EASTERN MICHAUD FLATS CONTAMINATION
POCATELLO, BANNOCK AND POWER COUNTIES, IDAHO
ATSDR uses a conservative approach to determine whether levels of air pollution indicate a past,present, or future health hazard. The following discussion describes this methodology, anddocuments how it was applied to the levels of contamination measured in the EMF study area. Theremainder of this section provides an overview of the large volume of data collected in the EMFstudy area, and appendices to this report present more detailed analyses.
ATSDR generally follows a two-step methodology to comment on public health issues related to airpollution. First, ATSDR obtains representative environmental monitoring data for the site ofconcern and compiles a comprehensive list of site-related contaminants. Second, ATSDR useshealth-based comparison values to identify those contaminants that do not have a realistic possibilityof causing adverse health effects. For the remaining contaminants, ATSDR reviews recent scientificstudies to determine whether the extent of environmental contamination indicates a public healthhazard.
The health-based comparison values used in this report are concentrations of contaminants that thecurrent public health literature suggest are "safe" or "harmless." These comparison values are quiteconservative, because they include ample safety factors that account for most sensitive populations. ATSDR typically uses comparison values as follows: If a contaminant is never found at levelsgreater than its comparison value, ATSDR concludes the levels of corresponding contamination are"safe" or "harmless." If, however, a contaminant is found at levels greater than its comparisonvalue, ATSDR designates the pollutant as a contaminant of concern and examines potential humanexposures in greater detail. Because comparison values are based on extremely conservativeassumptions, the presence of concentrations greater than comparison values does not necessarilysuggest that adverse health effects will occur among exposed populations. More information on thecomparison values used in this report can be found in Appendix B.
In the case of particulate matter, however, some scientists argue that adverse health effects can occuramong sensitive populations even when ambient air concentrations are lower than the health-basedcomparison value used in this report (i.e., EPA's actual and proposed National Ambient Air QualityStandards). In other words, levels of contamination below the health-based comparison value might,in fact, not be "safe" or "harmless" to certain sensitive populations. The sidebar on the abovereviews additional information on the selection of health-based comparison values for particulatematter, and Section IV.E comments on this issue further.
The following analyses identify air pollutants for the EMF study area (Section IV.B), describe how these pollutants disperse throughout the area (Section IV.C), review site-specific studies that have measured levels of air pollution (Section IV.D), and finally comment on the public health implications of inhalation exposures to air pollution in the EMF study area (Section IV.E).
To identify site-related contaminants for the EMF study area, ATSDR consulted with EPA, IDEQ,the Shoshone-Bannock Tribes, FMC, and Simplot to obtain reports that characterize air emissionsfrom the two phosphorous processing facilities. The reports ATSDR obtained indicate that eitherFMC or Simplot, or both facilities, are suspected of emitting at least the following pollutants into theair (Bechtel 1996; Bechtel 1998; FMC 1999a, 1999b, 1999c; IDEQ 1999a; USEPA 1999d):
As an example of emissions data for these facilities, Table 1 presents the air emissions data thatFMC and Simplot reported to EPA's Toxic Release Inventory (TRI) for calendar years 1997 and1998. The TRI database is an important source of "right-to-know" information, or information thatpeople can access about the releases of toxic chemicals in their communities. Because the accuracyof TRI emissions data are not known, ATSDR based its findings of this health consultation on thelevels of chemicals that were measured in the ambient air, rather than focusing strictly on theemissions data. It is important to note that a large volume of air quality measurements are availablefor almost every pollutant listed above and in Table 1, and the evaluations of ambient air monitoringdata presented later in this section consider the pollutants that FMC and Simplot emit in greatestquantities.
Also noteworthy is the fact that the TRI data do not show that many different operations at FMC andSimplot emit pollutants to the air. Some pollutants are released from elevated sources, like stacks,and others from ground-level sources, like waste ponds. Several studies have reported estimates ofchemical-specific emissions from FMC and Simplot (Bechtel 1996; IDEQ 1999a). Thoughestimated emission rates are somewhat uncertain, they do provide insight into the relative impacts ofvarious sources on air quality. As Table 2 shows, studies have estimated that, in recent years, FMCand Simplot released 727 and 135 tons of particulate matter to the air in a calendar year,respectively (IDEQ 1999a; USEPA 1999a). The data in Table 2 are interpreted in greater detailbelow.
It is expected that emission rates from these facilities likely have varied from year to year, as a resultof changes in production demands, installation and operation of different pollution controls, use ofores from various sources, and other factors. As examples, particulate emissions from Simplotdecreased considerably in 1991, after the facility began to receive ore in a slurry pipeline, instead ofby rail car (Bechtel 1996); similarly, particulate emissions from FMC decreased after the facilityinstalled new scrubbers on its calciners in 1992 (Severson 1999), and
FMC is currently implementing controls at many other specific emissions sources. The emissionsfrom these facilities will likely continue to decrease in the future, due to pollution control plansrecently adopted by EPA and IDEQ (FR 1999; IDEQ 1999a). In fact, FMC has informed ATSDRthat its ongoing emissions controls projects are expected to result in a 67% reduction in particulateemissions.
In addition to FMC and Simplot, other industrial and non-industrial sources throughout the EMFstudy area release many of the pollutants listed above. For example, the Bannock Paving Company,which was known to emit particulate matter, metals, and other pollutants, operated on a leasedportion of the FMC property. These operations reportedly ceased on March 12, 1995, and BannockPaving Company moved to another location in Pocatello later in the year (Bechtel 1996). Furthermore, aircraft, trains, automobiles, residential wood burning, and agricultural operations allemit particulate matter to the atmosphere (IDEQ 1999a). These other sources, many of which arefound throughout Chubbuck, Pocatello, and the Fort Hall Indian Reservation, undoubtedlycontribute to air pollution in the EMF study area.
For perspective on the relative amounts of particulate matter released by FMC, Simplot, and othersources, Table 2 presents selected findings from recent emissions inventories for particulate matter(IDEQ 1999a; USEPA 1999a). The table indicates that particulate emissions from FMC andSimplot account for a considerable portion of the overall emissions for the EMF study area. To afirst approximation, therefore, emissions from these facilities also account for a considerable portionof the airborne particulate matter in the EMF study area, but the relative impacts of these facilitieson air quality certainly vary from location to location. Also noteworthy is the fact that the emissionsinventories suggest that FMC might release more than five times more particulate matter to the airthan does Simplot.
Though this health consultation evaluates many different pollutants that FMC and Simplot emit,much of this document focuses on the facilities' emissions of particulate matter--a class ofpollutants consisting of solid particles and liquid droplets in the air. The sidebar on the followingpage provides definitions of, and relevant background information for, particulate matter.
Although the FMC and Simplot facilities have emitted pollutants in varying quantities over theyears, it does not necessarily follow that residents have been continuously exposed to the site-relatedpollutants. Local meteorological conditions determine whether emissions from the facilities rapidlydisperse in the air or gradually accumulate to potentially unhealthy levels. To understand how theselocal conditions affect levels of air pollution, ATSDR reviewed several studies that evaluated how emissions from FMC and Simplot disperse in the atmosphere(Bechtel 1993; USEPA 1999d; OMNI 1991a; TRC 1993). These studies identified manymeteorological conditions that affect local air pollution, but two factors--surface windpatterns and stagnation episodes (or inversions)--appear to have the strongest impact on air pollution in the EMF study area:
Surface winds. Not surprisingly, the wind direction plays a very important role on airquality issues in the EMF study area: winds blow emissions from the facilities to"downwind" locations, including parts of Chubbuck, Pocatello, and the Fort Hall IndianReservation. According to wind direction measurements both at the Pocatello Airport (seeFigure 2) and near FMC's main process operations, the prevailing wind direction atlocations immediately north of the industrial complex is from the southwest to the northeast(USEPA 1999d; TRC 1993). This wind pattern suggests that emissions from the facilitiesgenerally, but not always, blow toward the northeast. Somewhat consistent with thisprevailing wind direction is the fact that community members have often reported seeing "adense brown cloud" extend from near the FMC and Simplot facilities to locations as far as 5miles to the north (Sho-ban 1989).
Though wind patterns observed at the Pocatello Airport exhibit consistent trends from yearto year, prevailing wind patterns are considerably different at other locations in the EMFstudy area. For instance, a meteorological station operated near the Simplot facility hasfrequently observed winds blowing from the southeast to the northwest--a wind directionrarely observed at the Pocatello Airport (Bechtel 1996). Moreover, prevailing wind patternsin the Portneuf River valley, where the city of Pocatello is located, are also expected to havea strong southeasterly component, due largely to influences from local terrain (TRC 1993). In fact, IDEQ recently observed a prevailing southeasterly wind pattern at its meteorologicalmonitoring station near downtown Pocatello (IDEQ 1999a).
Two studies have reported noteworthy associations between certain wind conditions andlevels of air pollution at locations downwind of the FMC and Simplot facilities. Morespecifically, roughly 75% of the highest PM10 concentrations measured by IDEQ atlocations northeast of the FMC and Simplot facilities have occurred when relatively strongwinds (i.e., 24-hour average wind speed greater than 9 miles per hour) blow from thesouthwest (IDEQ 1999a). Further, an ongoing study at the EMF site indicates that thehighest concentrations of PM10 at a location directly across the street from the FMC facilityare associated with winds blowing from FMC toward the monitors (USEPA 1999d). SectionIV.D comments on these studies further.
Stagnation conditions (inversions). Some of the highest levels of air pollution in the EMFstudy area have occurred during stagnation conditions (IDEQ 1999a). In fact, a particularlysevere stagnation episode occurred in December 1999, as ATSDR was preparing an earlierrelease of this health consultation. In general, these stagnation conditions, which arecharacterized by a calm atmosphere, light and variable winds, little or no precipitation, andnear ground-level inversions, are typically observed in the winter, but they are observedinfrequently. In fact, in some years, stagnation episodes have not occurred at all in the EMFstudy area. During the infrequent stagnation periods, however, emissions from FMC,Simplot, and other local sources become trapped in the lowest levels of the atmosphere. When stagnation conditions persist or are severe, air pollution throughout this area can reachpotentially unhealthy levels.
Some researchers have characterized the specific meteorological conditions that areassociated with the infrequent inversions. For instance, IDEQ has reported that thewintertime inversions generally occur on days with "temperatures near or below freezing;relative humidities above 70 percent; and multi-day meteorologically stagnant conditions"(IDEQ 1998b). Consistent with this observation, EPA has reported that the inversions occurprimarily during "very specific and rare meteorological conditions--cold stagnant winterdays with relative high humidity" (USEPA 1999a). As discussed in greater detail in SectionIV.D, the aforementioned stagnation conditions are a major factor in the infrequent pollutionepisodes, or days when airborne particulate matter in much of Chubbuck, Pocatello, and the Fort Hall Indian Reservation reach unusually high levels.
It should be noted that ATSDR has reviewed several dispersion modeling studies (studies thatsimulate the transport of emissions in the atmosphere) for the EMF study area (Bechtel 1993; IDEQ1991; OMNI 1991b; TRC 1993). Though these studies provide insight into levels of air pollutionin locations where monitoring has not been conducted, the dispersion modeling results can be highlyuncertain and are limited by the accuracy of critical inputs, particularly the emission rates from thephosphate processing plants. Perhaps the only consistent finding among these studies, however, isthat modeled concentrations of PM10 are highest in the immediate vicinity of FMC and Simplot andthat trace levels of site-related contaminants are predicted to occur throughout the EMF study area,including at locations in the cities of Chubbuck and Pocatello, at the Fort Hall Agency, and inunincorporated areas between these locations.
