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




Fort Devens Sudbury Training Annex (STA) is classified as ATSDR's category of no apparent public health hazard. Few exposures have been identified, and those do not exceed health comparison values. However, a few STA study areas are classified as indeterminate public health hazards because limited groundwater data are available.

The Army has used STA since 1942. Past uses include training, research and development, and munitions storage, testing, and disposal. Current activities at STA are limited testing, and storage by a variety of organizations. An Air Force radar facility, a Federal Emergency Management Agency office and housing units are on post.

STA was listed on the National Priorities List in 1990. Seventy study areas have been investigated for potential environmental contamination resulting from past disposal practices. Those investigations have shown localized surface soil, surface water and sediment contamination. The levels of metals, pesticides, herbicides, volatile and semivolatile organic compounds, explosives and polychlorinated biphenyls detected are not expected to cause illness because human exposures are infrequent, if they occur at all. Physical hazards also exist on post. Limited groundwater contamination has been detected at STA. Additional groundwater investigations are planned to determine whether or not contaminants could migrate toward drinking water supplies.

In the past, people may have been exposed to contaminants if they ate fish from Puffer Pond or drank groundwater from on-post potable wells. However, continued exposure is unlikely because Puffer Pond is posted for catch and release fishing only and STA potable wells are no longer used for drinking water. People who trespass on STA could be exposed to localized surface soil contamination and physical hazards. The Army has maintained and improved institutional controls restricting access to STA. The Army is further evaluating possible groundwater contaminant movement toward private drinking water supplies.

Community members are concerned about potential exposure to contamination at STA and about associated health effects. Others are specifically concerned about an increase in the number of children receiving special education. The health outcome data reviewed, coupled with limited human exposure to site contaminants, does not suggest that STA is contributing to illnesses in the community.

As a result of the Agency's evaluation of available information, ATSDR recommends: (1) sampling private drinking water wells on Dawes Road in Stow; (2) continuing access restrictions on post; and (3) continuing posting of Puffer Pond for catch and release fishing only. The Army is implementing those recommendations.

An ongoing Public Health Action Plan for STA has been developed. ATSDR has discussed public health issues regarding environmental contamination with health professionals, community members, the Army, the Massachusetts Department of Environmental Protection (MDEP) and the Environmental Protection Agency (EPA). The Army is coordinating private well sampling on Dawes Road with MDEP and EPA. ATSDR will review that data and recommend followup public health actions as needed. Moreover, ATSDR will continue to evaluate relevant new environmental, toxicologic, or health outcome data to determine the need for additional public health actions at STA.


A. Site Description and History

The Fort Devens Sudbury Training Annex (which will be referred to as STA or Annex throughout this document) is 20 miles west of Boston and 2 miles northwest of the town of Sudbury in Middlesex County, Massachusetts (Figure 1) (All figures are in Appendix 1 unless otherwise stated). This installation covers about 2,750 acres and includes portions of the towns of Maynard, Hudson, Stow, and Sudbury. Hudson Road divides the facility into two sections. The smaller, southern section that covers about 289 acres includes a military family housing area and an area where cloth durability testing is performed. The Army considered selling all of the southern section in 1989, but subsequent listing of STA on the National Priorities List (February 1990) has delayed the sale indefinitely (1). The active operations in the larger northern section include an Air Force radar installation, the Federal Emergency Management Agency Regional Operations Center, and a guardhouse at the main gate.

The history of STA is divided into three parts: the Maynard Phase (1942 - 1957); the Natick Phase (1958 - 1982); and the Devens Phase (1983 - present). The chronological history and principal uses of the installation during each phase of operation are presented in the following paragraphs. More detailed information is presented in Table 1 and references (2) and (3) (All tables are in Appendix 2 unless otherwise stated).

Maynard Phase: STA was originally acquired by the U.S. Government in the early 1940s and was known as the Maynard Ordnance Depot. During World War II, 50 earth-covered concrete igloos were constructed for ammunition storage. After the war, the Annex became the Maynard Ordnance Test Station. The principal use was ammunition storage, testing and disposal and related support activities.

