MARAMA Mid-Atlantic Regional Air Management Association

Update on MARAMA Ozone and PAMS Data Analysis

Background

MARAMA has an ongoing commitment to participate in analyzing air quality monitoring data, particularly data from PAMS stations. MARAMA seeks to integrate this analysis into a framework that will be useful to MARAMA agencies.

MARAMA's first PAMS analysis report was the 1995 Ozone Atlas for the Mid-Atlantic Region that presented the context of the ozone problem in the region and summarized PAMS data collection efforts. (The Ozone Atlas is available on MARAMA's web site.) One conclusion from this report was that the PAMS data showed that many of the most prevalent hydrocarbon compounds measured at PAMS sites in the Mid-Atlantic Region are linked to motor vehicle emissions.

In January 1998 the MARAMA Board authorized a project to analyze PAMS data and produce a report on the importance of mobile sources in the Mid-Atlantic Region. MARAMA issued an RFP and formed a technical committee to review responses.

As a result of that review, the project was split into two contracts, one with Sonoma Technology Inc., and the other with E.H. Pechan. Further, MARAMA reduced the scope of the original project because the bids received exceeded the available resources.

The following briefly summarizes the reports prepared to date under the Sonoma and Pechan contracts. In addition, two other reports are discussed. In early 1999 MARAMA also contracted with Sonoma for an analysis of EPA's modeling of toxic air pollutants in comparison to PAMS data. This document also summarizes a report on 8-hour ozone values prepared by the MARAMA staff. Finally, this document also describes work left to do and potential additional work that could be undertaken if additional funds are available.

1. Validation of PAMS VOC Data in the Mid-Atlantic Region, February 1999, H. Main and P. Roberts, Sonoma Technology, Inc.

   This project added considerably to the credibility of Mid-Atlantic PAMS data. For this project, Sonoma validated data from 12 PAMS sites for 1996 and 1997. Sonoma had previously validated 1995 data under a CRC-funded contract. They had also validated North Carolina data from "PAMS-like" stations under a separate contract. Some data from 1994 for Philadelphia and Washington, DC, had also been previously validated.

   After this report, the state and local agencies operating the PAMS sites made corrections, the revised data was posted in AIRS, and Sonoma is expected to update the NARSTO archive.

2. Characteristics of Volatile Organic Compounds in the Mid-Atlantic Region, March 1999, H. Main, S. Hurwitt, P. Roberts, Sonoma Technology, Inc.

   This report analyzes concentrations of various hydrocarbons at Photochemical Assessment Monitoring Stations (PAMS sites) in the Mid-Atlantic Region, principally for the year 1997. This report does not include the Corbin site in Northern Virginia, since Sonoma did not have data for that site at the time the report was prepared. However, MARAMA has signed a contract with Sonoma to include the Corbin data in a revised version of the report, which is due to be completed by January 2000.

   The report confirms that motor vehicles are the dominant source of the species of Volatile Organic Compounds (VOCs) measured at Mid-Atlantic PAMS sites, although some sites also indicate the presence of stationary sources of VOC. Total Non-Methane Organic Compound concentrations were statistically higher on mornings of days with peak ozone concentrations over 100 ppb.

   The report identifies all PAMS sites as "VOC limited." Mid-Atlantic PAMS sites in general are located in or downwind of major metropolitan areas. Combustion sources, including both motor vehicles and other fuel burning in metropolitan areas, are a rich source of NOx emissions. Because NOx emissions are high in urban areas, the amount of ozone produced may be limited by the availability of VOCs. This finding should be confirmed using other data analysis techniques before being used to influence control strategy selection.

   Recommendations for further analysis include the following:

   • Update this analyses to include the Corbin site in Northern Virginia (underway).

   • Investigate the VOC/NOx limitation issue at a broader range of monitoring sites using additional techniques.

   • Investigate potential stationary sources of VOCs identified at the following sites:
     • A nearby source of toluene impacts the New Brunswick site.

     • There were other sources of i-pentane and n-pentane at Rider College.

