Regulatory Memo From EPA Model Clearinghouse

January 22, 1997 MEMORANDUM

Subject: Nonmethane Organic Compounds (NMOC) and Nitrogen Oxides Monitoring Required for the Empirical Kinetics Modeling (EKMA) for use in Colorado

From: Ellen Billingham, AQM Specialist
EKMA Modeling Group

To: Air Quality Services, Inc

In response to your request on the above subject, the Model Clearinghouse has reviewed your position on the application of EKMA Colorado.

Guidance on the use of EKMA (Environmental Protection Agency, 1989) is silent on the use of default ratios in EKMA State Agency implementation plan (SIP) analyses. This implies that the use of state ratios is not generally recommended. Therefore, we endorse the approach you mention, which entails collecting NMOC and NO_x data in your area consistent with the monitoring guidance and requirements of the EKMA. We believe this represents the most defensible modeling procedure and is the approach most consistent with the 1989 guidance for using EKMA.

We understand, however, the concern you raise over the time constraints you are working under in your Scope of Work agreement with Lotsonox, Inc. We agree that, prior to the availability of NMOC and NO_x monitored data in your area, it may be necessary to resort to other procedures for estimating the NMOC/NO_x ratio for use in your OZIP modeling runs.

We agree that use of a default ratio is the only viable interim procedure open to Colorado for use in the southeastern part of the state. A question that arises as to what default value should be used. As you have noted, earlier guidance suggested "11.5:1" as a default value. More recent data compiled by Bates (1991) suggest typical 50th percent ratios to be lower (9.8:1) . This appears to be more in the middle of the distribution of ratios found in nearby cities. For example, the range of 25-75 percentile values of the ratios in Denver, Fort Collins, and Colorado Springs is 9.0-9.9; 10.4-11.1; 9.8-11.0 respectively. Thus, we recommend use of a default using most recent data. As such, any choice of a design ratio between the values of 9.0 and 11.0 which provides the best estimate of reductions needed is acceptable for use with the OZIP model. We recommend that you calculate a median value based on the ranges of these three representative locations.

Please do not hesitate to direct further questions or concerns to the EKMA Modeling group (ekma@shodor.org).

Developed by