PUBLIC HEALTH ASSESSMENT
BANGOR NAVAL SUBMARINE BASE
BANGOR ORDNANCE DISPOSAL (USNAVY)
SILVERDALE, KITSAP COUNTY, WASHINGTON
The following section discusses the various contaminantsof concern, how people might come into contact with thesecontaminants and the potential health effects that mayresult. In order for an exposure to these contaminants tooccur, all the elements of an exposure pathway must be inplace. Exposure pathways are divided into completed andpotential and can be current, past or future. A completedexposure pathway consists of five elements: source,environmental media/transport, point of exposure, route ofexposure and receptor population. A potential exposurepathway exists when some but not all of these five elementsare present and the potential exists that the missingelements have been present (past), are present (current) or will be present (future). The completedand potential exposure pathways for Bangor are given in Tables 2 and 3 below. Each pathway is then discussed in terms of the contaminants of concern and the potential health hazard posed.
Evaluating Non-cancer Risk
In order to evaluate the potential for non-cancer adverse health effects that may result fromexposure to contaminated media (i.e., air, water, soil, and sediment), a dose is estimated for eachcontaminant of concern. These doses are calculated for situations (scenarios) in which nearbyresidents or on-base workers might come into contact with the contaminated media. The estimateddose for each contaminant under each scenario is then compared to ATSDR's minimal risk level(MRL) or EPA's oral reference dose (RfD). MRLs and RfDs are doses below which non-canceradverse health effects are not expected to occur (so called "safe" doses). They are derived from toxiceffect levels obtained from human population and laboratory animal studies. These toxic effectlevels can be either the lowest-observed adverse effect level (LOAEL) or a no-observed adverseeffect level (NOAEL). In human or animal studies, the LOAEL is the lowest dose at which anadverse health effect is seen, while the NOAEL is the highest dose that did not result in any adversehealth effects.
Due to the uncertainty in these data, the toxic effect level is divided by "safety factors" giving thelower and more protective MRL or RfD. If a dose exceeds the MRL or RfD, this indicates only thepotential for adverse health effects. The magnitude of this potential can be inferred from the degreeto which this value is exceeded. If the estimated exposure dose is only slightly above the MRL orRfD, then that dose will fall well below the toxic effect level. The higher the estimated dose is above the MRL or RfD, the closer it will be to the toxic effect level.
Evaluating Cancer Risk
Some chemicals have the ability to cause cancer. Cancer risk is estimated by calculating a dosesimilar to that described above and multiplying it by a cancer potency factor, also known as thecancer slope factor. Some cancer potency factors are derived from human population data. Othersare derived from laboratory animal studies involving doses much higher than are encountered in theenvironment. Use of animal data require extrapolation of the cancer potency obtained from thesehigh dose studies down to real-world exposures. This process involves much uncertainty. Currentthinking suggests that a very small dose of a carcinogen will give a very small cancer risk. Cancerrisk estimates are, therefore, not yes/no answers but measures of chance (probability). Suchmeasures, however uncertain, are useful in determining the magnitude of a cancer threat since anylevel of a carcinogenic contaminant carries an associated risk. Some evidence suggests that certainchemicals considered to be carcinogenic must exceed a threshold of tolerance before initiatingcancer.
This document describes cancer risk qualitatively using terms like low, very low, slight and nosignificant increase in cancer risk. Cancer risk estimates are used to help environmental agenciesknow when to take action to clean up hazardous waste sites or conduct health actions. Cancer riskestimates are theoretical. Actual risk may be as low as zero. These terms can be better understoodby considering the population size required for such an estimate to result in a single cancer case. Forexample, a low increase in cancer risk indicates an estimate in the range of one cancer case per tenthousand persons exposed over a lifetime. A very low estimate might result in one cancer case perseveral tens of thousands exposed over a lifetime and a slight estimate would require an exposedpopulation of several hundreds of thousands to result in a single case.
ATSDR considers theoretical cancer risk not to be significant when the cancer risk estimateresults in less than one cancer per ten thousand exposed over a lifetime. Cancer riskestimates less than one cancer per ten thousand are therefore considered to pose noapparent public health hazard. The reader should note that these estimates are for excess cancersthat might result in addition to those normally expected in an unexposed population.
Cancer is a common illness and its occurrence in a population increases with age. Depending on thetype of cancer, a population with no known environmental exposure could be expected to have asubstantial number of cancer cases. There are many different forms of cancer that result from avariety of causes; not all are fatal. Approximately one quarter to one third of people living in the United States will develop cancer at some point in their lives.
Multiple Exposure and Toxicological Mixtures
A person can be exposed by more than one pathway and to more than one chemical. Exposure tomultiple pathways occurs if a contaminant is present in more than one medium (i.e., air, soil,surface water, groundwater, and sediment). For example, the dose of a contaminant received fromdrinking water may be combined with the dose received from contact with that same contaminant insoil.
It is much more difficult, however, to assess exposure to multiple chemicals. In almost everysituation of environmental exposure, there are multiple contaminants to consider. The potentialexists for these chemicals to interact in the body and increase or decrease the potential for adversehealth effects. The vast number of chemicals in the environment make it impossible to measure all ofthe possible interactions between these chemicals. Individual cancer risk estimates can be addedsince they are measures of probability. When estimating non-cancer risk, however, similarities mustexist between the chemicals if the doses are to be added. Groups of chemicals that have similar toxiceffects can be added such as volatile organic compounds (VOCs) which cause liver toxicity.Polycyclic aromatic hydrocarbons (PAHs) are another group of chemicals that can be assessed asone added dose based on similarities in chemical structure and metabolites. Although somechemicals can interact to cause a toxic effect that is greater than the added effect, there is littleevidence demonstrating this at concentrations commonly found in the environment.
