Objective Traditional gas monitoring methods for oil and gas leakage fall short in meeting the growing demands for the safe operation and maintenance of smart pipeline networks, due to factors such as low detection sensitivity levels, short calibration intervals, frequent false alarms, limited ranges of detectable gases, susceptibility to environmental interference, and reliance on charged sensors. Consequently, there is an urgent need to develop monitoring sensors that are suitable for detecting oil and gas leakage in emerging complex scenarios.

Methods Considering various confined space scenarios, including primary and secondary seals, pontoons, tunnels, underground spaces, and laboratories within the context of oil and gas pipeline storage and transportation, a new type of diffusion sensor probe was designed, which is suitable for field deployment. This probe features all-fiber coupling, intrinsic safety, and networking capabilities. It is specifically aimed at simultaneously monitoring multiple selected parameters, including combustible gases (methane, propane, and butane), combustion-supporting gas (oxygen), and temperature. By leveraging spectral analysis for complex systems and employing a design approach focused on background noise removal, traditional tunable diode laser absorption spectroscopy (TDLAS) technology was integrated with partial least squares (PLS) algorithm, non-negative least squares (NNLS) algorithm, and artificial intelligence models. Utilizing the following central absorption spectra—oxygen at −761.0 nm, methane at −1,653.0 nm, propane at −1,686.1 nm, and butane at −1,686.5 nm—a generic technique was proposed to address interference caused by absorption spectrum aliasing during the measurements of alkane gas molecules. Furthermore, a multi-component gas concentration measuring sensor was developed, suitable for detecting oil and gas leaks in various scenarios.

Results The developed sensor system was implemented in the long-distance, safe, and stable real-time monitoring and analysis of multi-component gases and integrated temperature data. With the capability of monitoring multiple parameters at various points across a wide dynamic range, it effectively met gas monitoring requirements in diverse application scenarios. Experimental testing of the sensor demonstrated the lower explosive limits (LEL) for detecting combustible gases: 0.1％ for methane, 0.8％ for propane, and 0.9％ for butane. Additionally, the minimum volume fraction of oxygen was recorded at 3％. These results fell within the minimum detection standard requirements established for the oil and gas industry.

Conclusion This in-depth analysis of various factors related to sensor interference provides insights for enhancing fire risk prevention and warning capabilities in oil and gas pipeline networks across multiple scenarios. It supports the goal of making pipelines safe and green conduits of development and friendship.