Skip directly to search Skip directly to A to Z list Skip directly to site content





A release of a hazardous substance does not always result in human exposure, and humanexposure does not always result in adverse health effects.

This section of the public health assessment evaluates the estimated exposure doses for contaminantsof concern for completed and potential exposure pathways for potentially affected populations. Inthese evaluations, ATSDR considered the frequency and duration of the estimated exposures; forcases in which a population is affected by more than one exposure pathway, we also considered thecombinations of contaminants and exposure routes. This section also presents the potential healtheffects 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 thecontaminant-specific discussions.

Special Considerations of Women and Children

Women and children can sometimes be affected differently from the general population bycontaminants in the environment. Both tend to be smaller than the average person, which means theycan 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 exposuresexperienced by its mother, as well as exposure during pregnancy and lactation, can expose a fetus orinfant to chemicals through the placenta or in the mother's milk. Depending on the stage ofpregnancy, the nature of the chemical involved, and the dose of that chemical, fetal exposure canresult 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 differfrom an adult's exposure in many ways. A child drinks more fluids, eats more food, and breathesmore air per kilogram of body weight than an adult, and has a larger skin surface area in proportionto 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 behaviorsmay result in longer exposure durations and higher intake rates.

Children's metabolic pathways, especially in the first months after birth, are less developed thanthose 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 adultsare 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 highconcentrations of certain contaminants during this period. Also, young children have less ability toavoid hazards, because they lack knowledge and depend on adults for decisions that may affectchildren but not adults.

This public health assessment assesses risks to children exhibiting pica behavior (a craving forunnatural food like soil). Information on the incidence of soil pica behavior is limited. A studydescribed in an EPA document [117] showed that the incidence of soil pica behavior wasapproximately 16% among children from a rural black community in Mississippi. However, thisbehavior was described as a cultural practice among the community surveyed, so that communitymay not represent the general population. In five other studies, only one child out of more than 600ingested an amount of soil significantly greater than the range in other children. Although thesestudies did not include data for all populations and represented short-term ingestion only, it can beassumed 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, ormg/day) by children with soil pica behavior [117]. Ingestion rates between 1,000 and 10,000mg/day have been used to estimate exposure doses for pica children. In the PGDP public healthassessment, ATSDR assumed a soil ingestion rate of 2,000 mg/day for approximately 290 days peryear to represent pica behavior in children aged 1 to 3 years old. ATSDR believes that this is ahealth protective assumption and likely overestimated soil consumption.

In the following discussions, we will indicate whether women and children were, are, or may beexposed 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 theexposure pathways, exposure routes, affected population, and duration of exposure for eachcontaminant in a potential or completed exposure pathway. Contaminants that are only present inpotential exposure pathways are in italics. Note that exposure durations for metals in thegroundwater exposure pathway are assumed to be chronic (i.e., lasting 1 year or more): it is difficultto identify the specific numbers of years of exposure for the metals, because there have not beensufficient 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 musthave their pathway-specific exposure doses summed to represent a total dose. However, most of thecontaminants listed in Table 24 are not present in multiple exposure pathways. Of the 17contaminants listed, only arsenic, radioactive materials, thallium, uranium, and vanadium havemultiple pathways of exposure to the same population. The only population that could have beenexposed to these contaminants via more than one exposure pathway were pica children living withinthe 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 radioactivematerials together, because radiation doses from each isotope were summed to include a total dose topotentially exposed populations. Uranium, as a chemical toxin, is listed separately.

Table 25 gives an estimate of the number of people potentially exposed through each exposurepathway. Figure 7 shows the locations of those potential exposures. The number of personspotentially 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 peoplestopped 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 1980and 1990. This means that the 1990 Census data may underestimate the number of peoplepotentially exposed. (The people who left the area were most likely less than 65 years old, includinga few less than 6 years old.) Also, about 25 of these people have lived in this area since the plantbegan operation in 1952. (Refer to Appendix A.) Note that Table 25 does not include the surfacewater and biota exposure pathway: most people potentially exposed through that exposure pathwaywould 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 knowthe number or ages of hunters and fishers.)

It is important to remember that an exposed person would not necessarily experience adverse healtheffects. Tables 24 and 25 describe the potentially affected populations; they do not describe potentialhealth effects. The discussion of potential health effects for each contaminant are based on calculatedexposure doses for PGDP and documented health effects from human and animal studies. Specificcontaminants 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
Uranium Air Inhalation (Acute) Residents living less than 4 kilometers (2.5 miles) southeast of PGDP fence Past: 1960 accident

Potential future

Vanadium Groundwater



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
Children under 610-147-92-3
Women 15 to 44 years16-2012-152-3
People over 6512-148-93
Total 18 and older67-7249-5211-12
Total under 1823-2818-224-5
American Indian000
Source: [26]

Areas of Contamination and Potential Health Exposure
Figure 7. Areas of Contamination and Potential Health Exposure

Specific Substances


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 inindustry--mixed with other metals to form alloys or produced as antimony oxide. The alloys areused in lead storage batteries, solder, sheet and pipe metal, bearings, castings, ammunition, andpewter. 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 ofsoil (mg/kg) [?]. Concentrations of antimony were not uniformly distributed throughout off-siteareas. Instead, they were log-normally distributed, meaning that a few samples had highconcentrations while most had low concentrations. In fact, most off-site soil concentrations werebelow 5 mg/kg [45]. The highest concentration was found 2.5 miles (4 kilometers) northwest ofPGDP, at a location where wells were installed. (That sample may not be representative of surfacesoil 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 easternUnited States (less than 1 to 8.8 mg/kg [90]); it is also higher than the background concentrationreported for the PGDP area (0.21 mg/kg [119]).

ATSDR scientists used conservative assumptions to estimate exposure doses for exposure toantimony in off-site soil. The highest estimated exposure dose was 0.001 milligrams of antimony perkilogram 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 thespecific compound inhaled or ingested. Both gastrointestinal and pulmonary absorption, althoughgenerally low, are a function of compound solubility.

ATSDR has not developed a health guideline for ingestion of antimony, because available scientificstudies are lacking for this route of exposure [118]. EPA has developed a health guideline, called areference dose (RfD), for chronic oral exposure to antimony, which is 0.0004 mg/kg/day. Thereference dose is based on a lowest-observed-adverse-effect level (LOAEL) in rats, which hadshortened lifespans and changes in blood glucose levels after ingesting 0.35 mg/kg/day of antimonyin drinking water [120]. EPA derived the RfD by dividing the LOAEL for rats by an uncertaintyfactor of 1,000, because humans may be more sensitive than rats, some humans may be moresensitive than others, and there was no experimental level for rats where no adverse effects wereseen. Other studies in which rodents were exposed orally have reported effects on lifespan, glucoselevels, and cholesterol metabolism [118].

Acute exposure to antimony by humans who ingested antimony-contaminated lemonade (at anestimated 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 recoveredfrom this acute exposure within a few hours to several days [118,120]. One review of soil ingestionstudies 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 doseswere considerably lower than the lowest levels reported to cause adverse health effects in animalsand humans [118,120]. They were also lower than the acute toxicity screening level proposed forantimony [92]. Furthermore, we most likely overestimated actual doses, since we used extremelyconservative assumptions to estimate dose.

EPA's antimony health guideline is based on a drinking water study in rats. Antimony in soil isgenerally in a less soluble form than when it is in water. Consequently, people would absorb lessantimony from soil than from water. Even with conservative assumptions about exposure and rate ofabsorption from soil, exposure to antimony in off-site soils near PGDP is not expected to result inadverse health effects.