Natick Phase: In 1958, control of the Annex was transferred to the U.S. Army Natick Research and Development Command (NARADCOM). From 1958 to 1982 the Annex was used by other agencies and operators for a variety of purposes, including testing, training, and waste disposal.

Devens Phase: In 1982, Fort Devens acquired custody of the entire Annex. STA was used for training by the Massachusetts Fire Fighting Academy, Massachusetts State Police, and Massachusetts Army and Air Guard reserve and active duty units. The Massachusetts State Police and NARADCOM also use bunkers on the facility for storage. NARADCOM continues to use STA for field testing, which is described in the Land Use section of this document.

ATSDR reviewed the documents listed in Appendix 3 which describe previous investigations of potential environmental contamination at STA. As a result of those investigations, 70 study areas have been identified; 3 of which are designated as areas of contamination (A4, A7 and A9). However, in this public health assessment, the terms study area or site will be used interchangeably to describe any of the 70 areas being investigated at STA. Table 2 lists each study area by number, name, waste type, and location on STA (shown on Figure 2).

Descriptions of study areas of potential public health concern are included in Appendix 4. Those areas are either on STA, or on property bordering the installation where contaminants could have migrated off STA. Sites about which community members expressed concerns are also included. Investigations of study areas are described in the Remedial Investigation Report (4) and in Phase II reports (5,6). Detailed information regarding environmental contamination on STA to which people may be exposed is presented in the Environmental Contamination and Other Hazards and Pathways Analyses sections of this document.

B. Site Visit

On March 21-22, 1991, ATSDR staff members visited Ft. Devens and STA. They met with a Massachusetts Department of Public Health (MDPH) representative and requested health outcome data and demographic data. ATSDR personnel also attended the Technical Review Committee (TRC) meeting at the Ft. Devens NPL site and requested current site investigation information, demographic data, and historical land use data for STA from base personnel. ATSDR representatives presented a brief history of ATSDR and explained why the Agency was conducting a public health assessment. Following the TRC meeting, ATSDR representatives met with members of Four Town FOCUS (Families Organized to Clean Up Sites), at a member's home, to discuss community health concerns. Four Town FOCUS members described adverse health effects that they believe were caused by chemical releases from STA (7).

At a meeting with an Army representative, ATSDR staffers were briefed on STA history and land use. A tour of the on-site study areas at the Annex followed the meeting.

An ATSDR staff member also attended an STA TRC meeting on May 14, 1991, and presented a brief history of ATSDR. Representatives of the towns of Sudbury, Maynard, Stow and Hudson also attended the TRC meeting. At that time ATSDR requested information related to special education from the representatives of the school districts in the area. The governor's office and state Congressmen's offices also participated. Some recent sampling data and the Draft Master Environmental Plan (MEP) were presented by the Army. Copies of the data and the draft MEP were also requested.

ATSDR staff members again visited STA on October 5-7, 1992 to attend meetings and tour the site. ATSDR staff discussed the ongoing Puffer Pond fish sampling with Massachusetts Department of Public Health personnel. Staffers met with local community group members (Four Town FOCUS, Lake Boone Association, and Organization for the Assabet River) to discuss their comments on the ATSDR Initial Release Public Health Assessment for STA and any other site-related health concerns. Public availability sessions were held with area residents to gather additional community health concerns about STA. ATSDR staff also toured STA and areas bordering the installation.

On June 1-2, 1993, an ATSDR staff member attended a meeting held at Fort Devens to discuss EPA and Massachusetts Department of Environmental Protection (MADEP) comments on the Draft Remedial Investigation Report for STA. Future sampling plans at STA and improvements made by the Army to restrict access to the installation were also discussed.

C. Demographics, Land Use, and Natural Resource Use


Demographic data provide information on population characteristics of communities living near hazardous waste sites. Specific information about who might be exposed to site contaminants is an essential part of the public health assessment. Information on age distribution, ethnicity, and socioeconomic status of the potentially affected community may assist in identifying susceptible subpopulations and in interpreting relevant health outcome parameters of those exposed to environmental contamination. Specific demographic information relating to the population living on the Annex and in the four towns surrounding the installation are presented below for general information.

