Experimental Study on Enhanced Methane Detection Using an MEMS-Pyroelectric Sensor Integrated with a Wavelet Algorithm

Abstract

An optical sensing approach that balances portability with cost efficiency has been designed for the reliable monitoring of fugitive methane (CH 4 ) emissions. Employing a LiTaO 3 -based pyroelectric detector integrated with micro-electromechanical systems and a broad infrared source, the developed gas sensor adeptly measured CH 4 concentrations with a low limit of detection of about 5.6 ppm v and showed rapid response times with t 90 consistently under 3 s. Notably, the novelty of our method lies in its precise control and reduction of CH 4 levels, enhanced by wavelet denoising. This technique, optimized through meticulous grid search, effectively mitigated noise interference noticeable at CH 4 levels below 10 ppm v . Postdenoising, nonlinear regression analyses based on the modified Beer−Lambert equation returned R 2 values of 0.985 and 0.982 for the training and validation sets, respectively. In conclusion, this gas sensor has been shown to be able to meet the requirements for early warning of CH 4 leakage on the surface in various carbon capture, utilization, and storage projects such as enhanced oil or gas recovery projects using CO 2 injection.