PUBLIC HEALTH ASSESSMENT
PADUCAH GASEOUS DIFFUSION PLANT (U.S. DOE)
PADUCAH, MCCRACKEN COUNTY, KENTUCKY
A release of a hazardous substance does not always result in human exposure, and human exposure does not always result in adverse health effects.
This section of the public health assessment evaluates the estimated exposure doses for contaminants of concern for completed and potential exposure pathways for potentially affected populations. In these evaluations, ATSDR considered the frequency and duration of the estimated exposures; for cases in which a population is affected by more than one exposure pathway, we also considered the combinations of contaminants and exposure routes. This section also presents the potential health effects from each contaminant of concern in a completed exposure pathway.
We considered characteristics of the exposed populations--such as age, sex, nutritional status, genetics, lifestyle, and health status--that influence how a person absorbs, distributes, metabolizes, and excretes contaminants; and, where appropriate, these characteristics are included in the contaminant-specific discussions.
Special Considerations of Women and Children
Women and children can sometimes be affected differently from the general population by contaminants in the environment. Both tend to be smaller than the average person, which means they can be affected by smaller quantities of contaminants. The effect of hormonal variations, pregnancy, and lactation can change the way a woman's body responds to some substances. Past exposures experienced by its mother, as well as exposure during pregnancy and lactation, can expose a fetus or infant to chemicals through the placenta or in the mother's milk. Depending on the stage of pregnancy, the nature of the chemical involved, and the dose of that chemical, fetal exposure can result in problems like miscarriage, stillbirth, and birth defects.
ATSDR's Child Health Initiative recognizes that developing young people, whether fetuses, infants, or children, have unique vulnerabilities. Children are not small adults; a child's exposure can differ from an adult's exposure in many ways. A child drinks more fluids, eats more food, and breathes more air per kilogram of body weight than an adult, and has a larger skin surface area in proportion to body volume. A child's behavior and lifestyle also influence exposure. Children crawl on floors, put things in their mouths, play close to the ground, and spend more time outdoors. These behaviors may result in longer exposure durations and higher intake rates.
Children's metabolic pathways, especially in the first months after birth, are less developed than those of adults. In some cases, children are better able than adults to deal with environmental toxins, but in others, they are less able and more vulnerable. Some chemicals that are not toxins for adults are highly toxic to infants.
Children grow and develop rapidly in the first months and years of life. Some organ systems, especially the nervous and respiratory systems, can experience permanent damage if exposed to high concentrations of certain contaminants during this period. Also, young children have less ability to avoid hazards, because they lack knowledge and depend on adults for decisions that may affect children but not adults.
This public health assessment assesses risks to children exhibiting pica behavior (a craving for unnatural food like soil). Information on the incidence of soil pica behavior is limited. A study described in an EPA document [117] showed that the incidence of soil pica behavior was approximately 16% among children from a rural black community in Mississippi. However, this behavior was described as a cultural practice among the community surveyed, so that community may not represent the general population. In five other studies, only one child out of more than 600 ingested an amount of soil significantly greater than the range in other children. Although these studies did not include data for all populations and represented short-term ingestion only, it can be assumed that the incidence rate of soil pica behavior in the general population is low.
There is little information on the amount of soil ingested (measured in milligrams per day, or mg/day) by children with soil pica behavior [117]. Ingestion rates between 1,000 and 10,000 mg/day have been used to estimate exposure doses for pica children. In the PGDP public health assessment, ATSDR assumed a soil ingestion rate of 2,000 mg/day for approximately 290 days per year to represent pica behavior in children aged 1 to 3 years old. ATSDR believes that this is a health protective assumption and likely overestimated soil consumption.
In the following discussions, we will indicate whether women and children were, are, or may be exposed to contaminants of concern and discuss the possible health concerns related to these exposures.
Identifying Potentially Affected Groups
Table 24 summarizes the completed and potential exposure pathways. This table presents the exposure pathways, exposure routes, affected population, and duration of exposure for each contaminant in a potential or completed exposure pathway. Contaminants that are only present in potential exposure pathways are in italics. Note that exposure durations for metals in the groundwater exposure pathway are assumed to be chronic (i.e., lasting 1 year or more): it is difficult to identify the specific numbers of years of exposure for the metals, because there have not been sufficient metals analyses in most residential wells to determine long-term trends in concentration. Additionally, the metals have different rates of groundwater transport relative to trichloroethylene (TCE) and other volatile organic compounds.
Populations that may be exposed to specific contaminants via multiple exposure pathways must have their pathway-specific exposure doses summed to represent a total dose. However, most of the contaminants listed in Table 24 are not present in multiple exposure pathways. Of the 17 contaminants listed, only arsenic, radioactive materials, thallium, uranium, and vanadium have multiple pathways of exposure to the same population. The only population that could have been exposed to these contaminants via more than one exposure pathway were pica children living within the groundwater plume areas before 1988. Less than 1% of children exhibit pica behavior [117], and it is unknown if any pica children were present in those areas. Table 24 lists radioactive materials together, because radiation doses from each isotope were summed to include a total dose to potentially exposed populations. Uranium, as a chemical toxin, is listed separately.
Table 25 gives an estimate of the number of people potentially exposed through each exposure pathway. Figure 7 shows the locations of those potential exposures. The number of persons potentially exposed was determined using 1990 Census data and the exposure areas from Figure 7. The 1990 Census information is appropriate to use since 1990 is close to the time when people stopped using contaminated well water. Comparing 1990 Census data with 1980 Census data, however, shows that the number of people potentially exposed decreased by about 10 between 1980 and 1990. This means that the 1990 Census data may underestimate the number of people potentially exposed. (The people who left the area were most likely less than 65 years old, including a few less than 6 years old.) Also, about 25 of these people have lived in this area since the plant began operation in 1952. (Refer to Appendix A.) Note that Table 25 does not include the surface water and biota exposure pathway: most people potentially exposed through that exposure pathway would be hunters and fishers visiting the Western Kentucky Wildlife Management Area (WKWMA) and would not live near the site. (The census would not include these individuals, so we do not know the number or ages of hunters and fishers.)
It is important to remember that an exposed person would not necessarily experience adverse health effects. Tables 24 and 25 describe the potentially affected populations; they do not describe potential health effects. The discussion of potential health effects for each contaminant are based on calculated exposure doses for PGDP and documented health effects from human and animal studies. Specific contaminants are discussed in this section (Public Health Implications) of this report.
Table 24. Summary of completed and potential
exposure pathways for each contaminant of concern
(Potential contaminants, exposure pathways, and populations are in italics)
| Contaminant | Exposure pathway(s) | Exposure Route(s) | Potentially Affected Population(s) | Duration of Potential Exposure |
| Antimony | Soil | Ingestion and dermal contact | Children with pica behavior1 | Past, present, and future: 1 to 2 years2 |
| Arsenic | Groundwater | Ingestion | Adults and children routinely drinking water from well RW-2943 | Past only: chronic exposure4 |
| Soil | Ingestion and dermal contact | Children with pica behavior | Past, present, and future: 1 to 2 years2 | |
| Cadmium | Groundwater | Ingestion | Adults and children routinely drinking water from northeast and northwest plume areas | Past: unknown exposure4 |
| Chromium (tri- and hexavalent) | Groundwater | Ingestion | Adults and children routinely drinking water from northeast and northwest plume areas | Past: chronic exposure4 |
| Hydrogen Fluoride | Air | Inhalation | ||
| Acute (11/17/60) | Adults and children living < 500 meters (1,640 feet) southeast of PGDP fence | Past and potential future: < 4 hours; accidental releases | ||
| Chronic (1956) | Adults and children living along northern fence boundary | Past; maximum annual releases | ||
| Lead | Groundwater | Ingestion | Adults and children routinely drinking water from wells RW-113 and RW-297 | Past and potential current: chronic exposure4 |
| Manganese | Soil | Ingestion and dermal contact | Children with pica behavior | Past, present, and future: 1 to 2 years2 |
| Nitrate (Nitrite) | Groundwater | Ingestion | Children routinely drinking water from wells RW-002, RW-030, and RW-294 | Past: chronic exposure4 |
| Pentachloro-phenol | Groundwater | Ingestion | Unknown | Unknown |
| Polychlorinated biphenyls (PCBs) | Food (biota) | Ingestion | Children and adults who eat significant quantities of fish caught in Little Bayou Creek | Past, present, and potential future |
| Radioactive Materials5 | Air | Inhalation | Residents living < 500 meters (1,640 feet) north of PGDP fence | Past: 9 years (1954-1963) |
| Air | Inhalation | Residents living less than 4 kilometers (2.5 miles) southeast of PGDP fence | Past: 1960 accident | |
| Surface water | Ingestion | Workers and visitors in WKWMA | Past | |
| Groundwater | Ingestion (Tc99, U-234, U-238) | Adults & children routinely drinking from wells RW-002, RW-017, & RW-113 | Past: 5 to 15 years chronic (1973 - 1988) | |
| Biota | Ingestion | Adults & children who eat 20% of their intake of fish, game, fruits, & vegetables from areas near PGDP | Past, current and potential future | |
| Thallium | Groundwater | Ingestion | Unknown | Past |
| Surface water | Visitors to WKWMA | Past, current, future | ||
| Trichloro-ethylene | Groundwater | Ingestion, inhalation | Adults and children routinely drinking water from wells RW-002, RW-017, and RW-113 | Past: 5 to 15 years chronic exposure (1973-1988) |
| Uranium | Air | Inhalation (Acute) | Residents living less than 4 kilometers (2.5 miles) southeast of PGDP fence | Past: 1960 accident
Potential future |
| Vanadium | Groundwater
Soil |
Ingestion
Ingestion and dermal contact |
Adults and children routinely drinking
water from northeast and northwest plume areas
Children with pica behavior |
Past: chronic exposure4
Past, present, and future: 1 to 2 years2 |
| Vinyl chloride | Groundwater | Ingestion and inhalation | Adults and children routinely drinking water from northeast and northwest plume areas | Past and potential future: unknown duration |
| Zinc | Groundwater | Ingestion | Only children routinely drinking from well RW-113 | Past: chronic exposure4 |
| 1 Less than 1% of children aged
1 to 3 exhibit pica behavior. 2 Pica behavior may last for only 1 to 2 years for each child. 3 "RW-#" indicates a residential well and well number. 4 Chronic exposure is exposure for 1 year or more. There have not been sufficient metals analyses in most residential wells to determine long-term trends in concentration. Lead contamination may come from lead solder in plumbing, not PGDP releases. 5 This category includes uranium 234, 235, and 238; neptunium 237; plutonium 239; thorium 230; and other radioactive substances. |
||||
Table 25. Estimated number of persons potentially
exposed per exposure pathway based on 1990 Census data and potential exposure
pathways
| Population Description | Soil/Sediment Exposure pathway | Air Exposure pathway | Groundwater Exposure pathway |
| Total | 90-100 | 67-74 | 15-17 |
| Children under 6 | 10-14 | 7-9 | 2-3 |
| Women 15 to 44 years | 16-20 | 12-15 | 2-3 |
| People over 65 | 12-14 | 8-9 | 3 |
| Total 18 and older | 67-72 | 49-52 | 11-12 |
| Total under 18 | 23-28 | 18-22 | 4-5 |
| White | 90-100 | 67-74 | 15-17 |
| Black | 0 | 0 | 0 |
| American Indian | 0 | 0 | 0 |
| Asian | 0 | 0 | 0 |
| Hispanic | 0 | 0 | 0 |
| Other | 0 | 0 | 0 |
| Source: [26] | |||
Figure 7. Areas of Contamination and Potential Health Exposure
Potential exposures to antimony in off-site soil are not a public health hazard.