Though ATSDR considered conducting its own dispersion modeling analysis for the EMF studyarea, the Agency eventually decided to abandon such efforts after learning of the difficulties EPAencountered with modeling emissions from FMC. As evidence of this, EPA has recently reportedthat ". . .despite repeated efforts of EPA, with the assistance of the Tribes, IDEQ, and affectedindustry, the air quality models initially selected and approved by EPA for use in the Power-Bannock area PM10 non-attainment area, have continued to fail well-established performancecriteria in the vicinity of the FMC facility. . ." (USEPA 1999a). For this reason and many otherreasons, ATSDR decided that dispersion modeling results for the EMF site would undoubtedly beextremely uncertain and might possibly raise more questions than they would answer. As a result,the conclusions in this health consultation are based entirely on trends and patterns among the largevolume of available air monitoring data, which, as mentioned previously, characterize airconcentrations of the pollutants that FMC and Simplot emit in greatest quantities.
This section reviews the results of relevant ambient air monitoring studies, or studies of the air thatpeople breathe. Since various organizations have measured levels of air pollution in the EMF studyarea over the past 25 years, a large volume of ambient air monitoring data are available for reviewfor many locations in the EMF study area. To illustrate this, Figure 3 indicates the locations of themonitoring stations operated by IDEQ and the Shoshone-Bannock Tribes. Further, Appendix A ofthis report includes ATSDR's review of 12 different air monitoring studies conducted in this area.
Since each study has a limited scope, no single study is sufficient for understanding how levels of airpollution have changed throughout the EMF study area over the years. Combining the results fromthe many studies, however, provides an extensive and consistent account of air quality in this region. More specifically, the collective weight-of-evidence from these studies indicates the followinggeneral trends in air quality:
- The data clearly show that air pollution in the EMF study area, like the air in most urbancenters in the United States, contains many different components. However, most studies ofthe air in the EMF study area have focused on measuring levels of particulate matter, and thechemicals contained in particulate matter. The remainder of this section also focuses onthese pollutants.
- Air monitoring data collected from 1975 to the present have consistently shown thatconcentrations of particulate matter, when averaged over the long term, are highest in theimmediate vicinity of the FMC and Simplot facilities and gradually decrease with distancefrom the facilities. The most plausible explanation for this trend is that emissions from FMCand Simplot largely account for the higher levels of particulate matter in the facilities'vicinity, and this influence decreases with distance from the plants.
- Air monitoring data collected by the Shoshone-Bannock Tribes at a location across the streetfrom FMC has consistently shown the highest levels of certain types of air pollution in theentire EMF study area. Moreover, an extensive source apportionment study has quiteclearly identified air emissions from FMC as the source of the elevated levels of air pollutionat this location on the Fort Hall Indian Reservation (USEPA 1999d).
- At monitoring stations northeast of FMC and Simplot, concentrations of particulate matterbetween 1994 and the present were, on average, more than 30% lower than concentrationsmeasured prior to that time--a concentration trend that was found to be statisticallysignificant. Since the decreasing concentrations were observed at locations that used thesame PM10 sampling methods since the mid 1980s, ATSDR has ruled out the possibilitythat the downward trend is somehow influenced by use of multiple sampling methods withdiffering sensitivities.(1) Though installation of emission controls at FMC and Simplot, andimplementation of a residential wood combustion program have all been credited, to varyingdegrees, for causing the decreasing PM10 concentrations, the exact reason or reasons for this decline are not fully understood.
- Though long-term average concentrations of PM10 decreased in recent years, inversions canstill cause unhealthy levels of air pollution to occur in the EMF study area. As evidence ofthis, some of the highest levels of air pollution ever measured in the city of Pocatellooccurred during a particularly severe 6-day inversion in December 1999. IDEQ hasconcluded that "industrial sources are significant contributors" to the elevated levels of air pollution during inversions (IDEQ, 2000b).
- Despite the large volume of ambient air monitoring data currently available, important datagaps exist. Most notably, no monitoring has been conducted in areas on the Fort Hall IndianReservation north of FMC and Interstate 86, and chemical analysis of particulate filters has not been conducted routinely at most monitoring stations.
ATSDR's more detailed findings regarding the ambient air monitoring data are presented below,classified by pollutant. Selected supporting calculations are documented in appendices, as noted. The findings are based only on ambient air monitoring data collected from 1975 to the present. Without extensive data available for earlier years, ATSDR cannot make firm conclusions aboutlevels of air pollution in the EMF study area prior to 1975.
The following discussion does not comment on whether the ambient air monitoring data trendsindicate health hazards. Such analyses can be found in the "Public Health Implications" section, orSection IV.E.
Overview of Exposures to Particulate Matter: The Area of Impact. As a brief summary of theAgency's findings regarding exposures to particulate matter, Figure 4 indicates the area whereATSDR believes concentrations of PM10 or PM2.5, either over the short term (24-hour average) orthe long term (annual average), have exceeded health-based comparison values at least one timebetween 1975 and the present. ATSDR derived the area of impact in Figure 4 from the following observations:
- Between 1975 and the present, every one of IDEQ's monitoring stations in Pocatello andChubbuck have had at least one 24-hour average PM10 or PM2.5 concentration greater thanEPA's corresponding health-based standards. Since the highest concentrations at thesestations appear to be largely caused by prolonged stagnation conditions (IDEQ 1998b;IDEQ 1999a; USEPA 1999a), which tend to trap pollutants in the lowest levels of theatmosphere throughout the Portneuf Valley, ATSDR has reason to believe that airborneparticulate matter has reached potentially unhealthy levels throughout the city of Pocatelloon isolated occasions in the past and can continue to do so in the future. The area of impactin Figure 4 reflects this determination.
- As Appendices A and C explain, both annual average and 24-hour average concentrations ofPM10 at the Pocatello Sewage Treatment Plant have exceeded EPA's corresponding health-based standard periodically between 1975 and the present. Since the short-term elevatedlevels of PM10 are generally influenced by winds blowing from FMC and Simplot towardthe monitor, ATSDR has reason to believe that concentrations of PM10 in the areas betweenthe facilities and the Pocatello Sewage Treatment Plant have also reached potentiallyunhealthy levels. Moreover, since the concentrations measured at the Pocatello SewageTreatment Plant are likely representative of air quality for areas surrounding the monitor,ATSDR has reason to believe that levels of particulate matter roughly within 1 mile of thismonitoring station also exceeded health-based standards, though this finding is clearlysomewhat uncertain (as expanded upon below). The area of impact in Figure 4 reflects this finding.
- Since no air monitoring studies have been conducted in areas more than 1 mile north ofFMC and Simplot, north of the Pocatello Sewage Treatment Plant, or north of ChubbuckSchool, the northern extent of the area of impact in Figure 4 cannot be established with thedata currently available and, therefore, is unknown. Figure 4 reflects this finding by using adashed line to mark the northern extent of the area of impact and a caption to explain thesignificance of this finding. The lack of monitoring data in this part of the EMF study area is a critical data gap that needs to be filled.
Overall, ATSDR believes the area of impact shown in Figure 4 is a best estimate of the areas wherelevels of airborne particulate matter (whether PM10 or PM2.5, whether over the short term or thelong term) have exceeded health-based standards at some time between 1975 and the present. Giventhe fact that elevated concentrations of particulate matter have occurred throughout this area asrecently as December 1999, ATSDR believes that elevated concentrations will likely occur in thefuture unless the main emissions sources in the area are reduced. As documented above and in theAppendices to this report, marking the boundaries of the area of impact in Figure 4 involvesconsiderable uncertainty.
Recognizing this, ATSDR emphasizes that the boundaries should be viewed as a defensible estimateof the actual region were concentrations have exceeded health-based standards, and the boundaryshown might understate or overstate the actual area over which concentrations reached potentiallyunhealthy levels. In other words, some residents who live outside the shaded region in Figure 4might have been, and continue to be, exposed to levels of particulate matter higher than relevanthealth-based standards, and some residents who live within the shaded region might not have beenexposed to such levels.
Many different emissions sources are believed to contribute to the elevated levels of particulatematter in the EMF study area, but emissions from FMC and Simplot undoubtedly account for aconsiderable portion of the air pollution in this area, especially in areas immediately downwind ofthe facilities. A detailed source apportionment study, however, is not included in the scope of thishealth consultation.
More information on the short-term and long-term concentrations of PM10 and PM2.5 in the EMFstudy area follows:
PM10. The results of the many air quality studies performed in the EMF study area show thatambient air concentrations of PM10 have varied both with time and with location. The followingdiscussion comments on these temporal and spatial variations by answering two basic questionsabout airborne levels of PM10 near the EMF site. The questions address 24-hour averageconcentrations separate from annual average concentrations of PM10, since health-based standardshave been developed for both exposure durations. Responses to the following questions are a criticalinput to the "Public Health Implications" section of this document:
At what locations were 24-hour average PM10 concentrations higher thancorresponding health-based comparison values? The weight-of-evidence from theambient air monitoring studies suggests that 24-hour average concentrations of PM10throughout Chubbuck and Pocatello and in parts of the Fort Hall Indian Reservationperiodically exceeded health-based standards (i.e., 150 ug/m3) and have the potential to doso in the future. As noted earlier in this report, elevated concentrations near FMC andSimplot are generally associated with strong southwesterly winds that blow emissionstoward the monitors, and elevated levels in the Portneuf Valley are generally associated withstagnation conditions, during which emissions from FMC and Simplot and many othersources appear to affect air quality.
The exceedances were clearly most frequent and most severe in the immediate vicinity of theFMC and Simplot facilities. Specifically, EPA has reported that 24-hour average PM10concentrations measured at a location on the Fort Hall Indian Reservation north of FMC andsouth of Interstate 86 exceeded 150 ug/m3 up to 21 days in 1996 and 20 days in 1997(USEPA 1999a), but the exact spatial extent of this poor air quality is not known. Similarly,according to IDEQ's monitoring data, the number of days with PM10 concentrations abovehealth-based standards also varied from year to year: in some years, no exceedances wereobserved in Chubbuck and Pocatello at all; in other years, however, as many as 6exceedances likely occurred (IDEQ 1999a). Exceedances of PM10 air quality standardsoccurred in Pocatello as recently as December 31, 1999--a finding that is based on data thatIDEQ recently released to ATSDR (IDEQ 2000).
Appendix C.1 presents the evidence ATSDR considered in reaching its conclusion regarding24-hour average concentrations of PM10. Note, ATSDR considers the lack of monitoringdata on the Fort Hall Indian Reservation at locations north of Interstate 86 an important datagap that needs to be filled.
At what locations were annual average PM10 concentrations higher thancorresponding health-based comparison values? The weight-of-evidence suggests that,in at least one year between 1975 and the present, annual average PM10 concentrationsexceeded EPA's health-based comparison value (50 ug/m3) in parts of Chubbuck, Pocatello,and the Fort Hall Indian Reservation. The frequency with which annual average levelsexceeded health-based standards appears to decrease with distance from the EMF site.
Air monitoring studies sponsored by FMC, Simplot, and EPA all indicate that annualaverage PM10 concentrations have exceeded EPA's health-based standard in an areaimmediately north of FMC (Bechtel 1995; Hartman 1999; USEPA 1999a). ATSDRbelieves these studies, taken together, suggest that concentrations of PM10 likely exceededthe annual average air quality standard in a small area for at least the last 6 years, andprobably longer. According to EPA, trends in the ambient air monitoring data "pointconclusively to FMC as the source" of the elevated PM10 concentrations in the area betweenFMC and Interstate 86 (USEPA 1999d). Note, it is not known how far north of the facilitiesconcentrations exceeded health-based standards.