Exposures to arsenic in groundwater and potential exposures in off-site soil are not a publichealth hazard. Arsenic was also evaluated in surface water and was not identified as acontaminant of concern for that exposure pathway.

Arsenic is a naturally occurring element in our environment but additional arsenic often gets into theenvironment during copper and lead smelting, wood treating, and pesticide applications. It is in ourenvironment 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 tworesidential wells at a maximum concentration of 10 micrograms per liter of water (or 10 g/L).These wells were used for an unknown period of time in the past, possibly up to 35 years. ATSDR'sestimated 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 shownin Table 6).

Inorganic forms of arsenic predominate in groundwater (and soils) and are generally more toxic thanorganic forms [122]. When humans and other animals are exposed to inorganic arsenic, their bodieschange 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 1mg/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 thishappens, blood levels increase and adverse effects can occur. ATSDR's estimated doses forgroundwater and soil exposure pathways are lower than the levels needed to saturate detoxificationmechanisms in the body.

For the purposes of this Public Health Assessment, ATSDR assumed that arsenic in biota existed as100% inorganic arsenic. However, the predominant (80 to 99% of the total arsenic) forms of arsenicin 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 effectsof arsenic appear to have a threshold, or minimum effective dose. In addition, a growing body ofscientific evidence suggests that cancer may result from mechanisms other than direct attack ongenetic 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 skinand gastrointestinal effects--range from 0.014 to 0.05 mg/kg/day. These doses were estimated froma study of Taiwanese people who drank arsenic-contaminated water for 45 years [123,124]. EPAderived a health guideline of 0.0003 mg/kg/day based on skin effects (e.g., hyperpigmentation andkeratosis) and a cancer slope factor of 1.5 (mg/kg/day)-1 for skin cancer based on the Taiwanesestudy [120].

This study has limitations that one must consider when using it to evaluate public health hazard forPGDP 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-watersources of arsenic, genetic susceptibility, and poor nutritional status of the exposed population.Therefore, arsenic exposure may have been underestimated in the study, possibly leading tooverestimation of the number of new cancer cases predicted for incremental increase in exposuredose. Second, the cancer slope factor for arsenic is based on the conservative assumption that nothreshold exists for cancer. As discussed previously, arsenic carcinogenicity appears to have athreshold.

The amount of arsenic absorbed from the gastrointestinal tract or skin can vary widely; it dependslargely on the water solubility of the arsenic compounds (either organic or inorganic) present in theenvironment. 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 ofinorganic 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]. ATSDRscientists do not have specific information about the types of arsenic compounds (and theirsolubility) present in groundwater and soils off site of PGDP; therefore, we assumed for exposuredose calculations that all arsenic was water-soluble and 100% absorbed.

Despite these conservative assumptions, the estimated groundwater doses were lower than levelsshown to cause adverse effects in the Taiwanese study and considerably lower than levels requiredto saturate detoxification mechanisms in the body.

Arsenic was detected in off-site soil in the WKWMA, southwest of the PGDP security fence. Themaximum off-site concentration was 38 milligrams of arsenic per kilogram of soil. The normalrange of soil concentrations in the eastern United States is less than 0.1 to 73 mg/kg [121], andbackground for the Paducah area is reported as 12 mg/kg [119]. ATSDR's estimated dose for pastand current exposure to children who exhibit pica behavior (for the resident exposure scenario) was0.002 milligrams of arsenic per kilogram of body weight per day, which exceeded the healthguideline for chronic ingestion of arsenic. However, our estimated dose was lower than theprovisional 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 limitedextent (e.g., less than 50%) following ingestion. This is thought to be primarily because soils containarsenic in less-soluble forms [122]. More-soluble arsenic compounds may be 60% to 70% absorbedthrough 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 studiesof 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% ofarsenic was absorbed for either route of exposure. This assumption resulted in a dose estimate thatmost 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 Authorityproperty). Ingestion of water from this well is unlikely, because the well is a monitoring well onindustrial property. The analytical results for cadmium in residential wells were reported asnon-detects, but the detection limits were above ATSDR comparison values. However, exposuresestimated 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 coaland mineral fertilizers, contain some cadmium. Pure cadmium is a soft, silver-white metal. It is oftenfound as part of small particles in air. It does not have a distinct taste or smell; therefore, it is notpossible to taste or smell cadmium in water or air. In the United States most cadmium is extractedduring the production of other metals such as zinc, lead, and copper. It has many uses in industryand consumer products, mainly batteries, pigments, metal coatings, and plastics.

Food and cigarette smoke are the largest potential sources of cadmium exposure for members of thegeneral population. Average cadmium levels in U.S. foods range from 2 to 40 parts of cadmium perbillion 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 averagedietary intake of cadmium in adult Americans is about 0.0004 mg/kg/day; smokers receive anadditional amount--about 0.0004 mg/kg/day--from cigarettes [128].

Numerous studies indicate that the kidney is the main target organ of cadmium toxicity followingextended 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 numerousepidemiologic 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 tocadmium. The oral MRL is based on a lifetime accumulated threshold of 2,000 milligrams ofcadmium from dietary sources. The threshold is associated with kidney effects (proteinuria, orprotein 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 ingestionfrom food and water, respectively. The critical effect is significant proteinuria in humans chronicallyexposed to cadmium, using a no-observed-adverse-effect level (NOAEL) of 200 milligrams pergram (mg/g) wet weight in the renal cortex and a kinetic model assuming 2.5% or 5% absorptionfrom food or water, respectively, and 0.01% per day excretion [128].

A relevant consideration is whether the proteinuria caused by cadmium exposure should beconsidered an adverse effect. By itself, the increased excretion of low-molecular-weight proteins hasno adverse effect on health. However, several studies have indicated that increased excretion ofcalcium 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 onewell would have estimated exposure doses that could result in adverse health effects. This isunlikely, however, since the well was never used as a residential source and is located on anindustrial property.

There is a high degree of uncertainty surrounding the actual exposure doses for cadmium ingroundwater, given that samples from residential wells were below the detection limit. Even if weassume that cadmium was present at that detection limit in these residential wells, cadmium wouldnot pose a public health hazard.

Chromium, Hexavalent

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 thechromium metal--e.g., trivalent (III) chromium or hexavalent (VI) chromium. Chromium in theenvironment (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 bodiesuse insulin. Hexavalent chromium is considerably more toxic to humans than trivalent chromium.Hexavalent chromium is used in chrome plating, dye manufacturing, leather tanning, and woodpreservation, and was used as a corrosion inhibitor in the cooling towers at PGDP. Because themeasured groundwater analyses are not specific as to valence, we calculated exposure dosesassuming that measured concentrations are present as the more toxic hexavalent form.