The following evaluations do not rely solely on whether the estimated dose of a contaminant exceedsits health comparison value (i.e., MRL, RfD, cancer risk levels). Factors such as backgroundexposure, a growing scientific data base and the inherent uncertainty in assessing health risk areconsidered when formulating conclusions. These evaluations are based on current data and subjectto change should more data become available relative to the site and/or the toxic potential of the contaminants.
|#||Area of Exposure||Source||Contaminant||Media||Point of Exposure||Route of Exposure||Exposed Population|| Persons |
|1||Mountain View Road||OU-8||VOCs |
|Groundwater||Residential wells||Ingestion |
|Off-base Residents||75||Past||No||Very Low||No current or future pathway exists since Mountain View residents are supplied with Silverdale water.|
|Hood Canal/ |
|Sites OU-1,4,6,7||Ordnance compounds, phthalates||Fish/ Shellfish||Hood Canal, Cattail Lake Beaches||Ingestion||On and Off-base Residents||500||Past||No||No||Only trace levels of contaminants found in sediment and shellfish.|
|3||On-base Hazardous Waste Sites||All Operable Units||Ordnance compounds||Soil||All sites||Ingestion |
|Workers||1000||Past||No||Very low||Current exposure has been reduced or eliminated by remedial actions|
|#||Area of Potential Exposure||Source||Contaminant||Media||Point of Exposure||Route of Exposure||Exposed Population||Time||Comments|
|1||On-base Supply Wells||OU-2 and 4||Ordnance Compounds||Groundwater||Bangor supply wells||Ingestion |
|On-base residents & workers||Future||Wells in sea-level and aquifer unlikely to be impacted.|
|2||Vinland Supply Wells||OU-1||Ordnance Compounds||Groundwater||Vinland public & residential wells||Ingestion |
|Off-base Residents||Future||Groundwater is not moving in the direction of wells.|
|3||Olympic View/ |
|OU-2||Ordnance Compounds||Groundwater||Olympic View/Old Bangor public and residential wells||Ingestion |
|Off-base residents||Future||Contamination is still about a mile away from off-base wells.|
No apparent public health hazard exists for people in the Mountain View Road area who drank1,2-dichloroethane (1,2-DCA) contaminated residential well water prior to 1995. Chemicals ingroundwater originating from the Public Works area of Bangor have contaminated residentialwells in the Mountain View Road area. The primary contaminant of concern in residential wellsis 1,2-dichloroethane (1,2-DCA) although benzene has been detected at high levels in off-basemonitoring wells. No current exposure exists because the Navy provided hook up to Silverdalepublic water supply for all residences within the groundwater contamination zone.
One private well (PW-16) may have been contaminated with 1,3-dinitrobenzene (1,3-DNB) asearly as 1984. Although subsequent testing did not confirm this detection, exposure to 1,3-DNBwas also evaluated for users of this well. Exposure to 1,3-DNB the level initially detected in this well does not pose a health hazard.
Groundwater contamination in the Mountain View Road area was first reported in October 1984when 19.3 ppb 1,3-dinitrobenzene (1,3-DNB) was detected in residential well PW-16 (Figure 7).No contaminants were found in any of the other 14 residential wells sampled at that time. Subsequent sampling and analysis of off-base and on-base wells over the next year revealedconflicting results. A sample taken from PW-16 in July/August of 1985 showed 13.2 ppb 1,3-DNBaccording to one lab but no detections were reported by a second lab that analyzed the same sample.A similar discrepancy was reported from samples collected in November 1985.12 Although 1,3-DNBhas been found at Site F (OU-2), two rounds of groundwater samples at Site 25 (OU-3) found no1,3-DNB. Analysis of a sub-surface soil samples from a well boring at Site 25 did detect 1,3-DNBat 51 ppb.7 Site 25 is the location of a former wastewater discharge pond adjacent to the MountainView Road area.
In May 1986, investigations of a potential gasoline release in the Public Works facility bordering theMountain View Road area identified a significant amount of groundwater contamination in the area.Based on the amounts of free product detected in groundwater monitoring wells, it was determinedthat approximately 20,000 gallons of gasoline had been spilled since 1982. A recovery system wasinstalled in August 1986 to recover the gasoline from the groundwater. No residential wells weresampled at this time.4
In February 1994, analysis of a sample taken by the local health department from a new residentialwell (PW-01) detected volatile organic compounds (VOCs) and the herbicide bromacil. Additionalsampling in March 1994 confirmed these detections but no contaminants were found in seven otherresidential wells. Further investigation of groundwater in the Public Works and Mountain ViewRoad areas detected contaminants in both on and off-base groundwater.4 Maximum levels ofcontaminants of concern found in off-base monitoring and private wells are given in Appendix C,Tables C1 and C2 (see Figure 7 for well locations). High levels of these contaminants were found inon-base groundwater including 1,2-DCA at a maximum of 1,700 ppb and benzene at a maximum of9,800 ppb. During these investigations, the base provided bottled water to residents in the MountainView Road from April 1994 through the end of 1995 when residences were connected to Silverdalepublic water. In May 1997, a groundwater containment/treatment system became operationaljoining the existing gasoline recovery system.13
The extent of the off-base 1,2-DCA contaminant plume is illustrated in Figure 8. Contaminants fromthe gasoline spill follow the same path as the chlorinated solvents. However, it is likely that thechlorinated solvent plume originated earlier since virtually no chlorinated compounds have beenfound in soil suggesting that enough time has elapsed to allow any surface spills to completely washthrough soil into groundwater. Both contaminant plumes are confined within a narrow lateralboundary of the surface aquifer extending southeast from the Public Works area as far as ClearCreek Road. The surface aquifer in this area extends down as far as 250 feet before reaching aconfining clay layer. Contaminant depth increases as the plume extends off-base indicating verticalas well as lateral movement from the source.
A quarterly monitoring program of residential wells in the Mountain View Road area has beenongoing since the first quarter of 1995. Although the initial detection of 1,2-DCA came from aresidential well that was never in use, later sampling found declining levels of 1,2-DCA in two otherresidential wells (PW-08 and PW-04). Initial sampling of PW-08 and PW-04 in first quarter 1995found 1,2-DCA at 4.6 and 0.42 ppb, respectively. These results represent the maximum level of 1,2-DCA found in these or any residential well in the area. Low levels of other VOCs have beenintermittently detected in residential wells. Chart 2 below gives the sampling history for residentialwell PW-08.