The U.S. Army Natick Research and Development Laboratory Family Housing Office maintains a residential area on STA. Capehart Housing, comprised of 35 family residences, is on the southern portion of STA. The majority of residents are military personnel (75%) who work at the Natick facility, with the remainder working in Boston either as recruiters or ROTC instructors. None of the residents of this housing area work at STA. No other persons live full-time on post (8).

There are three former guest houses on STA, managed by the Fort Devens Property Office (S212, S214 and S215). All are currently not occupied. Two additional buildings, the former Installation Commander's House and the Deputy Commander's House, are unoccupied and in a state of disrepair.

There are four towns surrounding STA: Hudson, Sudbury, Maynard and Stow. Figures and tables with census information are in Appendix 5. Figure A-1 is a map delineating the census tracts in the towns with relation to the Annex. Table A-1 lists the 1980 and 1990 demographic characteristics of each town and of the individual census tracts bordering STA (9,10).

Hudson is the largest of the towns with a population in the 1980 Census of 16,408. Most of the residents of the town were white (98%). Seven percent of the population was under age five and nine percent was over age 64. Hudson had the lowest median age of the four towns at 29.2 years. The population in the Hudson census tract (3221) that borders STA was 3,720 in 1980 and 4262 in 1990.

Sudbury is the next largest town surrounding STA with a population of 14,027 in 1980. The town is comprised of two census tracts which both border STA. Ninety-eight percent of the population in Sudbury was white. Children under the age of five account for five percent of the population as do adults over the age of 64. Sudbury had the highest median age at 32.1 years. The population in Sudbury was 14,358 in 1990.

Maynard (census tract 3641) had a population of 9,590 in 1980. The population was 98 percent white. Six percent of children were under the age of five. Twelve percent of the population was over the age of 64. In 1990, the population in Maynard was 10,325.

The town of Stow (census tract 3231) had a population of 5,144 in 1980. Ninety-eight percent of the residents were white. Seven percent of the population was under age five and six percent were over age 64. This area had a 1990 population of 5,328.

Demographic characteristics of the community living around Lake Boon was further evaluated because of community interest. The area around Lake Boon is in block group 4 of census tract 3231 in Stow and block group 1 of census tract 3221 in Hudson. In the area surrounding Lake Boon, there are 1,402 people living in 543 housing units. Twenty-seven percent of the population was below the age of 18 years and four percent of the population was over age 65. The portion of block group 4 around Lake Boon had 581 people or 11 percent of the population in census tract 3231. In census tract 3221, 19 percent of the population resided in the portion of block group 1 that borders Lake Boon. See Appendix 5 for a full breakdown of the population figures for the census blocks surrounding Lake Boon (Figure A-1.1) in Table (A-1.1).

Land Use

Information about land use on STA and in bordering areas is important for determining who might be exposed to site-related environmental contamination. Various land use activities on and around STA are discussed below.

About 2,525 acres of the total STA acreage of 2,750 is managed as woodlands. Eighty percent of the woodlands support timber growth. Timber removal for construction purposes by the Army is limited. Past training and demolition activities were localized or area specific, thereby, minimally impacting the forested area. Studies have been conducted to evaluate STA for the production of harvestable quantities of timber. The Army has considered excessing or selling Annex property; however, no land will be excessed until completion of the installation restoration program activity (1). In the northern portion of STA, Puffer Pond has been used for recreational fishing.