In addition to the data collected in the vicinity of FMC and Simplot, IDEQ's monitoringdata suggest (1) that annual average PM10 concentrations at the Pocatello SewageTreatment Plant might have exceeded 50 ug/m3 in as many as 12 years between 1975 andthe present, and (2) that annual average PM10 levels at Chubbuck School might haveexceeded this level in 3 years or fewer during this same time frame. On the other hand,ATSDR does not believe that such elevated annual average levels occurred at either Garretand Gould or Idaho State University. As Appendix C.3 explains, these estimates are based,in part, on extrapolations of TSP monitoring data and therefore are somewhat uncertain. Appendix C.2 presents the evidence ATSDR considered in reaching its conclusion.
PM2.5. Though ambient air concentrations of particulate matter have been measured extensivelythroughout the Pocatello area, few studies have measured concentrations of fine particles, alsoknown as PM2.5. Nonetheless, the available PM2.5 monitoring studies characterize the sizedistribution of airborne particles typically observed in the EMF study area. Knowledge of theparticle size distribution, coupled with the PM10 and TSP measurements made over the years,provides insight into what PM2.5 concentrations might have been during times when this pollutantwas not actually measured.
Responses to the following two questions summarize ATSDR's findings regarding the levels ofPM2.5 that likely occurred in the EMF study area between 1975 and the present. Like the questionsin the review of PM10 concentrations, the following questions address 24-hour average and annualaverage concentrations separately. Responses to the following questions are a critical input to the"Public Health Implications" section of this document:
At what locations were 24-hour average PM2.5 concentrations higher thancorresponding health-based comparison values? To date, 24-hour average ambient airconcentrations of PM2.5 have been measured at several locations, including across the streetfrom FMC and at the Pocatello Sewage Treatment Plant, Chubbuck School, Idaho StateUniversity, and Garret and Gould. The most extensive PM2.5 monitoring effort conductedin the EMF study area to date has shown that 24-hour average ambient air concentrations ofPM2.5 across the street from FMC frequently exceeded health-based comparison values (i.e.,65 ug/m3) between October 1996 and September 1998 (USEPA 1999d). It is reasonable tobelieve that these exceedances occurred at this location prior to October 1996, even thoughmonitoring was not conducted during this time. It is not known how far north these elevatedPM2.5 concentrations occur.
In addition to the data collected across the street from FMC, IDEQ has measured 24-houraverage PM2.5 concentrations above health-based standards at all four of its monitoringstations. Since some of the elevated PM2.5 concentrations occurred as recently as December1999, ATSDR believes it is possible that elevated PM2.5 levels will continue to occur in thefuture unless sources of this pollutant are reduced. Unlike the trend observed across thestreet from FMC, the elevated 24-hour average concentrations of PM2.5 in Chubbuck andPocatello appear to occur infrequently, primarily during stagnation episodes or inversions.
Appendix D.1 presents the evidence ATSDR considered in reaching its conclusion regarding24-hour average concentrations of PM2.5. The lack of extensive PM2.5 monitoring dataon the Fort Hall Indian Reservation at locations north of Interstate 86 is an important datagap that needs to be filled.
At what locations were annual average PM2.5 concentrations higher thancorresponding health-based comparison values? The available monitoring data suggeststhat annual average levels of PM2.5 were highest in the immediate vicinity of the EMFstudy area, with levels gradually decreasing with downwind distance. For instance, the mostextensive PM2.5 monitoring study to date has shown that annual average concentrations ofthis pollutant have exceeded, and continue to exceed, 15 ug/m3 at locations immediatelynorth of FMC. Based on a limited set of data collected by IDEQ in 1998 and 1999, annualaverage concentrations of PM2.5 currently do not exceed health-based standards throughoutChubbuck and Pocatello.
The weight-of-evidence suggests that, in the years before the PM2.5 monitoring studies wereconducted, annual average PM2.5 concentrations likely exceeded EPA's health-basedcomparison value (15 ug/m3) in much of Chubbuck and Pocatello and in parts of the FortHall Indian Reservation. As Appendix D.2 explains, this finding is based primarily onextrapolations of PM10 monitoring data, using defensible estimates of PM2.5/PM10 ratios. In other words, this finding is somewhat uncertain since it is based on estimated--notmeasured--concentrations of PM2.5.
Appendix D.2 presents the evidence ATSDR considered in reaching its conclusion regarding annual average concentrations of PM2.5.
Ionic species in particulate matter. Since studies have linked inhalation exposure of acid aerosolsto an increased incidence of adverse health effects among sensitive populations, ATSDR obtainedand reviewed ambient air monitoring data for several ionic species. These data were found forammonium, chloride, fluoride, nitrate, potassium ion, and sulfate (Bechtel 1996; IDEQ 1999b). Ofthese species, the highest peak concentrations observed to date were for ammonium (42.75 ug/m3),nitrate (27.15 ug/m3), and sulfate (83.9 ug/m3) (IDEQ 1999b). Interestingly, these three peakconcentrations all occurred at the Idaho State University monitoring station--the IDEQ stationlocated furthest from the FMC and Simplot facilities.
The fact that the highest concentrations of these ions occurred far from FMC and Simplot does notnecessarily imply that emissions from these facilities contributed little to the measured levels. To thecontrary, the data trends are consistent with the hypothesis that emissions from the two facilitiesaccounted for a considerable portion of the measured concentrations. For example, IDEQ hasestimated that emissions of sulfur dioxide from FMC and Simplot account for more than 93% of thetotal emissions of sulfur dioxide in the EMF study area (IDEQ 1999d). Since sulfur dioxideemissions are a precursor to ambient sulfate ions, and since FMC and Simplot clearly emit moresulfur dioxide to the air than all other sources in the area combined, it is reasonable to assume thatthe peak concentrations of sulfate at Idaho State University can be attributed, to a large extent, toemissions from the phosphate processing plants. Moreover, given the fact that it takes time forairborne sulfur dioxide to react and form sulfates, it is not surprising that the highest sulfateconcentrations have been observed at the monitoring station located furthest from FMC and Simplot. Regardless of the source of these ions, however, Section IV.E evaluates whether these elevatedconcentrations present a public health hazard.
Though never measured at the levels observed for ammonium, nitrate, and sulfate, fluoride wasconsistently detected in air samples, particularly those collected in close vicinity to Simplot, a knownsource of fluoride emissions. For example, the RI reported that the highest concentrations offluoride were measured at the three stations located around the perimeter of Simplot. The highestconcentrations for these stations were 13.14 ug/m3, 11.29 ug/m3, and 10.92 ug/m3; averageconcentrations were not reported for these stations (Bechtel 1996). All of the samples from IDEQ'snetwork that were selected for chemical analyses had concentrations lower than those measuredduring the RI. The "Public Health Implications" section reviews the fluoride concentrations ingreater detail.
ATSDR reviewed the available monitoring data for the two remaining ionic species (chloride andpotassium ion), but both species were measured at considerably lower levels than the other ionicspecies discussed above. More specifically, concentrations of chloride and potassium ion in the 72valid samples collected were all less than 2.0 ug/m3. A brief toxicological evaluation is presentedfor these ions in the "Public Health Implications" section.
Phosphorous compounds (phosphorous, phosphate, phosphine, phosphorous pentoxide). Since both FMC and Simplot process vast quantities of phosphorous every year, ATSDR carefullyexamined the measured ambient air concentrations of various forms of phosphorous. To date,ambient air monitoring studies conducted in the EMF study area have measured levels of totalphosphorous in particulate matter as well as levels of phosphate ion (PO43-). However, no studieshave characterized ambient air concentrations of phosphorous pentoxide--a pollutant known to beemitted by FMC (Bechtel 1993). Though ATSDR identified emissions estimates and dispersionmodeling results for phosphorous pentoxide, the lack of ambient air monitoring data appears to bedue to the lack of approved sampling and analytical methods for this compound. As a result, theactual levels of phosphorous pentoxide that people might have breathed, and might continue tobreathe, are not known.
ATSDR does not consider this a critical data gap in the health consultation, however, sincephosphorous pentoxide is known to react rapidly in air to form phosphate ion (USEPA 1999b). Dueto this reaction, phosphorous pentoxide emitted by FMC will partly, if not entirely, transform tophosphate ion by the time the emissions reach residential areas. Thus, ATSDR believes evaluatingambient air concentrations of total phosphorous and of phosphate ion will adequately address thecommunity concerns regarding emissions of phosphorous pentoxide.
Not surprisingly, concentrations of total phosphorous were consistently found to be highest in areasclosest to FMC and Simplot.(2) For example, according to the RI, average concentrations of totalphosphorous at a monitoring location immediately north of FMC were more than five times higherthan average concentrations measured at any of the six other monitoring locations (Bechtel 1996). The magnitude of total phosphorous concentrations also varied with time: sometimes phosphorouswas not detected in 24-hour average samples, and other times it was detected at concentrations ashigh as 26.8 ug/m3 (Bechtel 1996; USEPA 1999d). The highest long-term average concentrationof total phosphorous reported to date is 5.45 ug/m3, at a location immediately north of FMC andbased on nearly 1 year of routine sampling (Bechtel 1996). Though neither ATSDR nor EPA havepublished health-based comparison values for total phosphorous, the "Public Health Implications"section of this report carefully reviews available toxicological data for this metal.
ATSDR also reviewed data available on concentrations of phosphate ion, which were measured byIDEQ using ion chromatography. Data trends for phosphate ion were quite similar to thosediscussed above for phosphorous. However, because these measurements were not conductedroutinely, representative average concentrations of phosphate ion cannot be calculated and comparedto the average phosphorous concentrations. Nonetheless, the sporadic measurement of phosphateion concentrations provides some insight into the magnitude of concentrations that have beenobserved in the area. More specifically, at the Pocatello Sewage Treatment Plant, 38 24-houraverage measurements of phosphate ion have been made over a 5-year period, of which, half hadphosphate ion concentrations between 10 and 50 ug/m3 and the other half had phosphate ionconcentrations lower than this range (IDEQ 1999b). Of the more limited phosphate ionmeasurements at IDEQ's three other monitoring stations, which are all in residential neighborhoods,no concentrations of phosphate ion were found to exceed 10 ug/m3. The "Public HealthImplications" section of this report comments on the significance of these measurements.
Finally, ATSDR obtained and reviewed emissions and monitoring data for phosphine, an inorganicform of phosphorous that is released from FMC's on-site waste management ponds (Bechtel 1998b;FMC 1999a, 1999b, 1999c, 1999d, 2000a). Unlike the data available for the other chemicalsemitted by FMC and Simplot, no off-site ambient air monitoring data are available for phosphine,thus greatly limiting ATSDR's ability to evaluate past and current exposures. Nonetheless, ATSDRhas learned that FMC has developed "pond management standards" that include provisions foremissions monitoring, fenceline air monitoring, and "a response action plan to ensure that the publicwill not be exposed to phosphine . . . levels that exceed federal guidelines" (Bechtel 1998). Thesemanagement standards reportedly have been reviewed and approved by both EPA and the Shoshone-Bannock Tribes (Bechtel 1998). ATSDR reviewed a limited set of phosphine sampling data thatFMC collected, which indicated that phosphine concentrations measured at the facility fencelineusing an OSHA-approved sampling and analytical method ranged from nondetect to 101 ppb(Bechtel 1998). Subsequent continuous measurements have shown phosphine concentrations at theedge of on-site ponds to range from nondetect to 2,310 ppb (FMC 1999a, 1999b, 1999c), andmeasurements of phosphine air concentrations at the facility fenceline on four occasions havereportedly exceeded 1.0 ppm: 1.90 ppm on October 6, 1999; 1.10 ppm on October 23, 1999; 2.50ppm on November 15, 1999; and 3.16 ppm on November 16, 1999 (FMC 1999d, 2000). Thesefenceline measurements were collected using "hand-held monitors and Draegers" and not usingmethods approved by federal agencies (OSHA has an approved phosphine sampling method). ATSDR reviewed additional phosphine monitoring data, but they were collected using a hand-helddevice that is known to report "false positive" detects for phosphine and, thus, are not included inthis health consultation. Section IV.E reviews the significance of the measured phosphineconcentrations, but ATSDR notes that the available data for this pollutant are limited.
