PUBLIC HEALTH ASSESSMENT

AIR AND GROUNDWATER PATHWAYS ANALYSES
AGANA POWER PLANT
MONGMONG, GUAM

September 22, 2004

Figures

Figure 1: Site Location, Agana Power Plant

Figure 2: Agana Power Plant Surface Geology

Figure 3: Agana Power Plant Surface Soils

Figure 4: Agana Power Plant Drainage Areas

Figure 5: Locations of Drinking Water Supply Wells Within a One-Mile Radius of Agana Power Plant

Figure 6a: Elevation Map Locations

Figure 6b: Elevation Map

Figure 7: 1945 Military Areas Near Agana Power Plant

Figure 8: Selected Former Ambient Air Monitoring Locations on Guam

Figure 9: Agana Swamp Wetlands Map

Appendix A: Possible Exposure Situations for PCBs From the Agana Power Plant

Exposure Situation and ATSDR's Conclusion	Hazard Category*	ATSDR's Recommendations
*Breathing In Contaminants* ATSDR considered two approaches when researching this issue: we evaluated measurements of air pollution made on Guam, and we assessed the available data for use in a computer model to estimate potential air quality impacts from the former APP's emissions. The available ambient air monitoring data suggest that sulfur dioxide and particulate levels in neighborhoods downwind from the former APP did not reach levels of health concern. The currently available information, however, is insufficient for an evaluation of past emissions from the power plant.	Past: indeterminate public health hazard Current: no public health hazard	No recommendations for this exposure situation.
*Groundwater and Drinking Water* The Guam Waterworks Authority routinely monitors the drinking water supply, and the results of monitoring tests have consistently met all Guam and federal safe drinking water standards. Some very low concentrations of VOCs and pesticides were detected during routine monitoring, but at levels well below their maximum contaminant levels for drinking water. Monitoring of supply wells in the Mongmong area has periodically detected bacterial contaminants at levels that exceed Guam and federal safe drinking water standards.	Past and current: no public health hazard	No recommendations for this exposure situation. The Navy continues to investigate the potential for groundwater contamination downgradient of APP.
* See Appendix B.

Appendix B: ATSDR Hazard Categories

Category	Definition	Criteria
A. Urgent public health hazard	This category is used for sites that pose an urgent public health hazard as the result of short-term exposures to hazardous substances.	Evidence exists that exposures have occurred, are occurring, or are likely to occur in the future AND Estimated exposures are to a substance(s) at concentrations in the environment that, upon short-term exposures, can cause adverse health effects to any segment of the receptor population AND/OR Community-specific health outcome data indicate that the site has had an adverse impact on human health that requires rapid intervention AND/OR Physical hazards at the site pose an imminent risk of physical injury
B. Public health hazard	This category is used for sites that pose a public health hazard as the result of long-term exposures to hazardous substances.	Evidence exists that exposures have occurred, are occurring, or are likely to occur in the future AND Estimated exposures are to a substance(s) at concentrations in the environment that, upon long-term exposures, can cause adverse health effects to any segment of the receptor population AND/OR Community-specific health outcome data indicate that the site has had an adverse impact on human health that requires intervention
C. Indeterminate (potential) public health hazard	This category is used for sites with incomplete information.	Limited available data do not indicate that humans are being or have been exposed to levels of contamination that would be expected to cause adverse health effects; data or information are not available for all environmental media to which humans may be exposed AND There are insufficient or no community-specific health outcome data to indicate that the site has had an adverse impact on human health
D. No apparent public health hazard	This category is used for sites where human exposure to contaminated media is occurring or has occurred in the past, but where that exposure is below a level of health hazard.	Exposures do not exceed an ATSDR chronic MRL or other comparable value AND Data are available for all environmental media to which humans are being exposed AND There are no community-specific health outcome data to indicate that the site has had an adverse impact on human health
E. No public health hazard	This category is used for sites that do not pose a public health hazard.	There is no evidence of current or past human exposure to contaminated media AND Future exposures to contaminated media are not likely to occur AND There are no community-specific health outcome data to indicate that the site has had an adverse impact on human health

Appendix C: Information on How ATSDR Assesses Exposure

What does "exposure" mean?

