LOCAL SCALE ATMOSPHERIC TRANSPORT MODELING FOR THE QUANTIFICATION OF METHANE EMISSIONS FROM INDUSTRIAL AND LANDFILL SITES

Accurate atmospheric monitoring of methane (CH4) emissions from industrial and waste processing sites such as landfills is essential to reduce their emissions. The monitoring of CH4 emissions from these sites often relies on local scale atmospheric transport models to simulate concentrations associated to the plume from a source with a given emission rate. In recent years, a growing interest in estimating CH4 emissions has led to the development of different atmospheric inversion frameworks based on stationary or mobile measurements and on transport modeling. The most commonly used model for atmospheric inversions at local-scale is the simple Gaussian model. Here, we investigate the power of a more complex model and compare the results with the Gaussian one. Large Eddy Simulations (LES) models are a promising tool for simulating the realistic CH4 concentration plumes from industrial and landfill sites. They are more versatile than the simple Gaussian model, because they can simulate the time evolution of the turbulent flow and the effects of a complex topography, and can be used not only for simulating the averaged data, but also for realizing high frequency “virtual measurements”. In this study, we utilized a LES model to simulate the CH4 plume dispersion from a point source controlled CH4 release in an industrial environment and from multiple complex area sources in a landfill. We characterize the LES of flow and plume dispersion by studying their convergence properties and their performance. The in-situ atmospheric CH4 mole fraction measurements from stationary and mobile sensors from the controlled methane release experiments are used as a benchmark of the LES model. For the landfill site, we use in-situ mobile measurements from a campaign and the LES simulations of plumes from the potential emission sources to estimate the total methane emission of the landfill. The inversion results for the landfill site with LES simulations are compared with those estimated by using a Gaussian plume model. The order of magnitude of the estimated landfill emissions obtained with the two transport models are found to be comparable, but the details of the transport are different when we deal with a complex topography. Future studies will be carried out to better exploit the high frequency data produced with LES. At this stage, we use only a coarse grid resolution for LES simulations of landfill plumes due to high computational cost, a grid sensitivity analysis will be conducted for better characterization of flow and plume dispersion.