Concentrations of chromium in the water from off-site groundwater monitoring wells ranged from40 to 270 g/L, which exceeded the comparison value of 30 g/L for hexavalent chromium.However, none of these samples were taken from residential drinking water wells. The maximumconcentration of chromium in residential wells was 20 g/L, which is lower than the comparisonvalue. Because not all residential wells were tested, ATSDR scientists assumed that maximum levelsin off-site wells near untested residential wells represented possible human exposure levels. Theestimated doses for ingestion of chromium in residential wells, assuming exposure to maximumconcentrations in nearby off-site wells, were 0.008 mg/kg/day for an adult and 0.021 mg/kg/day fora child. These doses exceeded health guidelines for hexavalent chromium. If the maximumconcentration measured in residential wells was used, the estimated doses would be 0.0006mg/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 ofpossible exposure doses (shown below) whose lower bound was maximum measured concentrationsin 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
Child0.002 mg/kg/day (or 0.03 mg/day) to 0.021 mg/kg/day (or 0.27 mg/day)
Adult0.0006 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 dataare insufficient or too contradictory to establish minimum levels of effect (e.g., LOAELs). Becausechromium is an essential nutrient in the body, the National Research Council has established a rangeof "estimated safe and adequate daily dietary intakes" (ESADDIs) for chromium. The range is 50 to200 micrograms (g) per day [135]. The upper end of this range, 200 g/day, has been adopted byATSDR 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 guidelineestablished by EPA for chronic ingestion of chromium VI. EPA's reference dose for chronic oralexposure, based on animal studies, is 0.003 mg/kg/day [120].

The estimated groundwater doses were slightly above ATSDR's interim guideline for "safe andadequate" intakes. As previously stated, these estimates are very conservative, because they werecalculated assuming exposure to maximum concentrations in wells near residential wells, ratherthan the residential wells themselves, and because they assumed that all chromium was present inthe (more toxic) hexavalent form. Exposure doses based on maximum concentrations measured inresidential wells are within the "safe and adequate" intake range. Therefore, ATSDR scientistsconclude that ingestion of chromium in off-site groundwater (drinking water) wells is not expectedto result in adverse human health effects.

Hydrogen Fluoride

Historically, chronic (long-term) exposures to hydrogen fluoride (HF) happened as a result ofreleases during normal process operations; acute (short-term) HF exposures happened as aresult 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 annualuranium hexafluoride releases and HF concentrations at the site perimeter. We calculatedexposure doses for potentially affected residents living north of PGDP (based on prevailing winddirections). 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. Themost serious accident (November 17,1960) created potential exposures to the southeast ofBuilding 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 bedetermined if that accident posed a public health hazard to an individual. Other accidentalreleases 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 isused as a catalyst, as a fluorinating agent, in making fluorine and aluminum fluoride, as an additivein 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, itis called hydrofluoric acid. Hydrofluoric acid is dangerous to humans, because it can burn the skinand eyes. At first, exposure to hydrofluoric acid may not look like a chemical burn. Skin may onlyappear 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 tissueloss can occur. In the worst cases, people who get a large amount of hydrofluoric acid on their skincan 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 healtheffects for breathing moderate amounts of HF for several months are not well known, but rats thatbreathed HF for several months suffered kidney damage and nervous system changes, such aslearning problems. If you breathe HF or fluoride-containing dust for several years, changes in yourbones (called skeletal fluorosis) can occur.

Studies have been conducted to determine if fluoride causes cancer in people who live in areas withfluoridated water or naturally high levels of fluoride in drinking water, or people who may beexposed to fluorides at work. The studies have not found an association between fluoride and cancerin people.

ATSDR's provisional screening value for intermediate exposure (15 to 364 days) is 0.010milligrams 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 ppbcomparison value for air is more than 100 times lower than exposures that caused mild irritation ofthe 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 HFin air for a 40-hour work week made up of 8-hour work days is 2.5 mg/m3 (3 parts per million, or3,000 ppb). The 12 ppb provisional screening value for air concentrations of HF is more than 250times lower than OSHA's occupational level.

Air releases of HF have occurred at the PGDP site. Because there is a strong correlation betweenuranium releases and ambient air concentrations of HF at this site, ATSDR assumed that the largestannual HF release coincided with the highest annual uranium release, which was in 1956. We usedthe estimated HF air concentration for 1956 to evaluate the health impacts of chronic exposure toHF 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 highestestimated annual average HF concentration in air (28 ppb for 1956) was at the "perimeter northmonitoring station." As such, the perimeter north monitoring station represents the point ofmaximum off-site exposure; however, no one lives at this location. The closest residence is about1,500 meters (almost a mile) from the source, about 500 meters (1,640 feet) from the perimeternorth monitoring station. The concentration of HF at the nearest residence was estimated atapproximately 22 ppb. The annual average concentrations for 1955 and 1956 are about two timesgreater than annual average concentrations for other years. If actual exposures to HF occurred at 22ppb, then mild adverse health effects may have resulted. Because our assumptions were soconservative, though, we believe that people were exposed to lower average air concentrations andno adverse health effects would have resulted. Additionally, it should be noted that the exposureassumptions and modeling used to estimate historical air levels were very conservative and mostlikely overestimated air concentrations. Past, current, and future long-term exposure to HFreleased 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. Theestimated 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 occupationalstandards--but occupational standards are not meant to protect sensitive populations (e.g., childrenand the elderly). If sensitive people were exposed at these levels, they may have experienced adversehealth effects (e.g., irritation of the eyes, nose, and throat). Because of the uncertainty associatedwith historical events (e.g., amounts of material released, modeling, location of off-site individualsduring accidents), past exposure to estimated maximum air concentrations poses an indeterminatepublic health hazard.


Past exposure to lead in three residential drinking water wells may have increased the likelihoodof neurological effects in young children, and thus posed a public health hazard. Currentexposure may still be occurring if the source of the lead was from pipes and plumbing asopposed to groundwater.

Lead is a naturally occurring element found in the earth's crust [89]. It is used in a variety ofproducts and industrial processes, which can release it into the environment. Lead can be introducedto 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 exposespeople to lead. Lead in soil can contaminate groundwater and surface water under certainenvironmental conditions. Pollution or use of lead solder in water delivery and household plumbingsystems can increase levels of lead in drinking water.

Lead was detected in groundwater near the site. Samples from 12 residential wells near PGDP hadconcentrations of lead ranging from 10 to 110 g/L [45]. Due to the locations of the wells with thehighest levels, lead did not appear to be related to PGDP. The lead concentration was 10 g/L innine of the residential wells, 100 g/L in one well, and 110 g/L in one well. There was one readingof 290 g/L in another residential well, but that reading could not be replicated; with the highreading included, concentrations in this single well averaged 103 g/L. Other off-site monitoringwells north of Ogden Landing Road near the North-South Diversion Ditch and southwest of the sitenear the inactive landfill had concentrations ranging from 10 g/L to 210 g/L. The highestconcentrations were near the drainage ditch north of the site. Most of the off-site wells were sampledfor lead only once. The background concentration of lead in groundwater for the PGDP area is 10g/L [119].

It has long been known that lead exposure can have harmful effects. Young children and fetuseshave 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 andchildren receive higher doses from any given level of environmental lead than do adults, becausethey have a greater absorption capacity for lead than adults,. Therefore, age is an importantdeterminant 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 liter0.00050.002
40 micrograms per liter0.0010.004
60 micrograms per liter0.0020.006
100 micrograms per liter0.0030.01
290 micrograms per liter0.010.02

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 between0.001 and 0.01 mg/kg/day produce minor changes in blood cells. Harmful effects in animals areseen 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 thatmay be seen. (This is illustrated in Figure 8, below.) Blood levels of lead can be elevated bysustained exposure to contaminated soil, food, air, or drinking water. Neurological effects are themost 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 andPrevention considers a child to have an elevated blood lead level if the amount of lead in his or herblood is 10 micrograms per deciliter (g/dL) or higher [137].