The provision of Silverdale public water to residents in the Mountain View Road area haseliminated exposure to groundwater contamination originating in the Public Works area of Bangor.However, residents using wells PW-08 and PW-04 as a drinking water source prior to the provision of alternate water were exposed to 1,2-DCA through ingestion, inhalation and dermal contact.
Non-cancer adverse health effects are not expected to result from past exposure of residents to the maximum level of 1,2-DCA ever found in Mountain View Road drinking water wells. No current exposure pathway exists since residents in affected area have been connected to the Silverdale public water supply. The estimated dose calculated for a resident exposed to the maximum detected level of 1,2-DCA in drinking water (4.6 ppb in PW-08) is 500-fold below ATSDR's intermediate minimal risk level (MRL).17
However, residential wells PW-08 and PW-04 have showndeclining levels of 1,2-DCA since testing began in April1994. This trend indicates that levels of 1,2-DCA could have been higher prior to initial testing. It isalso important to note that no chronic MRLs or RfDs are available for 1,2-DCA. Chronic MRLsand RfDs are usually lower than intermediate or acute MRLs because they are derived from studiesin which the dose was given over a longer periods of time. The toxicity of 1,2-DCA is discussed further in the Toxicological Evaluation section (see page 34).
EPA has classified 1,2-DCA as a Group B2 probable human carcinogen based on evidence of blood vessel tumors that developed in rats given high doses of 1,2-DCA. Although there is no supporting evidence that 1,2-DCA can cause cancer in humans,a cancer potency factor was derived by EPA from the animal study mentioned above.18 Using this cancer potency factor (called a slope factor), a very low cancer risk was estimated for a child growing to adulthood over a 30-year period while exposed to the maximum detected level of 1,2-DCA. This level which over-estimates the actual exposures presents no apparent public health hazard.
The source of the 1,2-DCA contamination of private wells in the Mountain View Road area isassumed to be near the highest levels of 1,2-DCA found in groundwater at the Public Works facility.However, no information has been located with which to estimate the quantity or date of anydischarge of these types of chemicals. Although levels of 1,2-DCA could have been higher inMountain View Road residential wells prior to initial testing, there is no current information topredict what those levels might have been.
It is possible, but not certain, that residents using private well PW-16 were exposed to 1,3-DNB indrinking water. Although the initial detection of 1,3-DNB in this well was not confirmed, thechemical was detected once more in this same well and in on-base monitoring wells. Available datawere insufficient to conclude that these detections were lab errors. Therefore, residents using wellPW-16 prior to the provision of alternate water are assumed to have been exposed to 1,3-DNB. Thisexposure would have been primarily through ingestion but dermal absorption may also havecontributed. Inhalation of 1,3-DNB volatilizing from water is not expected to be a significant route of exposure.
Adverse health effects are not likely to result from past exposure of residents to 1,3-DNB inresidential well PW-16. No current exposure pathway exists since residents using PW-16 have beenconnected to the Silverdale public water supply. The toxicity of 1,3-DNB is discussed further in the Toxicological Evaluation section (see page 36).
There is no evidence to suggest that 1,3-DNB is carcinogenic.
No apparent public health hazard exists for residents living in the Mountain View Road area whodrank contaminated groundwater that originated at the Bangor Public Works facility (OU-8).Residents have been connected to the Silverdale public water supply since December 1995.
Exposure of residents to 1,3-dinitrobenzene was evaluated based on unconfirmed detections of thisordnance compound in residential well (PW-16). The maximum levels of 1.3-dinitrobenzenedetected is not likely to result in adverse health effects.
Mountain View Road residents should not use private drinking water wells for domestic purposesuntil groundwater contaminant levels have declined below a level of health concern for long termuse. These wells are currently being monitored for VOC contamination but should be tested forbromacil prior to resumption of use as a domestic water supply.
No public health hazard exists for base personnel or off-base shellfish harvesters from exposureto contaminants in fish and shellfish taken from Cattail Lake or Hood Canal beaches. Samplingand analysis of sediment, fish and shellfish along Hood Canal and Cattail over the past 10 yearshas detected very little contamination. Low levels of polycyclic aromatic hydrocarbons (PAHs)were found in sediments along with trace levels of phthalates and ordnance compounds inshellfish from Hood Canal and trout from Cattail Lake. The beaches below the former landfill atFloral Point (Site B) are currently not open to shell fishing. Although sediment and shellfishsampling indicate that the landfill is not a contaminant source, further sampling of shellfishshould be conducted prior to reopening of this beach.
Analysis of fish and shellfish samples taken along Hood Canal and Cattail Lake from October 1988through October 1998 have shown minimal contamination. Contaminants of concern given inAppendix C, Table C3, were selected based on detection for organics and exceedance of backgroundlevels for inorganics. Sediment sampling along Hood Canal has consistently shown only sparse andlow level contamination.
Sampling of shellfish and sediment along Hood Canal and Cattail Lake dates back to1988 when samples were gathered in support of the RI/FS for Site A (OU-1). Further sampling of Hood Canal sediment and shellfish was conducted in 1991 and 1992 as part of the RI/FS for Site 26 (OU-7).11 Samples were collected from seven areas along Hood Canal from Floral Point south to Carlson Spit (Figure 9). The ROD for OU-7 signed in April 1996 determined that four of these areas (Floral Point, Marginal Wharf, K/B Dock and Service Pier) warranted further sampling. Two more rounds of sediment and shellfish samples were taken from these areas in September/October 1996 and September/October 1998. Shellfish sampling during these last two rounds was limited to the Floral Point area. In 1997, a 12 inch vegetative soil cover was installed over the former landfill in areas of chemical contamination.19
Polycyclic aromatic hydrocarbons (PAHs) were the primary contaminants found in Hood Canalsediments during RI/FS sampling. The most recent sampling of sediment in 1998 showeddecreasing concentrations of PAHs at Marginal Wharf, K/B Dock and Service Pier. Sampling ofFloral Point sediment and shellfish in 1996 and 1998 was based primarily on the potential for metalsto move from the former landfill via groundwater to beach sediments. Maximum concentrations ofmercury (3.1 ppm) and lead (629 ppm) had been previously detected in subsurface soil samplestaken at the landfill as part of the RI/FS. Also of concern were elevated levels of PCBs (maximumof 6.4 ppm) and pesticides also found in landfill soils at this time. Therefore, analysis of clam tissuefrom Floral Point during these later sampling rounds included metals, pesticides and PCBs.