     • At Philadelphia, the i-pentane, n-pentane, and 2-methylpentane concentrations showed a split in the relationships, indicating either two unique sources of these hydrocarbons or an analytical issue.

     • At Philadelphia, good correlations between decane and toluene, the xylenes, and 1,2,4-trimethylbenzene were probably indicative of an industrial signature.

     • At Lums Pond, high propane and propene concentrations were probably from an upwind industrial complex.

     • At Essex, the good correlation between noane and styrene may be indicative of an industrial source of these species.

     • At Aldino and Camden, hydrocarbons associated with industrial sources (e.g., toluene and 2,2,4-trimethylpentane) showed significant correlations.

   • CO is a good tracer for motor vehicle emissions, but it is only measured at Camden. Morning CO concentrations of CO at Camden were significantly lower on weekends than weekdays. Sonoma recommends that CO be measured at all PAMS sites, with ppb level accuracy.

   • NOy data is useful in assessing VOC/NOx limitation but is not routinely collected except in North Carolina. Sonoma recommends that improving NOx and NOy data collection should be a high priority.

   • Sonoma recommends that all agencies speciate as many compounds as possible and report this data to AIRS.

3. Technical Assistance to MARAMA on EPA Cumulative Exposure Project (CEP): Review of Air Toxics Modeling, January 1999, P.T. Roberts, H.H. Main, S.B. Hurwitt, Sonoma Technology, Inc.

   This report is a review of toxics modeling conducted by EPA under the Cumulative Exposure Project, comparing model-estimated concentrations of various toxic air pollutants with ambient concentrations in Philadelphia, New Brunswick (NJ), and Baltimore. Ambient concentrations were for 1996, while the model estimates were for 1990, so the comparison illustrates analyses that could be performed when 1996 modeling is completed. In addition, the ambient data is from PAMS monitoring sites, which only operate during the summer, while the model estimates are annual averages. Morning data were used for the comparisons, because a comparison of summer morning data to 24-hour, year round canister data for 8 HAPs at Philadelphia showed good agreement.

   Sonoma identified three overall issues in reviewing EPA's modeling:

   • It is inappropriate to rely on model results at the census-tract level. The model is more reliable over larger areas.

   • The expected model performance and uncertainty of model estimates will vary widely for individual hazardous air pollutants (HAPs). HAPs vary in their emission, chemical reactivity, transport and dispersion, and background characteristics, and all these factors influence model performance.

   • EPA modeling guidance for dispersion models indicates that short-term concentrations should be within a factor of two of monitored values, and annual average model performance is expected to be better than short-term estimates. EPA's own model performance evaluation performed as part of the CEP compared available ambient data for 19 HAPs with model results and found that 68% of the comparisons were within a factor of three. Sonoma raises the question of whether this is acceptable performance.

   Sonoma's comparison of monitored and modeled concentrations in Philadelphia, New Brunswick, and Baltimore led them to several conclusions, including those listed below. Of the HAPs investigated in this study, benzene, formaldehyde, and carbon tetrachloride are the most important with respect to cancer risk based on ambient data. Measured concentrations of these HAPs are well above cancer risk benchmarks (1 in 1 million risk levels).
   • Measured concentrations of species highly associated with motor vehicle emissions (e.g., benzene, toluene, and xylenes) were typically within a factor of two or three of the model's estimates. (The ambient concentration was usually lower than predicted by the model. This is consistent with recent studies showing that benzene has declined in the 1990s due to the use of reformulated gasoline.)

   • PAMS carbonyle compound concentrations were typically about twice those modeled. The model chemistry may be inadequate to reflect the secondary formation of carbonyles (e.g., formaldehyde).

   • Measured and modeled concentrations of methylchloride and carbon tetrachloride concentraations were similar, and were typically at about the level the model assumed as a background concentration. (Data only for Philadelphia.)

   • Several differences between modeled and monitored concentrations may indicate problems with emissions factors or activity data used to estimate emissions:
     • Ethylene dibromide was monitored at 20 times the level predicted by the model. The model may be missing key sources. (Data only for Philadelphia.)