Antimony is a metal that occurs naturally at low levels in the earth's crust. It is used in industry--mixed with other metals to form alloys or produced as antimony oxide. The alloys are used in lead storage batteries, solder, sheet and pipe metal, bearings, castings, ammunition, and pewter. The oxide is added to cloth and plastic to make them more fire-resistant [118].
Off-site soil concentrations of antimony ranged from 1 to 50 milligrams of antimony per kilogram of soil (mg/kg) [?]. Concentrations of antimony were not uniformly distributed throughout off-site areas. Instead, they were log-normally distributed, meaning that a few samples had high concentrations while most had low concentrations. In fact, most off-site soil concentrations were below 5 mg/kg [45]. The highest concentration was found 2.5 miles (4 kilometers) northwest of PGDP, at a location where wells were installed. (That sample may not be representative of surface soil samples, and the higher concentrations may not be a potential source of exposure to humans.) The maximum concentration is well above the reported range of antimony in soil for the eastern United States (less than 1 to 8.8 mg/kg [90]); it is also higher than the background concentration reported for the PGDP area (0.21 mg/kg [119]).
ATSDR scientists used conservative assumptions to estimate exposure doses for exposure to antimony in off-site soil. The highest estimated exposure dose was 0.001 milligrams of antimony per kilogram of body weight per day (mg/kg/day) for a child who exhibits pica behavior (see Table 15A). The absorption and toxicity of antimony depend on the physical and chemical state of the specific compound inhaled or ingested. Both gastrointestinal and pulmonary absorption, although generally low, are a function of compound solubility.
ATSDR has not developed a health guideline for ingestion of antimony, because available scientific studies are lacking for this route of exposure [118]. EPA has developed a health guideline, called a reference dose (RfD), for chronic oral exposure to antimony, which is 0.0004 mg/kg/day. The reference dose is based on a lowest-observed-adverse-effect level (LOAEL) in rats, which had shortened lifespans and changes in blood glucose levels after ingesting 0.35 mg/kg/day of antimony in drinking water [120]. EPA derived the RfD by dividing the LOAEL for rats by an uncertainty factor of 1,000, because humans may be more sensitive than rats, some humans may be more sensitive than others, and there was no experimental level for rats where no adverse effects were seen. Other studies in which rodents were exposed orally have reported effects on lifespan, glucose levels, and cholesterol metabolism [118].
Acute exposure to antimony by humans who ingested antimony-contaminated lemonade (at an estimated dose of 0.5 mg/kg for a 70-kilogram adult who ingested 300 milliliters of lemonade) resulted in burning stomach pains, nausea, and vomiting [118,120]. Most exposed people recovered from this acute exposure within a few hours to several days [118,120]. One review of soil ingestion studies proposed an acute toxicity screening dose of 0.528 mg/kg/day for antimony exposure via soil for young children who exhibit pica behavior [92].
Although ATSDR's estimated exposure doses slightly exceeded EPA's health guideline, the doses were considerably lower than the lowest levels reported to cause adverse health effects in animals and humans [118,120]. They were also lower than the acute toxicity screening level proposed for antimony [92]. Furthermore, we most likely overestimated actual doses, since we used extremely conservative assumptions to estimate dose.
EPA's antimony health guideline is based on a drinking water study in rats. Antimony in soil is generally in a less soluble form than when it is in water. Consequently, people would absorb less antimony from soil than from water. Even with conservative assumptions about exposure and rate of absorption from soil, exposure to antimony in off-site soils near PGDP is not expected to result in adverse health effects.
Exposures to arsenic in groundwater and potential exposures in off-site soil are not a public health hazard. Arsenic was also evaluated in surface water and was not identified as a contaminant of concern for that exposure pathway.
Arsenic is a naturally occurring element in our environment but additional arsenic often gets into the environment during copper and lead smelting, wood treating, and pesticide applications. It is in our environment in both the organic form (combined with carbon and hydrogen) and the inorganic form (combined with other elements, like oxygen, chlorine, or sulfur) [121]. Arsenic was found in two residential wells at a maximum concentration of 10 micrograms per liter of water (or 10 µg/L). These wells were used for an unknown period of time in the past, possibly up to 35 years. ATSDR's estimated doses, which assumes daily chronic exposure, for past groundwater exposure to adults (0.003 mg/kg/day) and children (0.007 mg/kg/day) exceeded health guidelines for arsenic (as shown in Table 6).
Inorganic forms of arsenic predominate in groundwater (and soils) and are generally more toxic than organic forms [122]. When humans and other animals are exposed to inorganic arsenic, their bodies change it to the much less toxic methylated organic form, which is readily excreted from the body. This methylation process is effective as long as the dose of inorganic arsenic remains below 0.2 to 1 mg/day [121]. In other words, people can tolerate a certain level of arsenic without adverse effects. At higher levels, the body's capacity to detoxify arsenic can be exceeded or saturated. When this happens, blood levels increase and adverse effects can occur. ATSDR's estimated doses for groundwater and soil exposure pathways are lower than the levels needed to saturate detoxification mechanisms in the body.
For the purposes of this Public Health Assessment, ATSDR assumed that arsenic in biota existed as 100% inorganic arsenic. However, the predominant (80 to 99% of the total arsenic) forms of arsenic in fish and shellfish are organic arsenicals, which have been found to be essentially nontoxic [121].
Saturation of the body's detoxification mechanism may explain why non-cancer and cancer effects of arsenic appear to have a threshold, or minimum effective dose. In addition, a growing body of scientific evidence suggests that cancer may result from mechanisms other than direct attack on genetic material, which suggests that carcinogenicity from arsenic exposure has a threshold [121].
The lowest doses of arsenic shown to cause human toxicity from chronic ingestion--namely skin and gastrointestinal effects--range from 0.014 to 0.05 mg/kg/day. These doses were estimated from a study of Taiwanese people who drank arsenic-contaminated water for 45 years [123,124]. EPA derived a health guideline of 0.0003 mg/kg/day based on skin effects (e.g., hyperpigmentation and keratosis) and a cancer slope factor of 1.5 (mg/kg/day)-1 for skin cancer based on the Taiwanese study [120].
This study has limitations that one must consider when using it to evaluate public health hazard for PGDP residents. First, it reported an association between arsenic in drinking water and skin cancer, but failed to account for potential confounding factors, including exposure to other non-water sources of arsenic, genetic susceptibility, and poor nutritional status of the exposed population. Therefore, arsenic exposure may have been underestimated in the study, possibly leading to overestimation of the number of new cancer cases predicted for incremental increase in exposure dose. Second, the cancer slope factor for arsenic is based on the conservative assumption that no threshold exists for cancer. As discussed previously, arsenic carcinogenicity appears to have a threshold.
The amount of arsenic absorbed from the gastrointestinal tract or skin can vary widely; it depends largely on the water solubility of the arsenic compounds (either organic or inorganic) present in the environment. It is often assumed that most arsenic in drinking water and soil is inorganic [125,121]. Studies of the bioavailability of arsenic from drinking water indicate that water-soluble forms of inorganic arsenic are almost completely absorbed (e.g., at least 95%) from the gastrointestinal tract, while less-soluble compounds are absorbed to a lesser extent (e.g., up to 30%) [122]. ATSDR scientists do not have specific information about the types of arsenic compounds (and their solubility) present in groundwater and soils off site of PGDP; therefore, we assumed for exposure dose calculations that all arsenic was water-soluble and 100% absorbed.
Despite these conservative assumptions, the estimated groundwater doses were lower than levels shown to cause adverse effects in the Taiwanese study and considerably lower than levels required to saturate detoxification mechanisms in the body.
Arsenic was detected in off-site soil in the WKWMA, southwest of the PGDP security fence. The maximum off-site concentration was 38 milligrams of arsenic per kilogram of soil. The normal range of soil concentrations in the eastern United States is less than 0.1 to 73 mg/kg [121], and background for the Paducah area is reported as 12 mg/kg [119]. ATSDR's estimated dose for past and current exposure to children who exhibit pica behavior (for the resident exposure scenario) was 0.002 milligrams of arsenic per kilogram of body weight per day, which exceeded the health guideline for chronic ingestion of arsenic. However, our estimated dose was lower than the provisional acute toxicity screening dose (0.005 mg/kg/day) for acute effects (e.g., throat irritation, nausea, and vomiting) in young children who exhibit pica behavior [92].
Studies indicate that arsenic in soils is absorbed from the gastrointestinal tract of humans to a limited extent (e.g., less than 50%) following ingestion. This is thought to be primarily because soils contain arsenic in less-soluble forms [122]. More-soluble arsenic compounds may be 60% to 70% absorbed through the gastrointestinal tract [126], but less-soluble forms are absorbed to about half that degree [122]. Dermal absorption of arsenic in soils is minimal compared to ingestion. According to studies of monkeys and humans, arsenic absorption from the skin ranges from 3.2% to 4.5% [122,127,99]. For ingestion of and dermal contact with soil, we made the conservative assumption that 80% of arsenic was absorbed for either route of exposure. This assumption resulted in a dose estimate that most likely overestimated actual doses.
To estimate soil exposure doses, ATSDR scientists used conservative assumptions that would overestimate exposure levels expected at the site. Conservative assumptions were used to be protective and to account for the uncertainty regarding actual exposure levels to off-site populations. Actual levels of exposure would be expected to be lower. Exposure to arsenic in off-site soil near PGDP is not expected to result in adverse human health effects, even to sensitive subpopulations exposed to the maximum soil concentration.
Cadmium was detected in one off-site groundwater well (on Tennessee Valley Authority property). Ingestion of water from this well is unlikely, because the well is a monitoring well on industrial property. The analytical results for cadmium in residential wells were reported as non-detects, but the detection limits were above ATSDR comparison values. However, exposures estimated using the detection limits do not pose a public health hazard.
Cadmium is an element that occurs naturally in the earth's crust. All soils and rocks, including coal and mineral fertilizers, contain some cadmium. Pure cadmium is a soft, silver-white metal. It is often found as part of small particles in air. It does not have a distinct taste or smell; therefore, it is not possible to taste or smell cadmium in water or air. In the United States most cadmium is extracted during the production of other metals such as zinc, lead, and copper. It has many uses in industry and consumer products, mainly batteries, pigments, metal coatings, and plastics.
Food and cigarette smoke are the largest potential sources of cadmium exposure for members of the general population. Average cadmium levels in U.S. foods range from 2 to 40 parts of cadmium per billion parts of food (ppb). Average cadmium levels in cigarettes range from 1,000 to 3,000 ppb. The level of cadmium in most drinking water supplies is less than 1 ppb. The current average dietary intake of cadmium in adult Americans is about 0.0004 mg/kg/day; smokers receive an additional amount--about 0.0004 mg/kg/day--from cigarettes [128].
Numerous studies indicate that the kidney is the main target organ of cadmium toxicity following extended oral exposure to cadmium, with effects similar to those seen following inhalation exposure [128]. Elevated incidences of kidney effects (tubular proteinuria) have been found in numerous epidemiologic studies conducted on residents of cadmium-polluted areas in Japan [129,130], Belgium [131,132], and China [133].
