| Les Analysis of Mixing Effects on Sour Gas Flare Emissions Under Variable Wind Conditions | ||
| Philip J. Smith, University of Utah, 155 S. 1452 E. Rm 380, Salt Lake City, UT 84112, Christoper M. Thurston, Dept. of Chemical Engineering, University of Utah, 50 So. Central Campus Drive, 3290 MEB, Salt Lake City, UT 84112-9203 and Allan Chambers, Carbon and Energy Management, Alberta Research Council, 250 Karl Clark Rd, Edmonton, AB T6N 1E4, Canada Hydrogen sulphide (H2S) is a common component of gas streams flared in Western Canada and thus the level of mixing and combustion efficiency of these flares under various operating conditions is a cause for focused analysis. Buoyancy-driven combustion plumes like fires and flares have been particularly difficult to simulate with traditional computational fluid dynamics (CFD) Reynolds-Averaged Navier-Stokes (RANS) approaches because of the large-scale mixing due to vortical coherent structures in these flames that are not readily reduced to steady-state CFD calculations with RANS. The availability of massively parallel computers using hundreds to thousands of processors has recently brought Large Eddy Simulation (LES) within reach of industrial flare applications. This sour gas flare simulation couples hydrocarbon and sulphur chemistry with fluid mixing and radiant heat transport to predict how sulphur compounds are distributed in the flare plume, and how these compounds are locally mixed and dispersed into the environment. The impact of H2S on flare combustion inefficiency and overall flare emissions is analyzed by using LES simulation and hierarchal validation assessment. A range of flare operating conditions have been studied to give insight into the impact of operating conditions on sour flare performance. The combustion inefficiency of two industrial scale sour gas flares and one sweet gas flare was measured using DIAL equipment in Alberta during 2003 and 2005 to examine sensitivity to H2S concentration and crosswind velocity. These measurements are used for validation analysis. A simulation test matrix of computations has been completed that spans wind velocities from 0-15 m/s, H2S concentrations from 0-30%, and fuel exit velocities from 1-15 m/s. Information from these simulations have been analyzed to: -track combustion and sulfur efficiency in the flares and as compared to the DIAL measurements, -track mass fluxes of CH4, H2S, SO2, SO3, and elemental sulfur out of the computation domain. Extended Abstract Status: Not Uploaded | ||