Abstract Description: As a Gaussian dispersion model, AERMOD is built on the basic dispersion formular that the result concentration is a proportional function of the source emission rate, regardless of source type, stack parameters, building downwash, meteorological data, terrain effects, etc. With the addition of various tiers of screening techniques of NO2 to NOx conversion, such as ARM2, OLM, PVMRM, GRSM, etc., the proportionality is no longer always valid. However, it is still expected that the result concentration is still a monotonic increasing function of the source emission rate. In other words, it is expected that higher emission rate will always result in higher model predicted concentration at any given receptor.
In a recent modeling analysis, it was noticed that at a low NO2 to NOx in-stack ratio, one of a series of sensitivity test on emission rates shows an unusual result that a lower emission rate input resulted in higher model predicted concentration, with PVMRM invoked. The research was presented at the AWMA modeling conference in November 2024, and the research was not complete yet. This presentation will present further developments of this finding and exploring the reason of this non-monotonic behavior of PVMRM, and discuss good practices when encounter similar situations.
