Public Health Consultation
Appendix D: Modeling approach to identify the area most likely to be affected by emissions from AES Greenidge
We used an approach called dispersion modeling to predict the area most likely affected by emissions from the AES Greenidge facility. Dispersion models are mathematical tools that predict the relationship between pollutant emissions from a source and resulting air quality by incorporating factors that affect pollutant release and movement in air. We considered using a kind of model often called a refined dispersion model, however this type of model requires site-specific meteorological data (e.g., wind speed and wind direction) and these data are not available for the AES facility area. Therefore, we chose to use a less complex model, known as a screening model, which does not require site-specific meteorological data. Screening models use information about the emission source, local terrain features (such as elevation) and land uses features (such as rural or urban) and the model output provides a rough, often conservative, estimate of the magnitude and distance of facility impacts without considering site-specific wind patterns. The model we used, which is called "SCREEN3," was developed by the US Environmental Protection Agency.
A limitation with using a screening model is that the model output does not provide any information on the direction (e.g., north, south) in which the estimated impact occurs. Therefore, we evaluated whether any local meteorological data could be incorporated in our approach. In consultation with New York Department of Environmental Conservation (NYS DEC) staff, we determined that data from a meteorological station at Cornell University would be appropriate to use. The Cornell station is located near Cayuga Lake, approximately 12 miles east of Seneca Lake. Both lakes are similar in size and shape, both have a general north-south orientation and the terrain characteristics near both lakes are similar. Specifically, both lakes are located in valleys with general north-south orientation and in both locations the terrain rises sharply to the west and east. Therefore, wind flow patterns are expected to be similar in both locations.
We worked with NYS DEC staff to combine the screening model with meteorological data from Cornell and terrain characteristics near the AES facility to qualitatively estimate facility impacts. Specifically, we used the screening model to provide a distance estimate of the impacts for locations where the terrain is relatively flat (north and south of the facility). We used the terrain data to help inform us of areas where terrain height might restrict pollutant dispersal (such as east and west). We used the Cornell meteorological data to tell us the area most likely affected by the emissions. The outcome of screening model provided an estimated range of pollutant concentrations by distance. We used this information to identify three different study areas, based on relative impacts from the AES facility, to assist in stratifying population exposures.
Additionally, we evaluated the potential for nearby electrical generating facilities, which primarily use coal, to impact the community surrounding AES Greenidge.
NYS DEC provided us with a diagram developed from data collected in 2001 at the Cornell station, indicating wind direction, frequency and speed (see Figure 1 (PDF, 25KB, 1pg.)). This diagram, called a wind rose, shows that winds coming from the south to southeast direction were most frequent, and winds from the west to northwest were somewhat less frequent. That is, the winds tend to follow the channeling of Cayuga Lake. Winds near the AES facility would be expected to follow the orientation of Seneca Lake.
Because the terrain north, northeast and south of AES Greenidge is relatively flat (because much of this area is the lake surface with adjacent shore) we used the screening type model to estimate the range of potential significant long-term impact. Applying the wind frequency information from Cayuga Lake to Seneca Lake indicates that there would be a strong southerly wind component, and therefore the region north of the AES Greenidge facility would likely be the most impacted area. Combining the screening model results, terrain characteristics and wind information for the region north of the facility yielded an estimated distance of impacts of approximately 20 miles. Applying the same method for the region south of the facility yielded an impact range estimated to be up to 10 miles south of the facility. The potential for impacts west or east from the facility would be relatively small due to the low frequency of winds from these directions and the channeling effect due to the sharp rise in terrain elevation. Figure 2 (PDF, 59KB, 1pg.) shows the estimated area of likely long-term impact and the location of the Cornell meteorological station. The AES facility has tall stacks (250 feet); therefore the estimated impacts of stack emissions close to the facility are relatively low (except for during certain weather conditions that decrease transport of the pollutants away from the facility). However, because of the tall stacks, areas as far away as 30 miles could experience some impacts from the facility.
To stratify exposure levels we used the screening model results and wind direction data to identify three areas of higher potential exposure, defined by distance from the facility and relative magnitude of estimated impact. The three exposure areas that we identified are those that include most of the area estimated to experience greater than 75% and within 50% to 75% and less than 50% of the estimated maximum impact. The higher potential exposure area, within which greater than 75% of the estimated maximum impact may occur, extends approximately 9 miles north, south and east of the facility. The moderate potential exposure area, within which 50 to 75% of the estimated maximum impact may occur, extends about another 5 miles to the north and east of the facility. The lower potential exposure area, within which less than 50% of the estimated maximum impact may occur, extends about another 6 miles north of the facility (see Figure 2) (PDF, 59KB, 1pg.).
Our evaluation of the surrounding electrical generating facilities indicated that these facility contributions would be negligible compared to the AES Greenidge contribution to the community.