People can only be exposed to a chemical if they come in contact with it. Contact (exposure) can occur by breathing, eating, or drinking a substance containing the contaminant or by touching a substance containing the contaminant with one's skin.

How do ATSDR scientists determine which exposure situations and contaminants to evaluate?

ATSDR's public health assessments are exposure-driven. Chemical contaminants disposed of or released into the environment have the potential to cause adverse health effects under certain conditions. That said, a release does not always result in exposure.

ATSDR scientists evaluate site conditions to determine if people could have been (a past scenario), are (a current scenario), or will be (a future scenario) exposed to site-related contaminants. When evaluating exposure pathways, ATSDR identifies whether exposure to contaminated media (soil, water, air, waste, or biota) has occurred, is occurring, or will occur through ingestion, dermal (skin) contact, or inhalation.

If exposure was or is possible, ATSDR applies a weight-of-evidence approach as to whether people might develop adverse health effects. First, ATSDR scientists select contaminants for further evaluation by comparing them against health-based values. Comparison values are developed by ATSDR from scientific literature available on exposure and health effects. These comparison values are derived for each of the different media and reflect the estimated contaminant concentration that is not likely to cause adverse health effects for a given chemical, assuming a standard daily contact rate (e.g., amount of water or soil consumed or amount of air breathed) and body weight. Comparison values are not thresholds for adverse health effects. ATSDR comparison values establish contaminant concentrations many times lower than levels at which no effects have been observed in experimental animals or human epidemiological studies. Some of the comparison values used by ATSDR scientists include ATSDR's environmental media evaluation guides (EMEG) and reference dose media guides (RMEG), EPA's cancer slope factors and reference doses (RfDs), and ATSDR's minimal risk levels (MRLs).

If someone is exposed, will they get sick?

Exposure does not always result in harmful health effects. The type and severity of health effects an individual experiences because of contact with a contaminant depend on the exposure concentration (how much); the frequency, duration, or both of the exposure or exposures (how long); the route or pathway of exposure (breathing, eating, drinking, or skin contact); and the multiplicity of exposure (the combination of contaminants involved). Once an individual is exposed, characteristics such as age, sex, nutritional status, genetics, lifestyle, and health status influence how he or she absorbs, distributes, metabolizes, and excretes the contaminant. Together, these factors and characteristics determine the health effects that occur as a result of exposure to a contaminant in the environment.

There is considerable uncertainty about the true level of exposure to environmental contamination. To account for this uncertainty and to be protective of public health, ATSDR scientists typically use high-end, worst-case exposure level estimates as the basis for determining whether adverse health effects are possible. These estimated levels usually are much higher than those to which people are really exposed. If they indicate that adverse health effects are possible, ATSDR performs a more detailed site-specific review of exposure combined with scientific information from the toxicological and epidemiological literature on health effects. Considering all of this weight-of-evidence information, ATSDR decides whether adverse health effects are likely.

A discussion outlining the difference between ATSDR's public health assessment process and EPA's risk assessment process is presented in Appendix H.

List of Comparison Values

Comparison values represent medium-specific contaminant concentrations that are used to select contaminants for further evaluation to determine the possibility of adverse public health effects. The conclusion that a contaminant exceeds the comparison value does not mean that it will cause adverse health effects.

Cancer Risk Evaluation Guide (CREG)
CREGS are estimated contaminant concentrations that would be expected to cause no more than one excess cancer in a million (10-6) persons exposed over their lifetime. ATSDR's CREGs are calculated from EPA's cancer potency factors (CPFs).

Maximum Contaminant Level (MCL)
An MCL is a drinking water standard established by EPA. It is the maximum permissible level of a contaminant in water that is delivered to the free-flowing outlet. MCLs are considered protective of public health over a lifetime (70 years) for individuals consuming 2 liters of water per day.

Environmental Media Evaluation Guide (EMEG)
An EMEG is an estimate of daily human exposure to a chemical (in mg/kg/day) that is likely to be without noncarcinogenic health effects over a specified duration of exposure, to include acute, intermediate, and chronic exposures. EMEGs are based on ATSDR minimal risk levels (MRLs), which consider body weight and ingestion rates.

