Detection of Methane Leaks via Drone in Release Trials: Set-Up of the Measurement System for Flux Quantification

In the oil and gas sectors, as well as in waste landfills, the commitment to greater sustainability is leading to increased efforts in the search for methane leaks, both to avoid the emission of a major greenhouse gas and to enable greater fuel recovery. For rapid leak detection and flow estimation, drone-mounted sensors are used, which require a balanced configuration of the detection and measurement system, adequate for the specific sensor used. In the present work, the search for methane leaks is carried out using a tunable diode laser absorption spectrometer (TDLAS) mounted on a drone. Once the survey is carried out, the data obtained feed the algorithms necessary for estimating the methane flow using the mass balance approach. Various algorithms are tested in the background measurement phases and in the actual detection phase, integrated with each other in order to constitute a single balanced set-up for the estimation of the flow emitted. The research methodology adopted is that of field testing through controlled releases of methane. Three different flows are released to simulate different emission intensities: 0.054, 1.91 and 95.9 kg/h. Various data configurations are developed in order to capture the set-up that best represents the emission situation. The results show that for the correction of methane background errors, the threshold that best fits appears to be the one that combines an initial application of the 2σ threshold on the mean values with the subsequent application of the new 2σ threshold calculated on the remaining values. Among the detection algorithms, however, the use of a threshold of the 75th percentile on a series of 25 consecutive readings to ascertain the presence of methane is reported as an optimal result. For a sustainable approach to become truly practicable, it is necessary to have effective and reliable measurement systems. In this context, the integrated use of the highlighted algorithms allows for a greater identification of false positives which are therefore excluded both from the physical search for the leak and from the flow estimation calculations, arriving at a more consistent quantification, especially in the presence of low-emission flows.