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




The water table at STA is generally shallow as indicated byswamps, bogs, and waterholes. The surficial aquifer is composedof unconsolidated outwash (glacial drift deposited by meltwaterstreams beyond an active glacier or "loose rocks") underlying thelowlands of the site. Because of the high permeability of itsdeposits, the surficial aquifer constitutes the principalaquifer. The thickness of the outwash averages about 50 feetwith a maximum of 100 feet (1). Depth to the water table is lessthan 15 feet, with the groundwater gradient following the areatopography (2). The bedrock is overlain by till, a compact,unsorted mixture of clay, silt, gravel, and boulders. Thethickness of the till ranges from a few inches to at least 40feet in thickness. The thickness of the bedrock cannot beaccurately defined, but one formation (Marlboro) is more than2,500 feet in thickness.

Regional wells, in the most favorable areas, have yields of from100 to 150 gallons per minute (gpm). Groundwater divides (thecrest of the water table with flow going in opposite directionson either side) correspond to surface water divides in theunconfined surficial aquifer. A divide transects Area A10, withflows reported to be in a westerly direction toward White Pond. Flows on the east side of this divide are eastward toward StearnsMill Pond and Hop and Marlboro Brooks. Groundwater flow in thenorthern portion of STA is toward or parallel to the flow ofTaylor and Honey Brook, except in the far southwest where flow istoward Boon and White Ponds (3). Groundwater recharge areas(areas of water movement into the aquifer) are shown in Figure 4(see Appendix 1).

Groundwater and surface water flow at STA is furthercharacterized into 7 distinct watersheds (see Figure 3 inAppendix 1) (3). Study areas within each watershed are listed inTable 6 (see Appendix 2). Potential transport of environmentalcontaminants in groundwater and surface water from these studyareas is influenced by watershed characteristics. Evaluatingsuch factors is essential for determining potential exposurepathways at STA. A general description of the watershedsfollows.

The Watershed lA area encompasses the watershed of Taylor Brook above its junction with Honey Brook. It receives flow from off-site areas, including Taylor Brook from its source to Cutting Pond, from Vose Pond, and the northern parts of the Annex adjacent to the Town of Maynard which were formerly part of the Annex. Puffer Pond itself is included in this group both because it receives discharge of groundwater from under the sites west of it, and because it is to be the site of a bioaccumulation study during which sediment, surface water, and fish tissue will be analyzed.

This part of the Taylor Brook watershed receives runoff and groundwater flow from two major groups of sites. These are Al, A2, P42, and P45 around Taylor Brook, and P11, P13, P23, and P26 which discharge to Honey Brook. This area covers much of the north end of the Annex with the exception of the small areas draining directly to the Assabet River (Watersheds 3 and 4). All the sites within this area lie on areas of sandy glacial outwash within short distances of wetland or surface streams, and the well-drained soils and shallow water tables would imply rapid impacts on groundwater, adjoining wetlands, surface water, and sediments, if contamination is present.

A small part of the Annex adjacent to Sudbury Road and the Massachusetts Firefighting Academy drains south into White Pond, which has no outlet. White Pond is used as a municipal water supply source by the Town of Maynard. The pond discharges underground south to Hop Brook. The entire area of the detached portion of STA where all Watershed 2 sites are located also drains into Hop Brook, although A11, A12, P36, and P37, and possibly parts of P28 may discharge first to Marlboro Brook, a small tributary of Hop Brook. Hop Brook discharges to Sudbury River, unlike most of the Annex, which drains to Assabet River.

A small unnamed tributary of the Assabet River starts at a wetland just north of White Pond Road and west of the former Massachusetts Fire Fighting Academy (study area P13) and discharges directly to the Assabet River. The entire tributary is only 2,000 feet in length, but it runs between or adjacent to a number of sites to be investigated including study areas A8, P9, P10 and the newly listed P57 at the site of the demolished Building S449.

This drainage area is difficult to delineate, as the small unnamed tributary originated in a large area of marsh with very little topographic relief, south of the hill on which the U.S. Air Force Weather Radar Laboratory is sited. Relative volumes of flow suggest that this tributary drains much less of this marsh than does Honey Brook to the east. The two sites within this watershed are either definitely draining towards this brook (Study Area A6), or, most probably discharging to groundwater that enters this brook (study area P22). To further characterize groundwater flow patterns in this and other STA watersheds, a facility-wide groundwater model is planned flow model using MODFLOW, a 3-dimensional USGS model. It will help establish groundwater flow patterns, volumes of groundwater flow off site through various media, and the relative proportions of flow. It is also capable of providing confirmation of probable mass transport of contaminants from specific sites to supplement individual site studies.

Part of the main portion of the Annex on its southwest side is very flat and has extensive wetlands. Some of these appear to discharge via groundwater to Lake Boon, southwest of the site. Lake Boon, in turn, discharges over a dam into the Assabet River.

On the east side of the main part of STA, some sites are located in areas that drain into a large area of swamp within which are Crystal Lake (formerly Bottomless Pond) and Willis Pond. This area drains toward Run Brook, which in turn drains into Hop Brook and the Sudbury River. Three sites, P1, P2, and P3, lie on or adjacent to a hill of ground moraine (till), which slopes and drains toward the swamp.


  1. U.S. Environmental Protection Agency. Hazard Ranking SystemEvaluation, Sudbury Annex Facility, Sudbury, MA. December1988.

  2. ATSDR. Record of Activity for telephone communication: Richard Collins, ATSDR, with Christine Williams,Environmental Management Office, Fort Devens, September 22,1993.

  3. Ecology and Environment, Inc. Technical Plan Addenda, PhaseII Site Inspections, Remedial Investigation, Fort DevensSudbury Training Annex, Massachusetts. June 1993.


Comparison values for ATSDR public health assessments arecontaminant concentrations in specific media used to selectcontaminants for further evaluation. The values provide guidelines for estimating the dose at which health effects mightbe observed. Comparison values and the units used to quantitatecontaminant concentrations that appear in the EnvironmentalContamination and Other Hazards and the Public HealthImplications sections of this public health assessment are listedand described in the following paragraphs.

Comparison Values
* CREG = Cancer Risk Evaluation Guide
* EMEG = Environmental Media Evaluation Guide
* MCL = Maximum Contaminant Level (µg/L)
* MRL = Minimal Risk Level (mg/kg/day)
* RfD = Reference Dose (mg/kg/day)
* RMEG = Reference Dose Media Evaluation Guide
* ppm = milligrams per liter (mg/L water)
   milligrams per kilogram (mg/kg soil)
* ppb = micrograms per liter (µg/L water)
   micrograms per kilogram (µg/kg soil)
* kg = kilogram
* mg = milligram
* µg = microgram
* L = liter

Cancer Risk Evaluation Guides (CREGs) are estimated contaminant concentrations that would be expected to cause no more than one excess cancer in a million (10E-6) persons exposed over a lifetime. CREGs are calculated from EPA's cancer slope factors.

Environmental Media Evaluation Guides (EMEGs) are based on ATSDRminimal risk levels (MRLs) and factor in body weight andingestion rates. An EMEG is a contaminant concentration inwater, soil, or air at which daily human exposure is unlikely toresult in adverse noncancerous effects.

Maximum Contaminant Levels (MCLs) are contaminant concentrationsthat EPA deems protective of public health (considering theavailability and economics of water treatment technology) over alifetime (70 years) at an exposure rate of 2 liters of water perday (for an adult).

A Minimal Risk Level (MRL) is an estimate of daily human exposureto a chemical (in mg/kg/day) that is likely to be without anappreciable risk of deleterious effects (noncancer) over aspecified duration of exposure. MRLs are calculated usinginformation from human and animal studies and are reported foracute (< 14 days), intermediate (15-364 days), and chronic (> 365days) exposures. MRLs are published in ATSDR ToxicologicalProfiles for specific chemicals.

EPA's Reference Dose (RfD) is an estimate of the daily exposureto a contaminant unlikely to cause adverse noncancerous healtheffects.

Reference Dose Media Evaluation Guides are derived by ATSDR fromEPA oral reference doses and factor in body weight and ingestionrates. An RMEG is a contaminant concentration in water or soilat which daily human exposure is unlikely to cause adversenoncancerous health effects.