An evaluation of Hood Canal sediments released by DOH in July 1997 concluded that Floral Point was the only beach threatened from upland sources of contamination. Although sediment data was deemed inadequate at Floral Point, it was recommended that this beach remain closed to shellfish harvesting based on the proximity of the former landfill.20 It is important to note that this evaluation was based on comparisons with Washington State Sediment Management Standards (SMS). The SMS standards adopted by Ecology use ecological endpoints (i.e., benthic organisms). While these standards may also be protective of human health, the following assessment used tissue data to estimate exposure via ingestion of shellfish.
Very little contamination is evident in sediment and fish/shellfish in any area of Hood Canal orCattail Lake. Previous assessments have shown decreasing levels of PAHs in sediments and onlytrace contaminants in shellfish. A dose was estimated for anglers and shellfish harvesters exposed tothe maximum levels of contaminants of concern from ingestion of fish from Cattail Lake andshellfish from Hood Canal. None of the estimated doses exceeded respective RfDs or MRLs. Thepotential for exposure to multiple contaminants was estimated by assuming a "worst-case" scenarioof a fish eater exposed to the maximum levels of each contaminant found in fish or shellfishregardless of source. A combined dose was calculated and determined to be below a level of concernwhen compared with a "combined" reference dose or hazard index.
No adverse health effects are expected to result from ingestion of shellfish from Hood Canal.Past soil sampling indicates that the former Floral Point landfill is a potential source of contaminantsmetals, PCBs and pesticides in beach sediment and shellfish. The lack of significant contaminationon the beach could be explained by the low mobility of PCBs and pesticides in groundwater and theinfrequent detection of mercury (2 of 32 samples) and lead (5 of 30 samples) in subsurface soil. It isuseful to note that groundwater samples taken near the landfill did not show particularly high levelsof these soil contaminants.
No adverse health effects are expected to result from ingestion of fish from Cattail Lake. Theamount of 1,2,3-trinitrobenzene (9 ppb) detected in one of two composite samples of cutthroat trouttaken from Cattail Lake October 1988 is below a level of health concern. No other contaminantswere found in either of these samples. However, these two samples represent the only fish analyzedfrom Cattail Lake. Site A continues to represent a potential source of contaminants for Cattail Lakevia discharge of groundwater containing RDX. Although RDX, as well as most other ordnancecompounds, is not expected to bioaccumulate in fish to any great extent, this potential has not been well studied.
No apparent public health hazard exists from ingestion of fish from Cattail Lake or shellfish from Hood Canal. Analysis of fish, shellfish and sediment samples indicate very little contamination.
Shellfish sampling near Floral Point on Hood Canal has been limited to littleneck clams. If the beachbelow the former landfill is to be reopened for shellfish harvesting, other species that might beavailable for harvest should be analyzed for metals, pesticides, PCBs and PAHs. These data shouldbe gathered and evaluated by DOH prior to reopening of the beach for shell fishing. More shellfishsampling is scheduled for the Floral point area as part of the Record of Decision.
No apparent public health hazard exists for workers who may have been exposed tocontaminants in on-base surface soil prior to remedial activities. Sites-A (OU-1) and Site D(OU-6) posed the greatest chance for exposure of workers to soil contaminants at levels of healthconcern. Past exposure of workers to ordnance compounds in surface soil during demolition andburning of ordnance is estimated to have resulted in only a low risk for cancer. No current healthhazard exists, however, as soil remediation in areas of concern is either ongoing or complete.
Contamination of soil, sediment and surface water has been identified at several sites throughout Bangor. Maximum levels of contaminants of concern detected in surface soil for all operable units is given in Appendix C, Table C4. Extensive sampling of surface water and sediment in Hood Canal, Cattail Lake and Vinland Creek found only trace levels of contamination in these media. This pathway analysis, therefore, considers only exposure to soil. Ingestion of fish from Cattail Lake or shellfish from Hood Canal is evaluated in Pathway 2.
In the past, base personnel may have been exposed to contaminants in soil through accidentalingestion, skin contact and inhalation of dust. Current exposure is assumed to be negligible based onthe ongoing or completed remediation at all sites. The primary areas of surface soil contaminationinclude Site A and Site D where the highest levels of ordnance compounds were found. Site F (OU-2) lagoon soils contained high levels of ordnance compounds but were only available for contact byworkers during transfer to Site A for disposal. Operable Unit -3 contained the highest levels ofarsenic and antimony both of which are listed as contaminants of concern.
The health assessment for workers exposed to contaminants in on-base surface soil focused on SitesA and D as these were the most contaminated sites. It was assumed that a worker could have beenexposed to the upper-bound, average concentration for each contaminant of concern found in surfacesoil at Sites A and D. A maximum value was used for 2,6-DNT as no upper-bound average wasavailable.
No current public health hazard exists for workers exposed to soil at any hazardous waste siteson base. A low risk for non-cancer adverse health effects exists for past exposure of workers overlong periods of time to soil at Sites A and D.
The maximum level or upper confidence limit (95 UCL) of each contaminant of concern found inon-base surface soil was used to estimate a dose for a worker exposed for 250 days/year over a 25year period. This exposure duration represents the entire length of demilitarization activities at SitesA and D. These doses were then compared with their respective oral reference dose (RfD). Theestimated dose for TNT was about 4-fold higher than its respective RfD. No other estimated doses for contaminants of concern in surface soil exceeded their RfDs.