     • Measured ethylene dichloride was at least an order of magnitude lower than modeled. Emissions estimates may be overstated. (Data only for Philadelphia.)

   • Ambient variability in concentrations is often much greater than reflected in the model.

   • Some pollutants are measured at very low concentrations, near the detection limits of monitoring equipment.

   Since EPA plans to conduct another round of toxics modeling using the ASPEN model and 1996 emissions estimates, the methods demonstrated in this report may be useful in assessing the results of the EPA 1996 modeling. EPA plans to do some work comparing their model results with monitored data.

4. Mobile Source Emission Estimation Methods in the Mid-Atlantic Region, Draft, July 1999, E.H. Pechan & Associates, Inc., and "Emissions Data Analysis" memo of October 8, 1999, to Susan Wierman from Jim Wilson and Maureen Mullen of E.H. Pechan.

   The draft report is a brief summary of the mobile source inventory methods used by the agencies in the Mid-Atlantic region. The report focused on ten attributes determined to be important in estimating travel and in using EPA's MOBILE5 model to estimate emissions.

   At the time the report was prepared, 1990 data was available for some areas, while 1996 data had been completed for others, so comparisons between the areas are difficult in some respects. However, it appears that the emissions estimation method differences among areas are as significant an influence on highway vehicle emissions estimates as are the control programs that individual areas have adopted. Some of the inter-city differences represent differences in the vehicle fleet and travel patterns in the areas of interest, but there are also analytical differences in assigning VMT among vehicle types, in allocating VMT by roadway type, and in estimating vehicle speeds by roadway type.

   This type of variation in the data can affect the results of comparisons between ambient data measured at PAMS sites and emissions data from mobile sources. The report recommends using both local and metropolitan area emissions data in such comparisons, especially where the fleet composition and allocations of VMT to roadway types differs markedly from county to county within the metropolitan area.

   The report recommends using PAMS monitors in Baltimore and Washington for initial comparison to mobile source emissions data.

   The memo compares mobile source emissions estimates for the major metropolitan areas in the MARAMA region from the 1995 OTAG inventory, the NET inventory for 1996, and SIP/Rate of Progress Plan inventories prepared by the States for 1996. The three VOC inventories are generally similar, but when emissions are divided into vehicle types, there are more differences between them. In addition, NOx emissions tend to be higher in the SIP inventories than the OTAG or NET inventories. This comparison highlighted several important considerations. NOx emissions are significantly higher when inventories account for the effects of a heavy duty diesel vehicle emissions control defeat device. The distribution of vehicle types can have an impact on emissions as well, and further changes are expected in the future as additional areas update this information. Vehicle distribution can also affect speciation of VOCs for comparison to PAMS data.

   This comparison among three types of mobile source inventories will aid in placing Pechan's forthcoming comparison of emissions data and PAMS data in context. There have been other comparisons that have used the OTAG data or the NET data, but MARAMA is not presently aware of other analysis that has compared PAMS data to SIP inventories.

5. Comparison of 8-hr and 1-hr Ozone Exceedances in the Mid-Atlantic Region, July 1999, S.H. Alcorn, P.T. Roberts, H.H. Main, Sonoma Technology, Inc.

   This report evaluates Mid-Atlantic air quality in 1997 using both the 1-hour and 8-hour ozone standards. The report concludes:

   • The 8-hour standard is exceeded more often than the 1-hour standard. In 1997, 8-hour ozone pollution levels were higher than 85 ppb on four times as many days as 1-hour levels were over 125 ppb.

   • The 8-hour standard is exceeded more widely than the 1-hour standard. In 1997, 89% of monitoring sites in the Mid-Atlantic region recorded 8-hr values over 85 ppb, while only 36% had 1-hr values over 125 ppb.

   • The 8-hr standard is exceeded both earlier and later in the summer. All of the days in 1997 when the 1-hour ozone level exceeded 125 ppm occurred in June and July. On the other hand, a significant number of days when 8-hour ozone exceeded 85 ppb occurred in May, August, and September.

