August 19, 1992 MEMORANDUM SUBJECT: Transcontinental Gas Pipe Line Corporation's Use of a Nonguideline Approach for Estimating the Amount of NO that is Converted to NO2 FROM: Dean A. Wilson, Meteorologist Techniques Evaluation Section, SRAB (MD-14) TO: Marcia Spink, Chief Air Programs Branch, Region III (3AT10) In response to your request, the Model Clearinghouse has reviewed the use of the proposed nonguideline technique for estimating annual average nitrogen dioxide (NO2) concentrations at the Company's compressor station near Unionville, VA. We have also discussed other relevant information regarding this source and regulatory aspects of the analysis with Todd Ellsworth of your staff. It is our understanding that the source is located in a rural area of Virginia and that background concentrations and contributions from other sources are essentially zero. Consequently, for purposes of comparison with the National Ambient Air Quality Standards (NAAQS), the concentration estimates need only reflect emissions from the compressor station. These estimates, using screening procedures recommended in the Guideline on Air Quality Models (Revised) (Guideline), were found to be above the NAAQS. Estimates using the Company's exponential decay procedures are below the NAAQS. Also, it appears from the State's December 12, 1991 letter to you that the highest concentrations were found to be associated with terrain interaction and downwash phenomena. I asked Mr. Shao-Hang Chu of our Branch, who has considerable experience and expertise in atmospheric nitrogen oxides (NOx) dispersion modeling and chemistry, to look at the procedure proposed by the Company for estimating NO2 levels. Mr. Chu agrees that the proposed five minute half life for NO to be converted to NO2 is conservative in most cases, as supported by the several literature studies that the company cites. At the same time Mr. Chu did note that there really is not enough data to clearly support the Company's assumption that 10 percent of the NOx is initially released as NO2. As you have recognized, the Guideline recommends a 3-tiered screening approach to modeling NO2. Apparently the source was unable to use the third level screen, application of the ozone limiting method on an hourly basis, due to lack of representative ambient NO2 and ozone data in the area. (Incidentally, there appears to be some conflict between the lack of ambient NO2 data and the previously mentioned assumption that the background ambient levels in the area are zero.) The important point to note here is that the Guideline recommendations are for screening procedures, and not refined models. The Guideline, in Section 3.2, also provides procedures for acceptance of a nonguideline model for a given situation. This Section essentially says that if there is no recommended refined model, then an alternative refined technique may be proposed for use. This proposed technique must be demonstrated to be applicable to the problem on a theoretical basis and that the data bases needed to operate the model are available. If these criteria are met, then either an on-site performance evaluation is necessary or a showing must be made that the proposed technique has not been shown to underpredict in similar circumstances. Given these premises, our analysis of the situation is as follows. The proposed technique while arguably conservative in most circumstances, is not a refined model, but yet another screening technique, apparently somewhat less conservative than the first two levels of screening recommended in the Guideline. The Guideline does not contain any provisions for acceptance of alternative screening techniques. We also note that the half life data contained in the literature comes almost exclusively from aircraft sampling of power plant plumes. There does not appear to be any data taken from situations involving terrain interaction of plume downwash. It should be noted that there is a wide range of half lives cited in the literature, ranging from 1 minute to 10,000 minutes. This suggests that an exponential decay mechanism may not be descriptive of the atmospheric chemistry involving NO2 formation and decay. This raises some questions about whether the Company's proposal is theoretically well founded, as required by the Guideline, Section 3.2. Given the above facts, if the action for this source were regulatory and the Environmental Protection Agency (EPA) were required to approve it, we would probably recommend against accepting the proposed technique. It is a bit unclear whether we can regard the modeling at this stage to be more investigative in nature, preliminary to State implementation plan (SIP) revision. If we could take this viewpoint, the Clearinghouse opinion is that the half life analysis suggests that there is some uncertainty on whether emission reductions are necessary and that more information should be obtained to make a more definitive conclusion. If the "more study" option is compatible with the EPA/State regulatory timeframe, then we suggest a couple of activities. As a minimum, the ambient NO2 and ozone data necessary to conduct the third level Guideline screen should be collected. This would require the collection of at least one year of such data, in an area representative of the ambient conditions at the source. Preferably, a network of NOx monitors should be installed to cover areas of expected high concentration. With careful planning, data from this latter option might be applicable in the "monitoring in lieu of modeling" approach described in Section 11.2.2 of the Guideline. If you have any questions, please contact me at 919-541- 5683. cc: S. Chu T. Ellsworth D. Grano J. Tikvart