline monitoring systems. Modern approaches rely on distributed sensor networks measuring pressure, flow rate, or acoustic signals. However, the effectiveness of such systems depends not only on sensor precision but also on optimal sensor placement within the network, especially under constrained budgets and complex pipeline geometries. The present paper addresses the problem of sensor placement optimization for detecting the maximum gas leak scenario in gas pipelines. The problem is formulated as a coupled physical–optimization framework, where gas flow and leakage dynamics are modeled using simplified pipeline flow equations, while sensor locations are selected to maximize leak detectability. Optimization criteria commonly include minimizing detection time and maximizing the scenario number. This work contributes to safer pipeline operation by providing a systematic methodology for sensor deployment tailored to high-impact leak scenarios. The proposed approach is applicable to the Jupiter 1000 Power-to-Gas installation in France, which is the leading French industrial demonstrator in this field, where the key aspect of the facility involves transporting hydrogen gas.