The Food and Drug Administration (FDA) action level is themaximum allowable level of poisonous and deleterious substancesin human food and animal feed.


Health Consultation

For the

Fort Devens Sudbury Training Annex
Sudbury, Massachusetts

Prepared by
Federal Programs Branch
Division of Health Assessment and Consulation

November 23, 1992

Background and Statement of Issues

A site tour of the Sudbury Training Annex (STA) NPL site was conducted by staff members from ATSDR's Fedeal Programs Branch (Lorna Bozeman, Gary Campbell, and Richard Collins) and Regional Operations (Susanne Simon) on October 7, 1992 (1). During that site tour, ATSDR staff members saw physical hazards that warrant prompt public health action.

  • Two abandoned buildings (Raytheon Building T106 [parachute building] and Raytheon Building T104), designated as sites P36 and P37 on the southern part of the post (Figure 1), are unsecured, contain asbestos (2) and abundant debris, and appear structurally unsound.

  • Evidence of trespassing at sites P36 and P37 was observed. Perimeter fencing is in disrepair and the Army has difficulty keeping trespassers of STA property.


Abandoned building T104 and T106 are unsecured (no doors), appear structurally unsound -- metal building with holes in the roof and walls -- and contain abundant debris (barbed wired, broken glass, scrap metal, and pieces of asbestos insulation that have falled off the walls and ceiling). Building T106 is a shed-type structure; one end is completely open. Part of the broken-down, chain-link fence remains along one side of T106. Building T104 is accessible through a garage door-size opening. Because the buildings are not maintained, they will continue to deteriorate. In addition to the physical hazards associated with the buildings, the nature and extent of environmental contamination at these sites are currently unknown. However, the Army plans to evaluate these areas during the remedial investigation of STA (3).

Health assessors have seen evidence of unauthorized use of the buildings and area. The walls of both buildings have graffiti, the windows are shattered, and bullet holes are in the sheet metal. Motorcycle tire tracks and remnants of fires set on the cement floor of building T106 were also seen. The buildings are 0.3 mile from the nearest residence. A dirt road provides direct access to the area from Moore Road in Sudbury (Figure 1). The Army has placed a cable and cement barricade across the road, but trespassers can easily enter the area on foot, or on motorcycles or off-road vehicles. During the site tour, we saw a person unload a motorcycle from a van at the access point. Approximately two weeks later, the Installation Restoration Program manager from STA visited the site and said that two children were skateboarding in T106 when he arrived (4).

Restricting access to several areas on post that are currently being investigated for possible contamination or that have physical hazards has been a chronic problem at STA. Attempts to limit access by placing cement barriers or cables across road have been ineffective (5). Although some contaminated areas appear secure (enclosed by 8-foot, chain-link fences), perimeter fencing is generally not well maintainted at STA. The Army is currently installing additional fencing along the northern part of STA, bordering the towns of Hudson and Stow; that action should further restrict access.


  • The hazards associated with abandoned buildings T104 and T106 make them an "attractive nuisance" of public health concern.

  • Although the Army has been diligent in its efforts to keep trespassers out of the area, the attempts have been unsuccessful.


  • To protect public health, ATSDR recommends that the Army demolish abandoned buildings T104 and T106.

  • Because both buildings contain asbestos, appropriate measures should be taken before and during demolition to prevent people from being exposed. Once the buildings are demolished, all debris should be removed.

  • Characterizations of the nature and extent of environmental contamination at these sites should be pursued as planned (6). ATSDR health assessors will evaluate additional data once it is available.

Lorna Bozeman, MS
Environmental Health Scientist


Richard Collins, MSEH, RS
Environmental Health Scientist


  1. Fort Devens Sudbury Training Annex Trip Report. Agency for Toxic Substances and Disease Registry. October 16, 1992.

  2. Contract for Asbestos Removal, Clean-up, and Disposal, Buildings T104 and T106, Sudbury Training Annex. October 16, 1986
  3. .
  4. U.S. Environmental Protection Agency, Region I and the U.S. Department of the Army. Final Federal Facilities Agreement, Sudbury Training Annex. November 1991.

  5. Jeff Waugh, STA Installation Restoration Program manager. Personal Communication. October 29, 1992

  6. Jeff Waugh, STA Installation Restoration Program manager. Personal Communication. October 7, 1992.

  7. OHM Remediation Services Corp. Final Master Environmental Plan, Fort Devens Sudbury Training Annex. January 1992.

Fort Devens Sudbury Training Annex - South Post
Figure 1. Fort Devens Sudbury Training Annex - South Post


There are three general forms of mercury: elemental, inorganic,and organic. The chemical form influences the absorption anddisposition of mercury in the body. Elemental mercury is notwell absorbed from the gastrointestinal tract, so it is not amajor source of mercury poisoning outside of occupationalsettings where workers may be exposed to mercury vapor in theair. Only about 7 percent of inorganic mercury in food isabsorbed while organic mercury is nearly completely absorbed. The toxic manifestations of inorganic mercury are renal whereasthose for organic mercury are neurological (1).

Food is the main source of exposure to mercury in non-occupationally exposed populations. Fish and other seafoodaccount for most of this exposure. Mercury in food ispredominately methylmercury (organic) and is toxic to the nervoussystem. Exposure to forms of mercury other than methylmercuryseldom reaches toxic concentrations in air, water, and food.

Two episodes of environmental mercury contamination showed thebehavioral effects of prenatal exposure to methylmercury. In avillage on the Minimata Bay in Japan, villagers were exposed tomethylmercury through the ingestion of contaminated fish duringthe 1950s. Cerebral palsy and mental retardation werewidespread. Congenital malformations were rarely seen. Theneurological problems seen in the villagers are called "MinimataDisease". Another outbreak occurred in Iraq in 1971 when seedgrain was accidentally treated with a methylmercury fungicide. The exposure here was to a much higher concentration and over ashorter period of time than that at Minimata Bay. The childrenwho were exposed prenatally had impaired mental development andproblems with speech and motor function (2).

Exposure to both elemental mercury or methylmercury can affectthe central nervous system. Elemental mercury exposure canresult in depression and emotional instability. Methylmercuryexposure can result in changes in vision, decrease incoordination, and impaired taste, smell, and hearing. Inorganicmercury primarily affects the kidneys causing a condition similarto acute tubular necrosis. The inorganic form can also irritatethe gastrointestinal tract causing ulceration (2).


1. Doull J, Klaassen CD and Amdur MO. Toxicology: The BasicScience of Poisons. New York: Macmillan Publishing Co., Inc. 1980.

2. Tarcher AB, ed. Principles and Practice of EnvironmentalMedicine. New York: Plenum Medical Book Company. 1992.


April 22, 1994

Department of the Army
Headquarters Fort Devens
Mr. Thomas Strunk
Fort Devens, Massachusetts 01433-5190

Dear Mr. Strunk:

The Agency for Toxic Substances and Disease Registry (ATDSR) recommends that the Army sample private drinking water wells on Dawes Road in Stow.

New groundwater data for Fort Devens Sudbury Training Annex study areas P31/P58 suggest that site contaminants could migrate to off-site drinking water wells. These data are reported in the March 1994 Ecology and Environment Phase II Site Investigations Report.

Three new monitoring wells were installed between study areas P31/P58 and private drinking water wells on Dawes Road. Groundwater samples collected from these wells in September and December 1993 contained metals at concentrations above drinking water standards. Unfiltered samples contained maximum levels of 91 ppb arsenic (the maximum contaminant level is 50 ppb) and 43 ppb lead (the action level is 15 ppb). Arsenic levels in filtered groundwater samples also exceed the standard. Trace levels of pesticides and the explosive 1,3-dinitrobenzene were also detected. No volatile organic compounds were found.

The six homes on Dawes Road are about 500 feet from study areas P31/P58 and are in the direction of groundwater flow. Water from the residential wells probably is not filtered before use. To determine if people are drinking water containing these chemicals, we recommend that you analyze duplicate unfiltered water samples from the private wells. Based on findings from P31/P58, we suggest that analyses include TCL organics and pesticides, TAL metals, herbicides and explosives.