ATSDR has derived a minimal risk level (MRL) of 0.0002 mg/kg/day for a chronic oral exposure to cadmium. The oral MRL is based on a lifetime accumulated threshold of 2,000 milligrams of cadmium from dietary sources. The threshold is associated with kidney effects (proteinuria, or protein in the urine) seen in residents of cadmium-polluted areas of Japan.
EPA has calculated oral chronic RfDs for cadmium of 0.001 and 0.0005 mg/kg/day for ingestion from food and water, respectively. The critical effect is significant proteinuria in humans chronically exposed to cadmium, using a no-observed-adverse-effect level (NOAEL) of 200 milligrams per gram (mg/g) wet weight in the renal cortex and a kinetic model assuming 2.5% or 5% absorption from food or water, respectively, and 0.01% per day excretion [128].
A relevant consideration is whether the proteinuria caused by cadmium exposure should be considered an adverse effect. By itself, the increased excretion of low-molecular-weight proteins has no adverse effect on health. However, several studies have indicated that increased excretion of calcium also occurs with cadmium-induced kidney damage. This can lead to an adverse effect (osteoporosis), particularly in postmenopausal women.
Hypothetically, children who drink groundwater with cadmium at the concentration detected in one well would have estimated exposure doses that could result in adverse health effects. This is unlikely, however, since the well was never used as a residential source and is located on an industrial property.
There is a high degree of uncertainty surrounding the actual exposure doses for cadmium in groundwater, given that samples from residential wells were below the detection limit. Even if we assume that cadmium was present at that detection limit in these residential wells, cadmium would not pose a public health hazard.
Exposures to hexavalent chromium in off-site groundwater are not a public health hazard.
Chromium is a naturally occurring element found in rocks, animals, plants, soil, and volcanic gases. Chromium occurs in the environment in several forms depending on the valence state of the chromium metal--e.g., trivalent (III) chromium or hexavalent (VI) chromium. Chromium in the environment (e.g., soil, water) and the body is more commonly trivalent than hexavalent [134]. Trivalent chromium is an essential nutrient in the human diet. It helps us regulate how our bodies use insulin. Hexavalent chromium is considerably more toxic to humans than trivalent chromium. Hexavalent chromium is used in chrome plating, dye manufacturing, leather tanning, and wood preservation, and was used as a corrosion inhibitor in the cooling towers at PGDP. Because the measured groundwater analyses are not specific as to valence, we calculated exposure doses assuming that measured concentrations are present as the more toxic hexavalent form.
Concentrations of chromium in the water from off-site groundwater monitoring wells ranged from 40 to 270 µg/L, which exceeded the comparison value of 30 µg/L for hexavalent chromium. However, none of these samples were taken from residential drinking water wells. The maximum concentration of chromium in residential wells was 20 µg/L, which is lower than the comparison value. Because not all residential wells were tested, ATSDR scientists assumed that maximum levels in off-site wells near untested residential wells represented possible human exposure levels. The estimated doses for ingestion of chromium in residential wells, assuming exposure to maximum concentrations in nearby off-site wells, were 0.008 mg/kg/day for an adult and 0.021 mg/kg/day for a child. These doses exceeded health guidelines for hexavalent chromium. If the maximum concentration measured in residential wells was used, the estimated doses would be 0.0006 mg/kg/day for an adult and 0.002 mg/kg/day for a child. (This equates to 0.04 mg/day for a 70-kilogram adult and 0.03 mg/day for a 13-kilogram child.) Therefore, we considered a range of possible exposure doses (shown below) whose lower bound was maximum measured concentrations in residential wells and whose upper bound was maximum concentrations in non-residential wells.
Table 26. Range of Possible Exposure Doses to Hexavalent Chromium
| Person | Lower-Bound to Upper-Bound Estimated Dose |
| Child | 0.002 mg/kg/day (or 0.03 mg/day) to 0.021 mg/kg/day (or 0.27 mg/day) |
| Adult | 0.0006 mg/kg/day (or 0.04 mg/day) to 0.008 mg/kg/day (or 0.56 mg/day) |
ATSDR has not established a health guideline for ingestion of chromium, because the available data are insufficient or too contradictory to establish minimum levels of effect (e.g., LOAELs). Because chromium is an essential nutrient in the body, the National Research Council has established a range of "estimated safe and adequate daily dietary intakes" (ESADDIs) for chromium. The range is 50 to 200 micrograms (g) per day [135]. The upper end of this range, 200 µg/day, has been adopted by ATSDR as an interim guideline for oral exposure to chromium VI and chromium III compounds [134]. This interim guideline is equivalent to an exposure dose of 0.003 mg/kg/day for a 70-kilogram adult, and 0.02 mg/kg/day for a 13-kilogram child. It is similar to the health guideline established by EPA for chronic ingestion of chromium VI. EPA's reference dose for chronic oral exposure, based on animal studies, is 0.003 mg/kg/day [120].
The estimated groundwater doses were slightly above ATSDR's interim guideline for "safe and adequate" intakes. As previously stated, these estimates are very conservative, because they were calculated assuming exposure to maximum concentrations in wells near residential wells, rather than the residential wells themselves, and because they assumed that all chromium was present in the (more toxic) hexavalent form. Exposure doses based on maximum concentrations measured in residential wells are within the "safe and adequate" intake range. Therefore, ATSDR scientists conclude that ingestion of chromium in off-site groundwater (drinking water) wells is not expected to result in adverse human health effects.
Historically, chronic (long-term) exposures to hydrogen fluoride (HF) happened as a result of releases during normal process operations; acute (short-term) HF exposures happened as a result of accidents or controlled releases. (See Appendix F for details on HF releases).
To estimate doses from long-term exposure to HF, we used a correlation between annual uranium hexafluoride releases and HF concentrations at the site perimeter. We calculated exposure doses for potentially affected residents living north of PGDP (based on prevailing wind directions). Long-term HF exposures are not a public health hazard at PGDP.
We estimated acute HF exposure doses using accident records and air dispersion modeling. The most serious accident (November 17,1960) created potential exposures to the southeast of Building C-333. If a sensitive person was exposed to HF at the level modeled for that accident, we expect, that person would experience adverse health effects; however, due to uncertainties (e.g., quantities released, modeling, locations of individuals at time of accidents), it cannot be determined if that accident posed a public health hazard to an individual. Other accidental releases involved smaller quantities and probably did not affect the off-site population.
HF is a colorless fuming gas or liquid that is made up of a hydrogen ion and a fluoride ion. HF is used as a catalyst, as a fluorinating agent, in making fluorine and aluminum fluoride, as an additive in rocket fuel, and for the refining of uranium.
HF is an irritant. It is very soluble in water. It dissolves easily in any water in the air or other media (including skin, the upper respiratory tract, eyes, plants, and soil). When HF is dissolved in water, it is called hydrofluoric acid. Hydrofluoric acid is dangerous to humans, because it can burn the skin and eyes. At first, exposure to hydrofluoric acid may not look like a chemical burn. Skin may only appear red, and may not be painful at first. Damage to the skin can occur over several hours or days, and deep, painful wounds can develop. When not treated properly, serious skin damage and tissue loss can occur. In the worst cases, people who get a large amount of hydrofluoric acid on their skin can die when the fluoride affects the lungs and/or heart.
Breathing in a large amount of HF can harm the lungs and heart and cause death. The human health effects for breathing moderate amounts of HF for several months are not well known, but rats that breathed HF for several months suffered kidney damage and nervous system changes, such as learning problems. If you breathe HF or fluoride-containing dust for several years, changes in your bones (called skeletal fluorosis) can occur.
Studies have been conducted to determine if fluoride causes cancer in people who live in areas with fluoridated water or naturally high levels of fluoride in drinking water, or people who may be exposed to fluorides at work. The studies have not found an association between fluoride and cancer in people.
ATSDR's provisional screening value for intermediate exposure (15 to 364 days) is 0.010 milligrams per cubic meter (mg/m3), or 12 ppb, for air and 0.06 mg/kg/day for oral exposure. Concentrations below these values are not expected to cause adverse health effects. The 12 ppb comparison value for air is more than 100 times lower than exposures that caused mild irritation of the nasal passages in human volunteers exposed for 10 days [63]. The highest average level (time-weighted average) allowed by the Occupational Safety and Health Administration (OSHA) for HF in air for a 40-hour work week made up of 8-hour work days is 2.5 mg/m3 (3 parts per million, or 3,000 ppb). The 12 ppb provisional screening value for air concentrations of HF is more than 250 times lower than OSHA's occupational level.
Air releases of HF have occurred at the PGDP site. Because there is a strong correlation between uranium releases and ambient air concentrations of HF at this site, ATSDR assumed that the largest annual HF release coincided with the highest annual uranium release, which was in 1956. We used the estimated HF air concentration for 1956 to evaluate the health impacts of chronic exposure to HF under normal operating conditions. All of the estimated annual average HF concentrations at the "one north monitoring station" (approximately 1 mile, or 1.6 kilometers, from the site perimeter) were below ATSDR's provisional screening value (see Appendix F, Figure F-2). The highest estimated annual average HF concentration in air (28 ppb for 1956) was at the "perimeter north monitoring station." As such, the perimeter north monitoring station represents the point of maximum off-site exposure; however, no one lives at this location. The closest residence is about 1,500 meters (almost a mile) from the source, about 500 meters (1,640 feet) from the perimeter north monitoring station. The concentration of HF at the nearest residence was estimated at approximately 22 ppb. The annual average concentrations for 1955 and 1956 are about two times greater than annual average concentrations for other years. If actual exposures to HF occurred at 22 ppb, then mild adverse health effects may have resulted. Because our assumptions were so conservative, though, we believe that people were exposed to lower average air concentrations and no adverse health effects would have resulted. Additionally, it should be noted that the exposure assumptions and modeling used to estimate historical air levels were very conservative and most likely overestimated air concentrations. Past, current, and future long-term exposure to HF released during the normal operations of the facility does not pose a public health hazard.
ATSDR used the November 17, 1960 accident data to estimate an acute exposure dose to HF. The estimated maximum acute off-site air concentration for HF was 2.0 to 4.5 parts per million (ppm) for 2 to 4 hours. These concentrations are close to the level of acceptable occupational standards--but occupational standards are not meant to protect sensitive populations (e.g., children and the elderly). If sensitive people were exposed at these levels, they may have experienced adverse health effects (e.g., irritation of the eyes, nose, and throat). Because of the uncertainty associated with historical events (e.g., amounts of material released, modeling, location of off-site individuals during accidents), past exposure to estimated maximum air concentrations poses an indeterminate public health hazard.
Past exposure to lead in three residential drinking water wells may have increased the likelihood of neurological effects in young children, and thus posed a public health hazard. Current exposure may still be occurring if the source of the lead was from pipes and plumbing as opposed to groundwater.
Lead is a naturally occurring element found in the earth's crust [89]. It is used in a variety of products and industrial processes, which can release it into the environment. Lead can be introduced to soil through exhaust from leaded gas fumes from vehicles, spillage of leaded paint or paint chips, or application of a variety of leaded products. Ingesting and inhaling contaminated soils exposes people to lead. Lead in soil can contaminate groundwater and surface water under certain environmental conditions. Pollution or use of lead solder in water delivery and household plumbing systems can increase levels of lead in drinking water.