Although the primary mission of the installation has been to train active duty and Army Reserve personnel to support various Army units, the installation has been closed to training activities since October 1992. Since that time, a small group of organizations had been granted permission to conduct occasional training. Currently, however, no training is allowed on post. Other local groups lease portions of the property (11). Natick Laboratories (or NARADCOM) and the Massachusetts Fire Fighting Academy have had an agreement with Fort Devens for the exclusive use of about eight acres of land known as the POL (Petroleum, Oil, and Lubricant) Burn Area (Study Area A9) (12). The Army reported that A9 is no longer in use, but NARADCOM is permitted to use the Drop Zone (Area P26) for unmanned aircraft test flights and another area in the southern portion which is used for clothing exposure testing (13). About 780 acres was used by the Army as a field evaluation test course. The U.S. Air Force (USAF) leases 500 acres for radar instrumentation research, 80 acres for an antennae farm, and has some full-time staff at STA. The Federal Emergency Management Agency (FEMA) leases approximately 262 acres near the eastern boundary of STA. In the southern portion of STA, the Capehart family housing area occupies 18 acres and includes a small recreational area for children. In the northern portion of STA, about 850 acres contains concrete, earth-covered bunkers which had been used for ordnance storage in the past. The Army currently leases some of the bunkers to state and federal agencies for storage of equipment and supplies (14-15). The Tennessee Gas Company has an agreement with Fort Devens for a right-of-way easement for pipeline installation and maintenance at STA.

The Army is obligated by the National Historic Preservation Act of 1966 to identify cultural resources at STA and to ensure conservation of these resources through appropriate programs. In 1984, the Public Archaeology Laboratory, Inc. was contracted by the Army to conduct a cultural resource inventory. Twenty-five prehistoric cultural resources were identified during the survey. Three areas are being considered for registration with the National Register of Historic Places (12).

The area around STA is mainly residential interspersed with agricultural areas and light industry (16). The Green Meadow Elementary School and the Maynard Public High School are adjacent to property excessed by the Annex in 1977 (17,18). Fairbank Elementary School in Sudbury is within one mile east of the Annex (13). Boons Pond (or Lake Boon[e]) and White Pond are west of STA and between the northern and southern portions. White Pond is a surface water source of drinking water from Maynard. Boons Pond is used for fishing, as is the Assabet River which flows eastward and is located north of the Annex. Some of the excessed Army property that is now the Marlboro Sudbury State Forest is used for dirt biking.

Natural Resource Use

Surface water and groundwater flow in the STA northern portion discharges into the Assabet River Basin, which includes Boons Pond, White Pond, Puffer Pond, Cutting Pond, Vose Pond, Taylor Brook, Honey Brook and their tributaries. Off-site surface waters within the Sudbury River Basin are Willis Pond, and Crystal Lake. Surface water and groundwater in the southern portion, south and east of the ground water divide, discharges to the Sudbury River Basin, which includes Marlboro Brook and Hop Brook (14).

Most of the STA surface water flows toward the Assabet River, which borders the northwest edge of the facility (Figure 3). Taylor Brook, in the northern part of STA, is the major surface drainage flow from the Annex to the Assabet River. On the south side of STA, surface water flows toward the Sudbury River that eventually joins the Assabet River to form the Concord River to the east. Poorly drained sections or lowlands consisting of bogs, marshes, swamps, and small waterholes are found throughout STA. The largest area of lowlands is located southeast of the centralized drumlins (oval-shaped hills composed of crushed rock). Several wastewater treatment plants discharge into the Assabet River upstream from STA. There are smaller unnamed tributaries that feed into Taylor Brook or Honey Brook, Puffer Pond, and the Assabet River. Some of these tributaries originate off base. Surface water uses include drinking water supply for the town of Maynard (White Pond), fishing, recreation, aesthetic enjoyment, and navigation (12).

Surface water or shallow groundwater covers several areas of STA. Surface water and groundwater flow patterns are influenced by the two major drainage basins, the Assabet River Basin and the Sudbury River Basin. The basins are separated by a groundwater divide located in the southern portion of STA. While the patterns of drainage from hills to swamps are reflected in the flow of both surface and groundwater, the flow within the swamps is very slight and, in some places, can only be observed at times of high flow such as snow melt or heavy rain. The two major watersheds (the area drained by a river or river system), those of the Assabet and Sudbury Rivers, are connected across the swamp between Willis Pond and Taylor Creek during high water periods. These surface water and groundwater flow patterns can be further characterized into seven distinct watersheds (see Figure 3)(5). A description of these watersheds is provided in Appendix 6. The relative location of STA study areas to groundwater recharge areas (areas of water movement into the aquifer) for the buried valley aquifer are shown in Figure 4.

