Abstract Description: In the event of an accidental release of unauthorized emissions, facilities will need to report the excessive emissions as an emissions event. Air quality enforcement and reviewing agencies will require air dispersion modeling analysis to evaluate if the emissions event has posted any significant impact on the ambient air quality.
A case study on an emissions event from an unauthorized flaring event will be shared and discussed. AERMOD is used as the dispersion model platform for the analysis. Flare is not a direct source type of AERMOD, and it is typically modeled as a pseudo point with a prescribed set of stack exit temperature and velocity, the flare tip height as the stack height, and an equivalent stack diameter calculated based on the net heat release of the flared stream.
The goal is to show a modeled impact as low as possible, for the purpose of presenting a case to the reviewing agency that the emissions event has not pose a significant impact to the ambient air. Therefore, a tiered approach with different level of efforts is discussed.
1. Screening Modeling with AERSCREEN or SCREEN3 2. AERMOD with Pre-Processed Worst-Case Meteorological Data 3. Emissions Event Duration Specific Met Data Processing 4. Hourly Varying Emission Rate 5. Hourly Varying Stack Parameters
In the last level described above, AERMOD is capable of varying stack exit temperature and velocity in the hourly emission file. However, in this case, the net heat release from the flared stream is varying by hour and therefore the equivalent stack diameter. AERMOD cannot directly address the varying diameter, but it is possible to set it up indirectly, by creating a source for each hour. Programming could be involved in this step, to automatically generate a large number of sources and their varying emissions. A recommendation that AERMOD may consider to allow varying all source parameters in the hourly emission file, had been made to the Environmental Protection Agency (EPA). Based on communication with EPA, this update is being considered for next release of AERMOD.
With the refinements and iterations described above, a clear trend can be seen that the model results are becoming lower, however still being representative of the impact due to the actual emissions event. By presenting the representative result with more accuracy built into the reviewing agency, the facility can expect to avoid unnecessary excessive enforcement actions compared to if the results from an overly conservative model.
.jpg)