When the sampling is completed, please submit the entire data package to ATSDR for review. If contaminated levels in any water samples exceed primary drinking water standards, provide alternate water to all residents (6 homes) as a precautionary measure until data can be fully evaluated. ATSDR will review data and make recommendations regarding public health actions. If the private wells are not contaminated, quarterly sampling of P31/P58 monitoring wells should continue because increases in contaminant levels would pose a greater threat to off-site drinking water wells.

If you need additional information, please contact Ms. Susanne Simon, ATSDR Regional Representative, Region I, (617) 860-4619, or Ms. Lorna Bozeman, Environmental Health Scientist, (404) 639-6070. We appreciate your ongoing efforts to address public health issues at the Annex.

Sincerely yours,

Sally E. Shaver
Federal Programs Branch
Division of Health Assessement and Consultation

Diana Feireisel, USAEC
C. Todd Jones, USAEHA
Robert Lim, EPA
Debra Darby, MDEP
Jeff Purvis, MDPH
Stow Board of Health
Four Town FOCUS


Listed below are comments received from the public regarding theDecember 1993 Public Health Assessment for the Sudbury TrainingAnnex. We have not included comments on accuracy of statedfacts. If the accuracy of a statement was questioned, thestatement was verified or corrected.


Many of the questions raised by the public concernedcontamination levels. We emphasize that human exposure, notcontamination levels, is the focus of a health assessment. Ourconclusion is that people were not being exposed to contaminantsat levels that could pose health concerns. Most people were notbeing exposed at all, either through groundwater, soil, or thefood chain. Even those people who trespassed onto Annex propertywould not have come into contact with contaminated soil oftenenough to cause health consequences.


Comment 1: We recognize that ATSDR's role in the CERCLA processis solely to identify sites that may present an actual past orpresent threat to public health. We also recognize thesignificant differences between the Public Health Assessment andthe Risk Assessment, which is used to support the selection of aremedy. Both the Public Health Assessment and the RiskAssessment require identification of sources, pathways, andreceptors. The existing environmental data permit identificationof contaminant sources, and the demographics of the area allowidentification of potential receptors, both of which arediscussed in the Public Health Assessment. In order to showpublic health impact, however, the sources and the receptors mustbe linked by pathways. With a few exceptions, this has not beendone by ATSDR. In order to establish a completed pathway, it isnecessary to show, through analytical data or modeling, that acontaminant is present in the environmental medium (soil, air,fish, sediment, etc.) that the receptor contacts.

The public health assessment does present pathways analysis. Little potential exists for people to contact contaminatedenvironmental media on the Annex; therefore, there is limitedpublic health impact. Public health actions to reduce the fewways that people could continue to be exposed are to restricttrespassing, and post Puffer Pond as catch and release fishing. Those actions have been implemented. The other public healthconcern is metals in groundwater at study areas P31/P58. We haverecommended that the Army sample off-post private wells closestto those study areas.

Comment 2: The environmental data from the RemedialInvestigation at the Annex identify numerous sources of toxicmaterials, but the environmental data are not always adequate tocharacterize pathways that may be important at the site. Theinvestigations at the Annex have been focused on discrete areasof suspected contamination, with minimal effort to characterizethe entire site or to look at relationships between thecontaminated sources. Although we realize that this was not theintent of the Remedial Investigations conducted at this site, wefeel that ATSDR should have acknowledged the data gaps and theinadequacy of the existing data to characterize pathways.

The sampling strategy used to characterize environmentalcontamination at the Annex is appropriate. Sampling is notwarranted in areas where historical activities would not lead toenvironmental contamination. In areas where more data are neededto evaluate potential pathways, broader sampling is planned orhas been carried out.

Comment 3 (summary): What schedule will be used to review newdata on the Annex? How will ATSDR obtain notice of this data? We would like a clarification of this statement.

New data, primarily environmental, relating to public healthquestions at the Annex are reviewed when they become available. A Memorandum of Understanding requires that the Army provideATSDR with data as it is available. Through ATSDR regionalrepresentatives, EPA also provides copies of split sample datafor review. Please refer to the public health action plan inthis document for further activities.


Comment 4 (Land Use): We were told in the past that no chemicalswere stored any longer on the Annex. Which bunkers containthese materials and are they being stored compliant with federallaw?

The comment refers to the sentence, "The Army currently leasessome of the bunkers to state and federal agencies for storage ofequipment, supplies, chemicals, and herbicides (14).", which issomewhat misleading. Annex bunkers are used for storage ofequipment and supplies and the sentence has been reworded toreflect that use. Only a few bunkers contain any chemicals, andthose chemicals are limited to an isolated jug of lubricatingoil, or a drum of thinner, etc. Bunker 303 was used to storeUreabor, but that herbicide is no longer stored on the Annex. The Army is further investigating areas in and around bunkerswhere there is a potential for chemical contamination. Forspecific information about materials stored in Annex bunkers, thereader is referred to pages 8-8 to 8-22 of OHM RemediationServices Corporation, Final Site/Remedial Investigation Report,December 31, 1993.

Federal law pertaining to storage of chemicals is part of CERCLAas amended by SARA, Title III, Emergency Planning Community Rightto Know. There is no requirement for storage conditions, but thecommunity Local Emergency Planning Committee (LEPC) must beinformed if the quantity of a specific chemical being storedexceeds the threshold planning quantity listed in 40 CFR Part355.

Comment 5 (summary): The specific study areas with thedesignation of "indeterminate health assessment" should bedefined.

Study areas P2, A6/P22, A9/P12, and P31/P58 were classified asindeterminate public health hazards because limited groundwaterdata were available. Providing specific study area numbers inthe summary will provide little clarification to the reader. Details concerning groundwater issues relating to those studyareas are presented in sections entitled EnvironmentalContamination and Other Hazards and Pathways Analyses (pp 20-23,35-36).

Comment 6: The second paragraph of the summary should specifywhat materials were used in R&D.

The summary is an overview of the entire public health assessmentand not meant to be a detailed section. Specific informationabout research and development activities are presented inAppendix 2 (Tables 1 and 2) and in Appendix 4: Site History, Description, and Observations.

Comment 7 (background): A description of the types of materialsand wastes should be included.

Page 3 Paragraph 2 refers the reader to Table 1 and specificreferences where more detailed information is provided. Table 2also lists additional information related to types of materialsand wastes.

Comment 8 (summary): Herbicides should be added to the list ofcontaminants found at the Annex.

We have added herbicides and explosives to the summary. Specificinformation about these substances are discussed in theEnvironmental Contaminants and Other Hazards section of thepublic health assessment.

Comment 9: ATSDR did not address the issue of elevated levels ofDDT in the Annex.

We include DDT among the chemicals detected on the Annex atlevels of potential health concern. To prevent exposure tocontaminants on the Annex, we currently recommend that peoplerefrain from trespassing. In coordination with EPA andMassachusetts Department of Environmental Protection, the Army isfurther evaluating the nature and extent of contamination at theAnnex. Areas contaminated with chemicals at levels of healthconcern will be cleaned up before the property is reused.

Comment 10: In the section "Environmental Contamination andOther Hazards", and Tables 4,5,8,9, Comparison Values areprovided only for those chemicals that exceeded them. Even amongthese it is not totally clear what value, out of those possiblefor a chemical, was chosen. For example it is not clear why anRMEG of 30 ppm was chosen for DDT and a CREG of 2 ppm was chosenfor DDE, since DDT is also a carcinogen with a slope factor. TheComparison Value for Arsenic is not listed, presumably because nosoil concentration exceeded it. However, A9 is known to have anarea of excess arsenic that will be further explored. This isnot in the table for surface soil, and arsenic usually shows arisk even at background levels on many sites.

Recommendation: Confirm that appropriate comparison valuesare being used in the public health assessment. These areunderstood to be screening levels, which are usuallyconservative, so that a chemical with any potential for healtheffect on a site can be further explored. The fact that arsenicdoes not show up at all, and DDT has a value of 30 ppm placessome doubt on the conservatism in the assessment. Alternatively,provide an explanation as to how values are chosen.

All of the comparison values used for the public healthassessment are overly conservative considering potential exposurescenarios for the Annex. Based on former and current Annex landuse, possible exposures are most likely short term (trespassers). For that reason, comparison values for acute exposure were usedwhen available. Often, however, there were no comparison valuesfor acute exposure (e.g., DDE and arsenic). For such chemicals,we selected what we considered the most appropriate comparisonvalue, for the exposure scenario. In some cases, comparisonvalues for chronic exposure (i.e., daily exposure over alifetime), such as the CREGs, were used for screening. Clearly,such comparison values are overly protective for short termexposures that might be experienced by trespassers.