Effects of lead on children and adults - Lowest-observed-adverse-effect levelsThe 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 tolead at a concentration of 0.1 micrograms percubic meter (g/m3) in air, 200 microgramsper gram (g/g) of soil and dust, and from 2.4to 3.4 micrograms per day(g/day)--depending on age--in the diet[138]. We assumed exposure to childrenbecause they are particularly sensitive to theadverse effects of lead [89,136]. Adults, including pregnant women, were not and are not likely tohave elevated blood lead levels if they were exposed to the mean residential well waterconcentrations.

Whether we estimated blood lead levels from the model, or from slope factors, we found thatchildren drinking water from wells with lead concentrations less than 60 g/L were not likely toexperience adverse health effects from exposure. Water from the three wells containingapproximately 100 g/L could have raised blood levels above the action level of 10 g/dL inchildren under 4 years old while the wells were in use. Therefore, we conclude that blood levels inthe 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 ahalf-life of 25 days) [136]. However, some of the lead in blood can be taken up by the bones andremain there for decades [136]. Bone lead can be a source of blood lead under conditions that mightcause bone desorption, such as pregnancy, poor diet, or older age [136]. We recommend thatresidents 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). Theresidential exposure scenario had an estimated exposure dose for a child with pica behavior (achild who exhibits an abnormal appetite for soil) that exceeded ATSDR's screening value. Theestimated exposure doses for an adult and normal child were below ATSDR's screening value.Based on conservative exposure assumptions, ATSDR believes that manganese exposure dosesfrom off-site soil is not a public health hazard.

Manganese is a naturally occurring substance found in many types of rock. Pure manganese is asilver-colored metal, somewhat like iron in its physical and chemical properties. Manganese does notoccur in the environment as pure metal. Rather, it occurs combined with other chemicals, such asoxygen, sulfur, and chlorine.

Rocks containing high levels of manganese compounds are mined and used to produce manganesemetal, which is mixed with iron to make various types of steel. Some manganese compounds areused in batteries, ceramics, pesticides, and fertilizers; and in dietary supplements.

Ingesting a small amount of manganese each day is important in maintaining your health. Theamount of manganese in a normal human diet (about 2 to 9 mg/day) seems to be enough to meet aperson's daily need; however, no cases of illness from eating too little manganese have been reportedin humans. In animals, eating too little manganese can interfere with normal growth, boneformation, and reproduction.

Too much manganese can cause serious illness. Although there are some differences betweendifferent 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 andemotional disturbances, and body movements may become slow and clumsy. This combination ofsymptoms is a disease called manganism. Workers usually do not develop symptoms unless theyhave been exposed for many months or years at high levels. Manganism occurs because too muchmanganese permanently injures a part of the brain that helps control body movements. It is notcertain 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 averageingestion rate of 3.8 mg/day. Other estimates of daily intake for adults range from 2.0 to 8.8milligrams. Even though gastrointestinal absorption of manganese is low (3% to 5%), oral exposureis also the primary source of absorbed manganese [139].

Manganese intake among individuals varies greatly, depending upon dietary habits. For example, anaverage cup of tea may contain 0.4 and 1.3 milligrams of manganese [139]. Thus, someone whodrinks three cups of tea per day might receive up to 4 mg/day from this source alone, doubling his orher the average intake.

The Food and Nutrition Board of the National Research Council estimated the adequate and safeintake of manganese for adults at 2.5 to 5 mg/day [139]. It is possible that a significant proportion ofAmericans, especially women, are not consuming sufficient manganese, although no cases ofmanganese deficiency have been documented in humans. However, infants may be ingesting morethan the estimated safe and adequate dose for their age group (which is 0.7 to 1.0 mg/day), due tohigh 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 ormore) to manganese in soil. EPA has derived a chronic oral RfD of 0.14 mg/kg/day for manganesein the diet [140]. This value is equal to the average daily intake of manganese in the diet (10mg/day) that is considered adequate and safe. The RfD was derived assuming an average bodyweight of 70 kilograms. An uncertainty factor was not employed, because (1) the information usedto determine the RfD was taken from many large populations, (2) humans exert an efficienthomeostatic control over manganese such that body burdens are kept constant through variations indiet, (3) there are no sub-populations that are believed to be more sensitive to manganese at thislevel, and (4) manganese is an essential element, required for normal human growth andmaintenance of health.

When assessing exposure to manganese from drinking water or soil, EPA recommends, one shoulduse a modifying factor (an uncertainty factor based on professional judgement) of 3, based on someevidence that infants younger than 28 days have a higher uptake of manganese in liquids, excreteless absorbed manganese, and, as neonates, pass the absorbed manganese more easily through theblood-brain barrier. The resulting chronic oral RfD for manganese in water and soil would be 0.05mg/kg/day. The estimated exposure dose for a pica child is 0.1 mg/kg/day (assuming ingestion of 2grams of soil per day for 290 days per year). However, if one assumes that manganese in soilbehaves similarly to manganese in food (i.e., that its bioavailability is similar), then a comparisonvalue at or near 0.14 mg/kg/day would be deemed more appropriate, and the estimated exposuredose for a pica child would not exceed this value.

Nitrates and Nitrites

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 iseasily oxidized into nitrate, nitrate is the form that is typically found in groundwater and surfacewater. Nitrate is the primary source of nitrogen for plants. Wastes containing organic nitrogen aredecomposed in soil or water by bacteria to form ammonia. Ammonia is then oxidized to nitrite andnitrate. Agricultural and residential use of nitrogen-based fertilizers, nitrogenous wastes fromlivestock and poultry production, and urban sewage treatment systems have increased levels ofnitrate in soil and water. Certain plants (cauliflower, spinach, collard greens, broccoli, carrots, andother root vegetables) have a naturally higher nitrate content than other plant foods and can accountfor a large percentage of nitrate in the diet. Nitrate and nitrite compounds are also used for colorenhancement and preservation of processed meat products. Nitrate is used in foods to preventbotulism, 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 transportedthrough storm runoff or groundwater recharge. The most common route of exposure occurs throughdrinking contaminated water, eating vegetables with naturally high levels of nitrate, and eating foodspreserved with nitrate.

Nitrate was detected in off-site groundwater (in RW-002) once used for residential purposes at amaximum concentration of 29.2 milligrams per liter (mg/L) as total nitrate (NO3). ATSDR believesthat no one (not even infants or children) would have experienced adverse health effects fromexposure through drinking water, even if they consumed nitrate-impacted drinking water at themaximum concentration detected. Nitrate is not now present in residential wells and should not bepresent in the future. It should be noted that nitrate was detected in surface water at a maximumconcentration of 84.6 mg/L as NO3. If people consumed the contaminated surface water at themaximum detected level on a regular basis for an extended period of time, they might experienceadverse health effects. However, this exposure scenario is very unlikely.

ATSDR has developed Reference Dose Media Evaluation Guides (RMEGs) for chronic (1 year ormore) oral exposure to nitrate in water. Media concentrations less than the RMEG are unlikely topose a health threat. The chronic RMEGs for a child are 20 mg/L for nitrate-nitrogen (NO3-N) and90 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 MaximumContaminant Level Goal, or MCLG (10 mg/L for NO3-N) as a guideline to evaluate potential infantexposure.

