第一届未来大气科学论坛

A single-blind methane emission detection and quantification test was conducted at the Methane Emissions Technology Evaluation Center (METEC) using a novel optical gas imaging-based methane leak continuous monitoring system. This system can monitor methane emissions 24 hours a day and 7 days a week. It can send alarms along with detection reports when methane emissions are detected. Detection videos are also available and can help trace leak sources quickly. This system is powered by an in-house data analysis algorithm that enables the system to quantify the methane leak rate in real time. The test facility simulates real life outdoor industrial scenarios. The equipment monitored during this test includes simulated separators, tanks, and well heads. Controlled releases were conducted by the METEC operational team with the emissions points, leak rates and emission start and end times unknown to the testing participants. The data was first analyzed using the Advancing Development of Emissions Detection (ADED) continuous monitoring protocol R1.0. After that, a different methodology developed by the Brandt group from Stanford University was used to analyze the same data set to offer a different perspective on the performance of this system. During the data analyzing process, the influence of different methodologies on the perception and presentation of system performance was noticed. While one methodology may provide certain insights into a certain methane leak detection and quantification system’s performance, it’s also limited by the experimental design upon which the methodology was developed. A new performance metric was proposed in this study that incorporates multiple performance indicators into one unifying metric. This new metric incorporates detection speed, number of detection reports associated with a certain controlled release, degrees of overlap between controlled release periods and reported detection periods, and quantification performance all in one formula. At the same time, this new metric also attempts to accommodate different types of experimental setups, such as single emission point versus multiple emission points testing scenarios, so that it can be applied to a broad range of experimental designs.  This new metric offers a new way to understand and compare the overall performance of different types of methane leak detection and quantification systems. Results and implications of this new metric are presented and discussed.
 

甲烷,ogi,气体泄漏