Metals and other inorganics. Several ambient air monitoring studies have measuredconcentrations of metals and other inorganics in particulate matter at various locations in the EMFstudy area (Bechtel 1995; IDEQ 1999b; USEPA 1999d). Combined, these studies characterizeairborne levels of more than 40 metals and other inorganics--most of which are emitted by eitherFMC or Simplot, or by both facilities. Table 3 lists these elements and summarizes how themeasured concentrations compared to health-based comparison values. The table classifies the metals and other inorganics into three categories:
- 8 metals or inorganics were measured at levels exceeding their corresponding health-basedcomparison value on at least one occasion. For these elements, the frequency with whichconcentrations exceeded comparison values is summarized below, and the significance ofthese concentrations is reviewed in the "Public Health Implications" section of this report.
- 16 metals or inorganics were always measured at levels lower than their correspondinghealth-based comparison values. Thus, the monitoring data suggest that ambient airconcentrations of these 16 elements have not reached "unsafe" or "unhealthy" levels in theEMF study area. Accordingly, toxicological evaluations of these metals and inorganics are not provided in this health consultation.
- 25 of the metals or inorganics that were measured during the air monitoring studies do nothave health-based comparison values published by ATSDR or EPA. Of these elements, six(calcium, carbon, phosphorous, potassium, silicon, and sulfur) had highest concentrationsgreater than 1.0 ug/m3, and the remaining 19 had concentrations lower than this level. The"Public Health Implications" puts the monitoring data for these 25 metals and inorganicsinto perspective.
As noted earlier, and described in detail in Appendix B, when ambient air concentrations of a givenpollutant exceed corresponding comparison values, this situation does not necessarily suggest thatadverse health effects will occur, but it rather suggests that concentrations of the pollutant should beevaluated in greater detail to make conclusions on public health implications. As a critical input tothe toxicological evaluations presented later in this report, the following list describes in greaterdetail the extent to which concentrations of 8 metals exceeded health-based comparison values. The"Public Health Implications" section of this report comments on the significance of the followingtrends.
Note, in the summaries below, results from three different studies were considered for identifying themaximum concentrations of metals and other inorganics (Bechtel 1995; IDEQ 1999b; USEPA1999d). Since one of these studies (IDEQ 1999b) did not routinely analyze filters for chemicalcomposition, ATSDR used only the data from the Remedial Investigation and the Fort Hall SourceApportionment Study to comment on average concentrations of metals and other inorganics.
Aluminum. Though concentrations of aluminum were measured in several air monitoringstudies, only two monitoring locations (monitoring station 6 from the RI, see Appendix A.2,and the "Primary" station in the Fort Hall Source Apportionment Study, see Appendix A.4)reported concentrations of aluminum greater than the metal's most conservative health-basedcomparison value (3.7 ug/m3). The average concentrations(3) of aluminum in PM10 at thesestations (0.15 ug/m3 and 0.85 ug/m3), however, were considerably lower than thecomparison value. Concentrations of aluminum measured at all other monitoring stationswere also considerably lower than the comparison value as well.
Arsenic. Three air monitoring studies indicated that concentrations of arsenic have recently,and frequently, exceeded the most conservative health-based comparison value (0.0002ug/m3) (Bechtel 1995; IDEQ 1999b). Average concentrations in PM10 measured duringthe RI ranged from 0.000502 to 0.00127 ug/m3 (Becthel 1995). Moreover, at the"Primary" station in the Fort Hall Source Apportionment Study, the highest annual averageconcentration of arsenic in PM10 was 0.0012 ug/m3. This study clearly showed that theelevated metals concentration at the "Primary" station were caused primarily by emissionsfrom FMC. Concentrations measured both in the immediate vicinity of the EMF site and innearby residential areas, therefore, were found to be higher than the most conservativehealth-based comparison value.
Barium. Of the numerous reported concentrations of barium that ATSDR reviewed, onlyone concentration--from a sample collected by IDEQ at the Pocatello Sewage TreatmentPlant in 1991--exceeded the corresponding most conservative health-based comparisonvalue. This one concentration (0.57 ug/m3) was only marginally higher than thecorresponding comparison value (0.51 ug/m3). At all other monitoring locations, everyconcentration of barium reported was considerably lower than the comparison value.
Beryllium. Of the many studies that measured ambient air concentrations of metals, onlythe RI measured ambient levels of beryllium (Bechtel 1995). As Appendix A.2 shows,every concentration of beryllium measured at six of the seven monitoring locations in thisstudy was lower than the corresponding health-based comparison value (0.0004 ug/m3). Station 2, on the other hand, which was located immediately north of FMC in anunpopulated area, had a single concentration in TSP higher than this comparison value. Theaverage concentration of beryllium in PM10 at this station (0.000179 ug/m3), however, waslower than the health-based comparison value.
Cadmium. Every study that has conducted speciated particulate monitoring in the EMFstudy area has reported both highest and average concentrations of cadmium at levelsexceeding the most conservative health-based comparison value (0.0006 ug/m3). This trendwas observed for every monitoring station in the RI (see Appendix A.2), for the Shoshone-Bannock monitors (see Appendix A.4), and for the IDEQ air monitoring network (seeAppendix A.9). The highest average cadmium concentration in PM10 (0.035 ug/m3) wasobserved at the "Primary" station in the Fort Hall Source Apportionment Study; thecadmium detected at this station was shown to originate primarily from FMC's emissions(USEPA 1999d). The levels of cadmium measured at stations closer to the FMC andSimplot facilities were consistently higher than the levels measured at stations further fromthe industrial complex.
Chromium. Three studies have routinely analyzed particulate filters to measureconcentrations of chromium (Bechtel 1995; IDEQ 1999b; USEPA 1999d). Interpretingthese ambient air monitoring data, however, is complicated by the fact that chromium isoften found in two different states (hexavalent and trivalent). These states have entirelydifferent implications from a toxicological perspective. As an initial screening, ATSDRcompared the measured concentrations of chromium to the most conservative health-basedcomparison value for the metal, which happens to be for the hexavalent state (0.00008ug/m3). This initial screening found that highest and average concentrations of chromiumat every sampling location, whether in residential neighborhoods or in close proximity to theEMF study area, exceeded the comparison value for hexavalent chromium. The highestaverage concentration of total chromium in PM10 (0.029 ug/m3) was observed at the"Primary" station in the Fort Hall Source Apportionment Study. Moreover, concentrationsof chromium at locations along the perimeter of FMC and Simplot were consistently higherthan those at downwind monitoring locations.
Manganese. Concentrations of manganese were measured in three studies, but only a smallsubset of the concentrations reported in two of these studies exceeded the correspondinghealth-based comparison value (0.04 ug/m3). As Appendix A.2 describes, data collectedduring the RI indicate that ambient air concentrations of manganese in TSP exceeded thecomparison value on at least one occasion at six of the seven monitoring locations, includingat the two monitoring stations near residential neighborhoods. At all seven monitoringstations, however, the average concentrations of manganese in PM10 were notably lowerthan the comparison value. Consistent with this trend, monitoring data collected by IDEQindicate that concentrations of manganese in PM10 generally exceeded the health-basedcomparison value on days when particulate concentrations were high, but the IDEQ data areinsufficient for calculating average concentrations. In the Fort Hall Source ApportionmentStudy, manganese never exceeded its comparison value in the fine fraction of particulatematter; in the coarse fraction, however, one sample had a manganese concentration (0.067ug/m3) greater than the comparison value. The average levels of manganese in PM10 duringthe Fort Hall Source Apportionment Study were lower than the comparison value.
Vanadium. Ambient levels of vanadium in the vicinity of the EMF study area have beenroutinely measured during three different sampling efforts. Two of the sampling effortsnever detected the metal at levels higher than the most conservative comparison value (0.2ug/m3). The RI, on the other hand, reported several concentrations in TSP at levels higherthan the comparison value, but only in unpopulated areas in the immediate vicinity of FMCand Simplot. At all seven monitoring stations that operated during the RI, averageconcentrations of vanadium in PM10 were lower than the comparison value.
Sulfur Dioxide. For more than 20 years, IDEQ has measured ambient air concentrations of sulfurdioxide in the EMF study area. Specifically, IDEQ monitored sulfur dioxide levels at the PocatelloSewage Treatment Plant from 1977 to the present and at Garret and Gould from 1994 to thepresent. Overall, every annual average concentration of sulfur dioxide at both monitoring locationswas less than EPA's health-based air quality standard (an annual average concentration of 0.03ppm). However, a subset of 24-hour average concentrations of sulfur dioxide at the PocatelloSewage Treatment Plant were higher than EPA's corresponding health-based standard (a 24-houraverage concentration of 0.14 ppm) at least once a year, but not more than six times a year, from1977 to 1985 (IDHW 1988).(4) Since IDEQ's sulfur dioxide monitoring prior to 1994 was limited toone sampling location, however, the area over which elevated sulfur dioxide concentrations occurredin the past is not known, but is likely limited to the immediate vicinity of the monitors at thePocatello Sewage Treatment Plant. Since 1985, concentrations of sulfur dioxide measured by IDEQhave not exceeded health-based comparison values. Therefore, the data suggest that 24-houraverage concentrations of sulfur dioxide exceeded health-based standards in a limited geographicarea periodically between 1977 and 1985, but not again since. The "Public Health Implications"section of this report puts the past elevated concentrations of sulfur dioxide into perspective.
Other pollutants. In addition to the pollutants listed above, ATSDR obtained and reviewedinformation characterizing ambient air concentrations of other pollutants. However, most air qualitystudies conducted in the Pocatello area have focused on particulate matter, and relatively few studieshave measured concentrations of other pollutants, like volatile organic compounds. Nonetheless,recent reports by IDEQ indicate that concentrations of carbon monoxide, nitrogen dioxide, andozone in Power and Bannock Counties are lower than EPA's corresponding health-based standards(IDEQ 1998a).
More specifically, IDEQ has conducted fairly extensive sampling for nitrogen dioxide at its Garretand Gould monitoring station in Pocatello (see Figure 3). Over the course of 5 years of sampling(from 1994 to 1999), annual average concentrations of nitrogen dioxide were always roughly one-third of EPA's health-based NAAQS of 0.053 ppm. Further, IDEQ has measured ozoneconcentrations in the EMF study area, but only during special studies conducted in the wintermonths, when ozone levels are typically at their lowest. All ozone concentrations measured duringthese studies were less than half of EPA's one-hour average health-based standard of 0.120 ppm, butthe extent and timing of sampling are extremely limited.
Finally, ATSDR gathered data on air quality measurements of hydrogen cyanide, a chemicalreleased to the air primarily by the waste-management ponds at FMC. The data obtained byATSDR indicate that monitoring for hydrogen cyanide has been performed only within the FMCproperty boundary, and no off-site monitoring data are available. The limited on-site data suggestthat air concentrations of hydrogen cyanide at the FMC fenceline range from nondetects to as high as430 ppb (Bechtel 1998). More recent monitoring at on-site locations along the perimeter of thewaste management ponds has revealed hydrogen cyanide concentrations ranging from nondetects to990 ppb (FMC 1999a, 1999b, 1999c, 1999d, 2000). FMC continues to monitor emissions andoff-site transport of hydrogen cyanide as part of its "pond management plan," which both EPA andthe Shoshone-Bannock Tribes have approved. Though implementation of this plan provides somelevel of comfort that off-site concentrations of hydrogen cyanide do not reach levels of healthconcern, ATSDR notes that only limited monitoring data are available to support such a conclusion.