Lead was detected in groundwater near the site. Samples from 12 residential wells near PGDP had concentrations of lead ranging from 10 to 110 µg/L [45]. Due to the locations of the wells with the highest levels, lead did not appear to be related to PGDP. The lead concentration was 10 µg/L in nine of the residential wells, 100 µg/L in one well, and 110 µg/L in one well. There was one reading of 290 µg/L in another residential well, but that reading could not be replicated; with the high reading included, concentrations in this single well averaged 103 µg/L. Other off-site monitoring wells north of Ogden Landing Road near the North-South Diversion Ditch and southwest of the site near the inactive landfill had concentrations ranging from 10 µg/L to 210 µg/L. The highest concentrations were near the drainage ditch north of the site. Most of the off-site wells were sampled for lead only once. The background concentration of lead in groundwater for the PGDP area is 10 µg/L [119].
It has long been known that lead exposure can have harmful effects. Young children and fetuses have been the main focus of health effects research, since they are the most sensitive individuals; however, adults exposed to lead can also experience adverse health effects [136]. Infants and children receive higher doses from any given level of environmental lead than do adults, because they have a greater absorption capacity for lead than adults,. Therefore, age is an important determinant of exposure dose for a given concentration of lead in drinking water (as shown in Table 27 below).
Table 27. Estimated lead doses in adults and infants from various water concentrations
| Lead concentration in water | Estimated dose range in milligrams per kilogram per day | |
| Adults | Infants | |
|
15 micrograms per liter |
0.0005 |
0.002 |
|
40 micrograms per liter |
0.001 |
0.004 |
|
60 micrograms per liter |
0.002 |
0.006 |
|
100 micrograms per liter |
0.003 |
0.01 |
|
290 micrograms per liter |
0.01 |
0.02 |
ATSDR reviewed 122 studies of human and animal exposures to various doses of lead. In general, exposure doses below 0.001 mg/kg/day do not harm humans or animals. Exposure doses between 0.001 and 0.01 mg/kg/day produce minor changes in blood cells. Harmful effects in animals are seen when doses reach and exceed 0.01 mg/kg/day [89].
For humans, there is a correlation between the levels of lead in blood and the harmful effects that may be seen. (This is illustrated in Figure 8, below.) Blood levels of lead can be elevated by sustained exposure to contaminated soil, food, air, or drinking water. Neurological effects are the most important health effects from exposure in childhood or during gestation (i.e., in the uterus). Changes in blood cells serve as indicators of exposure. The Centers for Disease Control and Prevention considers a child to have an elevated blood lead level if the amount of lead in his or her blood is 10 micrograms per deciliter (µg/dL) or higher [137].
The
relationship between blood lead level and lead concentration in environmental
media is determined by several factors, including the chemical and physical
form of lead, the lead particle size, and the age of the person exposed [136].
Scientists at ATSDR and EPA have developed a model for estimating blood lead
in children based on the lead bioavailability generally observed at hazardous
waste sites. This model is called the Integrated Exposure Uptake Biokinetic
(IEUBK) Model for Lead in Children [138]. ATSDR scientists estimated blood levels
for children drinking water from residential wells near PGDP using this model.
We also estimated blood lead levels using EPA's slope factors for lead [89,120].
Adult blood lead concentration is less affected by lead concentration in environmental
media. To estimate adult blood lead levels from environmental media, we used
EPA's slope factors only [89].
The most contaminated residential wells near PGDP have been closed, and residents that relied on them are now using alternate water sources. (This is assuming that the source of the lead was the groundwater and not the residential piping and plumbing.) To estimate past blood levels for exposure to water from the residential wells, we made the conservative assumption that people were simultaneously exposed to lead in several environmental media (water, air, soil, and food). This is a valid assumption, because lead was detected in various off-site media: although levels were below environmental comparison (screening) values, all media would contribute to the body burden of lead.
We assumed that children were exposed to lead at a concentration of 0.1 micrograms per cubic meter (µg/m3) in air, 200 micrograms per gram (µg/g) of soil and dust, and from 2.4 to 3.4 micrograms per day (µg/day)--depending on age--in the diet [138]. We assumed exposure to children because they are particularly sensitive to the adverse effects of lead [89,136]. Adults, including pregnant women, were not and are not likely to have elevated blood lead levels if they were exposed to the mean residential well water concentrations.
Whether we estimated blood lead levels from the model, or from slope factors, we found that children drinking water from wells with lead concentrations less than 60 µg/L were not likely to experience adverse health effects from exposure. Water from the three wells containing approximately 100 µg/L could have raised blood levels above the action level of 10 µg/dL in children under 4 years old while the wells were in use. Therefore, we conclude that blood levels in the past may have been sufficient to have marginal effects on hearing, intelligence quotient (IQ), and growth in young children using these wells (as illustrated in Figure 8).
After exposure ends, blood lead level and the likelihood of harmful effects, declines with time (at a half-life of 25 days) [136]. However, some of the lead in blood can be taken up by the bones and remain there for decades [136]. Bone lead can be a source of blood lead under conditions that might cause bone desorption, such as pregnancy, poor diet, or older age [136]. We recommend that residents who are concerned about lead in their drinking water have their wells tested.
Manganese was detected in off-site soil at levels ranging from 34 to 4,020 mg/kg (ppm). The residential exposure scenario had an estimated exposure dose for a child with pica behavior (a child who exhibits an abnormal appetite for soil) that exceeded ATSDR's screening value. The estimated exposure doses for an adult and normal child were below ATSDR's screening value. Based on conservative exposure assumptions, ATSDR believes that manganese exposure doses from off-site soil is not a public health hazard.
Manganese is a naturally occurring substance found in many types of rock. Pure manganese is a silver-colored metal, somewhat like iron in its physical and chemical properties. Manganese does not occur in the environment as pure metal. Rather, it occurs combined with other chemicals, such as oxygen, sulfur, and chlorine.
Rocks containing high levels of manganese compounds are mined and used to produce manganese metal, which is mixed with iron to make various types of steel. Some manganese compounds are used in batteries, ceramics, pesticides, and fertilizers; and in dietary supplements.
Ingesting a small amount of manganese each day is important in maintaining your health. The amount of manganese in a normal human diet (about 2 to 9 mg/day) seems to be enough to meet a person's daily need; however, no cases of illness from eating too little manganese have been reported in humans. In animals, eating too little manganese can interfere with normal growth, bone formation, and reproduction.
Too much manganese can cause serious illness. Although there are some differences between different kinds of manganese, most manganese compounds seem to cause the same effects. Manganese miners or steel workers inhaling high levels of manganese dust may have mental and emotional disturbances, and body movements may become slow and clumsy. This combination of symptoms is a disease called manganism. Workers usually do not develop symptoms unless they have been exposed for many months or years at high levels. Manganism occurs because too much manganese permanently injures a part of the brain that helps control body movements. It is not certain whether eating or drinking too much manganese can cause manganism [139].
There is little evidence to suggest that cancer is a major concern for people exposed to manganese. EPA does not classify manganese as a human carcinogen.
The most significant exposure to manganese for the general population is from food, with an average ingestion rate of 3.8 mg/day. Other estimates of daily intake for adults range from 2.0 to 8.8 milligrams. Even though gastrointestinal absorption of manganese is low (3% to 5%), oral exposure is also the primary source of absorbed manganese [139].
Manganese intake among individuals varies greatly, depending upon dietary habits. For example, an average cup of tea may contain 0.4 and 1.3 milligrams of manganese [139]. Thus, someone who drinks three cups of tea per day might receive up to 4 mg/day from this source alone, doubling his or her the average intake.
The Food and Nutrition Board of the National Research Council estimated the adequate and safe intake of manganese for adults at 2.5 to 5 mg/day [139]. It is possible that a significant proportion of Americans, especially women, are not consuming sufficient manganese, although no cases of manganese deficiency have been documented in humans. However, infants may be ingesting more than the estimated safe and adequate dose for their age group (which is 0.7 to 1.0 mg/day), due to high manganese levels in prepared infant foods and formulas [139].
ATSDR has derived a provisional MRL of 0.07 mg/kg/day for a chronic oral exposure (365 days or more) to manganese in soil. EPA has derived a chronic oral RfD of 0.14 mg/kg/day for manganese in the diet [140]. This value is equal to the average daily intake of manganese in the diet (10 mg/day) that is considered adequate and safe. The RfD was derived assuming an average body weight of 70 kilograms. An uncertainty factor was not employed, because (1) the information used to determine the RfD was taken from many large populations, (2) humans exert an efficient homeostatic control over manganese such that body burdens are kept constant through variations in diet, (3) there are no sub-populations that are believed to be more sensitive to manganese at this level, and (4) manganese is an essential element, required for normal human growth and maintenance of health.
When assessing exposure to manganese from drinking water or soil, EPA recommends, one should use a modifying factor (an uncertainty factor based on professional judgement) of 3, based on some evidence that infants younger than 28 days have a higher uptake of manganese in liquids, excrete less absorbed manganese, and, as neonates, pass the absorbed manganese more easily through the blood-brain barrier. The resulting chronic oral RfD for manganese in water and soil would be 0.05 mg/kg/day. The estimated exposure dose for a pica child is 0.1 mg/kg/day (assuming ingestion of 2 grams of soil per day for 290 days per year). However, if one assumes that manganese in soil behaves similarly to manganese in food (i.e., that its bioavailability is similar), then a comparison value at or near 0.14 mg/kg/day would be deemed more appropriate, and the estimated exposure dose for a pica child would not exceed this value.
Exposures to nitrate from PGDP sources are not a public health hazard.
Nitrate and nitrite are naturally occurring compounds, part of the nitrogen cycle. Because nitrite is easily oxidized into nitrate, nitrate is the form that is typically found in groundwater and surface water. Nitrate is the primary source of nitrogen for plants. Wastes containing organic nitrogen are decomposed in soil or water by bacteria to form ammonia. Ammonia is then oxidized to nitrite and nitrate. Agricultural and residential use of nitrogen-based fertilizers, nitrogenous wastes from livestock and poultry production, and urban sewage treatment systems have increased levels of nitrate in soil and water. Certain plants (cauliflower, spinach, collard greens, broccoli, carrots, and other root vegetables) have a naturally higher nitrate content than other plant foods and can account for a large percentage of nitrate in the diet. Nitrate and nitrite compounds are also used for color enhancement and preservation of processed meat products. Nitrate is used in foods to prevent botulism, a life-threatening food-borne illness.
Nitrate-containing compounds are water soluble, which means that they can be carried in water. Thus, nitrate can enter drinking water supplies through surface water runoff, home sewage systems, agricultural fields, and groundwater recharge.
In agricultural areas, a seasonal pattern of increased nitrate levels in drinking water has been seen. This increase occurs most often in spring, when fertilizers are applied and nitrate is transported through storm runoff or groundwater recharge. The most common route of exposure occurs through drinking contaminated water, eating vegetables with naturally high levels of nitrate, and eating foods preserved with nitrate.
Nitrate was detected in off-site groundwater (in RW-002) once used for residential purposes at a maximum concentration of 29.2 milligrams per liter (mg/L) as total nitrate (NO3). ATSDR believes that no one (not even infants or children) would have experienced adverse health effects from exposure through drinking water, even if they consumed nitrate-impacted drinking water at the maximum concentration detected. Nitrate is not now present in residential wells and should not be present in the future. It should be noted that nitrate was detected in surface water at a maximum concentration of 84.6 mg/L as NO3. If people consumed the contaminated surface water at the maximum detected level on a regular basis for an extended period of time, they might experience adverse health effects. However, this exposure scenario is very unlikely.