The RfD for TNT is based on liver toxicity seen in ratsgiven high doses of TNT in a capsule over a 25 weekperiod. The lowest-observable adverse effect level(LOAEL) noted in this study is approximately 270 timeshigher than the estimated dose calculated for pastexposure of a worker to TNT in soil. Since this estimateddose exceeds the RfD, but is well below the LOAEL,adverse health effects are not likely. The RfD for TNTwas set 1000 times lower than this LOAEL in order toprovide added safety for the protection of public health.18The toxicity of TNT and other ordnance compounds isdiscussed further in the Toxicological Evaluation section(see page 35).
The estimated doses for each contaminant of concern were added to provide an overall estimate ofnon-cancer health risk. This combined exposure did not result in a significant increase in health risk. It is important to note that some chemicals can cause more than an additive effect when combined, inwhich case this assessment will underestimate the potential for adverse health effects. There is littleevidence, however, of such synergistic effects between chemicals at levels commonly found in theenvironment.
The non-cancer risk calculated for TNT applies to exposures of intermediate as wells as long-termduration (i.e., several weeks or longer). This risk is primarily due to high levels of TNT in surfacesoil at Sites A and D. It is important to note that this risk is based on a "worst-case" scenario and ishighly dependant upon duration of exposure and work practices. Those workers using personalprotective equipment or who were not frequently involved in demilitarization activities at Sites A and D would have had significantly less exposure.
A low cancer risk was estimated for workers exposed under the same scenario used above anddoes not pose a public health hazard. Several ordnance compounds including RDX, TNT, 2,4 and2,6-DNT and arsenic contributed to this risk estimate. Each of these chemicals have been classifiedby EPA as known, probable or possible human carcinogens (see Appendix C for class descriptionsand Table C1 for specific classifications).18 The majority of this risk is attributable to the ordnancecompounds with a small percentage coming from arsenic. Also, included in this estimate wasdi(2-ethylhexyl) phthalate (DEHP) which, along with other similar chemicals called phthalates, arecommon environmental and laboratory contaminants due to their presence in plastics. Thecontribution of DEHP to the overall cancer risk is negligible.
Cancer risk estimates made in this health assessment utilized a cancer potency factor (CPF) derivedby EPA from animal data. The relevance of cancer caused in laboratory animals at high doses forhumans exposed to much lower levels found in the environment is questionable. Such animal data isconsidered to be much stronger when supported by evidence of cancer in humans. There is littleevidence, however, to show that any of these ordnance compounds can cause cancer in humans. Inorder to relate these high dose animal exposures to lower environmental exposures, estimates aremade using mathematical equations. These mathematical equations are used to derive a CPF thatcan be used to estimate risk. A discussion of cancer risk estimation is given in the introduction of thePathways Analysis section (see page 12). The toxicity of ordnance compounds is discussed further inthe Toxicological Evaluation section of this document (see pages 35 and 36).
No current public health hazard exists for workers or other base personnel exposed to surfacesoil at hazardous waste sites located on-base. Remedial activity is either ongoing or completed ateach operable unit and has effectively reduced or eliminated exposure to below a level of health concern.
No apparent public health hazard existed for past exposure of workers to contaminants in on-basesurface soil primarily at Sites A and D. Exposures of several weeks or more to ordnance compoundsin soil during burning and demolition at these sites represents a low increase in risk for cancer . The magnitude of this risk would have been significantly reduced by protective clothing and other gear.
No recommendations are necessary to reduce exposure to soil contaminants as remedial measures have already been taken.
Groundwater contaminants at Site A (OU-1) are not expected to impact off-site drinking waterwells in the nearby community of Vinland located on the norther border of the base. Ordnancecompounds detected in the shallow aquifer are expected to move primarily towards Cattail Lake. No contaminants have been found in the deeper sea-level aquifer. Continued monitoring of MW-28will ensure detection of any groundwater contaminants moving towards drinking water wells in Vinland.
Site A was used to burn and explode ordnance between 1962 and 1986 resulting in contamination ofsoil, surface water and groundwater (Figure 10). Groundwater contamination at Site A has beenfound in both the perched and shallow aquifers.1 The perched aquifer is subject to large seasonalvariation since its sole recharge source is rain infiltration. Groundwater moves north from thisaquifer and surfaces during the wet months in a marsh prior to discharge into Vinalnd Creek Waterin the perched zone can also move vertically into the shallow aquifer. Groundwater in the shallowaquifer around Site A moves west/northwest discharging to Cattail Lake and Hood Canal. Thedeeper, sea level aquifer is a major regional water bearing unit that extends across the KitsapPeninsula and discharges to both Hood Canal and Liberty Bay. It is separated from the shallowaquifer in the area of Site A by 140 to160 feet of dense, low permeability silt.1
RDX and TNT are the primary contaminants in the perched aquifers while lower levels of other ordnance compounds have also been detected. RDX is the primary contaminant in the shallow aquifer although trace levels of TNT have also been found. No ordnance compounds have been detected in the sea-level aquifer. The maximum level of RDX ever detected in monitoring wells at Site A is 1,000 ppb found in MW-48. The most recent sampling of Site A monitoring and extraction wells in the Spring of 1999 found RDX at a maximum of 500 ppb in extraction well EW-7 (Figure 11).21
The community of Vinland is located on the northern border of the base within 1500 feet of Site A.There are currently two municipal supply wells serving approximately 2,400 Vinland residents withan additional five wells available for emergency supply. A well survey conducted during the RI/FSprocess located private and public supply wells in Vinland area in both the shallow and sea levelaquifers (Figure 12). Two municipal wells were located by this survey directly north of Site A in thesea level aquifer. One of these wells (PUD-1) is located on the border of the base while the other(PUD-2) is located 1500 feet to the north. These two municipal supply wells along with some privateresidential wells were sampled in August 1989. A slight detection of TNT was recorded for PUD-2(0.01 ppb) along with 0.02 ppb RDX in residential well 5L1. Both of these reported values, however,were below their respective sample detection limits. These wells were apparently sampled again inOctober 1989 but the results of this sampling were not located.1 It is not clear whether PUD-1 andPUD-2 are still in service as part of the Vinland public water supply. No ordnance compounds were found in any other off-site wells.