RMEGs are media-specific chemical comparison values derived from EPA's RfDs. RfDs arehealth-based guidelines for non-cancer effects. An RfD is an estimate of the amount of a chemicalthat a person can be exposed to, on a daily basis, that is not anticipated to cause adverse healtheffects over a person's lifetime. MCLGs, which EPA sets after reviewing health effects studies, arethe maximum levels of contaminants in drinking water at which no known or anticipated adverseeffect on the health of persons would occur, and that allow an adequate margin of safety. MCLGsare non-enforceable public health goals. When determining an MCLG, EPA considers the risk thatsensitive sub-populations (infants, children, the elderly, and those with compromised immunesystems) will experience various adverse health effects. For chemicals that can cause adversenon-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 10mg/L. (This regulation does not cover private wells.) If the results of a water analysis are reported asNO3 (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 biologicalconversion 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 aremore sensitive to nitrate for several reasons. They consume more water relative to their body weightthan adults, and the hemoglobin in an infant's blood (called fetal hemoglobin) is more easilychanged into methemoglobin than an adult's hemoglobin. Also, an infant's digestive system is lessacidic, which enhances the conversion of nitrate to nitrite. The two most common symptoms relatedto 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. Illnessand death caused by methemoglobinemia are not always recognized, so methemoglobinemia'soccurrence may be under-reported.

Families with infants should use an alternate water supply if their well is known to contain elevatedlevels of nitrate. When preparing infant formula, families should use nitrate-free water. If a privatewell 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 causedsymptomatic methemoglobinemia in children.

Nitrates can react with other substances to form N-nitroso compounds. Some of these N-nitrosocompounds 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 ofnitrate or nitrite causes cancer.

Based on the information presented above, nitrate concentrations detected in off-site groundwater arenot 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 inresidential wells (approximately 50 g/L) was five times higher than ATSDR's comparison value(10 g/L). Even if concentrations are assumed to be 50 g/L, the resulting exposures are not apublic health hazard.

Pentachlorophenol is a man-made substance that was used widely as a pesticide, herbicide, andwood preserver [141]. Pentachlorophenol by itself is slightly water soluble. However, technical-grade pentachlorophenol that is used as a pesticide or wood preserver typically contains othercontaminants, such as chlorinated dibenzodioxins, that are not as soluble. One of these chlorinateddibenzodioxins, octochlorodibenzodioxin (OCDD), is 189 million times less soluble in water thanpentachlorophenol [141,142]. Environmental contamination at most industrial sites containstechnical-grade as opposed to pure-grade pentachlorophenol. When waste technical-gradepentachlorophenol seeps into the soil and migrates downward toward the groundwater, OCDDcomes 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 ofsoil on site, but are not detected in samples taken at depths greater than 3 feet [45]. In order forpentachlorophenol in soil to reach groundwater under PGDP, it must travel through 30 to 100 feetof silt and clay; by then, it is essentially free of less-soluble dioxin contaminants, which have sorbedto soils. Therefore, pentachlorophenol in groundwater is essentially the same as pure-gradepentachlorophenol.

Pentachlorophenol was detected in one off-site monitoring well, at a maximum concentration of 8 g/L. It was not detected in any off-site residential wells; however, the well sampling could notdetect pentachlorophenol at concentrations below 50 g/L, which is higher than the comparisonvalue used to select contaminants of concern. ATSDR scientists used this detection limit to estimateexposure 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 oralexposure to pentachlorophenol. This guideline is based on observations of increased serum levels ofliver enzymes in rats, which is considered suggestive of liver toxicity [141]. When rats were givenfood contaminated with either technical-grade or pure pentachlorophenol, those receiving 1 to 25mg/kg/day of technical-grade product showed signs of liver injury. ATSDR based its healthguideline on the lowest dose (1.2 mg/kg/day) of technical-grade pentachlorophenol shown to causeliver injury, because it is likely that most hazardous waste sites contain technical-grade as opposedto pure pentachlorophenol. We applied an uncertainty factor of 1,000 to this lowest dose, becausehumans may be more sensitive to pentachlorophenol than rats, because some humans are moresensitive than others, and because the animal study involved intermediate-duration (rather thanchronic) exposure. The only effect caused by the pure pentachlorophenol in this study was anincreased 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 drinkingwater, we could derive a tentative, site-specific oral health guideline for chronic duration based onthe highest dose (5 mg/kg/day) that failed to cause liver injury in rats. If, as above, we divided by anuncertainty factor of 1,000 to account for differences in sensitivity and exposure duration, thistentative health guideline would be 0.005 mg/kg/day. ATSDR's estimated exposure doses are equalto or lower than this health guideline. Exposure to pentachlorophenol in groundwater at thedetection 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 andpentachlorophenol containing lower levels of dioxins than technical-grade. The doses required toproduce cancers in these studies were at least 3,000 times higher than the maximum doses ATSDRestimated for ingestion of drinking water from the residence near PGDP [120]. The types of livertumor observed in these rats are also associated with dioxin exposure; the hemangiosarcomas arenot.

From this information, EPA derived a cancer slope factor of 0.12 mg/kg/day based on all tumorscombined [120]. However, there is no clear evidence from high occupational exposures thatpentachlorophenol causes cancer in humans [141]. Therefore, there is even less likelihood that lowerenvironmental exposures could produce these effects. There is also no evidence of humanangiosarcomas among people exposed to pentachlorophenol [141]. Even if we assume that cancer isa possibility for humans, and we consider maximum estimated exposure doses to be equal to theresidential well detection limit, cancer effects are not likely for people who may have ingestedpentachlorophenol-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.

Polychlorinated Biphenyls

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 individualchemicals, 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 intransformers, 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, scientistsfound PCBs in ambient air, soil, water, and sediment, even though there are no known naturalsources of PCBs in the environment. EPA banned the production of PCBs in 1978. Traces of PCBscan still be found in the tissues of wildlife, domestic animals, and people--PCBs have chemical andphysical properties that make them persistent in the environment and readily accumulate in the fattytissues 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) stillcontain PCBs. Old electrical appliances may release PCBs when they get hot and contaminate insideair. 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. PCBspresent in sediment may enter the aquatic food chain and smaller fish, which in turn, become PCBsources for larger fish. Birds and land predators, such as man, may be exposed to PCBs when theyeat contaminated biota. At each step in the "food chain," PCBs that have accumulated in theanimals' fatty tissues can appear in greater concentration, or "bioconcentrate," in the species that eatthem. PCBs were found in fish sampled from several locations in Little Bayou Creek, and to a muchlesser extent in fish sampled from Bayou Creek. PCB levels in deer tissue were extremely low anddo not pose any threat. More recent (1997) samples from deer have been below the detection limit inmultiple tissues (muscle, liver, fat, and mammary).

Kentucky has issued a health advisory regarding consumption of specific species of fish from LittleBayou Creek. The PGDP 1989 Environmental Report indicated that total PCB concentrations infish from Little Bayou Creek averaged approximately 5 micrograms of PCB per gram of fillet(g/g); see Table 18A(2). The highest average total PCB concentrations (17.95 g/g) were reportedin sunfish collected from Outfall #11, which is part of the Little Bayou Creek area. The total PCBconcentrations in fish tissue from Outfall #11, based on samples from three sunfish, were more thanthree 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 wecalculated the average total PCB concentration for the Little Bayou Creek area. Total PCBconcentrations in fish from Bayou Creek were approximately five times lower (averageapproximately 1 g/g) than fish from Little Bayou Creek.