ATSDR has developed Reference Dose Media Evaluation Guides (RMEGs) for chronic (1 year or more) oral exposure to nitrate in water. Media concentrations less than the RMEG are unlikely to pose a health threat. The chronic RMEGs for a child are 20 mg/L for nitrate-nitrogen (NO3-N) and 90 mg/L for NO3; for adults, the chronic RMEGs are 60 mg/L for NO3-N and 270 mg/L for NO3. The RMEG for nitrate is not protective of infants, so ATSDR recommends using EPA's Maximum Contaminant Level Goal, or MCLG (10 mg/L for NO3-N) as a guideline to evaluate potential infant exposure.
RMEGs are media-specific chemical comparison values derived from EPA's RfDs. RfDs are health-based guidelines for non-cancer effects. An RfD is an estimate of the amount of a chemical that a person can be exposed to, on a daily basis, that is not anticipated to cause adverse health effects over a person's lifetime. MCLGs, which EPA sets after reviewing health effects studies, are the maximum levels of contaminants in drinking water at which no known or anticipated adverse effect on the health of persons would occur, and that allow an adequate margin of safety. MCLGs are non-enforceable public health goals. When determining an MCLG, EPA considers the risk that sensitive sub-populations (infants, children, the elderly, and those with compromised immune systems) will experience various adverse health effects. For chemicals that can cause adverse non-cancer health effects, MCLGs are based on RfDs.
EPA requires that the amount of nitrate (as NO3-N) in public drinking water supplies not exceed 10 mg/L. (This regulation does not cover private wells.) If the results of a water analysis are reported as NO3 (total nitrate) instead of NO3-N, the equivalent value would be 45 mg/L.
Nitrate can affect the blood's ability to carry oxygen. Nitrate's acute toxicity is due to its biological conversion to nitrite, which oxidizes ferrous iron in the hemoglobin producing methemoglobin. Methemoglobin interferes with the oxygen transport system in the blood. Methemoglobinemia (blue-baby syndrome) is caused by high levels of nitrite (or indirectly, nitrate) in the blood. Infants are more sensitive to nitrate for several reasons. They consume more water relative to their body weight than adults, and the hemoglobin in an infant's blood (called fetal hemoglobin) is more easily changed into methemoglobin than an adult's hemoglobin. Also, an infant's digestive system is less acidic, which enhances the conversion of nitrate to nitrite. The two most common symptoms related to the consumption of water with high levels of nitrate are methemoglobinemia and acute diarrhea. Fatalities from methemoglobinemia occur infrequently and are most common in rural areas. Illness and death caused by methemoglobinemia are not always recognized, so methemoglobinemia's occurrence may be under-reported.
Families with infants should use an alternate water supply if their well is known to contain elevated levels of nitrate. When preparing infant formula, families should use nitrate-free water. If a private well is used, it should be inspected for proper construction and tested for nitrate and bacteria levels. Foods containing nitrate, as well as sausage preserved with nitrate and nitrite, have caused symptomatic methemoglobinemia in children.
Nitrates can react with other substances to form N-nitroso compounds. Some of these N-nitroso compounds have caused cancer in animals. However, the mechanism for this is not well defined. Human and experimental animal studies have failed to provide conclusive evidence that ingestion of nitrate or nitrite causes cancer.
Based on the information presented above, nitrate concentrations detected in off-site groundwater are not expected to cause an adverse public health effect in adults, infants, or children.
Pentachlorophenol was not detected in any off-site drinking water wells, but the detection limit in residential wells (approximately 50 µg/L) was five times higher than ATSDR's comparison value (10 µg/L). Even if concentrations are assumed to be 50 µg/L, the resulting exposures are not a public health hazard.
Pentachlorophenol is a man-made substance that was used widely as a pesticide, herbicide, and wood preserver [141]. Pentachlorophenol by itself is slightly water soluble. However, technical-grade pentachlorophenol that is used as a pesticide or wood preserver typically contains other contaminants, such as chlorinated dibenzodioxins, that are not as soluble. One of these chlorinated dibenzodioxins, octochlorodibenzodioxin (OCDD), is 189 million times less soluble in water than pentachlorophenol [141,142]. Environmental contamination at most industrial sites contains technical-grade as opposed to pure-grade pentachlorophenol. When waste technical-grade pentachlorophenol seeps into the soil and migrates downward toward the groundwater, OCDD comes out of solution and remains in the surface soils. This has apparently occurred at PGDP, because OCDD and other chlorinated dibenzodioxins are present at low levels in the top 3 feet of soil on site, but are not detected in samples taken at depths greater than 3 feet [45]. In order for pentachlorophenol in soil to reach groundwater under PGDP, it must travel through 30 to 100 feet of silt and clay; by then, it is essentially free of less-soluble dioxin contaminants, which have sorbed to soils. Therefore, pentachlorophenol in groundwater is essentially the same as pure-grade pentachlorophenol.
Pentachlorophenol was detected in one off-site monitoring well, at a maximum concentration of 8 µg/L. It was not detected in any off-site residential wells; however, the well sampling could not detect pentachlorophenol at concentrations below 50 µg/L, which is higher than the comparison value used to select contaminants of concern. ATSDR scientists used this detection limit to estimate exposure doses of 0.005 mg/kg/day for a child and 0.001 mg/kg/day for an adult.
ATSDR has developed a health guideline (0.001 mg/kg/day) for intermediate-duration oral exposure to pentachlorophenol. This guideline is based on observations of increased serum levels of liver enzymes in rats, which is considered suggestive of liver toxicity [141]. When rats were given food contaminated with either technical-grade or pure pentachlorophenol, those receiving 1 to 25 mg/kg/day of technical-grade product showed signs of liver injury. ATSDR based its health guideline on the lowest dose (1.2 mg/kg/day) of technical-grade pentachlorophenol shown to cause liver injury, because it is likely that most hazardous waste sites contain technical-grade as opposed to pure pentachlorophenol. We applied an uncertainty factor of 1,000 to this lowest dose, because humans may be more sensitive to pentachlorophenol than rats, because some humans are more sensitive than others, and because the animal study involved intermediate-duration (rather than chronic) exposure. The only effect caused by the pure pentachlorophenol in this study was an increased liver concentration of an enzyme needed to eliminate pentachlorophenol in the rat's urine. This effect was observed at doses above 5 mg/kg/day.
It is more likely that people near PGDP were exposed to pure, rather than technical-grade, pentachlorophenol. To evaluate health effects from exposure to pure pentachlorophenol in drinking water, we could derive a tentative, site-specific oral health guideline for chronic duration based on the highest dose (5 mg/kg/day) that failed to cause liver injury in rats. If, as above, we divided by an uncertainty factor of 1,000 to account for differences in sensitivity and exposure duration, this tentative health guideline would be 0.005 mg/kg/day. ATSDR's estimated exposure doses are equal to or lower than this health guideline. Exposure to pentachlorophenol in groundwater at the detection limit concentrations is not expected to result in adverse health effects.
EPA classifies pentachlorophenol as a probable human carcinogen (a cancer-causing substance). The classification is based on studies of rats that developed liver cancer and hemangiosarcoma (blood vessel tumors) after being exposed to technical-grade pentachlorophenol and pentachlorophenol containing lower levels of dioxins than technical-grade. The doses required to produce cancers in these studies were at least 3,000 times higher than the maximum doses ATSDR estimated for ingestion of drinking water from the residence near PGDP [120]. The types of liver tumor observed in these rats are also associated with dioxin exposure; the hemangiosarcomas are not.
From this information, EPA derived a cancer slope factor of 0.12 mg/kg/day based on all tumors combined [120]. However, there is no clear evidence from high occupational exposures that pentachlorophenol causes cancer in humans [141]. Therefore, there is even less likelihood that lower environmental exposures could produce these effects. There is also no evidence of human angiosarcomas among people exposed to pentachlorophenol [141]. Even if we assume that cancer is a possibility for humans, and we consider maximum estimated exposure doses to be equal to the residential well detection limit, cancer effects are not likely for people who may have ingested pentachlorophenol-contaminated water in the past.
ATSDR scientists conclude that past ingestion of pentachlorophenol in off-site groundwater (drinking water) is not expected to cause adverse human health effects.
Exposure to polychlorinated biphenyls (PCBs) through consumption of biota (fish and deer) from the WKWMA is not a public health hazard.
PCBs are a group of man-made chlorinated organic compounds that contain hundreds of individual chemicals, called congeners, with varying toxicities. PCBs can be liquids or solids; they are oily, colorless to light yellow, tasteless, and odorless. They are difficult to burn and are good insulators. These properties once made them useful for a variety of purposes: coolants and lubricators in transformers, capacitors, and other electrical equipment; additives in paint, plastics, newspaper print, and dyes; extenders in pesticides; and heat transfer and hydraulic fluids. During the 1970s, scientists found PCBs in ambient air, soil, water, and sediment, even though there are no known natural sources of PCBs in the environment. EPA banned the production of PCBs in 1978. Traces of PCBs can still be found in the tissues of wildlife, domestic animals, and people--PCBs have chemical and physical properties that make them persistent in the environment and readily accumulate in the fatty tissues of organisms. Overall, levels of PCBs in the environment have been declining since 1978 [101,143].
Although PCBs are no longer made in the United States, people can still be exposed to them. Transformers are useful for several decades, and many older transformers (and capacitors) still contain PCBs. Old electrical appliances may release PCBs when they get hot and contaminate inside air. Discarded capacitors and transformers can release PCBs into the environment from landfills. Heavy electrical power consumers, such as PGDP, are sources of environmental PCBs.
PCBs are poorly soluble in water and tend to adsorb onto sediments in lakes and streams. PCBs present in sediment may enter the aquatic food chain and smaller fish, which in turn, become PCB sources for larger fish. Birds and land predators, such as man, may be exposed to PCBs when they eat contaminated biota. At each step in the "food chain," PCBs that have accumulated in the animals' fatty tissues can appear in greater concentration, or "bioconcentrate," in the species that eat them. PCBs were found in fish sampled from several locations in Little Bayou Creek, and to a much lesser extent in fish sampled from Bayou Creek. PCB levels in deer tissue were extremely low and do not pose any threat. More recent (1997) samples from deer have been below the detection limit in multiple tissues (muscle, liver, fat, and mammary).
Kentucky has issued a health advisory regarding consumption of specific species of fish from Little Bayou Creek. The PGDP 1989 Environmental Report indicated that total PCB concentrations in fish from Little Bayou Creek averaged approximately 5 micrograms of PCB per gram of fillet (µg/g); see Table 18A(2). The highest average total PCB concentrations (17.95 µg/g) were reported in sunfish collected from Outfall #11, which is part of the Little Bayou Creek area. The total PCB concentrations in fish tissue from Outfall #11, based on samples from three sunfish, were more than three times greater than average total PCB concentrations from the Little Bayou Creek area. Outfall #11 data are limited, and seem not to be representative of the Little Bayou Creek area. Also, Outfall #11 is fenced and posted with warning signs. Accordingly, we did not use Outfall #11 data when we calculated the average total PCB concentration for the Little Bayou Creek area. Total PCB concentrations in fish from Bayou Creek were approximately five times lower (average approximately 1 µg/g) than fish from Little Bayou Creek.