Reference Media Evaluation Guide (RMEG)
ATSDR derives RMEGs from EPA's oral reference doses. The RMEG represents the concentration in water or soil at which daily human exposure is unlikely to result in adverse noncarcinogenic effects.

More information about the ATSDR evaluation process can be found in ATSDR's Public Health Assessment Guidance Manual at http://www.atsdr.cdc.gov/HAC/HAGM, or obtained by contacting ATSDR at 1-888-42ATSDR (1-888-422-8737).

Appendix D: Exposure Evaluation Methodology and Assumptions

This appendix details the assumptions and calculations that ATSDR used to estimate potential exposure levels from consumption of free-range animals (i.e., chickens, eggs, and swine [pigs], beef, and milk) obtained in the Mongmong area near APP. To account for uncertainty about how well the exposure factors represent the residents of Mongmong, ATSDR used health-protective assumptions to estimate the reasonable maximum exposure level for each relevant contaminant. The resulting estimates calculate daily exposure doses in milligrams of contaminant per kilogram body weight (mg/kg/day). They are intentionally protective of public health, likely overestimating the amount of chemical exposure from eating free-range animal products.

The PCBs were evaluated using the highest detected level of any Aroclor at a particular sampling location. For example, if Aroclors 1248, 1254, and 1260 were all detected at a location, ATSDR used the highest value of the three in the analysis.

Methods and Assumptions for Estimating PCB Concentrations in Free-Range Animals

ATSDR's evaluation of eating foods raised locally in the Mongmong area consisted of (1) conducting an extensive literature search on the potential for chemical contaminants to accumulate in locally raised and free-ranging animals under various exposure situations, (2) reviewing current research that estimates soil ingestion by free-range animals, (3) reviewing published studies or review articles that present bioconcentration and biotransfer factors for certain free-range animals, and (4) contacting agricultural industry experts and local government of Guam agencies to compile information about prevalence and consumption patterns of free-range animals on the island.

Through these various activities, ATSDR has established specific methods and assumptions for estimating the PCB concentration in free-range animals or animal products (e.g., eggs).

Formula Used To Calculate Concentrations in Animals

ATSDR derived concentrations of chemicals in animal tissue from concentrations of those chemicals in the soil using the following equation (see Table D-1 for calculated concentrations):

A_animal = Q_s × C_s × BA_animal

Where: A_animal = concentration of chemical in the animal (in mg/kg fresh weight) Q_s = quantity of soil ingested by the animal (in kg/day) C_s = chemical concentration in soil (mg/kg) BA_animal = biotransfer factor for that animal (in days/kg)¹

Table D-1. Estimated Concentrations in Meat, Dairy, and Eggs of Free-Ranging Animals Based on Detected Soil Concentrations
Location	Analyte	Soil Concentration	Chicken		Eggs		Pig
Location	Analyte	Soil Concentration	BTF*	Concentration (ppm) Fresh Weight	BTF*	Concentration (ppm) Fresh Weight	BTF*	Concentration (ppm) Fresh Weight
Agana Springs (Upland)	Aroclor 1260	391 ppm (average) 7,800 ppm (maximum)	0.032 0.032	0.3 5.5	12.8 12.8	110 2,197	0.049 0.049	7.1 141
APP Drainage Outfall (Upland Area)	Aroclor 1260	2.2 ppm (average) 25 ppm (maximum)	0.032 0.032	0.002 0.02	12.8 12.8	0.62 7.0	0.049 0.049	0.04 0.5
APP West Drainage Area	Aroclor 1254	3.7 ppm (average) 29 ppm (maximum)	0.032 0.032	0.003 0.02	12.8 12.8	1.0 8.2	0.049 0.049	0.07 0.5
Reference sample area	Aroclor 1254	0.05 ppm (average) 0.2 ppm (maximum)	0.032 0.032	0.00004 0.0001	12.8 12.8	0.01 0.06	0.049 0.049	0.0009 0.004
Residential property area	Aroclor 1254	0.06 ppm (average) 0.6 ppm (maximum)	0.032 0.032	0.00004 0.0004	12.8 12.8	0.02 0.2	0.049 0.049	0.001 0.01
Soil ingestion rates were based on those listed in EPA 1998a. Rates given were consistent with those found in other published documents. Chickens = 0.022 kg/day Pigs = 0.37 kg/day * Biotransfer factors (BTFs) were based on values given in EPA 1998a. Since no BTF for Aroclor 1260 was provided, the values for Aroclor 1254 were used. ppm = mg/kg