With regard to future land use, human exposures are not a concernbecause contaminated areas of the Annex will be cleaned up tolevels protective of public health.


Comment 11 (Groundwater Contamination): ATSDR states that theArmy is installing wells between P43A & B, this is incorrect.

The cited information is from the June 1993 Ecology andEnvironment, Inc., Technical Plan Addenda, Phase II SiteInspections, Remedial Investigations. That document states thatmonitoring wells will be placed between the alleged site(s)(Study Areas P43A and P43B) and the municipal well. However,those plans were modified in later versions of the Technical Plan(October 1993)/January 1994). Rather than install new monitoringwells, existing wells were resampled. The revised public healthassessment reflects changes in groundwater monitoring for studyareas P43A and P43B.

Comment 12: What is meant by 'explosives' in monitoring wells?Are these nitrate breakdown products? The term 'explosives'-implies a hazard that does not exist in groundwater.

Groundwater from monitoring well DM9A contained the explosives1,3,5-Trinitrobenzene and 2,4,6-Trinitrotoluene. The reader isreferred to page 19 and Table 6 of the public health assessment.

Comment 13: The groundwater pathway through on-site water supplynotes that selenium, iron, and manganese were the onlycontaminants detected in the on-site potable wells. Have these'contaminants' been linked to a site source? If not, they shouldnot be evaluated as site-related contaminants. Iron and manganeseare common in shallow wells. They are not highly toxic but theymight be indicators of leachate from buried metal. Has this beenevaluated?

Selenium, iron, and manganese are discussed because they weredetected in groundwater from on-site wells that were previouslyused as potable water sources. The concentrations of the metalsexceeded drinking water standards. Based on the findings ofenvironmental investigations at the Annex, it is likely thatthese substances are naturally occurring, rather than indicatorsof leachate from buried metal as you suggest. For clarity, theuse of the word contaminant in reference to those metals has beenchanged to metals.

Comment 14: The Onsite Water Supply section discusses theexposure doses of manganese that are expected to be safe, or tohave a health effect, by oral route. However, EPA providesdifferent health criteria for food (or soil) manganese intake andthat from water. The 2.5-5 mg/day amount from the NRC is basedon dietary manganese. The IRIS discusses manganese in water andgives the following levels as the NOAEL and LOAEL: 81.6-252.6 and1600-2300 µg/L.

Recommendation: Differentiate between dietary and wateroral ingestion of manganese when discussing health effectslevels.

Because the calculated oral exposure dose of manganese is so low,little information is gained by differentiating between dietaryand water oral ingestion of manganese. The oral exposure dosefrom water of 0.88 mg/day is conservative. It was calculatedusing the highest concentration of manganese measured in on-postpotable wells. Moreover, it assumes that 100% of the ingestedmanganese is absorbed. The 0.88 mg/kg oral dose is three timesless that the safe and adequate intake value of 2.5 mg/day. Thedose is also more than 1000 times less than the neurologicaleffects level of 980 mg/kg (59).

Comment 15: The local populace appears to be very concerned withthe contamination of groundwater in the RI sites A7 and A9, amongothers. The public health assessment does not indicate that anygroundwater contamination exceeded comparison values in A7 andthat only the two explosives exceeded them in A9. The HealthRisk Assessment shows on-site groundwater risks exceeding EPAcriteria for these two sites. For example, DCE on A9 shows a"maximum risk of 7E-5.

Recommendation: Reevaluate whether groundwatercontamination from monitoring wells on these RI sites is belowcomparison values for all but the two explosives. Consideringthe local concern over the RI sites, and their groundwatercontamination, perhaps some more discussion of what is abovecomparison values, but also their limited extent of migration,would be helpful.

We do not believe that groundwater contaminants at RI sites A7and A9 have affected drinking water supplies. Although thehealth risk assessment shows on-site groundwater risks from thesetwo sites, no one is drinking the on-site groundwater. Weevaluated whether groundwater contaminants from these RI sitescould migrate toward drinking water wells. The potential forsite contaminants to affect drinking water supplies in the areais low. The nearest potable wells are about a quarter mile fromA9. Additional groundwater monitoring was recommended and hasbeen done but the results are not yet available. We will reviewthe new groundwater data and make further public healthrecommendations if needed.

Comment 16: The taste and odor problems reported in Maynard WellNo. 3 are not associated with the Annex. Modeling of groundwaterflow in the area of the well should be done in conjunction withthe continuing studies to determine whether a potential pathwayexists between the contaminated Annex groundwater and the zone ofinfluence of the well. ATSDR should note that there is noevidence of a pathway, and should make recommendations regardingfuture use of this well.

The taste and odor problems reported in Maynard Well No. 3 arethe reasons why the well is currently closed. For future use,the well must comply with state drinking water standardsapplicable for public water systems. We state that the drinkingwater supply for the towns of Maynard, Hudson, and Sudbury havenot been affected by contamination from the Annex. Phase IIgroundwater data supports the conclusion. The bis (2-ethylhexyl)phthalate (BEHP) previously detected in groundwater frommonitoring wells nearest Maynard Well No. 3 has been attributedto improper sample handling either in the field or the laboratoryand is not site related. Maynard Well No. 3 did not contain BEHPfor any of the Phase II sampling rounds.

Comment 17: The absence of any environmental sampling andanalysis in off-site private wells is of concern. ATSDR shouldrecommend sampling of these wells as well as placement ofmonitoring wells between the private wells and the areas of knowncontamination. This type of study will permit a more confidentassessment of potential future health impact.

Monitoring wells are already in place between private wells andareas of known or potential groundwater contamination. We haverecommended sampling of private wells in areas where contaminatedgroundwater on-post could affect off-site private wells.

Comment 18: In its Initial Release of the public healthassessment, ATSDR recommended the Army conduct sampling of WhitesPond and private wells. The Army never did this, yet, ATSDRstates in the December 93 public health assessment that "The Armyhas implemented or is addressing all previous ATSDRrecommendations." This is untrue. Please explain thisinconsistency.

The statement is true. The Army has sampled White Pond and nocontamination was detected in either sediment or surface (52). The sampling is cited in both the initial release (pp 28, 45) andpublic comment release (pp 25) of the public health assessment. In addition, the Army has installed new monitoring wells in areaswhere drinking water supplies are hydraulically downgradient ofstudy areas with incomplete sampling information. If groundwatermonitoring data indicates that contaminants could affect off-sitewater supplies, we will recommend sampling those water supplies.

On March 30, 1994, we received new groundwater data for studyareas P31 and P58. Metals (arsenic, lead, aluminum, iron, andmanganese) exceed drinking water standards. The explosive 1,3-dinitrobenzene was detected in one monitoring well. Traces ofpesticides were also detected. Volatile organic compounds werenot detected. We called the Army on April 1, 1994 andrecommended off-post sampling of private wells on Dawes Road. The Army is already considering that action. A followup letterrecommending private well sampling was sent to the Army on April22, 1994 (see Appendix 11).

Comment 19: Conspicuous by its absence is any discussion of the'spring' located off Track Road in Maynard. The Maynard Board ofHealth posted this 'spring' several years ago to discourageresidents from obtaining water there.

The Track Road spring was not discussed because water samplescollected in June and October 1992 did not contain chemicals atlevels of health concern. The reader is referred to the datasummary for samples FWGW12 in Table 8-11, Site/RemedialInvestigation Report, Fort Devens Sudbury Training Annex. July1993 and January 1994 (4,15). To address the concern, a briefdiscussion of these findings has been added to the public healthassessment in the section entitled Environmental Contaminationand Other Hazards.


Comment 20: The Maynard Water transmission line is not locatedin areas of 'known contamination'. The statement that workersrepairing lines 'could' be exposed if breaks occurred in areas ofknown contamination is speculative. ATSDR should also note inits response to citizen concerns about the water line that theline is located approximately four feet below the ground surface,and at least four feet above the groundwater surface. Thetransmission line is under pressure, and lines under pressurewill leak out and not in. There is no evidence that soils underor adjacent to the line are contaminated.