In 1993 and 1994, fish tissue samples from Bayou Creek and Little Bayou Creek indicated that concentrations had decreased since 1989: they were about 10 times lower than 1989 reported values. Additionally, total PCB levels in fish tissue from Bayou Creek (0.143 µg/g) were about four times lower than concentrations detected in fish tissue from Little Bayou Creek (0.553 µg/g). Background samples, from Hinds Creek, did not contain detectable levels of PCBs. In 1993, 40% of the fish sampled from Bayou Creek did not contain detectable levels of PCBs. In 1994, that number was 20% (fewer fish were sampled in 1994, which may account for the difference). Also, in 1993 and 1994, several fish from Little Bayou Creek did not have detectable levels of PCBs.
In 1997, the Kentucky Division of Waste Management collected 20 sunfish from Little Bayou Creek and analyzed them for levels of PCBs in fillets. The average concentration of total PCBs (0.561 µg/g) in fish tissue from Little Bayou Creek was similar to the 1993-1994 results. Two of the twenty fish sampled did not have detectable levels of PCBs in fillets. Fish tissue results were not available for Bayou Creek in 1997.
ATSDR evaluated whether adults and children eating fish from either Bayou Creek or the Little Bayou Creek system could obtain PCB doses that would cause adverse health effects. ATSDR assumed that subsistence and recreational anglers got 20% of their total fish intake from the creeks for 30 years. The estimated exposure dose for children (assumed to be the children of subsistence/recreational anglers) was based on a 6-year exposure duration. The consumption rate for recreational anglers was 8 grams per day (g/day), which equates to 20 meals per year at 150 grams (5.3 ounces) per meal. For subsistence anglers, the consumption rate was 60 g/day; that equates to 150 meals per year at 150 grams (5.3 ounces) per meal. A recreational angler's child was assumed to consume 3 g/day, which equates to 20 meals per year at 50 grams (1.8 ounces) per meal. A subsistence angler's child was assumed to consume 8 g/day--equal to 150 meals per year at 50 grams (1.8 ounces) per meal. If a fish tissue sample was below the detection limit, we used the detection limit as the measured value for total averaged values. This is a conservative approach that could overestimate the exposure dose.
ATSDR believes that people are more likely to fish in Bayou Creek than in Little Bayou Creek. This is because of the posted fish advisories and more limited access to Little Bayou Creek. Additionally, Bayou Creek provides a better habitat for fish that people typically eat.
Ingestion or inhalation of PCBs at high exposure doses has been shown to cause skin irritations, such as chloracne and rashes, in animals and humans [101,120]. The doses required to produce such effects are quite high: daily, occupational exposure doses ranging from 0.07 and 0.14 mg/kg/day failed to produce adverse health effects in workers [101]. Reports of developmental effects from lower exposures are controversial and have not been verified [101].
Generally, humans appear to be less sensitive than other animals to the toxic effects of PCBs. In laboratory animals, PCBs have been shown to produce skin effects (similar to those seen in people exposed at high doses) as well as effects on the thyroid, immune system, liver, toenails, and eyelids. Of the laboratory animals tested (i.e., rabbits, minks, mice, rats, ferrets, and monkeys), the rhesus monkey appears to be the most sensitive [120]. PCBs have been shown to impair the monkey's immune system (in addition to producing skin, fingernail, and toenail effects), at doses as low as 0.005 mg/kg/day. This dose is almost 28 times lower than the dose shown not to harm people. ATSDR and EPA have developed a health guideline of 0.00002 mg/kg/day, based on adverse effects in monkeys [101,120].
Several human studies have reported that low level environmental PCB exposure during in utero or neonatal development can effect a child's neurologic system [144,145], immunologic system [146,147], or development [148,149,150]. However, reported study limitations include: unmeasured exposure concentrations, possible exposure to other neurotoxins (e.g., dioxins, mercury, lead, organochlorine pesticides), and inadequate control for confounding factors (e.g., birth weight and maternal smoking, alcohol, and drug use). These studies suggest, but do not prove, an association between prenatal/neonatal exposures to PCBs and neurologic, immunologic and developmental effects in young children. Therefore, it cannot be equivocally determined whether low level environmental PCB exposures affect prenatal or neonatal development.
Rats are the only laboratory species shown to develop cancer after ingesting PCBs [101,120]. The animals were administered doses of PCBs that were considerably higher than environmental doses. For example, the doses given rats in one study were equivalent to human doses of 0.35 to 3.0 mg/kg/day (which is 100 to 1,000 times higher than the estimated potential doses to children and adults in the PGDP area). In order to use animal data to predict whether humans are likely to develop cancer, we often assume that the relationship between PCB dose (administered) and cancer development is the same at high and low doses. We also assume that there is no dose at which there is no risk for cancer development. Many scientists believe that these assumptions are valid for substances that cause cancer by directly attacking (i.e., mutating) genetic material in all living cells, but the assumption is much less likely to hold for substances that cause cancer without directly attacking genetic material. PCBs are considered by many scientists to induce tumors (in rats) primarily through mechanisms that do not involve genetic mutation [120].
To evaluate the potential for cancer in humans using data from animal studies, scientists must make assumptions about the ways humans resemble or differ from the animal "models." EPA's standard methodology uses a "scaling factor" to account for differences in the size of the test animals (e.g., body weight, lung surface area) compared to humans, and a cancer slope factor to predict the likelihood of cancer developing per unit dose (measured in mg/kg/day) [151].
One approach, called physiologically based pharmacokinetic (PBPK) modeling, incorporates information about how a substance and its degradation products are chemically modified and distributed throughout the body following exposure. When PBPK models were used to compare how different animal species handle PCBs and their metabolites, many inconsistencies were found, making cross-species predictions highly uncertain [101].
These considerations may explain why there are no scientific reports of cancer in any animal species other than rats, not even in the sensitive rhesus monkeys, following exposure to PCBs. Also, PBPK modeling may explain the lack of conclusive reports of cancer in multiple studies of workers occupationally exposed to PCBs [101,120].
A recent study of more than 7,000 capacitor workers reported exposures to PCB air concentrations as high as 1,500 µg/m3. Workers in this study were employed for at least 3 months, and their health status was followed for an average of more than 30 years. Using the reported exposure levels, ATSDR estimated lifetime exposure doses of 0.0004 to 0.009 mg/kg/day for these workers. Using standard EPA methods to predict the likelihood of cancer at these doses, one would have expected to see additional cancers among 1,687 workers who received the highest PCB exposures. However, the study found no excess cancers of the liver or any other organ [143]. The estimated occupational exposure doses that failed to produce detectable increases in liver cancer were more than 5 to 185 times higher than the lifetime exposure doses estimated for subsistence and recreational fishers who could have ingested fish from waters near PGDP or been exposed to PCBs in off-site soils.
Therefore, ATSDR scientists conclude that exposure to PCBs from ingestion of deer and fish from Bayou and Little Bayou Creeks is not expected to result in adverse health effects.
Radioactive Materials (Radiation Exposures)
Radioactive materials, both naturally occurring and man-made, have been detected in all media at PGDP. The cumulative radiation dose from potential chronic exposure to those media is not a public health hazard. Potential acute exposures from an accident in 1960 are an indeterminate health hazard.
Radioactive material's concentrations and total annual quantities were reported for each medium (e.g., soil) by DOE (and formerly by the U.S. Energy Research and Development Administration or the U.S. Atomic Energy Commission). For media in which the off-site concentrations were not reported, ATSDR estimated the concentrations by using computer models. The predominant radioactive materials at PGDP are, and were in the past, uranium 234, uranium 235, uranium 238, and technetium 99. These contaminants were screened in all media. Other radioactive contaminants (e.g., thorium 230, plutonium 239, neptunium 237) were analyzed in some media and estimated in some cases; however, these other radioactive contaminants contributed approximately 10% or less to the exposures doses.
Table 28 lists maximum estimated annual committed effective doses for children and adults in different media. The potential exposure doses occurred at different times and in different places: realistically, the total doses from each exposure pathway should not be added together. Current exposure doses are much less than those estimated for the first 10 years of plant operations.
The potential health effects from each radioactive material per route of exposure were reviewed. Also, the potential health effects from estimated radiation doses from all routes of exposure were considered, as were organ doses. The estimated radiation exposure dose from all media for any year of plant operation does not exceed 500 millirems (mrem), or 5 millisieverts (mSv), except for a potential acute exposure in 1960. The International Commission on Radiological Protection (ICRP) recommends, for annual committed effective dose to the general population, a limit of 100 mrem (1 mSv) above background [152]. Before 1990, the ICRP recommendation was 500 mrem (5 mSv) per year. Although the ICRP recommendations were lowered for chronic exposure over a 70-year life span, no adverse health effects have been seen at the estimated chronic exposure doses for PGDP, and no apparent increased cancer risk would be expected [152,154].
Table 28. Maximum estimated annual committed effective doses for radiation exposure near PGDP
| Exposure pathway | Route of Exposure | Maximum Estimated Annual Committed Effective Dose for Children | Maximum Estimated Annual Committed Effective Dose for Adults |
| Groundwater1 | Ingestion | 7 mrem (0.07 mSv) | 7 mrem (0.07 mSv) |
| Surface water2 | Ingestion | 2.0 mrem (0.02 mSv) | 0.8 mrem (0.01 mSv) |
| Soil/sediment3 | Ingestion | 9 mrem (0.09 mSv) (pica child) | 0.6 mrem (0.01 mSv) (workers) |
| Food/biota4 | Ingestion | 0.4 mrem (0.00 mSv) | 0.7 mrem (0.01 mSv) |
| Air5 | Inhalation (chronic)
Inhalation (acute) |
340 mrem (3.4 mSv)
500 - 1500 mrem (5 to 15 mSv) |
340 mrem (3.4 mSv)
500 - 1500 mrem (5 to 15 mSv) |
| 1 The maximum concentration of
technetium 99 used in the calculation was detected in 1988, before the first
well was taken out of service. 2 The maximum concentrations used in the calculation were detected in 1959 and 1960. 3 Most of these samples were collected and analyzed in the 1990s. 4 Most of these samples were collected and analyzed in the 1990s. 5 The concentrations used in this model for chronic exposure were estimated for 1956. The quantities used to model acute exposure were reported for the 1960 accidental release. |
|||
| Key: mrem = millirems; mSv = millisieverts | |||
ATSDR concludes that past or current chronic exposure to radioactive materials in off-site media from normal plant operations is not expected to result in adverse human health effects.
For potential acute exposure, there are many uncertainties involved in determining estimated doses, including quantities released, the duration of the release, and the exact location of individuals at the time of the accident. Epidemiological worker studies of chronic exposures to uranium dust suggest, but do not confirm, evidence of adverse health effects, primarily malignant and non-malignant lung diseases. However, these workers were chronically exposed to higher levels of insoluble uranium than estimated exposure doses calculated for past accidents. Animal studies on rats investigated acute exposures to uranyl nitrate (a more-soluble form) and reported an increased frequency of lung tumors and osteosarcomas. However, the doses in these studies were substantially higher than the estimated exposure doses from the 1960 accident and the experiment did not provide enough information for confident extrapolation of risk coefficients to humans [153]. Because of the uncertainties in the release quantities and whether the airborne exposure pathway was complete during this accident, ATSDR scientists concluded that the 1960 accident posed an indeterminate health hazard. If an individual was exposed to the maximum estimated exposure and using EPA's cancer risk coefficients [154], we would predict a moderate increased cancer risk.
For more information on uranium, refer to the discussion for that element (below).
Exposure to thallium in off-site surface water and groundwater is not a public health hazard.