Table D-1 (continued). Estimated Concentrations in Meat, Dairy, and Eggs of Free-Ranging Animals Based on Detected Soil Concentrations
Location	Analyte	Soil Concentration	Beef		Milk
Location	Analyte	Soil Concentration	BTF*	Concentration (ppm) Fresh Weight	BTF*	Concentration (ppm) Fresh Weight
Agana Springs (Upland)	Aroclor 1260	391 ppm (average) 7,800 ppm (maximum)	0.04 0.04	7.9 158	0.01 0.01	2 40
APP Drainage Outfall (Upland Area)	Aroclor 1260	2.2 ppm (average) 25 ppm (maximum)	0.04 0.04	0.04 0.51	0.01 0.01	0.01 0.13
APP West Drainage Area	Aroclor 1254	3.7 ppm (average) 29 ppm (maximum)	0.04 0.04	0.07 0.59	0.01 0.01	0.02 0.15
Reference sample area	Aroclor 1254	0.05 ppm (average) 0.2 ppm (maximum)	0.04 0.04	0.001 0.004	0.01 0.01	0.0002 0.001
Residential property area	Aroclor 1254	0.06 ppm (average) 0.6 ppm (maximum)	0.04 0.04	0.001 0.01	0.01 0.01	0.0003 0.003
Soil ingestion rates were based on those listed in EPA 1998a. Rates given were consistent with those found in other published documents. Beef = 0.5 kg/day Dairy cows = 0.4 kg/day * Biotransfer factors (BTFs) were based on values given in EPA 1998a. Since no BTF for Aroclor 1260 was provided, the values for Aroclor 1254 were used. ppm = mg/kg

Methods and Assumptions for Calculating PCB Human Dose

Using the estimated PCB concentrations in animals, ATSDR calculated a human dose of PCBs. The concentrations in animals were derived as described above. They were based on the mean values reported for surface soil at the three locations where PCB concentrations in at least one sample were above ATSDR's screening level for PCBs-specifically for Aroclor 1254 and for the residential locations that were sampled. For samples where PCBs were not detected, a value of one-half the reporting limit was used to contribute to the average.

To calculate cancer risk estimates based on the human dose, ATSDR used EPA's cancer slope factor for oral doses and then compared those estimates to a risk level of 0.0001 (i.e., 1 in 10,000). We used that risk level because it reflects realistic human exposures; however, risk levels are theoretical and actual risks may be as low as zero.

To screen non-cancer effects, we compared the estimated doses to ATSDR's minimum risk level (MRL) for PCBs (ATSDR specifically used Aroclor 1254, the only PCB for which there is an MRL: 0.00002 mg/kg/day). The MRL is derived from the lowest observed adverse effect level (LOAEL) identified in the scientific literature; to obtain the MRL, ATSDR divided the LOAEL by a safety factor to protect sensitive groups and account for differences between humans and animals in response to exposure.

We also reviewed the scientific evidence on PCBs to determine the likelihood of adverse health effects in children, pregnant women, and other adults.

Formula Used To Calculate Potential Exposure Dose

The estimated exposure doses are calculated using the following equation. (The terms used in this equation are spelled out in Table D-2.)

Table D-2 lists the assumptions ATSDR used in calculating doses

Table D-2. Exposure Dose Assumptions
Parameter	Abbreviation	Child	Adult
Chemical concentration	C	Specific to location	Specific to location
Ingestion rate	IR	Chicken: 0.074 kg/day Eggs: 0.051 kg/day Pig: 0.041 kg/day Beef : 0.084 kg/day Milk: 1.05	Chicken: 0.147 kg/day Eggs: 0.102 kg/day Pig: 0.082 kg/day Beef: 0.168 kg/day Milk: 2.1
Exposure frequency	EF	50 days/year (1x/week)	50 days/year (1x/week)
Exposure duration	ED	10 years	40 years
Body weight	BW	15.3 kg	72 kg
Averaging time	AT	365 days	365 days
Averaging duration, carcinogens	AD	70	70
Averaging duration, non-carcinogens	AD	10 years (i.e., the ED)	40 years (i.e., the ED)
Target dose (EPA's reference dose for PCBs)	TD	0.00002 mg/kg/day	0.00002 mg/kg/day
Note: Ingestion rates are based on the 95th percentile (kg/day) per capita intake for each food item (EPA 1997).