On October 6, 1992, we met with personnel from the MaynardDepartment of Public Works and discussed the location of thewater transmission line. On a map provided at that meeting, thewater line is shown near Annex study areas P31, P40, A5, P11,P13, P10, and A8. For that reason we state in the public healthassessment that . . . "Workers who repair the pipe could beexposed to chemical contamination if breaks in the water mainoccur in areas with known contamination." We have also added thecomment that . . . " The Maynard Board of Selectmen state thatthe Maynard water transmission line is not located in areas ofknown contamination. The transmission line is under pressure,and lines under pressure will leak out and not in. The waterdistribution line is located approximately four feet below theground surface, and at least four feet above the groundwatersurface." The measures currently employed for repairing andflushing the water main are appropriate. However, we maintainthat repair work should be coordinated with Annex Site Health andSafety personnel. Those personnel are knowledgeable aboutspecific chemical hazards in various areas on post. If breaks inthe water transmission line occur in such areas, then analysis ofwater from the Maynard distribution point should include site-specific chemicals.

Comment 21: My concern is that I have not found informationwhich says that the water which the people of Maynard aredrinking has been tested and has been found safe to drink. Theassumption appears to be that the standard state mandated testsperformed by the Maynard water department are sufficient toinsure water quality. My assumption is that state mandated testswould not be designed to test for the exotic compounds whichmight come from the Annex.

On page 47, a community member asks: "b. The Maynard water linefrom White Pond to Summer Hill crosses STA (Sudbury TrainingAnnex). Breaks in the line could contaminate the water supply." The answer is sufficient to address the question of "breaks" inthe line and cautions about hazards to workers. The answer doesnot address the possibility of contamination from small,undetectable leaks in the line. Considering the topology of STA,it is highly unlikely that the line runs up hill from White'spond all the way across STA. Small leaks at seams indowngradient areas could siphon in contamination when negativepressures are experienced inside the line. Given that harmfullevels of toxics described in the public health assessment aremeasured in parts per billion, it would not take a very largeleak to introduce significant contamination in areas where theline runs under or close to toxic sites. I strongly urge ATSDRto recommend that the Maynard water supply be monitored where itexits STA.

Siphoning of contaminants into the water transmission line willnot occur. As previous discussed in comment 20, the watertransmission line is under pressure, and lines under pressurewill leak out and not in. Further, the Maynard Board ofSelectmen state that the water transmission line is not incontaminated areas of the Annex. As a precaution, however, ATSDRrecommended that workers coordinate repairs to the watertransmission line with Annex Site Health and Safety personnel. Those personnel are familiar with study areas on the Annex wherethere is chemical contamination. If breaks in the water line dooccur in such areas, we recommended testing water at the Maynarddistribution point for specific chemicals detected in those studyareas.

Comment 22: The environmental characterization data that thePublic Health Assessment relies upon are not complete for eachpotential pathway identified by ATSDR. Of primary concern is therelative absence of surface soil quality data, particularly inareas of the Annex where trespassers may contact surface soils. Of equal concern is the scarcity of soil quality data in thevicinity of the Town of Maynard water transmission line fromWhite's Pond. The Public Health Assessment identifies soil inthis area as a potential pathway (for water department workersrepairing the line), but there are no data to permitquantification of potential risks to these workers. There isalso a lack of adequate analytical data for stream sediments,although these are identified as potential pathways in theAssessment.

Discussions of surface soil data focus on Annex study areas wheretrespassing is known to occur, where community members expressedspecific concerns, or in areas with public access. We recognizethat there are other areas on the Annex where trespassers couldcontact surface soil that have not been discussed. Thediscussion makes the point that there are areas of the Annex thatcurrently have surface soil contamination and physical hazards. The nature and extent of the that contamination is beingevaluated. Clean up measures protective of public health will beimplemented before the property is suitable for reuse. In theinterim, people can avoid potential exposure to areas of surfacesoil contamination and physical hazards, by NOT trespassing.

We do not know if the water transmission line on the Annex is inareas of chemical contamination. Although the Maynard Board ofSelectmen state that it is not, a map, provided in a meeting withthe Maynard Department of Public Works (October 6, 1992), showsthe water main near some Annex study areas. As a precaution werecommend that workers who repair breaks in the line coordinatewith Site Health and Safety personnel. The safety personnel arefamiliar with specific areas of chemical contamination and canensure that repair workers avoid potential chemical exposurewhile working on post.

Evaluation of available stream sediment data was included in theDecember 1993 public health assessment. We also stated that moresampling was planned in Phase II investigations. The Phase IIdata now available shows metals and pesticides in some on-poststream sediments and the Army plans to expand sampling in thoseareas. The Phase II data do not change the public healthimplications of the surface water and sediments pathway aspreviously stated in the December 1993 public health assessment. These are that people would not come into contact with surfacewater or sediment often enough (given the limited nature, bothconcentration and extent, of contamination) to cause adversehealth consequences.

Comment 23: The Surface Soil Contamination section appears tostate that only certain sites out of all the areas of concern onthe installation were evaluated in the public health assessment. Even if this is appropriate for a public health assessment, onesite that went on to an RI is not listed as having beenevaluated. This is A4.

Recommendation: Even if A4 was shown to be lacking in significant risk, it should have been on the list as having been evaluated. If all the sites of concern, and data so far collected on them, were not evaluated in the public health assessment, state why. Also state if these other sites will be evaluated in the future, before the land is turned over to other public uses.

We have reviewed data for all Annex study areas and continue toreview new information relating to public health issues. Although area A4 was not specifically discussed, we haverecommended that people not trespass on the Annex because thereis surface soil contamination and physical hazards in many of theon-site study areas; A4 among them. We walked over A4 during ourOctober 1992 site visit. The area is vegetated and contains someconstruction debris. Cleanup is planned for Annex study areaswhere surface soil contaminants are at levels of health concern. Environmental cleanup will be completed before the Army releasesAnnex property for reuse.


Comment 24 (Land Use): Please state the date that Puffer Pond wasposted. Also, permits for fishing are not obtained at the Annexgate.

The catch and release fishing policy at Puffer Pond wasimplemented in April 1992 and was stated on page 27 of theDecember 1993 public health assessment. According to Fort DevensPhysical Security personnel, the Annex has been closed torecreational use since the summer of 1993. Permits to fish areno longer available. The public health assessment has beenrevised.

Comment 25: What is the significance of mercury in one fish fromPuffer Pond? Mercury is a widespread contaminant in inland waterbodies in the northeast, especially those that have lostbuffering capacity due to acid precipitation. If the mercury inPuffer Pond fish cannot be linked to a contaminant source on theAnnex, this metal should not be considered a site-relatedcontaminant.

We do not state that mercury is a site-related contaminant. Thesignificance of the mercury is that it was detected in one chainpickerel at a concentration exceeding the Food and DrugAdministration (FDA) action level. For that reason, the pond hasbeen posted catch and release fishing only. We concur that theposting is an appropriate public health action at this time. Because of the limitation of previous fish sampling, additionalsampling has been done in both Puffer Pond and the off-postMinister's Pond. The results of that study, which is being doneby OHM Corporation, are not yet available.

The mercury phenomenon may be widespread throughout the state inlow alkalinity water bodies (Whitney Pond, Quabbin Reservoir,Echo Lake, and Walden Pond). Mercury may be available to fishfrom the substrates of these lakes or through atmosphericdeposition (64, 99,66,64, 99).

Comment 26: Is there any evidence that mercury or othercontaminants have migrated from the Annex to Puffer Pond? Onpage 33, the report notes that contaminants could not migrate toVose Pond, Cutting Pond, or Crystal Lake, but nothing in the RIor the Assessment indicates that contaminants in fact can migrateto Puffer Pond.

Mercury is discussed in the response to Comment 25. There issome question about the source of arsenic detected in surfacewater and sediments from Puffer Pond. The arsenic could be fromnatural sources, general sources such as the use of arsenic-basedherbicides on the Annex, or from the bunker area west of thepond. Further investigations are planned and should revealpotentially site-related sources of arsenic to Puffer Pond (40).