Extensive information on pollutants other than those listed above are not readily available for theEMF study area. However, the previous summary reviews air quality data for a very large subset ofpollutants released by FMC and Simplot, especially those released in greatest quantities.
This section evaluates the public health implications of the levels of air pollution in the EMF studyarea. In general, the ambient air monitoring data described in the previous section indicate that alarge segment of the population throughout the EMF study area have, at some time since 1975, beenexposed to some site-related air contaminants, including PM10, PM2.5, and the various constituentof these airborne particles (e.g., metals, fluorides, phosphoric acid, sulfuric acid). This sectionprovides a public health context to the exposures that have occurred to individuals who live near theEMF study area, including residents of Chubbuck, Pocatello, and the Fort Hall Indian Reservation. It is important to note that ambient air monitoring levels are used in this health consultation as asurrogate for exposure in the EMF study area. Actual individual exposure to air pollutants isdetermined by a complex interplay between human activity, including the locations where time isspent, housing characteristics (as they influence penetration of outdoor pollutants), and other factors.
This section opens by providing relevant background information on the many studies that havebeen conducted in other parts of the country to determine public health implications associated withexposures to particulate matter. Following this general background discussion are detailed healthevaluations for the following six categories of site-related contaminants:
- Particulate matter: exposures to PM10 and PM2.5 are evaluated, with a greater emphasis placed on evaluating the potential PM2.5 exposures.
- Sulfates: exposures to sulfates measured in the EMF study area are evaluated.
- Acid Aerosols: exposures to several ionic species (other than sulfates) are considered, including an evaluation of exposures to phosphoric acid.
- Metals and inorganics: exposures to the 8 metals with at least one concentration greater thanits comparison value (see Table 3) are evaluated in detail, and exposures to other metals and inorganics are also briefly discussed.
- Sulfur dioxide: exposures to sulfur dioxide are reviewed and evaluated.
- Phosphine and hydrogen cyanide: potential exposures to these chemicals are brieflyreviewed.
For contaminants that are believed to have reached levels that might be associated with adversehealth effects, the following discussion identifies populations that are believed to be at the greatestrisk. For reference, Appendix B explains some of the health-based comparison values andguidelines that were used to evaluate the public health implications of exposures in the EMF studyarea. It is important to note that there is some scientific debate regarding the levels of PM2.5 orPM10 that are considered protective for all segments of the population. Threshold concentrationsfor PM 2.5 or PM 10 (i.e., a level below which no adverse health effects are likely) have not beenestablished within the scientific literature.
As a result, EPA's PM10 standard and proposed PM2.5 standard may not be protective of all sensitive subpopulations, though it is generally believed that the proposed annual PM2.5 standard is protective of the general population and probably many of the sensitive subpopulations. However, when establishing the PM2.5 standards, EPA intended for the annual average and 24-hour levels to work as a dual standard. That is, the 24-hour standard alone does not protect against short-term health effects but the two standards working in concert are protective. Therefore, EPA set a value of 40 ug/m3 (termed an air quality index, or AQI) as a rough surrogate for the general level of protection provided by the two standards in combination. For more information regarding EPA's use of AQIs, see the notice in the Federal Register, Volume 64, No. 149, page 42542, Wednesday, August 4, 1999.
The following evaluation of the public health implications of exposures to PM incorporates theunderstanding that there are no currently established PM thresholds and the understanding of thedual nature of the PM2.5 standards.
Relevant Background Information on Health Implications of Exposures to PM and RelatedConstituents. Over the past 20 years, numerous investigators have researched the public healthimplications of inhalation exposures to PM. The following discussion reviews this large volume ofresearch, which provided a basis for much of the evaluations presented later in this section.
Prior to 1987, EPA enforced health-based standards that regulated ambient air concentrations oftotal suspended particulates, or TSP. By 1987, a growing amount of research had shown that theparticles of greatest health concern were actually PM10, which, at the time, were shown to becapable of penetrating into sensitive regions of the respiratory tract. Consequently, EPA and thestates took action in 1987 to monitor and regulate ambient levels of PM10. Since 1987, hundreds ofadditional studies (mostly epidemiological) have been published on the health effects of PM. Thesestudies generally suggest that adverse health effects in children and other sensitive populations havebeen associated with exposure to particle levels well below that allowed by EPA's PM10 standard(USEPA 1997). Moreover, it is generally believed that fine particles (PM2.5) can penetrate into thelungs more deeply than PM10 and that fine particles are more likely to contribute to adverse healtheffects than coarse particles (i.e., particles larger than 2.5 microns, but smaller than 10 microns).
According to the various studies on PM, many health effects were found to be associated withPM2.5 exposures or with PM2.5 exposures coupled with exposures to other pollutants (USEPA1997). A partial list of these health effects follows:
- premature death
- respiratory-related hospital admissions and emergency room visits
- aggravated asthma
- acute respiratory symptoms, including aggravated coughing and difficult or painfulbreathing
- chronic bronchitis
- decreased lung function that can be experienced as shortness of breath
These studies indicate that elderly, infants, and persons with chronic cardiopulmonary disease,influenza, or asthma, are most susceptible to mortality and serious morbidity effects from short-termacutely elevated exposures. Others are susceptible to less serious health effects such as transientincreases in respiratory symptoms, decreased lung function, or other physiological changes. Chronicexposure studies suggest relatively broad susceptibility to cumulative effects of long-term repeatedexposure to fine particulate pollution, resulting in substantive estimates of population loss of lifeexpectancy in highly polluted environments (Pope 2000). It is important to note that susceptibilitymay also be dependent on a number of exposure factors, including duration of exposure. The degreeto which an added particle burden may impact an individual will likely be affected by their age,health status, medication usage, and their overall susceptibility to PM inhalation exposures. Certainly, one factor that may promote increased risk in the older population is that, over theirlifespan, they may have had more exposure and hence more opportunity to accumulate particles ordamage their lungs (USEPA 1996). Current epidemiological research does not provide conclusiveevidence of an association between exposure to PM, in general, and cancer. However, since PM ismade up of various constituents, depending on the source(s), there are likely to be chemicalsincluded in PM that are potential carcinogens.
For reasons above, EPA proposed revisions to its PM standards in 1997 to include a primary(health-based) annual average PM2.5 standard of 15 ug/m3 and a 24-hour PM2.5 standard of 65ug/m3 (USEPA 1997). EPA's scientific review concluded that fine particles are a better surrogatefor those components of PM most likely linked to mortality (death) and morbidity (disease) effects atlevels below the previous standard, while high concentrations of coarse fraction particles are linkedto effects such as aggravation of asthma (USEPA 1997).(5)
The body of scientific knowledge used to set the health-based PM2.5 standard consisted primarily ofepidemiological studies of communities exposed to elevated levels of PM--communities like thosein and around the EMF study area. These epidemiological studies found consistent associationsbetween exposure and adverse health effects both for short-term or acute PM exposure scenarios(i.e., usually measured in days) and for long-term or chronic exposure scenarios (i.e., usuallymeasured in years) (USEPA 1996). Chronic exposures are best measured using annual averagePM2.5 levels (concentrations above 15 ug/m3) for one or several years; whereas, acute exposures arebest measured by using the 24-hour average PM10 and PM2.5 levels (concentration above 150ug/m3 and 65 ug/m3, respectively). It should be noted that the epidemiological studies indicateincreased health risks associated with PM exposures, either alone or in combination with other airpollutants.
PM-related increases in individual health risks are small, but likely significant from an overallpublic health perspective because of the large numbers of individuals in susceptible risk groups thatare exposed to ambient PM (USEPA 1996). Although the epidemiological data provide support forthe associations mentioned above, an understanding of the underlying biological mechanisms has notyet emerged (USEPA 1996). Much of the toxicological findings related to PM are derived fromcontrolled exposure studies in humans and laboratory animals. These studies have most extensivelyfocused on acidic aerosols (a subclass of PM), namely sulfuric acid aerosols and various sulfates andnitrates, and have included characterization of acid aerosols effects on pulmonary mechanicalfunctions, lung particle clearance mechanisms, and other lung defense mechanisms (USEPA 1996). Controlled human exposures to PM constituents other than acid aerosols are limited. Laboratoryanimal studies and occupational exposure studies provide information on other PM substances,including metals, diesel emissions, crystalline silica, and other miscellaneous particles. Humanexposure studies of particles other than acid aerosols generally provide insufficient data to drawconclusions regarding health effects (USEPA 1996). A recent study (Godleski, et al. 2000), fundedby the Health Effects Institute (HEI), an independent and unbiased source of information, supportedby both public and private sources, found that concentrated airborne particles had adverse effects onthe electrical regulation of the heart in dogs with a pre-existing heart condition, while the impact onnormal dogs was not clear. Moreover, biological evidence indicates that urban combustion particlescan penetrate past the primary defense mechanisms of the lung, can elicit inflammatory changes inthe lung and systematically (throughout the body), contain a constituent (soluble transition metals)that by itself can be demonstrated to produce lung damage, can produce electrocardiogram changesincluding arrhythmia (heart irregularities), and can kill animals with pre-existing heart and lungdisease (Schwartz 1999). Human studies have also reported inflammatory changes, includingsystemic changes, and changes to cardiovascular risk factors (Schwarz 1999). Although scientificevidence has provided some clues into the biological mechanisms of how PM may elicit adversehealth effects in animals an humans, clear evidence of the exact mechanisms has not emerged.
In summary, the weight-of-epidemiological evidence suggests that ambient PM exposure hasaffected and continues to affect the public health of U.S. populations. However, a great deal ofuncertainty remains regarding many issues related to the overall scientific inquiry into the healtheffects of PM (USEPA 1996). Moreover, several viewpoints currently exist on how best to interpretthe epidemiological data: one sees PM exposure indicators as surrogate measures of complexambient air pollution mixtures and reported PM-related effects represent those of the overallmixture; another holds that reported PM-related effects are attributable to PM components (per se)of the air pollution mixture and reflect independent PM effects; and yet another suggests that PMcan be viewed both as a surrogate indicator as well as a specific cause of health effects. Whicheverthe case, reduction of PM exposure would be expected to lead to reductions in the frequency andseverity of PM-associated health effects (USEPA 1996).
PM2.5 and PM10 Exposures. ATSDR estimates that at least 53,710 persons have been exposedat some time between 1975 and the present to potentially unhealthy levels of either PM10 or PM2.5. This finding is based on census data and the area of impact shown in Figure 4. Of this exposedpopulation, ATSDR estimates that at least 12,129 persons (that is, 6,619 children 6 years andyounger and 5,510 adults aged 65 and older) are in subpopulations that may be sensitive to theeffects of exposure to PM. It is important to note that it is likely that these estimates either overstateor understate the actual population exposed to unhealthy levels of PM. As indicated in Figure 4,since levels of air pollution were not measured at locations north of the EMF study area, ATSDRcannot establish the northern extent of the area of impact.
The health concerns expressed by community members in the EMF study area (i.e., increasedincidence of asthma, upper respiratory illness, and heart disease) are reasonably consistent withadverse health outcomes reported in the epidemiological research for both acute and chronicexposures to PM2.5 and PM10 above health-based standards. However, the consistency betweenthe concerns and the epidemiological studies does not suggest that any given incident of these healthoutcomes is caused solely by inhalation exposures to PM2.5 or PM10. Rather, causality of anygiven disease is usually a result of multiple factors. For example, smoking is a strong risk factor formany lung and heart diseases. Therefore, smokers comprise another population group at likelyincreased risk for PM-related health effects (USEPA 1996).
