The 2010 1-hour NO2 National Ambient Air Quality Standard (NAAQS) is considerably more stringent than the longstanding annual standard, greatly reducing the margin for compliance. Consequently, accurate NO2 concentration estimates are needed which means that both accurate NOx concentration estimates, and accurate NO2/NOx ratios are needed. To evaluate AERMOD accuracy, an ambient NO2/NOx AERMOD performance assessment and model improvement project was conducted. First, a comprehensive dataset was collected based on 13 months of monitoring at a compressor station in Oklahoma, which included ambient monitoring of ozone (O3) and oxides of nitrogen (NOx and NO2) at four locations, meteorological data, and reciprocating engine emissions. Comparisons between the monitored data and AERMOD results showed NOx and NO2 were both over and under-predicted, sometimes beyond a factor-of-two. Evaluations were then carried out to determine if better NOx predictions could be obtained using PRIME2 building wake enhancements along with modified building dimensions. Better agreement with NOx observations was found using PRIME2 and alternate building dimensions.
Next, an evaluation of the method for computing the conversion of NOx to NO2 was conducted. The most sophisticated tier-3 Plume Volume Molar Ration Method (PVMRM) was used for this purpose. A comparison of computed to observed NOx/NO2 ratios at the four monitor stations showed PVMRN was tending to significantly overpredict that ratio. A detailed evaluation of the PVMRM theory was then conducted which revealed the following theoretical problems: 1) PVMRM calculations with and without building downwash effects produce identical plume volume estimates and hence identical NO2/NOx ratio estimates demonstrating model formulation problems; 2) for near-field impacts, PVMRM assumes that the plume volume is a cylinder instead of a cone leading to overestimates of plume volume; 3) PVMRM plume volume calculations are over-estimated for cases where building downwash effects are insignificant due to the actual plume dimensions not being used in the calculations. Minimum larger plume dimensions are used as default values at all downwind distances unless or until the actual plume dimensions are greater; 4) PVMRM plume volume calculations for scenarios where building downwash effects are important, do not use the PRIME computed plume dimensions.
This paper will discuss the importance of accurate NOx and NO2/NOx ratio estimates and provides evidence that these estimates are currently not sufficiently accurate using AERMOD and PVMRM. PVMRM is a necessary model option for NO2 compliance demonstrations however it contains model formulation errors. PVMRM should be revised as soon as practicable to provide improved model estimates and performance. Revisions detailed in this paper are critical to model performance. The paper will also show that improved accuracy for NOx estimates can be obtained using PRIME2 and alternate building dimensions.
Once PVMRM is updated and better agreement between AERMOD and NOx observations is confirmed, comparisons of AERMOD NO2 predictions versus observations should be performed. Based on limited analysis to date it appears that to get good agreement with NO2 observations, the NSUBZ factor currently used in PVMRM will have to be changed to a much lower value.
.jpg)