Comment 27 (summary): ATSDR neglects to address Taylor Brook andthe Assabet River, both are downstream of Puffer Pond. If fishin Puffer Pond are considered a hazard, then the public may be atrisk if they fish in Taylor Brook or the Assabet River. Theseshould also be posted.

Taylor Brook and the Assabet River are not mentioned in thesummary. However, those, and other area surface waters, arediscussed elsewhere in the document. Sampling of on-post surfacewater and sediment has been done and the findings are discussedin the Environmental Contamination and Other Hazards section ofthe public health assessment. Fish sampling in the Assabet Riverand Taylor Brook has not been done in connection withinvestigations for the Annex. However, the MassachusettsDepartment of Environmental Protection did sample fish in theAssabet River in 1985. Five stations were sampled betweenMaynard and Shrewbury. The highest concentration of mercurydetected in fish tissue was 0.3 ppm, which is below the FDAaction level of 1 ppm (98).

Tissue from a single chain pickerel in Puffer Pond containedmercury at a concentration exceeding the FDA action level. Although fish data for Puffer Pond are limited, the Army postedthe pond as catch and release fishing only. Because chainpickerel are territorial fish, remaining in fairly limited areas,they are likely to stay in Puffer Pond. Therefore, it is notnecessary to post areas downstream. The Army has done moresampling in Puffer Pond, but the results of the fish study arenot yet available.

Although elevated concentrations of arsenic do appear in some on-post areas, arsenic levels are close to background in samplestaken at the exit point of Taylor Brook from the Annex. (40) Thehighest arsenic concentration detected in sediment samples fromthe Assabet River was 140 ppm and appears to be related mainly tosources upstream of the Annex. Arsenic does not biomagnify inthe food chain and much of the arsenic in fish is an organic formthat is essentially nontoxic.

Surface water posting is recommended when fish sampling dataindicate that a fish consumption advisory is needed. Massachusetts has a fish toxics group that conducts suchevaluations in the state. The fish toxics group includes membersof the Massachusetts Department of Environmental Protection(Office of Watershed Management and Office of Research andStandards), the Massachusetts Department of Public Health, andthe Department of Fish, Wildlife and Environmental LawEnforcement.

Concerns regarding fish contamination in surface waters off-postshould be referred to the Massachusetts fish toxics group. TheMassachusetts Department of Environmental Protection, Office ofResearch and Standards, in Boston, is the point of contact forthe fish toxics group. The office telephone number is 617-292-5570. Data relating to Annex watersheds do not suggest that sitecontaminants have migrated to off-post portions of Taylor Brookor the Assabet River.

Comment 28: The presence of arsenic in Assabet River sedimentsshould be evaluated in light of the historical use of arsenicalsas pesticides in apple orchards.

As discussed in response to comment 27, elevated levels ofarsenic in surface water and sediment do not necessarily lead toarsenic contamination in fish. Arsenic does not biomagnify inthe food chain and much of the arsenic in fish is an organic formthat is essentially nontoxic.

Comment 29: Explain how the AWQC were used to evaluate thepotential for bioaccumulation in fish. The AWQC were developedusing bioaccumulation data among other factors, but they do notprovide a measure of bioaccumulation. A better estimate would bethe octanol-water partition coefficient (KoW) or theBioconcentration Factor (BCF).

By definition Ambient Water Quality Criteria (AWQC) are designedto provide adequate protection of aquatic life and its uses fromacute and chronic toxicity to animals, toxicity to plants, andbioaccumulation by aquatic organisms (56). Bioaccumulationincludes not only bioconcentration (the uptake of contaminantsfrom water only) but also predators accumulating toxic substancesby consuming prey.


Comment 30 (summary): People trespass on the Annex regularly. ATSDR should recommend that the Army post the site on the fencingthey are reportedly installing.

A community member brought up the concern at the January 20, 1994ATSDR public meeting at Maynard High School. ATSDR referred theposting request to the Army in a letter sent February 16, 1994.

Comment 31 (summary): Fences alone do not discourage people fromentering the site if they are not aware of the fact that it maybe hazardous to their health. We would have hoped that ATSDRwould have at minimum recommended the Army undertake an educationprogram to inform residents living adjacent to the Annex of thehazards of the site. None of the agencies have seen fit tocommunicate with the people living near the site via a mailing.

The Army held public meetings in January and April of 1994 todiscuss the Annex. ATSDR also held a public meeting in January1994 to discuss the hazards associated with the Annex. Duringthat meeting we recommended that people refrain from trespassingon the Annex because there are localized areas of surface soilcontamination and physical hazards.

Comment 32: There are inadequate analytical data for surfacesoils in all areas of the Annex that could be accessed bytrespassers.

We emphasize that the Army has improved Annex fencing andincreased patrols to discourage trespassing. To prevent exposureto surface soil contaminants and physical hazards on the Annex,people should not trespass.

Surface soil sampling on the Annex is biased toward areas withevidence of contamination or a history of activities that mighthave lead to chemical contamination. Such a sampling strategy isappropriate and adequate. In study areas where contaminants havebeen detected at levels of health concern, furthercharacterization is ongoing. Once the nature and extent ofchemical contamination around these particular sites is known,appropriate cleanup can be implemented.

Comment 33: Conclusion 4 notes that there has been 'limited'human exposure to contaminants. This has not been shown. Therehas been possible exposure to Annex contaminants, most likely bytrespassers and people who drank from the Track Road spring, butATSDR has not shown any evidence of exposure, 'limited' orotherwise.

It is unlikely that anyone has been exposed to contaminants atlevels of health concern even if they trespassed on Annexproperty. "Limited" refers to the small number of trespasserswho could contact areas with surface soil contamination on theAnnex. Such exposure, if it occurs at all, would be short term. It also is not very likely that people are eating sufficientquantities of contaminated fish from Puffer Pond to result inadverse health effects. With regard to the Track Road spring,samples from June and October 1992 did not contain chemicalcontaminants at levels of health concern.

Comment 34 (summary): Hunting that takes place on the site isnot addressed. Although it is illegal to hunt on the Army'sland, we are concerned that the animals may eat fish from thepond or otherwise be contaminated. Again, posting the site ascontaminated would help reduce this risk to the hunters.

As pointed out in the comment, hunting on the Annex is illegal. The Army has discouraged trespassing, including illegal hunting,by improving perimeter fencing and increasing the frequency ofon-post patrols. On-post contamination is not widespread and,considering the size of the Annex, game are not likely toconsistently feed in contaminated areas. Moreover, the types ofanimals that might eat fish from the pond (probably raccoons) arenot consumed by people.


We emphasize again that the focus of a public health assessmentis whether human exposure occurred or is occurring. If exposuredid not occur or is not occurring, as is the case with the Annex,then there is no question of health effects relating to exposure.

Comment 35: I have only lived here 5 years, yet I have heardartillery fire in the Annex, so particles could carry on thewind. Also dirt on contaminated soil is easily picked up duringthe summer and fall and with the wind ride through the air toreach surrounding people. I am concerned with wind and I thinkyou should be working backwards. Gather data first and searchfor the problem. Look for patterns. Track peoples health first. There is enough data for you to gather.

We do not think that people in residential areas are beingexposed to contaminated dust from the Annex. Although there issome soil contamination on the Annex, the contaminants are mainlyassociated with localized spills and are not widespread. Thewind would not carry contaminants very far because much of theAnnex is wooded. Moreover, the on-post training areas are fairlyremote and not close to residential areas.

It is not effective to work backwards from health outcome data toidentify a disease in a community and then attempt to find apoint of exposure. Diseases described in registry information ormedical records may have multiple causes including infectiousagents, genetic factors, household chemicals, industrialchemicals or the cause may be entirely unknown. Little or noinformation may be available in the data bases about theindividual case in terms of location or lifestyle. To workbackwards from health information to a list of chemicals at aparticular installation would introduce too many uncontrolledvariables into the analysis to get accurate results. Instead,the public health assessment focuses on whether people arecontacting environmental contaminants at levels of healthconcern. We do not think that such exposures are occurring atthe Annex.

Comment 36: The Public Health Assessment finding of no apparentpublic health hazard is based primarily on the absence ofincreased cancer incidence in the four towns between 1982 and1986.

The finding of no apparent public health hazard is based on noone being exposed to site contaminants. Annex study areas withchemical contamination are being further investigated and will becleaned up before the property is available for reuse.