Remedial efforts have been ongoing at Site A since December 1994. A leach basin was constructedin which burn area soils were placed. Leachate collected from rain infiltration is routed to aactivated carbon filtration system where ordnance is removed from the leachate before discharge intothe leach basin. Leachate will be directly discharge to this area once cleanup requirement are met. Inaddition to soil cleanup, a groundwater extraction system is removing contaminated groundwaterthat is also routed to the filtration system for ordnance removal before discharge to a drainage area east of Cattail Lake.22
The most recent sampling of monitoring wells in the burn area shows that RDX in the shallowaquifer remains high with a maximum of 200 ppb. However, sampling of shallow monitoring wellsMW-28 and MW-30 located on the north border of the base near Vinalnd have consistently detectedno contaminants or only trace levels of RDX and TNT. The trace detections in 1989 of TNT inPUD-2 and RDX in private well 5L-1 are not consistent with the expected movement ofgroundwater or the lack of significant detections in MW-28 and MW-30. Since the detections in 5L-1 and PUD-2 were below the method detection limit, verification would have been necessary toconfirm these results. Although these wells were apparently sampled again one week later, the datawere not located. It is possible that the detections in PUD-1 and 5L1 were the result of fieldcontamination or lab error.
It is unlikely that the small amount of TNT in the shallow aquifer at Site A could reach PUD-2which is located in the sea-level aquifer about 1,500 feet northeast of the base boundary. It is alsonot clear whether this well is still in service. Private well 5L-1 is located at a depth of 70 feet in theshallow aquifer and is approximately 500 feet downgradient from MW-28 which is screened atabout 80 feet. Regular sampling of MW-28 since 1991 has failed to detected any ordnancecompounds. However, detection limits in MW-28 sample analyses have risen from 0.02 ppb in1991 to 1.5 ppb during the most recent sampling in February 1999.
Ongoing remediation along with the lack of any off-base groundwater contamination indicates littlechance for future exposure of Vinland residents to contaminants in drinking water originating fromSite A. Shallow groundwater at Site A moves primarily to the west in the direction of Cattail Lake.The lack of any recent detections of TNT or RDX in monitoring wells MW-28 and MW-30indicates that drinking water wells in the Vinland area are not threatened by Site A contaminants.
Contaminants in groundwater at Site A do not appear to be moving in the direction of Vinland andare not likely to be a threat to drinking water wells along this border of the base.
Annual sampling of MW-28 as part of the ongoing remediation at Site A should ensure that anycontaminants moving in groundwater towards Vinland will be detected. Detection limits for samples analyzed from MW-28 should not exceed the current level of 1.5 ppb.
Site F (OU-2) is the location of a former unlined wastewater lagoon that received effluent fromdemilitarization activities from 1960 through 1970. These activities resulted in contamination of theshallow aquifer with ordnance compounds including RDX, TNT, 2,4-DNT and 2,6-DNT. Nocontaminants have been found in the deeper sea-level aquifer. Remedial efforts have been ongoing atSite-F since December 1994. The current containment and monitoring system in place at Site-F willeither prevent or detect further movement of contaminants, specifically RDX, in the shallow aquifer.Since off-base wells are still about one mile from the leading edge of the RDX plume, currentmonitoring will provide ample warning should the plume move beyond the containment system. Theinactive base supply well SO6 should be sampled to ensure that ordnance compounds have not reached the sea-level aquifer.
Site F is the location of a former unlined wastewater lagoon that received effluent fromdemilitarization activities from 1960 through 1970 (Figure 13). Effluent discharged to the lagoonconsisted primarily TNT, RDX, 2,4-DNT and 2,6-DNT and was commonly known as "pink water."The lagoon was excavated to an average depth of about five feet in the till overlying the shallowaquifer and had an overflow channel that ran south to an undetermined area. Contaminants inlagoon wastewater have penetrated into the surface aquifer below but have not been detected in thedeeper sea-level aquifer. This sea-level aquifer is protected by a thick (80-100 foot) layer of dense,low permeability silt located beneath the shallow aquifer that retards the downward movement ofcontaminants. Groundwater in the shallow aquifer flows northwest at a rate of 200 feet per year anddischarges via seeps to an unnamed creek that runs through the town of Old Bangor. Downwardmovement of groundwater through the aquitard separating the shallow and sea-level aquifers isestimated at one foot per year. The sea level aquifer is a major regional water bearing unit thatextends across the Kitsap Peninsula and discharges to both Hood Canal and Liberty Bay.2
The primary ordnance compounds detected in the shallow aquifer include RDX, TNT, 2,4-DNT and 2,6-DNT. Groundwater sampling during the RI/FS between September 1990 and April 1992 detected RDX, TNT and total DNT in the shallow aquifer at maximums of 7,120, 8,900 and 540 ppb, respectively. The most recent sampling of monitoring wells at Site F in January 1999 detected an overall maximum level of RDX at 1,300 ppb in MW-39, TNT at 5,800 ppb in MW-31 and total DNT at 358 ppb in MW-31.22
Remedial efforts have been ongoing at Site F since December 1994. Soil from the lagoon area wasexcavated, composted and returned prior to the installation of an infiltration barrier. A pump-and-treat groundwater remediation system was put in place to both contain and remove contaminants from the plume.
Groundwater monitoring and remediation of the shallow aquifer is ongoing at Site-F. Although highlevels of RDX, TNT and DNTs are still present near the former lagoon, only RDX has traveledsignificantly beyond this source area. RDX has never been detected in MW-65. MW-65 is thefurthest monitoring well from the source (approximately 4,200 feet northwest) and has been sampledsince May 1998 (Figure 14). However, MW-63, approximately 3,000 feet from the source,contained 120 ppb RDX during the most recent sampling in January 1999. MW-64, locatedbetween these two wells and about 3,500 feet from the source, showed 2.5 ppb RDX in the January1999 sampling and represents the outer edge of the RDX contaminant plume.23 MW-64 is estimatedto be 5,400 feet away from the northwest border of the base where the residential communities ofOlympic View and Old Bangor are located.