Thallium is an element that occurs naturally in the environment. Certain industrial processes (e.g., cement manufacturers, coal-burning power plants, and smelters) release thallium to the environment [155]. Environmental thallium is found chemically combined with other substances such as oxygen, sulfur, and halogens. Most of the chemical compounds are soluble in water. The general public is exposed to low levels of thallium through eating, smoking tobacco, and breathing second-hand tobacco smoke. The average person takes in about 2 micrograms of thallium per gram of food daily. Once ingested, thallium distributes throughout the human body; it can cross the placenta in pregnant women and be distributed to the developing fetus.
Thallium was detected in surface water near PGDP. The maximum thallium concentration in surface water was 5,260 µg/L in Bayou Creek near the inactive southwest landfill [45]. Using this maximum concentration, we estimated that incidental ingestion of water from Bayou Creek would result in an exposure dose of 0.001 mg/kg/day for adults and 0.002 mg/kg/day for children 1 to 6 years old.
Thallium was not found in drinking water wells, but the lowest level of analytical detection was 10 µg/L--higher than EPA's drinking water standard of 2 µg/L [155]. Therefore, we used the detection limit of 10 µg/L to estimate exposure doses. This gave us doses of 0.0003 mg/kg/day for an adult and 0.001 mg/kg/day for a child, assuming that these residential wells were the sole source of drinking water.
ATSDR has no health guideline for ingestion of thallium. EPA has RfDs for several thallium compounds. Each RfD covers a particular compound and is based on animal studies for that compound. For example, the RfD for thallium sulfate is based on a failure to observe harmful effects in rats that were administered as much as 0.25 mg/kg/day of thallium by gavage (stomach tube). EPA divided this number by an uncertainty factor of 3,000 to account for humans being more sensitive than rats to thallium, for some humans being more sensitive than others, and for a lack of chronic toxicity data; this gave EPA an RfD of 0.00008 mg/kg/day [120].
The thallium dose that did not cause toxicity to rats (i.e., 0.25 mg/kg/day) was 200 times higher than the maximum exposure dose that ATSDR estimated for surface water or groundwater ingestion, despite the fact that we used very conservative assumptions to estimate dose. If more realistic exposure assumptions were used, our estimated doses would be even lower. For example, our surface water dose is based on the assumption that a child ingests half a liter of maximally contaminated water a month for 6 years and an adult ingests this amount for 30 years. It is probably not very likely that a young child, who is under constant care by an adult, would consume these quantities of surface water at this maximum concentration. Likewise, it is unlikely that an adult would ingest a half liter of maximally contaminated surface water once a month for 30 years. Lastly, our groundwater doses are not based on measured concentrations in drinking water, but on levels of analytical detection. The actual levels in these wells were lower than the detection limit.
Therefore, ATSDR scientists conclude that ingestion of thallium in surface water from Bayou Creek or from drinking water wells located near PGDP is not expected to result in adverse human health effects.
Past exposure to TCE at levels found in well RW-002 was a public health hazard for children, because it increased the likelihood of neurological effects such as speech and hearing deficits. No public health hazard currently exists, because this residential well is no longer in use and the exposure pathway is incomplete.
TCE is a nonflammable, oily, colorless liquid that has a sweet odor and a sweet, burning taste. Years ago, TCE was used as an anaesthetic. It is now used as a solvent to remove grease from metal parts and to make other chemicals. It is heavier than water and has low solubility (up to one part TCE per thousand parts of water at room temperature) [156]. These qualities make TCE a troublesome contaminant at hazardous waste sites.When present in groundwater, TCE tends to settle into a layer at the bottom of the aquifer and then continuously dissolves into the groundwater. This may result in high levels of TCE in the aquifer for years after the original release of contamination has ended. This has happened at PGDP and is the reason why there was TCE contamination in private well water.
TCE contamination of groundwater beneath the PGDP facility and in nearby private wells was discovered in August 1988. TCE was detected at concentrations above ATSDR's comparison value in four off-site residential wells. Maximum levels of TCE, ranging from 20 to 43 µg/L, were found in three of the wells (RW-004, RW-017, RW-113); a maximum level of 960 µg/L was found in a fourth well (RW-002). As a result of this sampling, the Department of Energy (DOE) immediately provided bottled water to residents with contaminated well water--until they could be supplied with municipal water--and completely discontinued private well use.
Groundwater sampling was not conducted before 1988. Sampling conducted after 1988, when the wells were no longer used for drinking, revealed higher levels of TCE than were first detected in 1988. This finding was not unexpected, considering the results of groundwater modeling of contaminant movement from sources on site. Modeling results indicate that levels of TCE before 1988 were likely to have been lower than levels detected in 1988. ATSDR scientists cannot determine with certainty whether TCE was present in private wells, or at what levels, before 1988. At most, residents used water from the most contaminated well (RW-002) for 5 to 15 years. If these wells are used in the future, or if new wells are drilled into the plumes, the residents could be exposed to much higher concentrations of TCE than in 1988.
There are several reports of an increased occurrence of nervous system and developmental effects, and cancer, from ingestion and inhalation of TCE by animals and humans [156,157,158]. Human health studies suggest an increased incidence of cancer of various types (e.g., bladder, lymphoma, kidney, respiratory tract, cervix, skin, liver, and stomach) from exposure to TCE; however, no studies provide clear, unequivocal evidence that exposure is linked to increased cancer risk in humans [156,157,158]. The available studies suffer from inadequate characterization of exposure, small numbers of subjects, and the fact that subjects were likely exposed to other potentially carcinogenic chemicals. There is, however, sufficient evidence that TCE exposure results in cancer development in animals, although animal studies may not be relevant for evaluating health hazard to humans [156,158].
In 1989, EPA withdrew its cancer assessment for TCE, which was based primarily on animal studies conducted in 1990 and earlier, because more recent pharmacokinetic and mechanistic data for TCE became available [120,159]. An updated approach to TCE cancer assessment using existing animal data and state-of-the-science papers has been proposed [159]. The approach, though high-dose animal studies support it, does not appear entirely relevant for evaluating health hazard from human environmental exposure. There are several reasons for this. First, cancer in animals appears to result from species-specific mechanisms that are not entirely relevant to humans [156,158]. Second, the animals used in these studies were exposed to very high doses of TCE, several orders of magnitude higher than estimated for PGDP residents, and the overall death rate in the animal studies was high. The surviving animals were not likely to have been in good health and, therefore, would have been more susceptible to adverse effects from TCE exposure (like infections and illnesses) than healthy animals. Third, the overall findings from animal studies are inconsistent: some studies report an increased incidence of cancer, while an equal number do not report an increase at similar levels of exposure [156]. Fourth, the studies did not evaluate the effect of exposure to stabilizers and impurities in TCE; these things may also be carcinogenic. For these reasons, ATSDR scientists decided to focus on non-cancer effects of TCE.
ATSDR derived a health guideline of 0.1 ppm for intermediate-duration (15 to 364 days) exposure to TCE by inhalation. This guideline, equivalent to 0.15 mg/kg/day, is based on neurological and cardiac effects (e.g., decreased wakefulness and decreased post-exposure heart rate) in rats. The lowest dose that produced these effects was 50 ppm in air, which is equivalent to 77 mg/kg/day. The estimated dose for PGDP residents (using the maximum concentration from a drinking water well) was similar to the ATSDR health guideline and more than two orders of magnitude lower than the lowest effect level observed in animals.
ATSDR derived a health guideline of 0.2 mg/kg/day for ingestion of TCE based on an acute-duration (less than 14 days) study showing developmental and behavioral changes in mouse pups administered 50 mg/kg/day of TCE [160]. In this study, the TCE was dissolved in oil and administered by stomach tube (gavage) [156]. The findings of this study are not entirely relevant for evaluating health hazard for PGDP residents exposed to TCE in well water for several reasons. First, gavage doses in the animal study were administered as one large dose per day, while PGDP residents were likely to have been exposed to TCE in drinking water several times a day. (The body handles a single large dose much differently than it does a series of smaller doses.) Second, the total dose entering the body is higher and maintained for a longer time when TCE is dissolved in oil than when it is dissolved in water. Lastly, exposure to TCE in the animal study lasted less than 14 days, while maximum exposures to PGDP residents (from the RW-002 well) may have occurred over a period of 5 to 15 years. Despite these limitations, the findings are supported by other animal and human studies.
ATSDR's TCE Sub-Registry reports an excessive number of children aged 9 years old or younger with speech and hearing deficits [161]. Although the exposure levels of these children were not well characterized, the findings support the types of outcome seen in animals. Several studies of workers and community residents suggest a possible association between exposure to TCE (and other chemicals) and developmental outcomes [157,162,163,164]. However, none of the studies provide conclusive evidence for a causal relationship, largely because information about TCE exposure was incomplete and exposure to other chemicals was likely [156,158].
Collectively, the scientific data indicate that the developing nervous system in young animals and humans may be sensitive to the toxic effects of TCE [156]. It is not clear whether past exposures to TCE by PGDP residents with contaminated wells were sufficient to result in similar outcomes. In order to be protective of the most sensitive individuals, ATSDR concludes that past exposure to contaminated water from well RW-002 may have resulted in neurological effects in children chronically using this water prior to 1989.
Short-term (1-hour average) off-site uranium air concentrations were modeled for the time period corresponding to the 1960 accidental release. Estimated levels were above ATSDR's intermediate comparison values and occupational standards at the nearest residence (southeast of Building C-333). If people were exposed to the estimated air concentrations, they could have experienced adverse health effects. However, the accident occurred at 4:00 a.m. in mid-November, when people were most likely indoors and asleep. Because of uncertainties (e.g., quantities released, locations of individuals at the time of the accident) it cannot be determined if this accident posed a public health hazard.
Also, in the past, it has been reported that UF6 was released at night through jets on top of the process buildings to accelerate the reduction of UF6 concentration in the process gas system in order to perform maintenance and inspection on process gas equipment. These releases, called "midnight negatives", potentially contained significant quantities of uranium (and hydrogen fluoride); however, the quantities released and the frequency of releases are unknown. Therefore, it cannot be determined if these releases posed a public health hazard.
Long-term exposure to airborne uranium also occurred during the years 1954 to 1963, as a result of elevated operational emissions. Because the prevailing winds were from the south and southwest, the primary exposed population was residents living north and northeast of the facility. This population may also have some exposure to uranium via soil and/or groundwater exposure pathways. Even if chronic exposure to air, soil, and groundwater occurred simultaneously, adverse health effects are not expected.
Under current conditions, uranium air concentrations are not a public health hazard.
Uranium is a radioactive metal, which is naturally present in rocks, soil, groundwater, surface water, air, plants, and animals in small amounts. It contributes to a natural level of radiation in our environment, called background radiation. The amount of uranium in drinking water in the United States is generally less than 1 picocurie per liter (or approximately 1.5 µg/L) [165].
Natural uranium, enriched uranium, and depleted uranium are mixtures of primarily three uranium isotopes (U-238, U-235, and U-234; chemically similar but with a different number of neutrons). Natural uranium is, by weight, more than 99% U-238, 0.72% U-235, and 0.005% U-234. Enriched uranium is more than 0.72% U-235 by weight, and depleted uranium is less than 0.72% U-235 by weight. All three isotopes are radioactive but have different specific activities (that is, radioactivity per gram of material). U-238 has the lowest specific activity, and U-234 has the highest.
Uranium can harm people in two ways, as a chemical toxin and as a radioactive substance. (That is, its chemical and radioactive properties can both be harmful, and these two things are considered separately.) Because natural uranium produces very little radioactivity, the chemical effects of uranium are generally more harmful than the radioactive effects. However, more radioactive mixtures (like enriched uranium) can harm the kidney more than natural uranium due to the combined effects of chemical and radioactive properties.