In 1993 and 1994, fish tissue samples from Bayou Creek and Little Bayou Creek indicated thatconcentrations had decreased since 1989: they were about 10 times lower than 1989 reportedvalues. Additionally, total PCB levels in fish tissue from Bayou Creek (0.143 g/g) were about fourtimes lower than concentrations detected in fish tissue from Little Bayou Creek (0.553 g/g).Background samples, from Hinds Creek, did not contain detectable levels of PCBs. In 1993, 40% ofthe fish sampled from Bayou Creek did not contain detectable levels of PCBs. In 1994, that numberwas 20% (fewer fish were sampled in 1994, which may account for the difference). Also, in 1993and 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 BayouCreek and analyzed them for levels of PCBs in fillets. The average concentration of total PCBs(0.561 g/g) in fish tissue from Little Bayou Creek was similar to the 1993-1994 results. Two ofthe twenty fish sampled did not have detectable levels of PCBs in fillets. Fish tissue results were notavailable for Bayou Creek in 1997.

ATSDR evaluated whether adults and children eating fish from either Bayou Creek or the LittleBayou Creek system could obtain PCB doses that would cause adverse health effects. ATSDRassumed that subsistence and recreational anglers got 20% of their total fish intake from the creeksfor 30 years. The estimated exposure dose for children (assumed to be the children ofsubsistence/recreational anglers) was based on a 6-year exposure duration. The consumption rate forrecreational 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 to150 meals per year at 150 grams (5.3 ounces) per meal. A recreational angler's child was assumedto consume 3 g/day, which equates to 20 meals per year at 50 grams (1.8 ounces) per meal. Asubsistence angler's child was assumed to consume 8 g/day--equal to 150 meals per year at 50grams (1.8 ounces) per meal. If a fish tissue sample was below the detection limit, we used thedetection limit as the measured value for total averaged values. This is a conservative approach thatcould 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 sucheffects are quite high: daily, occupational exposure doses ranging from 0.07 and 0.14 mg/kg/dayfailed to produce adverse health effects in workers [101]. Reports of developmental effects fromlower 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. Inlaboratory animals, PCBs have been shown to produce skin effects (similar to those seen in peopleexposed 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 rhesusmonkey appears to be the most sensitive [120]. PCBs have been shown to impair the monkey'simmune system (in addition to producing skin, fingernail, and toenail effects), at doses as low as0.005 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 adverseeffects in monkeys [101,120].

Several human studies have reported that low level environmental PCB exposure during in utero orneonatal 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 weightand maternal smoking, alcohol, and drug use). These studies suggest, but do not prove, anassociation between prenatal/neonatal exposures to PCBs and neurologic, immunologic anddevelopmental effects in young children. Therefore, it cannot be equivocally determined whetherlow level environmental PCB exposures affect prenatal or neonatal development.

Rats are the only laboratory species shown to develop cancer after ingesting PCBs [101,120]. Theanimals 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 3.0mg/kg/day (which is 100 to 1,000 times higher than the estimated potential doses to children andadults in the PGDP area). In order to use animal data to predict whether humans are likely todevelop cancer, we often assume that the relationship between PCB dose (administered) and cancerdevelopment is the same at high and low doses. We also assume that there is no dose at which thereis no risk for cancer development. Many scientists believe that these assumptions are valid forsubstances 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 directlyattacking 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 makeassumptions about the ways humans resemble or differ from the animal "models." EPA's standardmethodology 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 thelikelihood of cancer developing per unit dose (measured in mg/kg/day) [151].

One approach, called physiologically based pharmacokinetic (PBPK) modeling, incorporatesinformation about how a substance and its degradation products are chemically modified anddistributed throughout the body following exposure. When PBPK models were used to compare howdifferent 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 speciesother than rats, not even in the sensitive rhesus monkeys, following exposure to PCBs. Also, PBPKmodeling may explain the lack of conclusive reports of cancer in multiple studies of workersoccupationally exposed to PCBs [101,120].

A recent study of more than 7,000 capacitor workers reported exposures to PCB air concentrationsas high as 1,500 g/m3. Workers in this study were employed for at least 3 months, and their healthstatus 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. Usingstandard EPA methods to predict the likelihood of cancer at these doses, one would have expected tosee additional cancers among 1,687 workers who received the highest PCB exposures. However, thestudy found no excess cancers of the liver or any other organ [143]. The estimated occupationalexposure doses that failed to produce detectable increases in liver cancer were more than 5 to 185times higher than the lifetime exposure doses estimated for subsistence and recreational fishers whocould 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 fromBayou 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 mediaat PGDP. The cumulative radiation dose from potential chronic exposure to those media is not apublic health hazard. Potential acute exposures from an accident in 1960 are an indeterminatehealth 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 orthe U.S. Atomic Energy Commission). For media in which the off-site concentrations were notreported, ATSDR estimated the concentrations by using computer models. The predominantradioactive 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 insome cases; however, these other radioactive contaminants contributed approximately 10% or lessto the exposures doses.

Table 28 lists maximum estimated annual committed effective doses for children and adults indifferent 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. Currentexposure 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 wereconsidered, as were organ doses. The estimated radiation exposure dose from all media for any yearof plant operation does not exceed 500 millirems (mrem), or 5 millisieverts (mSv), except for apotential 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 (1mSv) above background [152]. Before 1990, the ICRP recommendation was 500 mrem (5 mSv) peryear. Although the ICRP recommendations were lowered for chronic exposure over a 70-year lifespan, 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 mediafrom 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 thetime 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 lungdiseases. However, these workers were chronically exposed to higher levels of insoluble uraniumthan estimated exposure doses calculated for past accidents. Animal studies on rats investigatedacute exposures to uranyl nitrate (a more-soluble form) and reported an increased frequency of lungtumors and osteosarcomas. However, the doses in these studies were substantially higher than theestimated exposure doses from the 1960 accident and the experiment did not provide enoughinformation for confident extrapolation of risk coefficients to humans [153]. Because of theuncertainties in the release quantities and whether the airborne exposure pathway was completeduring this accident, ATSDR scientists concluded that the 1960 accident posed an indeterminatehealth hazard. If an individual was exposed to the maximum estimated exposure and using EPA'scancer 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 isexposed to low levels of thallium through eating, smoking tobacco, and breathing second-handtobacco 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 pregnantwomen and be distributed to the developing fetus.

Thallium was detected in surface water near PGDP. The maximum thallium concentration insurface water was 5,260 g/L in Bayou Creek near the inactive southwest landfill [45]. Using thismaximum concentration, we estimated that incidental ingestion of water from Bayou Creek wouldresult in an exposure dose of 0.001 mg/kg/day for adults and 0.002 mg/kg/day for children 1 to 6years old.

Thallium was not found in drinking water wells, but the lowest level of analytical detection was 10g/L--higher than EPA's drinking water standard of 2 g/L [155]. Therefore, we used the detectionlimit of 10 g/L to estimate exposure doses. This gave us doses of 0.0003 mg/kg/day for an adultand 0.001 mg/kg/day for a child, assuming that these residential wells were the sole source ofdrinking water.

ATSDR has no health guideline for ingestion of thallium. EPA has RfDs for several thalliumcompounds. Each RfD covers a particular compound and is based on animal studies for thatcompound. For example, the RfD for thallium sulfate is based on a failure to observe harmfuleffects in rats that were administered as much as 0.25 mg/kg/day of thallium by gavage (stomachtube). EPA divided this number by an uncertainty factor of 3,000 to account for humans being moresensitive than rats to thallium, for some humans being more sensitive than others, and for a lack ofchronic 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 higherthan the maximum exposure dose that ATSDR estimated for surface water or groundwateringestion, despite the fact that we used very conservative assumptions to estimate dose. If morerealistic 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 maximallycontaminated water a month for 6 years and an adult ingests this amount for 30 years. It is probablynot very likely that a young child, who is under constant care by an adult, would consume thesequantities of surface water at this maximum concentration. Likewise, it is unlikely that an adultwould 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 ofanalytical 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 Creekor from drinking water wells located near PGDP is not expected to result in adverse human healtheffects.


