Optimizing continuous monitoring sensor placement on oil and gas sites

We propose a generic, modular framework to optimize the placement of continuous monitoring sensors on oil and gas sites aiming to maximize the methane emissions detection efficiency. Our proposed framework substantially expands the problem scale compared to previous related studies and can be adapted for different objectives in sensor placement. This optimization framework is comprised of five steps: (1) simulate emission scenarios using site-specific wind and emission information; (2) set possible sensor locations under consideration of the site layout and any site-specific constraints; (3) simulate methane concentrations for each pair of emission scenario and possible sensor location; (4) determine emissions detection based on the site-specific simulated concentrations; and (5) select the best subset of sensor locations, under a given sensor budget, using genetic algorithms combined with Pareto optimization. We demonstrate the practicality and effectiveness of our framework through its application to an oil and gas emission testing facility with a large search space of possible sensor locations; a setting which is computationally infeasible to solve with commonly used mixed-integer linear programming formulations. Additionally, a case study illustrates the successful application of our algorithm to a real oil and gas site, showcasing its real-world applicability and effectiveness.