Air Toxic Study 2005 Executive Summary

Community Assessment

Spokane Air Toxic Study 2005
Executive Summary

An ambient air sampling study designed to assess airborne toxics was conducted in Spokane, WA during the 2005 calendar year. The goal of this community assessment program was to provide data that could be used to characterize human exposure levels to air toxics, better understand temporal and spatial trends of the air toxics, as well as, provide measurement data for air quality model evaluation. Samples were collected over the course of the year at four sites on a one-in-six day schedule. The targeted airborne toxic compounds include VOCs, carbonyls and metals contained in PM10. Ambient concentration levels have been tabulated and evaluated for EPA’s fifteen “core” urban air toxics. 

The highest annual average ambient concentrations of most air toxics were observed at the Crown Zellerbach (CZ) site. CZ is centrally located in a commercial/industrial zone. The exception was acetaldehyde and formaldehyde, which exhibited their maximum annual averages at the most northeasterly School District (SD) and Orchard School (OC) sites.  Lowest annual average air toxic levels were recorded at the most westerly Health District (HD) site. On a temporal basis, many air toxics exhibited elevated levels in the wintertime and lower ambient concentrations during the summer months. Summertime concentrations were larger for the carbonyls due to more favorable secondary formation during this period of the year. Tetrachloroethylene exhibited sporadic peaks throughout the year while the other chlorinated species (CCl4, CHCl3, and TCE) remained low in all seasons. Annual average concentrations for Spokane “core” air toxics in 2005 were as follows: benzene 0.24 ppbv; 1,3-butadiene 0.05 ppbv: carbontetrachloride 0.11 ppbv; chloroform 0.01 ppbv; tetrachloroethylene 0.04 ppbv; trichloroethylene 0.02 ppbv; acetaldehyde 1.4 ppbv; formaldehyde 2.0 ppbv; arsenic 0.9 ng/m3; beryllium 0.02 ng/m3; cadmium 0.2 ng/m3; chromium 4.8 ng/m3; lead 4.9 ng/m3; manganese 19 ng/m3; nickel 5.6 ng/m3; and, PM10 24.5 ug/m3. These levels are similar to those reported in other U.S. cities. 

We examined source-receptor relationships for the air toxics using several different techniques. Strong correlations, suggesting a common source, existed between benzene and 1,3-butadiene at all four sites. The same was true for acetaldehyde and formaldehyde.  With the trace metals, significant correlations between arsenic, cadmium and lead inferred a common source. There was a very high correlation between chromium and nickel in PM10 collected at the CZ site indicating a nearby source of these trace metals. 

An analysis of lead isotopes (206Pb, 207Pb, 208Pb) measured in PM10 provided information about lead sources in Spokane. The lead isotope ratios varied depending on season and meteorology. On warm, windy summer days, isotope measurements showed most of the lead in PM10 to originate from crustal sources, while during low ventilation periods in the winter, combustion processes contributed most of the airborne lead. On an annual average basis, the lead isotope ratios measured at the CZ site implied about a 50:50 mix of crustal and combustion sources for lead in Spokane’s particulate matter. 

Positive matrix factorization (PMF) on the combined Spokane data set yielded five factors.  There are two factors that have high aluminum and/or iron content that relate to crustal sources. Two other factors lack the crustal tracers and are rich in combustion related species (benzene, arsenic, cadmium and lead). The fifth factor contains chromium and nickel nearly exclusively and represents industrial emissions near the CZ site. The PMF analysis indicated roughly a 50:50 split annually in the crustal and combustion sources of PM10 in Spokane. 

In order to pinpoint specific VOC sources in the neighborhoods surrounding each of the four Spokane urban sites, we outfitted a mobile van with real-time sensors that allowed emission plumes to be easily identified. A Proton Reaction Transfer Mass Spectrometer (PTR-MS) was used to locate emission sources of benzene, acetaldehyde, several low molecular weight oxygenated solvents and certain BTEX (toluene, xylenes, etc) species.  For example, plumes emanating from auto repair shops included high concentrations of acetone and xylenes. A large source of styrene was recorded in the vicinity of Spokane’s Industrial Park east of the city. Particle bound polycyclic aromatic hydrocarbon (PPAH) concentrations and surface areas were measured along Spokane’s roadways using sensors in the mobile van. Background PPAH levels in Spokane were on the order of 7 ng/m3 compared to an average concentration of 81 ng/m3 measured on streets in Spokane. Chase experiments with the mobile unit clearly showed diesel vehicles to be a large source of PPAH. There was good agreement between PPAH concentration vs. surface area relationships for Spokane diesel vehicles and controlled diesel-chamber experiments. 

Fixed location sampling with the mobile van at the SD site during a 24-hr sampling day provided a temporal record of air toxic concentrations that can be used to better understand diurnal pollutant behavior.  Air quality in Spokane is predicted daily by Washington State University using the AIRPACT-3 model system. This model estimates benzene, 1,3-butadiene, acetaldehyde, formaldehyde and PM10 hourly concentrations in Spokane. The air quality model has 12 km resolution, which provided one cell that encompassed all four sites. Modeled concentrations for the air toxics listed above were compared with measured levels recorded on five sampling days in December 2005. This was a useful exercise in that it provided some insight into model shortcomings. For example, the model over predicts benzene and an evaluation of the reason for the difference indicated a problem with the model’s benzene emission inventory. In a similar vein, the model greatly under predicted the carbonyl concentrations. The reason for this is most likely unrealistically low boundary/initial conditions employed for acetaldehyde and formaldehyde in the model.

A final task in this community assessment program was to evaluate citizen exposure levels to air toxics in neighborhoods surrounding the four sampling sites. This was accomplished in two ways. First we conducted a simple screening test to see which of the air toxics exceeded a screening value based on cancer and/or non-cancer risk factors. This showed that of the 15 “core” air toxics all but chloroform, beryllium and lead exceeded the chronic screening value (CSV). In order to establish more quantitative exposure levels, we ran EPA’s Hazardous Air Pollutant Exposure Model (HAPEM5) using air quality data gather at each of the four sites. Modeled exposures for cadmium and nickel were below the CSV except for nickel in the neighborhood surrounding the CZ site. Thus, benzene, 1,3-butadiene, carbon tetrachloride, tetrachloroethylene, trichloroethylene, acetaldehyde, formaldehyde, arsenic, chromium, and manganese exceeded the health screening value in Spokane neighborhoods. 

In summary, a very successful community assessment study was conducted in Spokane that has provided a wealth of information about air toxic levels, sources and exposures in the region.