Although there has never been any contamination detected in the sea-level aquifer, no sampling data after 1992 was located for wells in this aquifer. As shown in Figure 14, base supply well SO6 is located about 1,500 feet northwest of MW-64 in the sea-level aquifer at a depth of 330 feet. This well has not been used as a drinking water supply since 1990 when sampling detected trace levels of VOCs.23 Off-base private and municipal supply wells are located in the communities of Old Bangor and Olympic View in both the shallow and sea-level aquifers downgradient from Site F. The contaminant plume is currently about one mile from these wells.
The groundwater containment system appears to be keeping the RDX plume from moving beyond the reintroduction wells as evidenced by the low RDX results in MW-64 and the consistent lack of RDX in MW-65 since September 1997. Regular sampling of monitoring wells located at the leading edge of the RDX plume should ensure that any advancement of the plume in the shallow aquifer will be detected. These monitoring wells lie between the plume and the reintroduction wells that serve to return treated water back into the ground. Since it is essential to keep the leading edge of the plume in front of the reintroduction wells, regular sampling of monitoring wells at the edge of the plume is required to ensure proper operation of the groundwater containment system. Recent recommended changes to the Compliance Performance Monitoring Plan for Site F include continued sampling of primary wells twice per year and secondary monitoring wells once every two years.24
The current containment and monitoring system in place at Site F will either prevent or detectfurther movement of contaminants, specifically RDX, in the shallow aquifer. Since off-base wellsare still about one mile from the leading edge of the RDX plume, current monitoring will provideample warning should the plume move beyond the containment system.
Although the sea-level aquifer is protected by an aquitard, there has been no sampling of this aquifernear Site F since 1992. The presence of inactive supply well SO6 provides a means to sample thesea-level aquifer at the leading edge of the plume.
Monitoring wells at the edge of the Site F contaminant plume should continue to be sampled forordnance compounds on a regular basis to ensure that the plume does not reach beyond thecontainment system and threaten shallow wells in the Old Bangor and Olympic View communities.
Bangor supply well SO6 is currently not in use. If this well is ever re-activated for use as a drinkingwater supply, it should be sampled for ordnance compounds in addition to any other required sampling.
All of the Bangor supply wells are located in the sea-level aquifer which is separated from theshallow aquifer by an aquitard of varying thickness. No groundwater contaminants have beendetected in the sea-level aquifer near Site F (OU-6) or Site D (OU-2) which are the only sitesproximal and upgradient of active on-base supply wells. A contaminant plume containing ordnancecompounds originating from Site F does exist in the shallow aquifer. The currently inactive basesupply well, SO6, is approximately 1,500 feet beyond the leading edge of this plume while theactive base supply wells are 2 miles away. Periodic sampling of SO6 will ensure that any migration of contaminants from shallow groundwater near Site F into the sea-level aquifer will be detected.
The locations of each Bangor public supply well is shown in Figure 6. Currently, only wells SO1,SO2, SO4 and SO9 are active sources of drinking water for the base. All other sources shown inFigure 6 are either abandoned, inactive or used as supplies for irrigation and/or fire prevention. Allof the on-base supply wells are located in the sea-level aquifer which is separated from the shallowaquifer by an aquitard of varying thickness.24 Based on the location of hazardous waste with respectto the active supply wells, only Site D and Site F are in a location that could potentially impact thesesources. Both of these sites have contaminant plumes associated with past demilitarization activities.
Each of the active supply wells are required to be sampled regularly for VOCs, metals othercontaminants. However, ordnance compounds are not part of this requirement. Low levels ofordnance compounds were detected in the perched aquifer at Site D but no ordnance compoundswere found in the shallow aquifer. A significant groundwater contaminant plume containingprimarily ordnance compounds does exist in the shallow aquifer at Site F. As discussed previously,RDX is the major contaminant of concern in this plume and has been detected as far as 3,500 feetfrom the source. The edge of this plume is still approximately two miles from the nearest active on-base supply well (SO1). No contaminants have been detected in the sea-level aquifer near Site F.The shallow and sea-level aquifers are separated by an aquitard that slows downward vertical migration of water and contaminants.
Groundwater contaminants have not been detected in the sea-level aquifer near Site F or Site Dwhich are the only sites proximal and upgradient of active on-base supply wells. The contaminantplume at Site F appears to be restricted to the shallow aquifer but has extended about 3,500 feetbeyond its source. An inactive supply well (SO6) is located in the sea-level aquifer about 1,500 feetdowngradient of the plume. Discussions with Bangor personnel indicate that VOCs were detected inthis well before its use a supply well was discontinued. These detections may have been the result ofimproper sampling.
SO6 is the most likely base supply well to be impacted if contaminants were to reach the sea-levelaquifer. If this well is ever put back into use as a drinking water supply, contaminant testing shouldbe conducted that includes ordnance compounds. In addition, periodic testing of this well wouldensure that any ordnance compounds that might reach the sea-level aquifer would be detected wellbefore they reach active, downgradient supply wells.
Contaminants have not been detected in the sea-level aquifer at any sites that could threaten on-basesupply wells. Active on-base supply wells are currently protected from ordnance compound in theshallow aquifer near Site F by distance (approximately two miles) and an aquitard that willsignificantly slow downward migration to the sea-level aquifer. One inactive on-base supply well,SO6, is close to the shallow aquifer plume associated with Site F. Sampling of this well wouldensure early detection of an ordnance compounds that move from this plume into the sea-level aquifer.
Bangor supply well SO6 is currently not in use. If this well is ever re-activated for use as a drinkingwater supply, it should be sampled for ordnance compounds in addition to any other required sampling.
1,2-Dichloroethane (1,2-DCA) is a volatile, organic chemical used as a cleaning solvent. It is mobilein groundwater and will volatilize from water during showering and other domestic uses. Therefore,exposure to 1,2-DCA in drinking water must consider the ingestion, inhalation and dermal routes.