Surface water and groundwater is used for potable water within a 3-mile radius of STA. Depending on location, on or off STA, within this 3-mile radius, drinking water is supplied by either domestic (private) wells, municipal (public or town) wells, surface water, bottled water, or a combination of these (4). Environmental contamination from STA study areas within each watershed is evaluated in terms of impact on downgradient on-site and off-site public and private water supplies.

On-Site Water Supply

The northern portion of STA has five on-site water supply wells; however, none are currently used for drinking water. The guest house well (GSHS) was previously used as a potable water supply for the three guest (transient) units. Those units are no longer occupied (19). The well appears to be completed in the surficial aquifer, based on sampling depth (20). The USAF radar facility well (USAF), which is 350 feet deep (14), was used for potable purposes until March, 1991. At that time, the selenium concentration (30 ppb) in the well water was above the Massachusetts Department of Environmental Protection standard of 10 ppb. Bottled water is now used by the Air Force (21). The guardhouse well (GDHS) at the Main Gate, is used for washing vehicles, watering the lawn, and for sanitary purposes (14). Water from the Civil Defense Well (CD1) is used only to cool generators at FEMA (14). According to the FEMA facilities manager, water used for cooling generators and sewage is either from Maynard's municipal system (99%) or pumped from Puffer Pond. Because of preferable taste, bottled spring water is used as a drinking water source (22). The supply well (CLHS) for the Deputy Commander's house is abandoned and the house is not occupied (23).

Since summer 1992, drinking water for the Capehart Housing area on the southern portion of STA has been supplied by the town of Hudson. The thirty-five houses had previously received water from the Maynard distribution system (24).

Off-Site Water Supply

Public Water Supply

Three of the nearby towns use groundwater from the surficial aquifer as part or all of their municipal (public or town) water supply (25). Sudbury, located 2.8 miles southeast of STA, has 8 municipal wells of about 60 feet deep that service approximately 14,000 residents (26). Hudson, four miles west of STA, has four municipal wells and one surface water source (Gates Pond) in Berlin, west of Hudson which service approximately 4,200 residents. Maynard, two miles north of STA, is currently drawing water from two municipal wells and one surface water source, White Pond, which is south of Hudson Road between the northern and southern portions of STA. The Maynard public water system services approximately 12,000 residents (27). Surface water sources are blended with groundwater sources in Hudson and Maynard before distribution (28). About 40,000 people use public or private drinking water supplies within three miles of STA (9).

Private Wells

Although the majority of Hudson, Maynard, and Sudbury residences are connected to municipal water systems, private wells are used in some areas. The town of Hudson has about 270 private wells in addition to the public water system (29). Some Hudson residences on Lake Boon, Concord Road and Parmenter Street have private wells (30). In Maynard, there are homes with wells on Riverside Park, a private road near the Assabet River bridge on White Pond Road (Old Lancaster Road) (31,32). There is also a home on Old Puffer Road and homes in the same vicinity on Parker Street (route 27) with private wells (33,34). In Sudbury there are a few homes with private wells around Willis Pond and along Maynard Road (route 27) near Willis Hill (35). All of the town of Stow has private wells serving approximately 5,500 residents (36). The depths of the private wells are unknown, but most are probably completed in the surficial aquifer (29).

D. Health Outcome Data

Health outcome data document health effects that occur in populations. The data can provide information on the general health status of the community living near a hazardous waste site. It can also provide information on patterns of specified outcomes. Some examples of health outcome databases are tumor registries, birth defects registries, and vital statistics. The specific health outcome data sources that were reviewed by ATSDR are listed below and are further discussed in the Public Health Implications section of this document.