The following discussion first evaluates the increased risks from exposures to PM2.5 (annualaverages) based on results from chronic mortality epidemiological studies and then evaluates theincreased risks from exposures to PM2.5 and PM10 (24-hour maximum values) based on resultsfrom acute mortality and morbidity epidemiological studies. The ambient air concentrations of PMreported in these epidemiological studies is compared to estimated and measured levels of PM in theEMF study area. The discussions present a qualitative evaluation of the data collected in the EMFstudy area and should provide context for understanding the risk of adverse health effects to personsexposed in the EMF study area.
Chronic Exposures to Annual Average PM2.5 Levels. Two large cohort studies, theHarvard Six-City Study (Dockery 1993) and the American Cancer Society Study (ACS)(Pope 1995), found an association between excess mortality in adults and increasing PM2.5concentrations in various cities and metropolitan areas of the United States (not including thePocatello area). More specifically, the Harvard Six-City Study showed a 31% increase inmortality for every 25 ug/m3 increase in PM2.5, and the ACS study showed a 17% increasein mortality for every 25 ug/m3 increase in PM2.5. The reported ranges of annual averagePM2.5 for the Harvard Six-City Study (HSCS) and the ACS study were 11-30 ug/m3(mean) and 9-34 ug/m3 (median), respectively, for the least to the highest levels of PM2.5 ina given city during the study period. The risks calculated above were based on the excessmortality between the least to the most polluted cities (USEPA 1996).
Given the importance of the HSCS and ACS studies, HEI funded a study to re-analyze theresults of the HSCS and ACS studies. The first major conclusion of the re-analysis studywas that the original results of these two studies was of high quality and that the independentanalysis of the data produced essentially the same results as the original studies. Moreover,the study tested the original results against a range of alternative variables and analyticmodels without substantially altering the original findings of an association betweenindicators of PM air pollution and mortality. In addition, an association between sulfurdioxide and mortality was observed and persisted when other possible confounding variableswere included; furthermore, when sulfur dioxide was included in models with fineparticulates or sulfate, the associations between these pollutants and mortality diminished. The study found relatively robust associations of mortality with fine particles, sulfates, andsulfur dioxide. The final interpretation by the researchers, related to their expanded analysisof the data, suggested that increased risk of mortality may be attributable to more than onecomponent of the complex mix of ambient air pollutants in urban areas of the United States(Krewski, et al. 2000).
These and other chronic exposure studies, taken together, suggest that there may be increasesin mortality in disease categories that are consistent with long-term exposure to airborneparticles and that at least some fraction of these deaths reflect cumulative PM impacts aboveand beyond those exerted by acute exposures events (USEPA 1996). Also important is thefact that the Harvard Six-City Study and the ACS study controlled for subject-specificinformation regarding other relevant risk factors (such as cigarette smoking, occupationalexposure, etc.); thus, these studies appear to provide reliable information about the effects oflong-term exposures to PM (USEPA 1996). Moreover, the findings of an independent re-analysis by the HEI of these studies only serves to strengthen the conclusions of the originalstudy and to show they were sound science. Overall, the weight-of-epidemiological datasuggests long-term, repeated PM exposure has been associated with increased population-based mortality rates as well as increased risk of mortality in broad-based cohorts or samplesof adults and children. Chronic exposures studies of PM suggest rather broad susceptibilityto cumulative effects of long-term repeated exposure. There is no evidence that increasedmortality risk is unique to any well-defined susceptible subgroup (Pope 2000).
Based on the epidemiological evidence, the extensive monitoring data available, and theestimates of historic levels of PM2.5, the community residing in the area of impact (seeFigure 4); that is, in the populated areas northeast of FMC and Simplot (i.e., between thePocatello Sewage Treatment Plant and Chubbuck School monitoring stations), may haveexperienced adverse health effects similar to those reported in the literature from chronicexposures to PM2.5 during several years between 1975 and 1993. Chronic exposures andthe resulting increased risk of adverse health effects to those residing in Pocatello during thissame time frame are also elevated but are likely to be less than those experienced by personsliving in areas between Chubbuck and the Pocatello Sewage Treatment Plant. As previouslyindicated, the numerous studies on PM suggest that the elderly, individuals with pre-existingheart or lung disease, children (not included in Harvard Six-City Study or ACS Study), andasthmatics are the most at risk for adverse health effects from chronic exposure to PM2.5.
The epidemiological evidence, results of monitoring data from the EMF study area from1994 to present (annual average PM10), and subsequent estimates of PM2.5 levels, indicatethat exposure to PM during this time frame within the area of impact were likely to result inonly minimal risks for adverse health effects for the general public and for probably manysensitive subpopulations. However, as previously indicated, there is no clear threshold levelfor PM. Therefore, some hypersensitive segments of the subpopulations residing in the EMFstudy area may have experienced adverse health effects from their long-term PM exposureduring the 1994 to present time frame.
Persons living on the Fort Hall Indian Reservation, especially areas of the reservation nearestto the FMC and Simplot facilities, have likely been and are still being exposed to annualaverage levels of PM2.5 and PM10 above levels of health concern; however, the actuallevels and areal extent of this exposure cannot be determined because of the lack ofmonitoring data north of the facilities and north of Interstate 86.
Acute Exposures to 24-Hour Average PM2.5 and PM10 Levels. Early indications thatfine particles are likely important contributors to observed PM-mortality and morbidity(disease) effects came from evaluations of past serious air pollution episodes in Britain andthe United States. The more severe episodes were characterized by several days of calmwinds, during which large coarse particles rapidly settled out of the atmosphere andconcentrations of fine mode particles dramatically increased (USEPA 1996). Thesemeteorological conditions have been reported on numerous occasions in the EMF study areasince 1975, the most recent being a severe 6-day inversion at the end of December 1999.
Most of the epidemiological studies of PM to date have focused on acute exposures (usuallydaily) and their association with various health end points; such as, mortality counts,hospitalizations, symptoms, and lung function. Unfortunately, until recently (following thepromulgation of the new proposed PM2.5 standards), there have been very little dailymonitoring of fine particles, and most of the studies used other methods of measuringparticulate concentrations (Pope 2000). The table on the following page provides asummary of the epidemiological evidence of health effects of acute exposure to PM (Pope2000).
Summary of Epidemiological Evidence of Health Effects of Acute Exposure to PM Air Pollutants (Adapted from Pope 2000) Health End Points Observed Association with PM Episodes of death and hospitalizations Elevated respiratory and cardiovascular mortality and hospitalizations. Mortality (death) Elevated daily respiratory and cardiovascular mortality counts. Effects persisted with various approaches to control for time trends, seasonality, and weather. Near-linear associations with little evidence of threshold. Hospitalization and other health-care visits Elevated hospitalizations, emergency room visits, and clinic/outpatient visits for respiratory and cardiovascular disease. Effects generally persisted with various approaches to control for time trends, seasonality, and weather. Symptoms/lung function Increased occurrence of lower respiratory symptoms, cough, and exacerbation of asthma. Only relatively weak associations with respiratory symptoms. Small, often significant declines in lung function.
The results of a major study in the United States that evaluated the association of short-termexposures to PM10 and other pollutants, as related to mortality and morbidity (as measuredby hospitalizations), was released in 2000 (Samet, et al. 2000). HEI's National Morbidity,Mortality, and Air Pollution Study (NMMAPS) used several new and innovative approachesto overcome some of the limitations of previous studies of daily exposures to air pollutantsand its relationship to death and hospitalizations. The approach used was to characterize theeffects of PM10 alone or in combination with gaseous air pollutants in a consistent way, in alarge number of cities, using the same statistical approach. The study looked at the effectsof PM10 and other pollutants on mortality in the 20 and 90 largest U.S. cities. In addition,the study looked at morbidity, as measured by daily PM10 effects on hospitalization amongthose 65 years of age and older, in 14 U.S. cities. The HEI concluded that the study hasmade substantial contribution in addressing major limitations of previous studies. Theresults of the 20 and 90 city mortality studies were generally consistent with an averageapproximate 0.5% increase in overall mortality for every 10 ug/m3 increase in PM10measured the day before death. This effect was slightly higher for deaths due to heart andlung disease than for total deaths. The PM10 effect on mortality also did not appear to beaffected by other pollutants in the model. The 14-city hospital admission study of persons65 years or older indicated that there was a consistent approximate 1% increase inadmissions for cardiovascular diseases and about a 2% increase in admissions forpneumonia and COPD for each 10 ug/m3 increase in PM10 (Samet, et al. 2000).
The results of these epidemiological studies suggest that the maximum 24-hour levels ofPM10 and PM2.5 in the EMF study area between 1975 and the present (see Table A-1)have exceeded concentrations, on numerous occasions, that are associated with adversehealth effects. The monitoring data and estimates suggest that the highest levels weredetected either near the FMC and Simplot facilities or in the City of Pocatello. These dataindicate that the population of Pocatello, because of the meteorological conditions that trappollutants in the Portneuf Valley during inversion conditions, was at a higher risk of adversehealth effects from acute levels of PM10 and PM2.5 than was the population of Chubbuck. However, this did not hold true during the December 1999 inversion, when the maximumPM2.5 levels for the same day (12/29/99), detected in Pocatello (119 ug/m3 at Garrett andGould) and in Chubbuck (110 ug/m3 at Chubbuck School) were not considerably different. The risks of combined chronic and acute adverse health effects for other years, duringthe1975 to present time frame, for persons residing in Chubbuck and between the PocatelloSewage Treatment Plant and Chubbuck would not be considered minimal.
According to the epidemiological literature, some of the adverse health effects associatedwith the range of maximum 24-hour levels of PM10 and 2.5 in the EMF study area,including the levels detected during the December 1999 inversion, are increased total acutemortality, increased hospital admissions for the elderly (>65 years) for lung and heartdisease, chronic obstructive pulmonary disease (COPD), pneumonia, ischemic heart disease,and increased respiratory symptoms (i.e., increased cough and decreased lung function)(USEPA 1996). Overall, the PM risk estimates from total mortality epidemiological studiessuggest that an increase of 10 ug/m3 in the 24-hour average PM10 level (or an increase of 5-6 ug/m3 in PM2.5) is associated with increased risks of adverse health effects of 0.5-1.5%(Pope 2000), with even higher risks possible for elderly sub-populations and for those withpre-existing respiratory conditions (USEPA 1996). Moreover, the levels of PM 2.5 detectedin the Chubbuck and Pocatello areas, during the December 1999 inversion, were about 2 to3 times higher than the AQI set by EPA (see previous discussion on the meaning of theAQI).
Persons living on the Fort Hall Indian Reservation, especially areas of the reservation nearestto the FMC and Simplot facilities, may have been and may still be exposed to maximum24-hour levels of PM10 and PM2.5 above levels of health concern; however, the actuallevels and areal extent of this exposure cannot be determined because of the lack ofmonitoring data north of the facilities (north of Interstate 86).
Sulfate Exposures. Some chronic epidemiological studies have shown that the annual mean levelsof sulfate (SO4-2), a subset of fine PM, to be associated with increased mortality in adults, increasedbronchitis in children, and decreased lung function in children (USEPA 1996). The two mainstudies (the Harvard Six-City Study and the ACS study) indicated that every 15 ug/m3 increase inannual average sulfate concentrations was associated with increases of 46 and 10%, respectively, inadult mortality (USEPA 1996). As previously indicated, annual average concentrations for sulfateion in the EMF study area are not available for comparison to the levels found in epidemiologicstudies associated with chronic adverse health effects.