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 theexposure pathway is incomplete.

TCE is a nonflammable, oily, colorless liquid that has a sweet odor and a sweet, burning taste. Yearsago, TCE was used as an anaesthetic. It is now used as a solvent to remove grease from metal partsand to make other chemicals. It is heavier than water and has low solubility (up to one part TCE perthousand parts of water at room temperature) [156]. These qualities make TCE a troublesomecontaminant at hazardous waste sites.When present in groundwater, TCE tends to settle into a layerat the bottom of the aquifer and then continuously dissolves into the groundwater. This may result inhigh 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 wellwater.

TCE contamination of groundwater beneath the PGDP facility and in nearby private wells wasdiscovered in August 1988. TCE was detected at concentrations above ATSDR's comparison valuein four off-site residential wells. Maximum levels of TCE, ranging from 20 to 43 g/L, were foundin three of the wells (RW-004, RW-017, RW-113); a maximum level of 960 g/L was found in afourth well (RW-002). As a result of this sampling, the Department of Energy (DOE) immediatelyprovided bottled water to residents with contaminated well water--until they could be supplied withmunicipal water--and completely discontinued private well use.

Groundwater sampling was not conducted before 1988. Sampling conducted after 1988, when thewells were no longer used for drinking, revealed higher levels of TCE than were first detected in1988. This finding was not unexpected, considering the results of groundwater modeling ofcontaminant movement from sources on site. Modeling results indicate that levels of TCE before1988 were likely to have been lower than levels detected in 1988. ATSDR scientists cannotdetermine 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 thesewells are used in the future, or if new wells are drilled into the plumes, the residents could beexposed 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]. Humanhealth 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, nostudies provide clear, unequivocal evidence that exposure is linked to increased cancer risk inhumans [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 potentiallycarcinogenic chemicals. There is, however, sufficient evidence that TCE exposure results in cancerdevelopment in animals, although animal studies may not be relevant for evaluating health hazard tohumans [156,158].

In 1989, EPA withdrew its cancer assessment for TCE, which was based primarily on animal studiesconducted in 1990 and earlier, because more recent pharmacokinetic and mechanistic data for TCEbecame available [120,159]. An updated approach to TCE cancer assessment using existing animaldata and state-of-the-science papers has been proposed [159]. The approach, though high-doseanimal studies support it, does not appear entirely relevant for evaluating health hazard from humanenvironmental exposure. There are several reasons for this. First, cancer in animals appears to resultfrom species-specific mechanisms that are not entirely relevant to humans [156,158]. Second, theanimals used in these studies were exposed to very high doses of TCE, several orders of magnitudehigher 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 beenmore susceptible to adverse effects from TCE exposure (like infections and illnesses) than healthyanimals. Third, the overall findings from animal studies are inconsistent: some studies report anincreased incidence of cancer, while an equal number do not report an increase at similar levels ofexposure [156]. Fourth, the studies did not evaluate the effect of exposure to stabilizers andimpurities in TCE; these things may also be carcinogenic. For these reasons, ATSDR scientistsdecided to focus on non-cancer effects of TCE.

ATSDR derived a health guideline of 0.1 ppm for intermediate-duration (15 to 364 days) exposureto TCE by inhalation. This guideline, equivalent to 0.15 mg/kg/day, is based on neurological andcardiac effects (e.g., decreased wakefulness and decreased post-exposure heart rate) in rats. Thelowest dose that produced these effects was 50 ppm in air, which is equivalent to 77 mg/kg/day. Theestimated 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 thelowest 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 pupsadministered 50 mg/kg/day of TCE [160]. In this study, the TCE was dissolved in oil andadministered by stomach tube (gavage) [156]. The findings of this study are not entirely relevant forevaluating 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 PGDPresidents were likely to have been exposed to TCE in drinking water several times a day. (The bodyhandles a single large dose much differently than it does a series of smaller doses.) Second, the totaldose entering the body is higher and maintained for a longer time when TCE is dissolved in oil thanwhen 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 aperiod of 5 to 15 years. Despite these limitations, the findings are supported by other animal andhuman studies.

ATSDR's TCE Sub-Registry reports an excessive number of children aged 9 years old or youngerwith speech and hearing deficits [161]. Although the exposure levels of these children were not wellcharacterized, the findings support the types of outcome seen in animals. Several studies of workersand community residents suggest a possible association between exposure to TCE (and otherchemicals) and developmental outcomes [157,162,163,164]. However, none of the studies provideconclusive evidence for a causal relationship, largely because information about TCE exposure wasincomplete and exposure to other chemicals was likely [156,158].

Collectively, the scientific data indicate that the developing nervous system in young animals andhumans may be sensitive to the toxic effects of TCE [156]. It is not clear whether past exposures toTCE by PGDP residents with contaminated wells were sufficient to result in similar outcomes. Inorder to be protective of the most sensitive individuals, ATSDR concludes that past exposure tocontaminated water from well RW-002 may have resulted in neurological effects in childrenchronically using this water prior to 1989.


Short-term (1-hour average) off-site uranium air concentrations were modeled for the timeperiod corresponding to the 1960 accidental release. Estimated levels were above ATSDR'sintermediate comparison values and occupational standards at the nearest residence (southeastof Building C-333). If people were exposed to the estimated air concentrations, they could haveexperienced 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 ifthis 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 theprocess buildings to accelerate the reduction of UF6 concentration in the process gas system inorder to perform maintenance and inspection on process gas equipment. These releases, called"midnight negatives", potentially contained significant quantities of uranium (and hydrogenfluoride); 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 aresult of elevated operational emissions. Because the prevailing winds were from the south andsouthwest, the primary exposed population was residents living north and northeast of thefacility. This population may also have some exposure to uranium via soil and/or groundwaterexposure pathways. Even if chronic exposure to air, soil, and groundwater occurredsimultaneously, 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 ourenvironment, called background radiation. The amount of uranium in drinking water in the UnitedStates is generally less than 1 picocurie per liter (or approximately 1.5 g/L) [165].

Natural uranium, enriched uranium, and depleted uranium are mixtures of primarily three uraniumisotopes (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. Enricheduranium is more than 0.72% U-235 by weight, and depleted uranium is less than 0.72% U-235 byweight. All three isotopes are radioactive but have different specific activities (that is, radioactivityper 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 consideredseparately.) Because natural uranium produces very little radioactivity, the chemical effects ofuranium are generally more harmful than the radioactive effects. However, more radioactivemixtures (like enriched uranium) can harm the kidney more than natural uranium due to thecombined effects of chemical and radioactive properties.

The kidney is the primary target organ for the chemical effects of ingested and inhaled uranium. Theextent of toxicity is determined primarily by exposure route, type of uranium compound, andsolubility of that compound. Ingested uranium compounds are generally less toxic to the kidneysthan inhaled uranium compounds, partly because uranium is poorly absorbed from the intestinaltract. Highly soluble uranium compounds are generally more toxic to the kidneys than less-solublecompounds 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 allexposure routes (inhalation, ingestion, and dermal).

