FEAST on MEET: Temporal and Spatial Modeling for Statistically Robust Evaluation of Leak Detection and Repair Programs in Oil and Gas

Abstract:

Private and governmental investments have rapidly accelerated the development of next-generation (next gen) leak detection solutions for oil and gas (O&G) operations. Unlike traditional solutions in most regulatory regimes, next gen solutions utilize downwind and optical sensors interpreted by analytics. These solutions also use deployment modalities differing substantially from traditional solutions, resulting in dramatic performance variation over environmental conditions and facility characteristics.

This study recaps protocol development, testing and modelling programs to assess next gen performance driven by a multi-university team. The models allow traditional and next generation methods to be compared by statistically assessing their impact on total emissions. This work includes three components:

1. a) Development of a temporally- and spatially resolved emissions model, the Methane Emissions Estimation Tool (MEET). Since O&G emissions are highly variable – and next generation solutions highly mobile – assessment of these solutions requires statistical emissions models that better model variability in time and space; annualized averages will not work. We recap the model structure and provide examples of results drawn from recent field studies.
2. b) Integration of a time-resolved leak detection simulator, the Fugitive Emissions Abatement Simulation Tool (FEAST) onto MEET. FEAST statistically simulates leak detection and repair (LDAR) programs. Using a defined emissions model, the efficacy of multiple LDAR programs, including reference programs, can be compared. We present an example using highly variable midstream emissions.
3. c) Finally, we recap recent progress from a controlled testing and field trials which will collect the necessary performance data for FEAST, and site emissions and activity data to populate the underlying MEET model. The testing program will provide a specific test case to compare controlled and field-testing results.