Several sources of health outcome data for the communities surrounding STA were reviewed. The Massachusetts Department of Public Health (MDPH) Community Assessment Unit provided cancer incidence data for the towns of Maynard, Stow, Sudbury and Hudson for the years of 1982-1986. The MDPH also provided the analysis of the rates for each town including their significance and the confidence interval around them. Those data appear in Tables A-2 through A-7 and Figures A-2 through A-14 in Appendix 5 (37). In addition, MDPH provided cancer mortality information for the four towns for the years 1987-1988. Although more recent health outcome data are now available, they were not reviewed because environmental data indicate little likelihood of people contacting contaminants at STA. Even if exposure did occur, people were not being exposed to contaminants at levels that could pose health concerns.

Because of a specific community health concern (see number 7 in the following section), ATSDR also reviewed information on the number of special needs students in the local school districts in Hudson, Sudbury, Maynard, and Stow (38-42). Tables 10 through 14 contain this information (see Appendix 2).


ATSDR staff members gathered community health concerns relating to environmental contamination at STA from a number of sources. ATSDR arranged meetings with area residents to discuss these issues. During a meeting on March 21, 1991, members of Four Town FOCUS described adverse health effects that they believe might be associated with chemical releases from STA (7). In a second meeting, held at the Maynard High School on October 5, 1992, additional community health concerns were expressed by representatives of Four Town FOCUS, the Lake Boon Association, the Organization for the Assabet River, and other area residents. The following day, ATSDR held public availability sessions (PAS) at the Massachusetts Fire Fighting Academy in Stow. Community members discussed health and site concerns with ATSDR staff during the one-on-one sessions (18).

Other community health concerns were raised by representatives of the towns of Stow, Maynard, Hudson, and Sudbury who attended the May 14, 1991 TRC meeting at the Annex (31). ATSDR also reviewed additional health concerns which are described in the Army's April 1992 Community Relations Plan. Those concerns were gathered during interviews conducted by representatives of Dames and Moore, EPA, and MADEP between May 15-17 and 20-23, 1991, with citizens living in the vicinity of the installation and community, political, and business leaders (1).

The eight community concerns listed below were identified from the previously described sources of information. Those concerns are addressed in the Community Health Concerns Evaluation section of this document.

1. Community members are concerned that analysis of samples from STA has not included all compounds that were previously used at the facility (such as explosives and chemical agents).

2. Many community members who live around Boon Pond (also called Lake Boon) are concerned about the effects of the eutrophication of the pond on their health. They are also concerned that explosives, particularly RDX, from the installation may be contributing to the eutrophication of Lake Boon.

3. Community members are concerned about potential contamination of area water supplies by past disposal practices at STA. Their specific concerns are:

a. The water in private wells in Stow and Hudson has not been tested for contaminants that were disposed of on STA.

b. The Maynard water line from White Pond to Summer Hill crosses STA. Breaks in the line could contaminate the water supply.

4. There is concern that area ponds -- on or near the Annex -- could be contaminated by chemicals that were dumped at the Annex. If contamination of surface water and sediment occurred, would edible fish be contaminated as well?

5. Community members are concerned that the public has access to potentially contaminated areas at STA. Specific concerns are:

a. Public access to contaminated areas on STA is not effectively controlled.

b. There is particular concern about A7, an area of contamination that was fenced in October 1991. Prior to that time, children rode dirt bikes in the area. Citizens were alarmed when environmental sampling at A7 was done by remedial workers, who were wearing "moon suits."

c. Community members are also concerned that portions of STA that were turned over to the state, and are now used as recreation areas, may be contaminated. The specific study areas identified were P7, P17, P31, P43A and P43B (see Figure 2 for site locations).

6. A community member was concerned that a cobalt 60 source was brought to STA and stored in a bunker.

7. The citizens expressed concern that the Boon Pond area has more children receiving special education than the other areas in the towns or the state.

8. Community members around the Boon Pond area report a variety of physical symptoms such as frequent nosebleeds, rashes, allergies, and possible seizures (episodes with a loss of motor control and amnesia for the event).