Studies using laboratory animals provide most of the evidence for kidney toxicity. ATSDR hasestablished intermediate (15 to 364 days) exposure health guidelines for inhalation of both solubleand insoluble uranium compounds. The guideline for insoluble uranium is 8 x 10-3 mg/m3. Thisguideline is based on structural changes (lesions) in kidneys of dogs exposed to uranium dioxidedust 6 hours a day, 6 days a week, for 5 weeks [166]. The health guideline for inhalation of solubleuranium is 4 x 10-4 mg/m3, based on kidney lesions in dogs exposed to uranium chloride in air 6hours a day, 6 days a week, for 1 year [167]. Neither study provided information about the size ofthe uranium particles used, so ATSDR based its guideline on the conservative assumption thaturanium 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-basedguideline for inhalation. On-site air concentrations would have been even higher, though it isuncertain whether on-site personnel were exposed to elevated air concentrations. ATSDR has notderived health-based guidelines for acute exposure. The estimated off-site air concentration exceededthe occupational standards for soluble and insoluble uranium compounds. If people were actuallyexposed to the estimated air concentrations, then a public health hazard existed. ATSDR does notbelieve 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 symptomsassociated with this accident (for either uranium or HF). Because of the lack of exposureinformation and considering that concentrations were derived from air dispersion modeling, weconclude 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 ismostly used as an alloying agent in steel production, although small amounts are also used inrubber, plastics, and ceramics [168]. Vanadium is a metallic element that occurs in six oxidationstates and numerous inorganic compounds. Vanadium's toxicity depends on its physical andchemical state, particularly on its valence state and solubility. Vanadium is poorly absorbed throughthe gut, but more readily absorbed through the lungs.

Vanadium was detected at a maximum concentration of 20 g/L in one off-site residential (drinkingwater) well. Although this concentration is lower than ATSDR's comparison value of 30 g/L, weselected vanadium as a contaminant of concern because maximum concentrations detected in off-sitemonitoring wells located near residential wells were as high as 210 g/L. ATSDR scientists usedthe maximum groundwater concentration (210 g/L) to estimate exposure (ingestion) doses of0.006 mg/kg/day for adults and 0.02 mg/kg/day for children. If we had used the concentrationsmeasured in residential wells, the estimated doses would have been an order of magnitudelower--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 astudy where rats were administered vanadium in their drinking water for 3 months [168]. In thisstudy, 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 wasconsidered the NOAEL, and is the basis for ATSDR's health guideline. The NOAEL for rats (0.3mg/kg/day) was divided by an uncertainty factor of 100, because humans are presumed to be moresensitive than rats to vanadium and because some humans are more sensitive than others. Theuncertainty factors may be overly conservative: scientific information suggests that humans areactually less sensitive than rats to ingested vanadium. Human volunteers who swallowed amaximum dose of 1.3 mg/kg/day of vanadium for 45 to 68 days showed no effects when tested forinjury to their liver, blood cells, or kidneys [168]. An adult would have to drink more than 4,500liters a day of water contaminated at the highest level in the residential well to take in the amount ofvanadium that did not cause adverse effects in the human volunteers. A child would have to take in650 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 averagevanadium content of soils in the United States is 200 mg/kg; vanadium seems to be most abundantin 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 picachild) 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 picachildren 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 theimplication that humans are less sensitive to ingested vanadium than rats, ATSDR does not expectadverse 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 soilis not expected to result in adverse human health effects for past, current, or future exposure tochildren or adults.

Vinyl Chloride

Potential past exposure to vinyl chloride in residential drinking water is an indeterminate publichealth hazard. It is not known whether anyone was exposed or at what levels due to inadequatedetection limits. No public health hazard currently exists, because no one is using theseresidential wells.

Vinyl chloride is a man-made substance used in the production of polyvinyl chloride (PVC) andother 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 increasedowngradient, depending on a number of factors, including the chemical characteristics of the soilthrough 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 onemonitoring well used to test for off-site groundwater contamination in the PGDP area. Themaximum concentration of vinyl chloride in this well was 110 g/L. The test well was located nearfour residential wells that were not found to contain vinyl chloride; however, the lower limits ofanalytical detection for these well samples were higher than EPA's Maximum Contaminant Level(MCL) of 2 g/L for public drinking water supplies. In addition, very few residential well watersamples (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-0021-500 g/L (four samples)
RW-0042-10 g/L (four samples)
RW-0174-10 g/L (three samples)
RW-11310 g/L (one sample)

The detection limit for RW-002 was 500 g/L on October 24, 1989, but the detection limit for thiswell was 1 g/L on August 14, 1990. Therefore, vinyl chloride was probably not a problem whenRW-002 was being used; however, there is some uncertainty due to variations in the TCE plumeconcentrations from seasonal factors. The lowest detection limits for wells RW-017 and RW-113were 4 and 10 g/L, respectively. Both values are above the MCL.

ATSDR's estimated ingestion doses, assuming exposure to the maximum concentration found in thetest 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 vinylchloride. This is based on a study of rats that developed liver toxicity from exposure to vinylchloride (in PVC) in their diet. The lowest dose at which adverse liver effects were observed--theLOAEL--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 sensitivethan 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 healthguideline and similar to the LOAEL (for children).

However, we are not certain whether people drank water from wells potentially contaminated withvinyl chloride. Therefore, the health hazard from past exposure to vinyl chloride cannot bedetermined. These wells are not currently being used. If we make the assumption that people wereexposed to vinyl chloride at maximum "detection limit" concentrations, then we conclude thatpeople may experience adverse health effects. ATSDR scientists recommend that detection limits fordegradation products of TCE, such as vinyl chloride, in groundwater analyses are low enough todetermine whether concentrations exceed health-based guidelines.

Given the lack of accurate concentration measurements for vinyl chloride in residential wells andexposure 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. Nopublic 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 foundin 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, andis also present in leafy vegetables, meat, poultry and fish.

Zinc was detected one time in one residential well, at a concentration of 5,090 g/L. ATSDR'sestimated ingestion doses, assuming exposure at this concentration, were 0.15 mg/kg/day for anadult 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, dairyproducts, and fibrous vegetables in the diet; and (for women) whether one is pregnant or nursinginfants. 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, moreactive elderly people consume closer to average levels. If a person's diet is low in zinc-containingfoods, 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 decreasedhigh-density (good) cholesterol levels, impaired immune function, and anemia [120]. These effectshave been observed at estimated total dietary intakes of 1 mg/kg/day (which is equivalent to 60mg/day for a 60-kg woman and 70 mg/day for a 70-kg man) and are the basis for EPA's healthguideline 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 causeadverse health effects and lower than the recommended dietary intake for adults. Also, the estimatedexposure dose for a child would decrease as the child developed into an adult (the estimated adultdose was 0.15 mg/kg/day), so the child dose does not represent a chronic exposure dose over alifetime.

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 The U.S. Government's Official Web PortalDepartment of Health and Human Services
Agency for Toxic Substances and Disease Registry, 4770 Buford Hwy NE, Atlanta, GA 30341
Contact CDC: 800-232-4636 / TTY: 888-232-6348

A-Z Index

  1. A
  2. B
  3. C
  4. D
  5. E
  6. F
  7. G
  8. H
  9. I
  10. J
  11. K
  12. L
  13. M
  14. N
  15. O
  16. P
  17. Q
  18. R
  19. S
  20. T
  21. U
  22. V
  23. W
  24. X
  25. Y
  26. Z
  27. #