The uncertainties in simulations of annually averaged concentrations of two air toxics (benzene and 1,3-butadiene) are estimated for two widely used U.S. air quality models, the Industrial Source Complex Short-Term, version 3, (ISCST3) model and the American Meteorological Society-Environmental Protection Agency Model (AERMOD). The effects of uncertainties in emissions input, meteorological input, and dispersion model parameters are investigated using Monte Carlo probabilistic uncertainty methods, which involve simultaneous random and independent perturbations of all inputs. The focus is on a 15 km ϫ 15 km domain in the Houston, Texas, ship channel area. Concentrations are calculated at hypothetical receptors located at the centroids of population census tracts. The model outputs that are analyzed are the maximum annually averaged maximum concentration at any single census tract or monitor as well as the annually averaged concentration averaged over the census tracts. The input emissions uncertainties are estimated to be about a factor of 3 (i.e., covering the 95% range) for each of several major categories. The uncertainties in meteorological inputs (such as wind speed) and dispersion model parameters (such as the vertical dispersion coefficient z ) also are estimated. The results show that the 95% range in predicted annually averaged concentrations is about a factor of 2-3 for the air toxics, with little variation by model. The input variables whose variations have the strongest effect on the predicted concentrations are on-road mobile sources and some industrial sources (dependent on chemical), as well as wind speed, surface roughness, and z . In most scenarios, the uncertainties of the emissions input group contribute more to the total uncertainty than do the uncertainties of the meteorological/dispersion input group.
Emissions of pollutants such as SO 2 and NOx from external combustion sources can vary widely depending on fuel sulfur content, load, and transient conditions such as startup, shutdown, and maintenance/malfunction. While monitoring will automatically reflect variability from both emissions and meteorological influences, dispersion modeling has been typically conducted with a single constant peak emission rate. To respond to the need to account for emissions variability in addressing probabilistic 1-hr ambient air quality standards for SO 2 and NO 2 , we have developed a statistical technique, the Emissions Variability Processor (EMVAP), which can account for emissions variability in dispersion modeling through Monte Carlo sampling from a specified frequency distribution of emission rates. Based upon initial AERMOD modeling of from 1 to 5 years of actual meteorological conditions, EMVAP is used as a postprocessor to AERMOD to simulate hundreds or even thousands of years of concentration predictions. This procedure uses emissions varied hourly with a Monte Carlo sampling process that is based upon the user-specified emissions distribution, from which a probabilistic estimate can be obtained of the controlling concentration. EMVAP can also accommodate an advanced Tier 2 NO 2 modeling technique that uses a varying ambient ratio method approach to determine the fraction of total oxides of nitrogen that are in the form of nitrogen dioxide. For the case of the 1-hr National Ambient Air Quality Standards (NAAQS, established for SO 2 and NO 2 ), a "critical value" can be defined as the highest hourly emission rate that would be simulated to satisfy the standard using air dispersion models assuming constant emissions throughout the simulation. The critical value can be used as the starting point for a procedure like EMVAP that evaluates the impact of emissions variability and uses this information to determine an appropriate value to use for a longer term (e.g., 30-day) average emission rate that would still provide protection for the NAAQS under consideration. This paper reports on the design of EMVAP and its evaluation on several field databases that demonstrate that EMVAP produces a suitably modest overestimation of design concentrations. We also provide an example of an EMVAP application that involves a case in which a new emission limitation needs to be considered for a hypothetical emission unit that has infrequent higher-than-normal SO 2 emissions.Implications: Emissions of pollutants from combustion sources can vary widely depending on fuel sulfur content, load, and transient conditions such as startup and shutdown. While monitoring will automatically reflect this variability on measured concentrations, dispersion modeling is typically conducted with a single peak emission rate assumed to occur continuously. To realistically account for emissions variability in addressing probabilistic 1-hr ambient air quality standards for SO 2 and NO 2 , the authors have developed a statistical technique, the Emissions Variabili...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.