Exposure Assessment Services
Exposure assessment, risk assessment, and risk management are the three major components of assessing risks from exposure to air pollutants. The staff of Sullivan Environmental is in the forefront of developing methods for exposure assessment and conducts research for the U.S. EPA and industrial clients aimed at identifying exposures from specific source categories. Risk assessment involves the review of average and worst case exposures in relation to cancer potency scores relative to threshold limit values for non-cancer endpoints. Through identification of the upper-bound risks by pollutant and source category, the priority sources and pollutants for subsequent control analyses can be identified. Risk management involves evaluating the most cost-effective and practical ways to reduce risk (average exposures and maximum exposures). Sullivan Environmental Consulting has been involved in all of these areas with exposure assessment as the primary area of expertise.
The initial stages of exposure assessment are often performed in a relatively simplistic manner. Conservative simplifying assumptions are used in the first steps of evaluation to determine the significance of the risk. If the risks are found to be high on this basis, the exposure assessment need is further refined prior to moving into risk management. This helps ensure that the connections between proposed controls and expected benefits are realistic. The following are some examples of how refinements can be made for the purpose of improving the usefulness of an exposure assessment. The three most common forms of exposure are chronic, acute, and accidental.
The goals of assessment are similar if the exposure study is a threshold or non-threshold evaluation. Locations in close proximity to key sources need to be modeled with as much specificity as possible in terms of the release characteristics of nearby sources. Simplifying assumptions in terms of spatial aggregation of sources, treatment of building downwash, and other site-specific features usually become of much greater significance as the distance between the source and receptor is decreased.
The diurnal features of the sources are critically important in all cases. Operations that are concentrated in the daytime, e.g., 7:30 A.M. through 3:30 P.M. shift, need to be modeled accordingly because of the strong bias towards restrictive dispersion conditions that produce worst-case conditions for near-ground sources during many nighttime periods. Especially for non-cancer endpoints with thresholds effects, it is essential to add background concentrations to the modeled concentrations for sources within the modeling domain. Our experience has shown that background is best characterized by season and wind direction quadrant.
Finally, if measured air quality data are available, model performance testing provides an opportunity to identify the strengths and weaknesses in the emissions and modeling treatments and the potential to improve model performance. Testing on this basis is recommended whenever sufficient measured air quality data are available. The above features were factored into the EPA Air Quality Integrated Management System (AIMS) developed by Sullivan Environmental and installed in the Baltimore metropolitan area in the mid 1990s as a prototype system for managing air quality at the urban-scale of analysis. This system also was used as one of the three urban air toxics studied conducted by EPA as required by the Clean Air Act of 1990, i.e. The Baltimore Air Toxics Study. A system such as AIMS offers the potential to provide efficient air quality management in the future.
Exposures associated with averaging times of 24 hours or less can be dominated by a few averaging periods per year that have emission rates that greatly exceed routine conditions. Examples:
- Batch operations at chemical facilities
- Pesticide applications onto agricultural fields
- Dump stack emissions from hazardous waste incinerators
- Upset conditions at industrial facilities
In all cases, there is a major limitation in terms of applying routine Gaussian dispersion models, i.e., how to define meteorological conditions when the emissions event of concern takes place. There are two major options. The most simplistic way, which is only appropriate as a default, is to assume that the worst-case emission event occurs every hour of the year. If the exposure assessment is screened on this basis, and there are no difficulties in meeting the exposure guidelines or standards, there is no need to refine the analysis any further. On the other hand, if the imposition of such conservative assumptions would result in additional controls or permit restrictions, the next appropriate step is to account for the variability in emission rates by using a Monte Carlo treatment.
The TOXST model was developed by Sullivan Environmental for the chemical industry and was later adopted by EPA in the form of the current model. This model provides a preferred option to refine the estimation of acute exposures. TOXST can be run with the current version of ISCST3 with modification to recognize the current version of the model. It provides the ability to simulate 200 to 2,000 years of operation to produce distributions of concentrations for receptors around the source. TOXST can account for batch emissions, mutually exclusive sources, and multiple hour averaging up to 24 hours. The next generation dispersion mode, AERMOD, has a connection established to support the use of TOXST.
One of the most serious concerns regarding toxic air pollutant exposures is the consequence of an emergency release. Exposure is not intended and may occur in dangerously high quantities. For example, a chlorine discharge from the rupture of a major storage tank can produce lethal effects at distances far from the point of release. Average health effects can be observed miles from the point of release. Sullivan Environmental has developed a modeling system for air quality emergencies, Risk Manager Pro TM, which computes sources terms (the temperature of the release, gas/liquid fractions, and emission rates) to support dispersion modeling of dense gas and non-dense gas releases. This system can also be used to estimate the benefits of mitigation measures such as dikes and sheltering of storage tanks within an enclosure. Another feature of Risk Manager Pro TM is the option to estimate a range of expected indoor concentrations as a function of time after release, which can be used to assess shelter-in-place versus evacuation options. A Monte Carlo feature similar to TOXST can also be used to account for the relatively low probability of a release and to decouple the forced conservative assumption that the worst-case accident occurs during the worst meteorology (i.e., with wind flow towards the most sensitive receptor and with most restrictive dispersion and transport conditions). The goal of all the preceding steps is to provide sufficient perspective on the likelihood of exposures to individuals in the vicinity of the point of release of various levels of exposure. When there is public concern about the safety of a particular industrial operation, a refined assessment analysis can serve to promote enhanced confidence by providing a more complete picture of the likelihood of exposures at various concentrations including the threshold concentrations of primary concern.