The kidney is the primary target organ for the chemical effects of ingested and inhaled uranium. The extent of toxicity is determined primarily by exposure route, type of uranium compound, and solubility of that compound. Ingested uranium compounds are generally less toxic to the kidneys than inhaled uranium compounds, partly because uranium is poorly absorbed from the intestinal tract. Highly soluble uranium compounds are generally more toxic to the kidneys than less-soluble compounds via ingestion, because the more-soluble compounds are more readily absorbed (that is, they pose a greater potential dose to the kidney). Absorption of uranium is low (less than 5%) by all exposure routes (inhalation, ingestion, and dermal).
Studies using laboratory animals provide most of the evidence for kidney toxicity. ATSDR has established intermediate (15 to 364 days) exposure health guidelines for inhalation of both soluble and insoluble uranium compounds. The guideline for insoluble uranium is 8 x 10-3 mg/m3. This guideline is based on structural changes (lesions) in kidneys of dogs exposed to uranium dioxide dust 6 hours a day, 6 days a week, for 5 weeks [166]. The health guideline for inhalation of soluble uranium is 4 x 10-4 mg/m3, based on kidney lesions in dogs exposed to uranium chloride in air 6 hours a day, 6 days a week, for 1 year [167]. Neither study provided information about the size of the uranium particles used, so ATSDR based its guideline on the conservative assumption that uranium particles were 2 microns or less in diameter.
The estimated 1-hour average off-site air concentration of uranium during the accident (approximately 4.3 ppm at the nearest residence) exceeded the intermediate-exposure health-based guideline for inhalation. On-site air concentrations would have been even higher, though it is uncertain whether on-site personnel were exposed to elevated air concentrations. ATSDR has not derived health-based guidelines for acute exposure. The estimated off-site air concentration exceeded the occupational standards for soluble and insoluble uranium compounds. If people were actually exposed to the estimated air concentrations, then a public health hazard existed. ATSDR does not believe that exposures occurred at this level, since the accident occurred at 4:00 a.m. in mid-November over a period of 4 hours.
Discussions with residents and site officials have not indicated any reports of acute symptoms associated with this accident (for either uranium or HF). Because of the lack of exposure information and considering that concentrations were derived from air dispersion modeling, we conclude that an indeterminate health hazard existed for uranium air concentrations in the past.
Exposure to vanadium from off-site groundwater and/or soil is not a public health hazard.
Vanadium is a naturally occurring element in the earth's crust, fuel oil, and coal. Vanadium is mostly used as an alloying agent in steel production, although small amounts are also used in rubber, plastics, and ceramics [168]. Vanadium is a metallic element that occurs in six oxidation states and numerous inorganic compounds. Vanadium's toxicity depends on its physical and chemical state, particularly on its valence state and solubility. Vanadium is poorly absorbed through the gut, but more readily absorbed through the lungs.
Vanadium was detected at a maximum concentration of 20 µg/L in one off-site residential (drinking water) well. Although this concentration is lower than ATSDR's comparison value of 30 µg/L, we selected vanadium as a contaminant of concern because maximum concentrations detected in off-site monitoring wells located near residential wells were as high as 210 µg/L. ATSDR scientists used the maximum groundwater concentration (210 µg/L) to estimate exposure (ingestion) doses of 0.006 mg/kg/day for adults and 0.02 mg/kg/day for children. If we had used the concentrations measured in residential wells, the estimated doses would have been an order of magnitude lower--0.0006 mg/kg/day for adults and 0.002 mg/kg/day for children.
ATSDR's intermediate screening value for ingestion of vanadium is 0.003 mg/kg/day based on a study where rats were administered vanadium in their drinking water for 3 months [168]. In this study, the treated rats showed mild changes to the kidneys at a minimum dose of 0.6 mg/kg/day, while no adverse effects were seen at the lower dose of 0.3 mg/kg/day. The lower dose was considered the NOAEL, and is the basis for ATSDR's health guideline. The NOAEL for rats (0.3 mg/kg/day) was divided by an uncertainty factor of 100, because humans are presumed to be more sensitive than rats to vanadium and because some humans are more sensitive than others. The uncertainty factors may be overly conservative: scientific information suggests that humans are actually less sensitive than rats to ingested vanadium. Human volunteers who swallowed a maximum dose of 1.3 mg/kg/day of vanadium for 45 to 68 days showed no effects when tested for injury to their liver, blood cells, or kidneys [168]. An adult would have to drink more than 4,500 liters a day of water contaminated at the highest level in the residential well to take in the amount of vanadium that did not cause adverse effects in the human volunteers. A child would have to take in 650 liters a day to take in this amount.
Vanadium was detected in off-site soil at levels ranging from 0.01 to 300 mg/kg (ppm). The average vanadium content of soils in the United States is 200 mg/kg; vanadium seems to be most abundant in the western United States. Using the 67th percentile concentration to calculate an exposure dose, we found only one exposure scenario had an estimated exposure dose (0.006 mg/kg/day for a pica child) that exceeded ATSDR's screening value for intermediate (15 to 364 days) oral exposure. ATSDR's screening value (0.003 mg/kg/day) is based on a drinking water study in rats. Water, though, tends to contain more-soluble forms of vanadium than do weathered soils. Consequently, less vanadium would be absorbed from soil than from water. The estimated exposure dose for pica children is approximately 200 times less than the dose given to human volunteers mentioned above. Based on the conservative exposure assumptions, vanadium's poor absorption from the gut, and the implication that humans are less sensitive to ingested vanadium than rats, ATSDR does not expect adverse health effects to result from ingestion of vanadium in soil.
Therefore, ATSDR concludes that ingestion of vanadium in off-site drinking water wells and/or soil is not expected to result in adverse human health effects for past, current, or future exposure to children or adults.
Potential past exposure to vinyl chloride in residential drinking water is an indeterminate public health hazard. It is not known whether anyone was exposed or at what levels due to inadequate detection limits. No public health hazard currently exists, because no one is using these residential wells.
Vinyl chloride is a man-made substance used in the production of polyvinyl chloride (PVC) and other plastic products. It is one of the substances generated when TCE breaks down in groundwater. As TCE degrades in groundwater, the resulting vinyl chloride concentration may increase downgradient, depending on a number of factors, including the chemical characteristics of the soil through which the contaminated groundwater travels and the distance traveled [169].
Vinyl chloride has not been detected in residential wells but was found in two samples from one monitoring well used to test for off-site groundwater contamination in the PGDP area. The maximum concentration of vinyl chloride in this well was 110 µg/L. The test well was located near four residential wells that were not found to contain vinyl chloride; however, the lower limits of analytical detection for these well samples were higher than EPA's Maximum Contaminant Level (MCL) of 2 µg/L for public drinking water supplies. In addition, very few residential well water samples (12 in all) were analyzed for vinyl chloride.
Table 29. Range of Sample Detection Limits for Vinyl Chloride in Residential Well Water
| Well | Range of Sample Detection Limits |
| RW-002 | 1-500 µg/L (four samples) |
| RW-004 | 2-10 µg/L (four samples) |
| RW-017 | 4-10 µg/L (three samples) |
| RW-113 | 10 µg/L (one sample) |
The detection limit for RW-002 was 500 µg/L on October 24, 1989, but the detection limit for this well was 1 µg/L on August 14, 1990. Therefore, vinyl chloride was probably not a problem when RW-002 was being used; however, there is some uncertainty due to variations in the TCE plume concentrations from seasonal factors. The lowest detection limits for wells RW-017 and RW-113 were 4 and 10 µg/L, respectively. Both values are above the MCL.
ATSDR's estimated ingestion doses, assuming exposure to the maximum concentration found in the test well, were 0.006 mg/kg/day for an adult and 0.02 mg/kg/day for a child.
ATSDR has developed a health guideline of 0.00002 mg/kg/day for chronic ingestion of vinyl chloride. This is based on a study of rats that developed liver toxicity from exposure to vinyl chloride (in PVC) in their diet. The lowest dose at which adverse liver effects were observed--the LOAEL--was 0.018 mg/kg/day. An uncertainty factor of 1,000 was applied to the LOAEL, because humans may be more sensitive than rats to vinyl chloride, some humans are more sensitive than others, and there was no dose level tested at which adverse effects were not observed [170]. ATSDR's estimated doses, based on maximum test well concentrations, were higher than the health guideline and similar to the LOAEL (for children).
However, we are not certain whether people drank water from wells potentially contaminated with vinyl chloride. Therefore, the health hazard from past exposure to vinyl chloride cannot be determined. These wells are not currently being used. If we make the assumption that people were exposed to vinyl chloride at maximum "detection limit" concentrations, then we conclude that people may experience adverse health effects. ATSDR scientists recommend that detection limits for degradation products of TCE, such as vinyl chloride, in groundwater analyses are low enough to determine whether concentrations exceed health-based guidelines.
Given the lack of accurate concentration measurements for vinyl chloride in residential wells and exposure information, we conclude that past exposure is an indeterminate health hazard.
Past exposure to zinc from one residential well near PGDP was not a public health hazard. No public health hazard currently exists, because this well is no longer being used.
Zinc is a naturally occurring element that is commonly used in industrial processes [171]. It is found in man-made products, such as metal alloys, dry cell batteries, metal beverage containers, and zinc-coated pipes. Zinc is also used in many over-the-counter medicines, sunblocks, and deodorants, and is also present in leafy vegetables, meat, poultry and fish.
Zinc was detected one time in one residential well, at a concentration of 5,090 µg/L. ATSDR's estimated ingestion doses, assuming exposure at this concentration, were 0.15 mg/kg/day for an adult and 0.39 mg/kg/day for a child.
Zinc is an essential element in the human diet [171]. Zinc deficiencies can produce loss of appetite, growth retardation, skin changes, slow healing of wounds, and depressed mental function in children [171]. The individual response to deficiency varies depending on age; the amount of meat, dairy products, and fibrous vegetables in the diet; and (for women) whether one is pregnant or nursing infants. The average American dietary intake is 15 mg/day for men and 12 mg/day for women [135]. If women are nursing infants less than 6 months old, they need to consume 19 mg/day. Elderly people generally consume lower amounts of zinc (7 to 10 mg/day), but healthier, more active elderly people consume closer to average levels. If a person's diet is low in zinc-containing foods, they may need to consume 36 to 45 mg/day to prevent deficiencies [171].
Long-term ingestion of excessive amounts of zinc can be related to toxicity, including decreased high-density (good) cholesterol levels, impaired immune function, and anemia [120]. These effects have been observed at estimated total dietary intakes of 1 mg/kg/day (which is equivalent to 60 mg/day for a 60-kg woman and 70 mg/day for a 70-kg man) and are the basis for EPA's health guideline of 0.3 mg/kg/day.
ATSDR's estimated exposure dose from the well water for a 13-kilogram child (0.392 mg/kg/day) exceeded the health guideline (0.3 mg/kg/day), but was lower than the lowest level shown to cause adverse health effects and lower than the recommended dietary intake for adults. Also, the estimated exposure dose for a child would decrease as the child developed into an adult (the estimated adult dose was 0.15 mg/kg/day), so the child dose does not represent a chronic exposure dose over a lifetime.
Therefore, ATSDR scientists do not expect adverse health effects to result from past exposure to zinc in drinking water from this residential well near PGDP.
Next Section Table
of Contents