Mountain View Road residents were exposed to 1,2-DCA in drinking water prior to the provision ofan alternate water supply. The dose estimated for a child exposed over several years to 1,2-DCA indrinking water at the maximum detected level of 4.6 ppb is more than 500-fold below ATSDRintermediate-MRL.16 This comparison indicates that no acute or intermediate adverse health effectsare anticipated at this level of exposure. Although a chronic oral MRL or RfD is not available, thelimited chronic exposure animal studies available indicate that LOAELs and NOAELs are wellabove the estimated intake given for this exposure scenario.
1,2-DCA is classified by EPA as a Group B2 probable human carcinogen based on adequate animalevidence and no human data.18 Rats given high oral doses of 1,2 -DCA showed increases in lung and blood vessel tumors.
2,4,6-Trinitrotoluene, commonly know as TNT, is an explosive chemical widely used in a variety of munitions and other explosive devices. The toxicity of TNT has been observed largely through the study of workers exposed in munitions plants during the First and Second World Wars. Damage to the liver and decreased red blood cell counts (anemia) are the most significant effects of TNT exposure and have been demonstrated in both munitions workers and laboratory animals. The ability of TNT and other organic nitrates to oxidize hemoglobin, decreasing the oxygen transport capability of the blood, supports the association between anemia and TNT exposure. Significant reproductive effects such as decreased testes weight have also been seen in animals at very high doses. This finding is supported by one reproductive study of munitions workers that showed sperm malformations.18,25
It is important to note the difference in exposure of munitions workers and those involved indemilitarization activities at Bangor. The assessment given for on-site workers (see page 24)considers exposure to TNT contaminated soil as opposed to direct exposure of munition workers topure TNT. This consideration results in a lower dose estimate for Bangor workers. Although thedose calculated for these workers exceeds the RfD, it is well below the actual toxic effect level uponwhich the RfD is based. This toxic effect level, known as a LOAEL, was the lowest dose thatshowed toxicity (liver damage) in a study of dogs fed high doses of TNT in their diet. As describedin the introduction of the Public Health Implications section, LOAELs are divided by "safetyfactors" to yield RfDs.
TNT is classified by the EPA as a Group C possible human carcinogen based on sufficient evidence of carcinogenicity in animals and no evidence in humans. Rats given very high doses of TNT in their diet developed bladder cancer. One recent study did show an increased risk for leukemia in a German population that lived near a munitions factory during the Second World War.18,25 However, this study is complicated by a lack of exposure characterization and the likely presence of other chemicals. Cancer risk associated with workers involved in demilitarization activities is discussed in the Public Health Implications section under completed exposure Pathway 3 (see page 24).
2,4 and 2,6-DNT are the primary isomers found in technical grade dinitrotoluene (DNT). Most of the DNT produced is used to manufacture other chemicals. However, it is also used in the manufacture of munitions which explains its presence at Bangor. The toxicity of DNT is has been largely observed from animal studies. DNT has displayed a wide range of toxicity in high dose animal studies that includes effects on the liver, kidneys, blood, immune, nervous, developmental and reproductive systems.26
There is some evidence in humans to support these effects at lower doses. Studies of workersoccupational exposed to DNT have reported general central nervous system effects such as dizzinessand headaches. One study found an increase in mortality from heart and circulatory disease inworkers exposed to DNT. A study of workers employed in a munitions factory during the SecondWorld War noted an increase in cyanosis and anemia although exposure was not quantified andprobably involved other chemicals. The ability of DNT and/or its metabolites to formmethemoglobin supports the finding of cyanosis and anemia in these workers.26
On-base workers could have been exposed to DNT through direct contact with surface soil duringpast demilitarization activities. Exposure of workers to the levels of 2,4 and 2,6-DNT found in soilat Bangor is estimated to result in doses much lower than those used in animal studies or predictedfor workers employed in manufacturing and munitions plants. In addition, estimated doses are wellbelow the RfDs and MRLs established for these two chemicals.
The EPA has classified the mixture of 2,4 and 2,6-DNT as a Group B2 probable human carcinogen based sufficient data in animals and no evidence in humans. High dose of this mixture caused liver tumors in rats.18 Cancer risk associated with workers involved in demilitarization activities is discussed in the Public Health Implications section under completed exposure Pathway 3 (see page 24).
Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX or Royal Demolition Explosive) is manufactured solely as an explosive compound. There is little evidence of RDX toxicity in humans although adverse effects have been found in laboratory animals. Adverse effects on the nervous system, liver, kidneys and blood have been noted in animals given high doses of RDX. An oral RfD was developed by EPA based on inflamation of the prostate gland seen in rats given high doses of RDX.18,27
On-base workers could have been exposed to RDX through direct contact with surface soil during past demilitarization activities. The dose estimated for a worker exposed to RDX in on-base surface soil is well below this RfD. The EPA has classified RDX as a Group C possible human carcinogen based sufficient data in animals and no evidence in humans. High dose of this mixture caused liver tumors in mice.18 Cancer risk associated with workers involved in demilitarization activities is discussed in the Public Health Implications section under completed exposure Pathway 3 (see page 24).
1,3-Dinitrobenzene (1,3-DNB) is manufactured as both an explosive and a chemical intermediate. The major targets of 1,3-dinitrobenzene toxicity are the blood and male reproductive systems as demonstrated in high dose animal studies. Although no human evidence is available to support these findings, they are consistent with the toxic endpoints observed for TNT and DNT.18,28
EPA has established an RfD for 1,3-DNB based on increased spleen weight in rats given high dosesof 1,3-DNB in drinking water.18 A dose estimated for an adult exposed over several years to themaximum detected level of 1,3-DNB is 5-fold higher than this RfD. However, this dose is still2,000-fold below the actual level that produced the adverse spleen effect. As described in theintroduction of the Public Health Implications section, "safety" factors are used to set the RfD wellbelow the actual toxic effect level (i.e., LOAEL). The relative risk of a dose that exceeds the RfD can be judged by how close it is to the actual toxic effect level.
EPA has not classified 1,3-DNB with respect to its carcinogenicity (Group D). No data was located relative to the ability of 1,3-DNB to cause cancer.