Table D-3 presents estimated PCB exposure doses for "non-cancer effects" from eating free-range chicken, eggs, and meat from swine (pigs), as well as beef and milk from dairy cows, in the three locations where PCBs were detected above the screening level.

Table D-3. Estimated PCB Exposure Dose Levels From Eating: Non-cancer Effects
Location	Food Item	Adult Dose (mg/kg/day)	Child Dose (mg/kg/day)	MRL (mg/kg/day)	Estimated PCB Concentration (ppm)
Agana Springs (Upland)	Chicken Eggs Pig Beef Milk (cow)	0.00008 0.02 0.0007 0.003 0.008	0.0002 0.05 0.002 0.006 0.02	0.00002 0.00002 0.00002 0.00002 0.00002	0.0008 0.31 0.02 0.02 0.006
APP Drainage Outfall (Upland Area)	Chicken Eggs Pig Beef Milk (cow)	0.0000006 0.0001 0.000004 0.00002 0.00004	0.000001 0.0003 0.00001 0.00004 0.00009	0.00002 0.00002 0.00002 0.00002 0.00002	0.002 0.62 0.04 0.05 0.01
APP West Drainage Area	Chicken Eggs Pig Beef Milk (cow)	0.0000008 0.0002 0.00001 0.00003 0.00008	0.000002 0.0005 0.00002 0.00006 0.0002	0.00002 0.00002 0.00002 0.00002 0.00002	0.003 1 0.07 0.08 0.02

Applying the worst case exposure scenario (i.e., using the maximum concentrations detected). The estimated doses for non-cancer effects from the consumption of eggs in the portion of Agana Springs (Upland) area where the old transformer pad was located exceeded EPA's reference dose (RfD). ATSDR also evaluated potential cancer effects from consuming free-range animals from the same sampling location-old transformer pad at Agana Springs (Upland). Although studies of human populations do not provide strong evidence that PCBs are associated with cancer in humans, there is evidence from animal studies that PCBs, especially certain PCB congeners, may cause cancer. ATSDR's cancer risk estimates were calculated using EPA's cancer slope factor for PCBs and compared to a risk level of 0.0001 (1 in 10,000). The the cancer estimates exceeded the comparison level for free-range chicken eggs at the old transformer pad at Agana Springs (Upland).

The doses that exceed ATSDR's MRL or cancer risk level largely reflect ATSDR's conservative approach used in estimating dose, and it is likely that actual doses would be considerably lower. ATSDR's conservative approach is also based on the uncertainty of health effects occurring at very low levels. Although most epidemiological studies do not support PCBs causing cancer or non-cancer effects at levels found in a typical U.S. mainland diet, ATSDR continues to evaluate whether very subtle health effects (e.g., liver function abnormalities, learning and other neurobehavioral effects, immune system disorders) may be present with current levels found in the world food supply. As an added protective measure against PCB exposure, residents living in the Mongmong area who consume free-range eggs from the small area near the old PCB transformer pad at Agana Springs (Upland) should consider alternate sources for their eggs. Additional guidance on the consumption of free-range animals from the three sampling areas is provided below.

ATSDR believes that our calculations and EPA's modeled RfDs are extremely health-protective. Levels found in the common food supply are similar to the levels modeled for meat and poultry in Mongmong. It also should be noted that the model used for this report is believed to overestimate the amount of PCBs in food for site conditions, increasing the margin of safety even more. The actual concentration in meat, poultry, and eggs may even be lower than found in the general U.S. food supply. ATSDR believes that the U.S. food supply is one of the safest in the world.

¹ The biotransfer factor is defined as the equilibrium concentration of a pollutant in tissue divided by the average daily intake of pollutant (Eduljee and Gair 1996).

Next Section Table of Contents