Community concerns received in response to the public comment period for the December 1993 Public Health Assessment are presented in Appendix 11. Those concerns were raised in letters from community members or during the January 20, 1994 ATSDR public meeting held at Maynard High School. In that meeting, ATSDR staff members gave an overview of the public health assessment and talked with community members about public health issues related to the site.


Findings from previous investigations of environmental contamination at STA have been reviewed by ATSDR. Data in the 1993 Remedial Investigation Report (4) build on the earlier studies and further characterize the extent of contamination at STA. Unless otherwise stated, the environmental data presented are from the Remedial Investigation Report (4). Findings in the Phase II site/remedial investigations are also discussed (40,6). The environmental data presented relate to possible exposure pathways (how people could be exposed to site contaminants) at STA. For further discussion of specific exposure pathways at STA, see the Pathways Analyses section of this document.

Contaminants discussed in subsequent sections of this public health assessment were evaluated to determine whether exposure to them has public health significance. ATSDR selects and discusses contaminants based on several factors: sample design, field and laboratory data quality, and comparison of chemical concentrations to levels that could cause cancer or other health effects. In addition, community health concerns are considered.

Tables 4, 6, 8 and 9 list contaminants detected in surface soil, groundwater, surface water and sediment at concentrations exceeding comparison values. The listing of a contaminant in those tables does not mean that it will cause adverse health effects if exposure occurs at the specified concentrations. It also does not mean that exposure has occurred or is occurring. Rather, the listing of a contaminant indicates that it will be discussed further in this public health assessment. The data tables do not include environmental data from Phase II investigations (40). However, specific Phase II data relating to potential exposure pathways are discussed in the text.

Comparison values for ATSDR public health assessments are contaminant concentrations in specific media that are used to select contaminants for further evaluation. ATSDR and other agencies developed the values to provide guidelines for estimating the media concentrations of a contaminant that are unlikely to cause adverse health effects, given a standard daily ingestion rate and standard body weight. See Appendix 7 for descriptions of the comparison values used in this document.

Environmental contamination relating to possible exposure pathways at STA is grouped by the type of medium (e.g. soil, groundwater, etc.) in which contaminants were detected. The media discussed are surface soil, groundwater, surface water, sediment, and biota (fish).

A. Surface Soil Contamination

Surface soil (0-6") sampling on STA is limited to areas with visibly disturbed soils, stained soil or surface soil debris. Additionally, surface soil was sampled in areas where drums were found (drum confirmation samples). Generally, samples were analyzed for volatile and semivolatile organic compounds, pesticides/PCBs, metals, and explosives. Samples from study areas with suspected mustard agent or herbicide contamination were also analyzed for thiodiglycol (the primary degradation product of mustard agent) or chlorinated herbicides. From a public health perspective, the sampling and analyses are appropriate for contaminants potentially present as a result of specific past activities.

Soil contamination data from selected on-site study areas (A1, A7, A9, A11, A12, P22, P32, P36, P37 and P45) were evaluated. Most of those areas are close to STA property boundaries and, reportedly, trespassing commonly occurs. Community members were also concerned about potential contamination at some of the sites (18). Surface soil data from off-site study areas where there is public access (P7, P17, P31, P34, P42, P43A, P43B, P58) were also evaluated. Table 3A (on-site) and 3B (off-site) provide detailed information about each of the study areas. Those study areas where contaminant concentrations in surface soil exceed comparison values are listed in Table 4. In general, there is limited surface soil contamination at STA with little likelihood of human exposure. Additional discussion is given in the Pathways Analyses section of this document.

Community members are concerned about possible exposure to site contaminants at A1, A7, P7, P17, P31 and P58. Study area A1 is less than one-half mile from the Green Meadow Elementary School in Maynard. Study areas P7, P17, P31 and P58 are on property excessed by STA in 1978 which is now part of the Sudbury Marlboro State Forest (4). Study area A7 was fenced in 1991, but was previously used by local residents for dirt biking and target shooting. Soil sampling included test pits at all five areas and surface soil at A1, A7 and P17. Test pits are excavations to determine possible sources of contamination identified by various field survey methods. Wastes were generally buried and there is little evidence of surface soil contamination. The only contaminants detected in surface soil at concentrations exceeding comparison values were from drum confirmation samples or where localized spills occurred (Table 4). The drums have been removed and additional sampling is planned in those areas to delineate the extend of soil contamination (5,6). Again, it is unlikely that human exposure to the limited surface soil contamination occurs. Further discussion is provided in the Pathways Analyses section of this document.