Comment 37: Our concern is that cancer incidence may not be themost appropriate measure of health status in the four towns. ATSDR has based its entire Assessment of a site that has been inoperation since the 1940s on four years of cancer incidence datafrom the early 1980s.

ATSDR based its public health assessment on little or no evidence(or likelihood) of human exposure to site contaminants.

Comment 38 (summary): All available health outcome data was notreviewed.

ATSDR reviewed cancer incidence data for 1982-1986 and cancermortality information for 1987-1988 because those data wereavailable to us at the time the initial release public healthassessment was written (1991-92). When the public commentrelease of the public health assessment was written (1992-93) noadditional health outcome data were reviewed because newenvironmental data indicated little likelihood of human exposure. The statement in the summary has been modified for clarity.

Comment 39 (Natural Resource Use): ATSDR did not evaluate allhealth outcome data. ATSDR only viewed 4 years of the 12 thatare available, that is only 1/3 of available data. This should becorrected.

We did not state that all health outcome data were reviewed. Thespecific data that we reviewed is listed in the public healthassessment. For clarity, a sentence has been added to theparagraph.

Comment 40: The Health Outcome Data Evaluation section statesthat if the "range included 100, then the difference is due tochance." This should state "probably due to chance".

In addition, the opposite problem should be explained more clearly. That is, what if multiple analyses are done, as they are here, and some turn up statistically significant. Some concern has already been raised at a public meeting over the "excess: pancreatic cancer in Sudbury, and liver cancer in Maynard. After this issue is raised in the public health assessment, it is not fully explained, only generalized in the statement on page 43 that states, "this analysis of the cancer incidence data around the Annex does not indicate any pattern of occurrence that is of increased public health concern at the present time." A member of the public may wonder how both conclusions are compatible - i.e., excess cancer and no excess cancer.

In addition, concern was raised over whether this should befollowed up in some way. The conclusions, recommendations andplans sections of the public health assessment do not addressthis.

Recommendation: Either the "excess" cancer that is shownneeds to be fully explained so that the public understands whythere is no need for follow up, or some sort of follow up shouldbe discussed.

An expanded discussion regarding cancer of the liver and pancreasis provided in the Health Outcome Evaluation Section (p. 43).

Comment 41: We recognize that Massachusetts has had a cancerregistry only since 1982 but there are other sources of healthoutcome data. In addition, ATSDR should have used the 1982-1988Cancer Registry Data, which has been available since July 1991.If six years of cancer incidence data are used, rates ofcolo-rectal cancer in Maynard and Stow and pancreatic cancer inSudbury are significantly higher than expected. There are otherdifferences that can be evaluated using six years' incidencedata, but for most of the reported cancers, the total numbers ofcases in the four towns are too low to evaluate statistically,even when the incidence data from the four towns are combined.

We are not suggesting that the differences in cancer incidenceover the six years is due to the Sudbury Annex, but arequestioning the use of such limited data, particularly when noattempt was made to correlate any of the cancer outcomes to thecontaminants present at the site. The ATSDR report could besummed as follows: "We didn't see increases in cancer in the fourtowns between 1982 and 1986, so we conclude that there are noimpacts from the site."

Even with additional years of registry data, the limitations ofthat type of information do not allow specific conclusions aboutcauses of the disease. Exposure pathway information must beconsidered. At the Annex, no evidence of exposure was seen;therefore, there is no reason to conclude the site has had animpact on the public health.

Comment 42: The Public Health Assessment makes no attempt tocorrelate contaminants found on the Annex with health outcome. There is no discussion of whether any of the site contaminants isassociated with the types of cancer addressed in the healthoutcome discussion. If the contaminants that receptors could havebeen exposed to through complete pathways are not associated withthe cancers that were discussed, ATSDR cannot use the absence ofincreased cancer incidence as evidence that there is no publichealth impact from the Annex.

We evaluated environmental information and analyzed the potentialfor people to be exposed to contaminants as a basis fordetermining the health impact of the Annex. The Public HealthAssessment makes no attempt to correlate contaminants found onthe Annex with health outcome (cancer or noncancer), becausethere is little evidence, or likelihood, of human exposure.

Comment 43: Based upon the analysis conducted in this report, wealso suggest that it is appropriate to state that the healthoutcome data "do not suggest that the Annex contributed toincreased cancer incidence between 1982 and 1986". There is nobasis for the broader statement in this report.

The conclusion has been modified.

Comment 44: As ATSDR knows, health impacts other than cancer canbe associated with exposure to environmental contaminants. Common contaminants such as mercury, lead, xylene, toluene, andnumerous others can cause harmful non-cancer effects in exposedpersons. In the Assessment ATSDR addressed the number ofchildren in special education classes in response to communityconcerns, and concluded that there were no significantdifferences between observed and expected in the four towns. Thiswas the only non-cancer impact that was given anything butcursory consideration in the Assessment.

The Maynard Board of Health reviewed the Massachusetts Departmentof Public Health death statistics for 1990, the latest year forwhich they are available. Age-adjusted death rates could not becalculated because the death statistics are not available in aformat that permits assessment of each individual case, andtherefore age at death.

However, the Board calculated Crude Death Rates for the sixcauses of death occurring with the greatest frequency in the fourtowns, with the exception of motor vehicle accidents and otheraccidental injury. The six causes as listed by the MDPH wereheart disease, all cancers, stroke, pneumonia and influenza,chronic obstructive pulmonary disease, and liver disease andcirrhosis. The calculated Crude Death Rates were compared tocalculated Crude Death rates for the state as a whole in 1990.With the exception of significantly increased death rates frompneumonia and influenza in Maynard and heart disease in Hudson,all rates for the four towns were significantly lower than thosefor Massachusetts. The presence of older than expectedpopulations in Maynard and Hudson may explain the causes of deathin these towns, but, as noted, we were unable to calculateage-adjusted rates. With the exception of Hudson, there weresignificantly fewer deaths from cancer in the four towns. InHudson, the rate is slightly lower than that of the state. Threetowns had very low deaths from liver diseases other than cancerin 1990 and Maynard had none.

We are not suggesting that any of these deaths are related to thesite. Our concern is that ATSDR did not review Crude Death Ratesas part of its evaluation. These data could be inspected forevery year since the Annex began operation. Other sources ofhealth outcome data include fetal death certificates, birthcertificates, death certificates and school health records. Werecognize that these records are often incomplete, and that deathcertificates may not include contributing and coexistingillnesses, but a review of other health outcome data would haveprovided additional confidence in the findings of the PublicHealth Assessment.

We did not examine the crude death rates because they areextremely limited in the information they provided. The majorfactor influencing the risk of death is age. Crude rates, whichrefer to the total population, may obscure the fact thatsubgroups of the population exhibit significant differences inrisk. In addition, the conditions examined above, such as heartdisease and COPD (Chronic Obstructive Pulmonary Disease), arechronic, multifactoral diseases. Therefore, it is not possiblein a general review to sort out effects from different lifestylefactors as opposed to environmental factors. The death rates donot adequately reflect the rate of new cases. Therefore, crudedeath rates offer no useful site-specific information.

With regard to death statistics data, individual cases are notneeded to calculate age-adjusted rates. The rates are determinedby the number of cases in each age group and the age structurefor a standard population.

As stated, special education rates were examined to provideinformation about a specific community concern. Other outcomedata were not examined because there is no evidence of exposureof public health significance and no other specific communityconcerns.

Comment 45: We are concerned that the public health assessmentdoes not answer the citizens' questions about the potentialhealth impacts of living near the Sudbury Annex. To this end, werecommend that ATSDR should review health outcome data other thancancer for the four towns, and should include data from the past50 years in this review.

It is our opinion that if there is an impact on public healthfrom the contaminants at the Annex, that impact would have beenevident in the past when active operations were occurring. Therewere more residents using private wells in the past and alsoresidents were more likely to have trespassed on the site. Anair pathway may have been of concern when the site was active. Itis our opinion that limiting a health outcome study to four yearsin the early 1980s is not appropriate, and that if there has beena health impact, it will be evident from death certificates andother available health outcome data records.

When there is no exposure there will be no impact on the publichealth. It is not appropriate to examine records from the pastfor multifactoral diseases and attempt to make an association, particularly when past exposure is not a public health issue.