Acute epidemiologic studies have associated sulfate exposures with increased hospitalizations and increased respiratory symptoms. The range of sulfate concentrations for these studies was 2-49 ug/m3. The five highest 24-hour sulfate ion concentrations detected at the IDEQ monitoring stations ranged from 18-73 ug/m3 for the STP monitor, 13-32 ug/m3 for the Chubbuck School monitor, 25-67 ug/m3 for the Garret and Gould monitor, and 26-84 ug/m3 for the ISU monitor. Based on these data and the results of the three epidemiological studies found in the literature, it can be reasonably assumed that persons, especially certain sensitive sub-populations residing in parts of Chubbuck and Pocatello, may have experienced an increased risk of adverse health effects during some of these days.
Acid Aerosol Exposures (including ionic species other than sulfates). Studies of past episodesof air pollution suggest that both acute and chronic health effects are associated with inhalationexposures to strongly acidic PM. For example, studies of historical pollution episodes, notably theLondon Fog episodes of the 1950's and early 1960's, indicate that acute exposures to extremelyelevated levels of acid aerosols may be associated with excess human mortality. Studies evaluatingpresent-day U.S. levels of acid aerosols have not found associations between acid aerosols and acuteand chronic mortality, but the series of hydrogen ion (H+) data used may not have spanned a longenough time frame to detect H+ associations. However, several morbidity studies have associated H+concentrations with increased bronchitis and reduced lung function in children and an increase inrespiratory hospital admissions (USEPA 1996). Furthermore, based on animal studies, it is knownthat sulfuric acid aerosols exert their action throughout the respiratory tract, with the site ofdeposition dependent upon the particle size and the response dependent on mass and numberconcentration of specific deposition sites (USEPA 1996). However, the animal studies on acidaerosols provide no evidence that ambient acidic PM components contribute to mortality andessentially no quantitative guidance as to ambient acidic PM levels at which mortality would beexpected to occur in either healthy or diseased humans. Furthermore, the effects seen in theseanimal studies were at acid levels that exceed worst-case ambient concentrations by more than anorder of magnitude (USEPA 1996).
Several acids, such as, sulfuric acid, phosphoric acid, and hydrofluoric acid, are know to be releasedfrom the phosphate plants. In addition, phosphorous pentoxide (a signature constituent of the FMCemissions) and sulfur dioxide can be transformed in the atmosphere into phosphoric acid andsulfuric acid, respectively. All of these acids are considered potential respiratory irritants. Theconcentrations of ammonium ion present in filter samples is indicative of the elevated levels ofammonia being released in the EMF study area. It is possible, under certain conditions, that thelevels of ammonia will neutralize all or some of the acids present in the ambient air thusameliorating their potential respiratory effects. Because hydrogen ion data are quite limited in theEMF study area, a more definitive conclusion regarding the acidic nature of the ambient air in theEMF study area and resulting health implications cannot be made.
The presence of other ionic species, such as chloride and potassium ions, detected in the filtersamples may be indicative of other acidic, basic, or other species (salts) that were present in theambient air. Since the concentrations of these ions present in the EMF study area are relativelysmall, however, it cannot be determined from the available data if they contribute more or less to theoverall acidity of the ambient air or are part of metallic or other salts that may have more importanttoxicological implications.
Exposures to Metals and Inorganics. The chemical analyses of filter samples performed duringthe RI, by the IDEQ, and by the Sho-Ban Tribe, present results for the elemental forms of metals andother inorganics. Therefore, the public health implications of exposure to the metals and otherinorganics detected must be made on this basis. However, it is likely that the elements detected andpresented in Table 3 were part of various compounds (either salts or covalently bound organicspecies of metals) which may be more or less toxic than the elemental species. However, it isimportant to note that scientific evidence indicates that different metallic salts show similar toxicity,whereas, more differences are found between elemental species with different valence states ormetals covalently bonded to organic species. In some cases, the public health implications for theseelements cannot be determined due to the paucity of studies for the elemental species. For example,the elements calcium, magnesium, and sodium were detected from filter samples; however, theywere likely in the ambient air in the form of various salts formed with other elements. The publichealth implications of these metallic compounds cannot be determined, since the true forms of themetals in ambient air are not known. In some cases, the toxicity of the metallic compounds inambient air may be greater (or less) than the elemental metal detected on a filter sample. Therefore,the toxicological evaluation of the individual elements below may overstate or understate thetoxicological significance of exposure to metallic compounds in the ambient air. Acceptableanalytical methods for determining the concentrations of metallic compounds in air have not beendeveloped.
The public health implications of silicon, bromine, carbon, and chloride ion cannot be determinedbecause they usually form other compounds of varying toxicological properties. For example,silicon in its crystalline forms has different toxicological significance than silicon in its amorphousform. The carbon fraction of ambient particulate matter consists of both elemental and organiccarbon. Elemental carbon, also know as carbon black or graphitic carbon, has a chemical structuresimilar to impure graphite and is emitted directly into the atmosphere predominantly duringcombustion. Organic carbon is either emitted directly by sources or can be formed in theatmosphere by chemical reactions of hydrocarbons. Soot is commonly represented as elementalcarbon, black carbon, or light absorbing carbon measured by thermal/optical or optical absorptiontechniques; however, soot has no firmly established definition (USEPA 1996).
The following discussion evaluates the public health implications of exposure to the eight metals that were detected above health-based comparison values: aluminum, arsenic, barium, beryllium, cadmium, chromium, manganese, and vanadium. As indicated above, only the public health implications of the elemental forms of these metals can be evaluated; these elemental forms are different from the species that may have been present in the ambient air. Furthermore, as previously indicated, the calculation of average annual metals concentrations and the reporting of 24-hour maximum levels were possible from the RI and Sho-Ban data. However, for the IDEQ data, only the maximum 24-hour levels were reported.
Aluminum. Elemental aluminum has not been classified as to its carcinogenicity. Theaverage concentrations of aluminum detected at the RI and Sho-Ban monitors were all belowlevels of public health concern. However, the maximum level of aluminum detected at theSho-Ban monitors (5.55 ug/m3) was above the chronic health comparison value (3.7 ug/m3)for non-carcinogenic health effects. The maximum level is more appropriately compared tolevels in the literature that have caused adverse health effects because of short-term or acuteexposures. The maximum levels of aluminum detected were compared to animal andhuman studies in the literature. Based on this evaluation, the levels detected in the EMFstudy area were about 540 and 1,260 times lower than the no-observed-adverse-effect level(NOAEL) and lowest-observed-adverse-effect level from animal studies (ATSDR 1999a);therefore, adverse health effects from short-term exposure to aluminum is not likely based onthe available data. The maximum concentrations of aluminum detected at monitors locatedin residential areas were below health-comparison values.
Arsenic. EPA has classified arsenic as a human carcinogen via the inhalation route. Basedon the highest average concentration of arsenic detected during the RI, exposure to arsenicwould result in a no apparent increase risk of cancer. The maximum 24-hour level detectedwas compared to studies in the literature that investigated the non-carcinogenic effects ofexposure to arsenic in animals and humans. Based on this comparison, the levels of arsenicin air were about 18,000 and 40,000 times lower than the NOAEL and the lowest-observed-adverse-effect level (LOAEL), respectively (ATSDR 2000a). Based on this analysis, it isunlikely that adverse health effects would result from short-term exposure to the levelsdetected in the EMF study area.
Barium. No studies were found in the literature regarding carcinogenic effects in humans oranimals after inhalation exposure to barium (ATSDR 1992a). The average concentrations ofbarium detected during the RI were well below the chronic health comparison value for allmonitoring stations. However, the maximum level detected for the IDEQ analysis ofselected filter samples was slightly above the chronic health comparison value of 0.51 ug/m3for non-carcinogenic health effects. Although there are not many studies in the literature forinhalation effects after exposure to barium, maximum levels of barium detected in the EMFstudy area were well below levels likely to result in adverse health effects from short-termexposures (ATSDR 1992a).
Beryllium. Beryllium is classified by EPA as a probable human carcinogen via theinhalation route. All of the average concentrations of beryllium detected during the RI werebelow the health-based comparison value for carcinogenic health effects. The maximumlevel of beryllium detected during the RI was at least 400,000 times lower than the lowestacute LOAEL for respiratory and other effects in animals (ATSDR 2000b). Therefore,adverse health effects from short-term exposure to the levels of beryllium detected in theEMF study area are not likely to occur.
Cadmium. EPA has classified cadmium as a probable human carcinogen via the inhalationroute. Based on the highest average concentration of cadmium detected from samples takenduring the RI and for the Sho-Ban monitoring, chronic exposure to cadmium would result inno apparent increased risk of cancer. The maximum level of cadmium detected during theRI, for the Sho-Ban monitoring, or during IDEQ's selective filter sampling, were evaluatedto determine potential non-carcinogenic health effects from acute exposures to cadmium. Based on this evaluation, the maximum levels of cadmium found in residential areas of theEMF study were at least 3,900 and 6,700 times lower than the lowest NOAEL and LOAEL,respectively, for less serious health effects found in animal studies (ATSDR 1999b). Fornon-residential areas (near the FMC facility), the maximum levels of cadmium were at least400 and 690 times lower than the lowest NOAEL and LOAEL, respectively, for less serioushealth effects found in animal studies (ATSDR 1999b). Moreover, for these same non-residential areas, the maximum levels of cadmium were at least 1,600 and 16,300 timeslower than the lowest NOAEL and LOAEL, respectively, for serious respiratory effectsfound in animal studies (ATSDR 1999b). Based on this analysis alone, exposure tocadmium detected in the EMF study area is not likely to result in adverse health effects. However, there are some uncertainties with this evaluation related to cadmium and othermetals. Please see the summary of the health effects of exposure to metals below for moredetails of these uncertainties.
Chromium. EPA considers hexavalent chromium to be a human carcinogen via theinhalation route; whereas, trivalent chromium has not been shown to be a carcinogen. Sincethe results from the RI are reported as total chromium, the concentrations of hexavalentchromium and trivalent chromium in the EMF study area are not known. Clearly, however,the relative quantity of hexavalent chromium cannot exceed the total chromium levels. Therefore, as a worst-case scenario of exposure, this analysis assumes that all of the totalchromium reported is hexavalent chromium--a highly conservative assumption.
The resulting evaluation of the levels of chromium detected in residential areas (monitoringstations # 3 and #4) for their carcinogenic health effects, indicate a no apparent increasedrisk of cancer. In addition, if the highest average level of total chromium detected in non-residential area (Sho-Ban monitors next to FMC) were evaluated for its carcinogenic healthrisks, the resulting analysis would indicate a low risk of cancer. However, it is likely thatthe actual risks are lower because all of the chromium is probably not predominantly in thehexavalent form.
For acute non-carcinogenic health effects, the maximum total chromium concentrationdetected in residential areas would be about 57 times lower that the lowest LOAEL for lessserious respiratory effects in studies of humans exposed to hexavalent chromium (ATSDR2000c). However, when compared to studies of animals exposed to the less toxic trivalentchromium, the maximum exposure levels in residential areas is about 25,000 times lowerthan the lowest LOAEL for less serious respiratory health effects (ATSDR 2000c). Themaximum total chromium concentration detected in non-residential areas of the EMF studyarea was from the Sho-Ban monitors. This level is about 10 times lower than the lowestLOAEL for less serious respiratory effects in humans exposed to hexavalent chromium(ATSDR 2000c). However, when compared to studies of animals exposed to the less toxictrivalent chromium, the maximum exposure levels in non-residential areas is about 4,500times lower than the lowest LOAEL for less serious respiratory health effects (ATSDR2000c).