   In addition, the report reviewed 1986-1997 data in comparison to the 8-hour standard. This review used 3-year averages of the annual fourth-highest daily maximum 8-hour values, so that the data is consistent with the form of the 8-hour ozone standard. The report concluded that there was no clear regional trend in 8-hour ozone values. Even looking at the sites with the most extreme changes, there was no improvement in air quality with respect to the new standard.

   This analysis could be updated using 1998 and 1999 data when they become available. It could also be integrated with a recent MARAMA staff report summarized below.

6. Summary of Mid-Atlantic 8-Hour Ozone Data, 1986-1997, October 1999, S.S.G. Wierman, S. Kayin, and E. Belyaev, Mid-Atlantic Regional Air Management Association.

   This report reviews 12 years of 8-hour ozone data for the Mid-Atlantic region. The data were obtained from EPA's AIRS system for each monitoring site in the region for 1986 through 1997.

   The data show continued widespread violations of the 8-hour ozone standard in all Mid-Atlantic States. Since the late 1980s, there has been a reduction in the number of very high exceedances, even though the total number of 8-hour exceedances remains high. At the state level, it is impossible to observe any obvious trend in recent data. Most of the monitors that continue to violate the standard have 3-year averages between 85 and 104 ppb, which EPA's Air Quality Index defines as "Unhealthy for Sensitive Groups.

   This report could be updated when data is available for 1998 and 1999.

7. "Reconciliation of an Emissions Inventory with PAMS Ambient Monitoring Data in the Mid-Atlantic Region," Draft, October 25, 1999, Till Stoeckenius and Michele Jimenez, ENVIRON International Corporation.

   This comparison of ambient data and emissions data is being undertaken for PAMS sites in Baltimore (Essex) and Washington, D.C. (McMillan). It will include a review and synthesis of previous studies to draw out information pertinent to our area. Emissions inventory data obtained from Virginia and Maryland will be adjusted to reflect 1997 and then compared to the PAMS data for that year. Even though only two PAMS sites will be reviewed, the literature review will apply region-wide, and if the techniques used prove acceptable, we could expand the analysis to other sites.

   A draft of this report is circulating for review by MARAMA agencies. Important conclusions are summarized here but may change as the report is finalized.

   The draft report concluded that the ratios of monitored Total Non Methane Hydrocarbons (TNMHC) to NOx were greater on average than corresponding ratios calculated by speciating emissions. This could be due to some combination of underestimation of VOC emissions and overestimation of NOx emissions or the influence of other factors which cause uncertainties in these comparisons. (These factors include temporal and spatial allocation of emissions, speciation of emissions, local atmospheric chemistry, and monitoring uncertainties.)

   The draft confirmed that gasoline-powered mobile sources dominate the McMillan Reservoir site. Comparison of individual species and species groups at McMillan indicated several key discrepancies between the mobile source emissions speciation profile and the ambient data. The abundance of aromatics is slightly over-predicted in the inventory relative to the ambient data. There was a greater fraction of light hydrocarbons in the ambient data than in the inventory, and much greater toluene, n-butane, and xylene in the inventory relative to the ambient data.

   For the Essex site, the presence of several large elevated point sources in the vicinity of the site complicated the analysis. It appeared that the agreement between the inventory and ambient data was better in Baltimore for mobile source-related species.

   The report also reviewed several previous studies that compared emissions and ambient data. Methods used in the studies varied, as did the time periods and locations. Three earlier studies also showed that TNMHC to NOx ratios were greater in ambient than emissions data. A recent LADCO analysis showed better agreement between the two except that light hydrocarbons were measured at PAMS sites in greater abundance than predicted by the inventory data. LADCO's results were consistent with tunnel studies.

   The draft report does not comment on the significance of these discrepancies for ozone modeling. In addition to expanding this analysis to other sites, follow-on analysis could include the development of updated source emissions profiles for the region, coupled with source-receptor analysis for key monitoring sites. Further analysis could also include a comparison of modeled and measured concentrations of key pollutants. In addition, an integrative report could be commissioned, pulling together both the Sonoma, Pechan, and ENVIRON work.