Comment 46: The toxicity profile for mercury included asAppendix 10 should reference the latest edition of Doull, et al.The fourth edition was published in 1991.

Doull, et al. fourth edition 1991 is the appropriate referencefor the information cited in Appendix 10.

Comment 47 (Demographics): Median age should be calculated forall towns.

The median age for each town is listed in Table A-1.

Comment 48 (Demographics): As the Annex has been on the NPLsince 1989 and ATSDR has been studying it since 1991, we arepuzzled that ATSDR was unable to determine who lived at theCapehart Housing and how many children. If it was difficult toobtain this information in the summer, why couldn't it beobtained during the other seasons?

Specific information regarding Capehart Housing residents variesover time but the occupancy rate is virtually %100. Theinformation presented in the December 1993 public healthassessment reflects the residential status information that wasavailable during the summer of 1993; the time period when theinquiry was made. Updated demographic information on CapehartHousing residents is not needed because they are NOT beingexposed to site contaminants.

Comment 49 (Land Use): The dates on the leases should be listed.

Army leases for the Annex vary over time. For details regardingproperty leases, the reader is referred to Chapter 3 of the DraftPhase II Site Investigations Report, Volume 1, March 1994prepared for the Army by Ecology and Environment, Inc.


Comment 50: The public health assessment does not address thefuture use of the Annex, perhaps the most important issue thecommunities face at this time.

Suggested reuse is open land for recreational use, but noagreement has been reached at this time. The Army plans to cleanup study areas on the Annex to levels protective of publichealth. The Army will retain control of the property until siteremediation is completed.

Comment 51 (Recommendations): This section is inadequate anddoes not address the future use of the Annex. ATSDR recommendsrestricting access, but this is unrealistic for the future.

At this time, plans for land reuse for the Annex are speculative. We recommend restricting access to the site because of currentsite conditions. Once environmental cleanup proceeds and plansfor future land use are clearly identified, we will modify publichealth recommendations as needed.


Comment 52 (Site Visit): ATSDR did meet with FOCUS and somemembers of other community groups on October 5, 1991; however todefine the Public Availability Session as a meeting with arearesidents is misleading. The only people who attended thismeeting were notified by FOCUS. ATSDR made no attempt toadvertise this meeting so that "the public" could attend. Wewould like to see an accurate description of the number of peoplewho attended these sessions" and their specific expressedconcerns.

ATSDR held several meetings with community members. A meetingwith representatives of area community groups was held on October5, 1992, not 1991. That meeting was arranged between ATSDR andcommunity group representatives and was not advertised. Approximately 10 people attended. Public Availability Sessions(PAS) were held the following day, October 6, 1992 from 10:00am -2:30pm and 4:30pm-8:00pm at the Capt. T. Dustin AlwardMassachusetts Firefighting Academy. Press releases announcingthe PAS were sent to all area newspapers, and local radio andtelevision stations. The Maynard Beacon and the Daily Sun(Marlboro-Hudson Daily Paper) printed articles announcing themeeting. ATSDR coordinated advertising with representatives ofFOCUS, the Lake Boone Association, and the Organization for theAssabet River. Those groups requested fliers announcing the PAS(250 were sent). Additional fliers (50) were mailed torepresentatives of local government for the four towns (boards ofselectmen, health departments, etc). Approximately 8 peopleattended. Community health concerns are listed on pages 12-13and addressed on pages 45 - 52.

Comment 53. Also, at the October 1991 meeting with FOCUS, ATSDRcommitted verbally to producing a public health assessment byNovember of 1991. The public health assessment was not produceduntil more than a year later. What data was evaluated after thismeeting?

The meeting was held in October 1992, not October 1991. In thatmeeting, Dr. Gary Campbell, ATSDR Army Unit Chief, stated that hecould not speculate when the public comment public healthassessment would be released. Data that were evaluated after themeeting are listed in the References and Appendix 3 of theDecember 17, 1993 public health assessment (public commentrelease).

Comment 54: In the section entitled "A note of explanation" ATSDR states that the public health assessment meets thestatutory criteria set out in CERCLA. This is incorrect.

The actual statement reads . . . "This document represents theAgency's best efforts, based on currently available information,to fulfill the statutory criteria set out in CERCLA section 104(i)(6) within a limited timeframe." That statement is true. Weare addressing public health issues at all sites proposed on theNational Priority List.

Comment 55 (Recommendations): Although ATSDR recommends"education of health professionals", they then state that "anATSDR staff member participated in the Massachusetts Departmentof Public Health Grand Rounds professional health educationseries. No additional health follow-up activities are indicatedat this time." At the Public Availability Session on January 20,1994, ATSDR representatives stated that the staff member was fromAtlanta. We would like to know how an ATSDR staff memberresiding in the Atlanta area has helped "educate" local healthprofessionals. This is absurd and is another example of what iswrong with ATSDR's process. This does not serve the public.

A physician from ATSDR headquarters in Atlanta flew toMassachusetts to present a grand rounds at the Framingham GeneralHospital on exposure to volatile organic compounds. Informational handouts were provided at the session and telephonenumbers of additional resources were provided. The MassachusettsDepartment of Public Health selects the format of the healthprofessional education because they have the expertise indetermining what are the best mechanisms for meeting those needsin the state. ATSDR provides support to that program asrequested and needed.

Comment 56: Why has ATSDR not been in contact with the townslocal emergency response teams? They are very concerned aboutthe potential of spills or contamination during clean upactivities. As the lead public health agency, ATSDR should be intouch with these professionals.

Emergency response activities relating to environmental cleanupat the Annex are coordinated by site remediation contractors. The reader is referred to the following documents for specificinformation about those activities; 1. OHM's Site-SpecificHealth and Safety Plan for Site and Remedial Investigations atthe Fort Devens Sudbury Training Annex dated April 13, 1992 and 2. Ecology and Environment's Health and Safety Plan Addendum,Phase II Site Inspections, Remedial Investigations dated January1994.

Comment 57: There are gaps in the data upon which the publichealth assessment conclusions are based. The logic behind theoverall conclusion, that no public health hazard exists, appearsquestionable in my mind for the following reasons:

1: Adequate community-specific health outcome data are limited at best;
2: Few people have been exposed, given that the area is restricted, so logically, it would be difficult, if not impossible, to determine the degree of public risk, let alone document negative health outcomes.

I question whether any adequate baseline data is available todetermine whether negative health outcomes exist or are due toenvironmental exposure over time. Nor is there a mechanism toidentify and track individuals who may have been exposed. Thenecessary data to substantiate a conclusion of "no public healthhazard" simply does not exist.

Given limited data and the lack of appropriate measures ofmortality and morbidity linked to environmental exposure, I amsurprised at the conclusion indicating no public health hazardfor the Annex. I am concerned about surface and groundwatercontamination. I think the general public is unaware of theprocess underway to test the site for possible health hazards aswell as the process involved in eventual land use. I wouldrecommend that a "public forum" be held in Maynard, withsufficient notice given to the general public, at which ATSDRwould:

  1. Outline the methodology and the conclusions of the report,
  2. Outline the overall process involved in conducting the health assessment and the data upon which conclusions were based,
  3. Outline the land use process and
  4. Provide time for public comment.

We concur that evaluating health outcome data in relation to thefew exposures that might have occurred at the Annex isunproductive. Instead, our approach is to identify ways in whichpeople could be exposed to site contaminants and take publichealth actions to eliminate those exposures.

The conclusion of no apparent public health hazard is appropriatebecause it is unlikely that people were exposed to sitecontaminants at levels of health concern. This conclusion isbased on a review of environmental data and historical activitiesat the Annex.

The potential for current exposure to site contaminants islimited to the few people who trespass on post. To avoidcontacting contaminants and physical hazards on the Annex, wehave recommended that people not trespass.

Future land use is not a public health concern becausecontaminated areas will be cleaned up before the property isreused, so people will not come into contact with contaminants.

We have already described at both public meetings and in thehealth assessment the methods used in the health assessmentprocess. In particular we have emphasized that the assessmentfocuses on whether or not people have come into contact withcontaminants, in other words, on whether there are completedexposure pathways. We will continue to address public healthissues at the site as needs arise.

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. #