PRCI ambient NO₂ AERMOD performance assessment and model improvement project: Modeled to observed comparison

On January 22, 2010, the U.S. Environmental Protection Agency (EPA) established a new one-hour probabilistic National Ambient Air Quality Standard (NAAQS) for nitrogen dioxide (NO₂) at 100 parts per billion (ppb) or approximately 188 micrograms per cubic meter (µg/m³). The 1-hr NO₂ NAAQS is considerably more stringent than the longstanding annual standard of 53 ppb or approximately 100 µg/m³. AERMOD is the EPA recommended dispersion modeling tool for predicting impacts from emission sources, including compressor station reciprocating engines. Existing datasets were previously reviewed to evaluate and improve AERMOD 1-hr NO₂ model performance. However, these datasets were not representative due to either an absence of accurate concurrent hourly emission inventories, being associated with larger electric generating units, shorter duration tracer studies, or were too site-specific (e.g., external influences such as terrain). To address this data gap, an intensive 13-month field data program was undertaken by Pipeline Research Council International (PRCI) to collect a comprehensive dataset including concurrent emissions, meteorological, and ambient ozone and nitrogen oxide measurements. This paper describes the Balko, OK natural gas compressor station host site and the design of the meteorological, ambient NO_x and O₃ monitors and continuous Parametric Emissions Monitoring System (PEMS). Two comparisons are made between the monitored data and AERMOD results. First, the measured onsite data (ambient air, meteorological, and emissions) were used as input to AERMOD to make comparisons between measured and modeled NO_x and NO₂. These comparisons showed that NO_x was both over and under predicted, sometimes beyond a factor-of-two. The second comparison uses the typical permitting approach where onsite data are not available to contrast how conservative input assumptions for emissions, background O₃, NO₂, and offsite meteorology cause significant over predictions especially for this site where engines have a low annual utilization rate and operate infrequently throughout the year.Implications: This project resulted following the promulgation of the new 1-hr NO₂ NAAQS and the challenges that natural gas pipeline compressor stations have in modeling intermittent emissions using AERMOD to demonstrate compliance with this standard. AERMOD was developed and validated with a primary focus on large sources with tall stacks, such as electric utility boilers. A recently collected dataset at a gas compressor station in Balko, Oklahoma shows that the use of permitted allowable emission values compared to actual hourly emission values causes over predictions using typical regulatory methods. Comparison of model results using actual hourly emission values to coincident ambient air monitoring values at the site shows that AERMOD over predicts near-field impacts due to simplifying assumptions including building downwash, plume dispersion, chemistry, and 1-hr invariant meteorological conditions.