Final Report October 2009 - Executive Summary
In July 2007, the New York State Department of Environmental Conservation (NYSDEC) initiated a year-long community air quality monitoring study in the town of Tonawanda (Erie County) to measure the concentration of air contaminants within the community and to evaluate the potential risk to public health. The Tonawanda Community Air Quality Study (hereinafter referred to as Study) was motivated by a number of critical factors: first and foremost, complaints received by NYSDEC from the community regarding odors and an overall compromised quality of life; second, the elevated ambient benzene concentrations sampled by a local community group and the NYSDEC; and third, the Tonawanda industrial area represents an excellent opportunity to assess the effectiveness of the current federal and state hazardous air pollution reduction strategies.
To address these issues, NYSDEC conducted monitoring, modeling and an inhalation risk assessment to estimate the risk posed by ambient concentrations of hazardous air pollutants (HAPs). The Study design allowed for the identification of results which could be used for risk management decisions and selection of options to reduce exposure to HAPs in the Tonawanda community. The Study design and findings were presented through a series of public meetings within the community to seek input, hear concerns and answer questions from the public and all interested parties in attendance. The findings from the Study have already resulted in a number of actions by NYSDEC and USEPA to evaluate and address potential sources of benzene emissions in the Study area.
Four air quality monitors were installed in and around the community in reference to the prevailing wind direction from the southwest. One monitor at Beaver Island State Park (BISP) was sited to establish background measurements of air toxics upwind of the industrial sources. Three monitors were placed downwind of the industrial sources in the Study area: Grand Island Boulevard Industrial (GIBI), Brookside Terrace Residential Site (BTRS) and Sheridan Park Water Tower (SPWT). The monitors collected 24-hour average ambient air concentrations of 56 air toxic pollutants on a one-in-six day schedule. All four monitors collected hourly average concentrations of fine particulate matter (PM2.5). Monitors placed at the BTRS site collected hourly average concentrations of sulfur dioxide (SO2) and carbon monoxide (CO). A meteorological station was placed at the BISP site to assess the local meteorology for the Study area and for use in assessing the sources influencing the air monitoring concentrations.
The GIBI monitoring site found significantly elevated concentrations of benzene and formaldehyde when compared to other areas of New York. The evaluation for benzene indicated higher daily concentrations of benzene when the wind originated from the direction of the largest known point source, Tonawanda Coke Corporation. The BTRS monitor, downwind from the industrial sources, also indicated more of an influence from the industrial sources than contributions from mobile sources in the area. The results for benzene at the other two ambient air quality monitors were similar to ambient air levels found in large urban areas such as New York City. The formaldehyde evaluation indicated that the measured concentrations were influenced by local area sources and mobile sources. The GIBI monitor reported the highest concentrations, much higher than the other study monitors and other monitors in the statewide network. The formaldehyde concentrations also appear to be influenced both by temperature and wind speed fluctuations with direct temperature correlations and an inverse wind speeds correlation.
A public health evaluation was conducted using NYSDEC derived health-based guideline concentrations and the results from the ambient air quality monitoring. The annual average concentration for five air toxics (1,3-butadiene, acetaldehyde, benzene, carbon tetrachloride, formaldehyde) exceeded the cancer risk screening level of one-in-one-million and one air toxic (acrolein) exceeded the non-cancer health-based comparison value.
A comprehensive inventory of sources for the Study area was prepared for use with two air dispersion models (Regional Air Impact Modeling Initiative (RAIMI) and AerMod) that are used by NYSDEC to evaluate the inhalation risk of exposure to HAPs from stationary and mobile sources. The predicted concentrations of the HAPs were modeled for the entire Study area and the results were compared to the monitored data and predictions from the 2002 National-scale Air Toxics Assessment (NATA).
The average ratios for ten air pollutants selected for the comparison of the RAIMI modeled predictions to the monitored concentrations were in close agreement (ranged from 0.58 to 1.25) across all four monitoring sites. However, an analysis of the site by site comparisons for benzene and formaldehyde revealed very poor agreement between the modeled and measured concentrations at the GIBI site.
The comparisons of the monitoring data to the 2002 NATA predictions indicated that the 2002 National Emission Inventory (NEI) used in the NATA model was fairly accurate for a number of HAPs. However, the NEI emissions inventory under reported acrolein emissions for the entire Tonawanda area and under reported 1,3-butadiene, benzene, ethylbenzene, formaldehyde, and propionaldehyde emissions for sources near the GIBI monitor. As a risk assessment screening tool designed to identify areas for further air pollution investigations, it would be preferable for the NATA modeled ambient concentrations to be similar to measured ambient concentrations for those air toxics that are identified as risk drivers.
The Study measured air quality in close proximity to the Tonawanda Coke Corporation in order to fill a data gap identified in the USEPA's Residual Risk Assessment for Coke Ovens which identified the lack of ambient monitoring information. Some HAPs known to be released from the facility were measured and an elevated concentration of benzene was observed at the Study area monitors. When compared to the USEPA's residual risk assessment, NYSDEC's modeling assessment, using a revised facility emissions inventory, resulted in larger predicted impacts within the community. Based on the assessment of the monitored and modeled data, the maximum individual cancer risk and population cancer risk associated with facility-wide emissions from the Tonawanda Coke Corporation exceeds an excess lifetime cancer risk of 10 in-one-million for the nine census tract Study area. Specific neighborhoods exceed a 100 in-one-million cancer risk level. Further work will be conducted on this issue by NYSDEC prior to USEPA's 2011 scheduled completion of a final residual risk assessment for the Coke Oven source category.
The Study design had several noteworthy strengths. The source attribution conclusions were derived from a weight-of-evidence approach rather than relying on a single result to achieve a conclusion. The source attribution assessment included evaluating meteorological information, emission releases through dispersion modeling and an in depth evaluation of the USEPA's residual risk assessment that included a site specific risk assessment for the Tonawanda Coke Corporation. A comprehensive emissions inventory was developed to further elucidate source contributions and emission reduction strategies for sources identified as contributing to elevated risk levels in the community. This information, coupled with the results between the upwind and downwind inhalation cancer risk values, provides a strong basis for further compliance monitoring and regulatory actions to reduce the inhalation cancer risk in the Tonawanda community.
In conclusion, the results of the Tonawanda Community Air Quality Study indicate that further work can be done to improve air quality in the community. Follow-up activities (e.g. increased compliance inspections and community observations) already have been implemented by the NYSDEC and the USEPA in an effort to improve air quality in the community. NYSDEC is continuing to monitor hourly benzene concentrations at the GIBI site to further evaluate the high levels of benzene measured at this monitoring site with a higher degree of temporal resolution.