For chronic non-carcinogenic health effects, the average concentration of total chromiumdetected in residential areas would be about 90 times lower than the lowest LOAEL for lessserious respiratory effects in humans exposed to hexavalent chromium (ATSDR 2000c). However, when compared to studies of humans exposed to the less toxic trivalent chromium,the maximum exposure levels in residential areas is about 3,300 lower than the lowestNOAEL for renal effects and about 90,000 times lower than the lowest LOAEL for lessserious respiratory health effects (ATSDR 2000c). The maximum total chromiumconcentration detected in non-residential areas of the EMF study area was from a samplefrom an RI monitor near the FMC and Simplot facilities. This level is about 115 timeslower than the lowest LOAEL for less serious respiratory effects in humans exposed tohexavalent chromium (ATSDR 2000c). However, when compared to studies of humansexposed to the less toxic trivalent chromium, the maximum exposure levels in non-residential areas is about 4,300 times lower than the lowest NOAEL for renal effects andabout 114,000 time lower than the lowest LOAEL for less serious respiratory health effects(ATSDR 2000c).
The actual hexavalent chromium levels in ambient air in the EMF study area areundoubtedly much lower than the total chromium levels used in the above evaluation. Inthis analysis, the actual estimates of health risk are likely closer to the estimates for studiesin which humans and animals were exposed to the less toxic trivalent chromium. Therefore,persons living in populated and non-populated areas of the EMF study are not likely toexperience adverse non-carcinogenic health effects from their short- or long-term exposuresto chromium.
Manganese. No studies were found in the literature regarding carcinogenic effects inhumans or animals after inhalation exposure to manganese (ATSDR 2000d). For non-carcinogenic health effects, the maximum level detected in the EMF study area (at theSewage Treatment Plant) was compared to animal and human studies in the literature. Based on this evaluation, the maximum level detected in the EMF study area were about11,600 times lower than the NOAEL for short-term adverse respiratory health effects foundin animal studies (ATSDR 2000). Based on this evaluation, the levels of manganesedetected in the EMF study are not likely to result in adverse health effects.
Vanadium. No studies were found in the literature regarding carcinogenic or chronic non-carcinogenic effects in humans or animals after inhalation exposure to vanadium (ATSDR1992b). For short-term non-carcinogenic health effects, the maximum levels detected in theEMF study area were compared to animal and human studies in the literature. Based on thethis evaluation, the maximum vanadium levels were about 75 times lower than the LOAELfor less serious respiratory effects in humans (i.e., bronchial irritation) (ATSDR 1992b). However, the maximum concentration detected was at the monitoring station located nearthe site perimeter and not in residential areas. Moreover, recent sampling at the siteperimeter did not indicate that the levels of vanadium were above acute health-comparisonvalues. The maximum levels detected in residential areas were below health comparisonvalues. Based on this evaluation, it is unlikely that exposures to vanadium in populatedareas of the EMF study would result in acute adverse health effects.
Summary of Metals Exposures. Although the above evaluation did not indicate a public health concern for individual metals, there is some uncertainty with this analysis. Current science provides little evidence as to whether the mix of these air contaminants may increase or decrease their toxicological effects because of cumulative exposures. Some of the metals (e.g., cadmium) were detected at levels in the fine fraction that were similar or greater than levels found in highly urbanized areas of the United States (ATSDR 1999). In addition, many of the metals detected in the EMF study area are transition metals.As indicated above, there is growing biological evidence that indicates that urban combustion particles (i.e., fine PM) can penetrate past the primary defense mechanisms of the lung, can elicit inflammatory changes in the lung and systematically (throughout the body), contain a constituent (soluble transition metals) that by itself can be demonstrated to produce lung damage, can produce electrocardiogram changes including arrhythmia (heart irregularities), and can kill animals with pre-existing heart and lung disease (Schwartz 1999). The extent to which the above evaluation of exposures to metals in the EMF study area is able to capture these concerns is not known. However, the epidemiological evidence (presented above) does indicate that PM, a measure of a mix of contaminants present in air, including all the metals detected in the EMF study area, is a good surrogate measure for estimating the short-term and long-term adverse cardiopulmonary health effects from exposure. From this standpoint, ATSDR evaluated and made definitive public health statements regarding the cumulative health effects of the exposure to the mix of metal contaminants present in the EMF study area as measured by PM.
Sulfur Dioxide Exposures. As previously indicated, annual average concentrations of sulfurdioxide at the Pocatello Sewage Treatment Plant have been below EPA's annual health-basedstandard since this monitoring station's inception. However, some 24-hour measurements of sulfurdioxide have exceeded EPA's health-based standard. In addition, the levels of sulfur dioxidedetected at the STP during the period 1977-1985 exceeded ATSDR's Minimal Risk Level (MRL)of 0.01 ppm at least once a year during that period. Moreover, the maximum levels detected forthese years indicate that levels of sulfur dioxide were 17-24 times higher than the MRL. Furthermore, ATSDR considers a concentration of sulfur dioxide of 0.1 ppm to be a minimalLOAEL (ATSDR 1998d). Available human controlled exposure studies indicate that sensitiveasthmatics may respond to concentrations of sulfur dioxide as low as 0.1 ppm. Healthy non-asthmatics respond to higher concentrations of sulfur dioxide (greater than or equal to 1.0 ppm). Factors that have been shown to exacerbate the respiratory effects of sulfur dioxide include exerciseand breathing of dry or cold air. Animal data support the human data on respiratory effects of sulfurdioxide (ATSDR 1998d).
As previously indicated, the only potentially unhealthy levels of sulfur dioxide measured in the EMFstudy area were detected at the Pocatello Sewage Treatment Plant during the years 1977 to 1985. Sulfur dioxide levels at this location did not exceed health-based comparison values from 1986 tothe present, neither did sulfur dioxide levels at Garret and Gould between 1994 and 1999. Based onthe available data, ATSDR suspects that the higher levels of sulfur dioxide from 1977 to 1985 wereconfined to areas in the immediate vicinity of the Pocatello Sewage Treatment Plant; however,ATSDR cannot rule out the possibility that certain sensitive individuals (i.e., asthmatics) were notexposed to sulfur dioxide at levels of health concern some time during this period. For theseindividuals, exposure to elevated levels of sulfur dioxide, along with elevated PM exposures, couldincrease the risk for adverse respiratory health effects. Since 1985, the levels of sulfur dioxidedetected at the STP have been below levels of public health concern.
Potential Exposure to Phosphine and Hydrogen Cyanide from FMC. Phosphine, a colorlessgas with a characteristic fish- or garlic-like odor, is a severe respiratory irritant. Gastrointestinal,respiratory, and central nervous system (CNS) effects have been noted in workers exposed to meanconcentrations less than 10 ppm (Jones 1964). EPA has insufficient information to classifyphosphine as to its potential as a human carcinogen (USEPA 1999b). NIOSH has a recommendedexposure limit (REL) for phosphine of 0.3 ppm (300 ppb) and a short-term exposure limit (STEL)of 1 ppm (1,000 ppb) (NIOSH 1994). The RELs are time-weighted average (TWA) concentrationsfor up to a 10-hour workday during a 40-hour workweek, and the STEL is a 15-minute TWAexposure that should not be exceeded anytime during the workday (NIOSH 1994). As previouslynoted, FMC has measured some phosphine concentrations at the ponds at levels above the STEL. However, the public health implications of these environmental levels in relation to the on-siteworkers is beyond the scope of this health consultation. Using OSHA-approved methods, themaximum level of phosphine detected at the fence line was 101 ppb--an average of the fence lineconcentrations was not available. Based on limited animal studies reported by EPA (USEPA1999b), short-term exposures (less than one year) to phosphine at the maximum levels detected atthe fence line are not likely to result in adverse respiratory health effects. The effects of chronicexposures (greater than one year) to phosphine are still unknown (USEPA 1999b). However,additional sampling for phosphine at the fence line using other, less reliable, methods have onseveral occasions indicated that phosphine levels may have exceeded the STEL. These measuredconcentrations, if correct, suggest that a passerby, offsite worker (not FMC or Simplot), or otherindividual in the area might suffer from adverse health effect if exposed to the peak levels ofphosphine for as little as 15 minutes.
Based on available data and knowledge of site-conditions, current exposures to the non-workerpublic would probably only be on an infrequent basis and for only a short duration. Therefore,based on limited environmental and scientific data alone, the occasional visitor to the area aroundthe FMC site would not experience any adverse respiratory health effects from exposure tophosphine at 101 ppb. However, fence line and possibly off-site concentrations of phosphine mayhave been higher in the past and may have reached levels of public health concern (i.e., above theSTEL) in the recent past, but the methods used may be unreliable. Therefore, the complete publichealth implication of off-site exposures to phosphine cannot be determined based on available data. Because of the toxicity of phosphine, continued operation of FMC's Pond Management Plan isneeded to ensure that emissions do not reach levels of health concern to the off-site non-workerpublic. Moreover, more monitoring at the fence line, using OSHA-approved methods, is needed.
The maximum concentration of hydrogen cyanide (HCN) detected at the ponds was 990 ppb or0.990 ppm. This level is almost five times lower than NIOSH's STEL (4.7 ppm)--NIOSH has notestablished a TLV-TWA guidance for HCN (ATSDR 1997b). The concentration of HCN at thefence line was compared to the lowest LOAELs reported in ATSDR's toxicological profile (ATSDR1997e). The maximum HCN concentration at the perimeter is about 15, 100, and 140 time belowthe lowest chronic, intermediate, and acute LOAEL, respectively. Therefore, based on the currentsite conditions, where it is likely that current exposures to the non-worker public would be on aninfrequent bases and for only a short time, it is not likely that adverse respiratory health effectswould occur from exposure to the maximum HCN level detected at the fence line. However, fenceline and possibly off-site concentrations of HCN may have been higher in the past. Therefore, thecomplete public health implication of off-site exposures to HCN cannot be determined based onavailable data. Because of the toxicity of HCN (albeit not as toxic as phosphine), continuedoperation of FMC's Pond Management Plan is needed to ensure that emissions do not reach levels of health concern to the off-site non-worker public.
1. ATSDR acknowledges that the statistically significant downward trend in PM10 concentrations might simply result from changes in meteorology, or even by chance. However, the fact that annual average PM10 concentrations over the last 5 years have remained lower than their pre-1994 levels suggests that the downward trend is not spurious. Ongoing review of air monitoring data from IDEQ's network can help confirm this hypothesis.
2. In this section, "total phosphorous" refers to the concentration of phosphorous measured by the x-ray fluorescence analytical method, which essentially measures all forms of phosphorous collected on particulate filters. Thus, "total phosphorous" includes phosphorous pentoxide, phosphoric acid, and other forms of the metal.
3. Average concentrations cited for the Fort Hall Source Apportionment Study are based on data from dichotomous samples. It is assumed that the sum of the average metal concentration in the fine fraction and the average metal concentration in the coarse fraction is equal to the average metal concentration in PM10.
4. It should be noted that EPA also has a 3-hour average air quality standard, but this standard is not based on adverse health effects. Ambient air concentrations of sulfur dioxide at the Pocatello Sewage Treatment Plant exceeded this 3-hour average standard only once in 1977, once in 1980, and twice in 1985.
5. A legal debate still surrounds EPA's promulgation of the PM2.5 standard. Regardless of the legal status of the standard, the authors of this report believe the epidemiological evidence considered in developing the standardis compelling and therefore use this epidemiological evidence to assess public health implications associated with PM2.5 exposures in the EMF study area.