B. Groundwater Contamination

Groundwater data were reviewed to determine if contaminants detected at study areas within STA watersheds could migrate toward potable water supplies. Groundwater contamination is discussed in terms of individual watersheds that flow toward public and private drinking water sources. Figure 5 and Table 5 provide additional information relating study areas within each watershed (groundwater flow) and potable water supplies. Contaminants potentially affecting potable water supplies both on STA and off site are discussed. Table 6 lists groundwater contaminants, from either monitoring wells or potable wells, that were detected at concentrations exceeding comparison values.

Additional groundwater monitoring has been done to further investigate potential movement of groundwater contamination from STA study areas toward off-site groundwater and potable water sources. As a followup to previously completed groundwater studies at STA, new monitoring wells were installed for Phase II Remedial Investigations (5,6). The new monitoring well locations are described in Table 7 and shown on Figure 6.

On-Site Water Supply

ATSDR reviewed groundwater sampling data for the five on-site wells previously used for drinking water (4). The drinking water wells were sampled for volatile organic compounds (VOCs), base neutral acid extractable compounds (BNAs), metals, and pesticides during June 1991, and June and October 1992. The location of these on-site wells is shown on Figure 5; the Guest House well (OS1), the U.S. Air Force Radar Facility well (OS2), the Guard House well (OS3), the Civil Defense well (OS4), and the Deputy Commander's House well (OS5). The wells have not been abandoned (grout sealed) and, although functional, they are currently not used as potable water sources. During the Phase II investigation, the U.S. Air Force well will be used as a non-potable water source for equipment decontamination and generator cooling water (5).

Substances detected in groundwater from on-site potable wells, at levels above comparison values, were selenium, manganese and iron (see Table 6). Several BNAs and VOCs were also detected but those substances were laboratory contaminants (4). Selenium was measured in groundwater from the USAF well at 30 µg/L in March 1991 (14); subsequent samples were not analyzed for this element. The Massachusetts standard for selenium is 10 µg/L for groundwater used as a potable water supply. Manganese was detected in groundwater samples from the Civil Defense (442 µg/L), Guest House (77 µg/L) and Deputy Commander's House (109 µg/L) wells. Water from the Civil Defense and Deputy Commander's House wells also contained iron. Although the manganese and iron concentrations exceed the drinking water standards, such levels commonly occur in groundwater/wetland systems regionally. The secondary drinking water standard for iron is primarily based on aesthetic (taste, odor, metal oxide laundry stains, etc.) rather than health criteria. The health implications of selenium, manganese, and iron in drinking water are discussed in the Toxicological Evaluation section of this document.

Off-Site Water Supply

Public Water Supply

In addition to the installation groundwater sampling, Maynard town well No. 3, and Maynard Public Works Department test well No. 14 were sampled. Town well No. 3 is north of STA (Figure 5, TW1). Test well No. 14 is on the edge of STA, downgradient and just south of town well No. 3. Those wells are in Watershed 1A (Table 5) and are downgradient of study areas P43A and P43B. Bis(2-ethylhexyl)phthalate (BEHP) was detected in test well No. 14 at concentrations of 90 µg/L in June 1992, 14 µg/L in October 1992 and not detected in September 1993 (4,40). BEHP was not detected in town well No. 3. Two monitoring wells (FWGW5 [GZA-MW1] and FWGW7 [GZA-MW3] [4]) adjacent to study areas P43A and P43B, are upgradient of Maynard town well No. 3 and test well No. 14. BEHP was detected at a concentration of 9.8 µg/L in the monitoring well (FWGW7) adjacent to study area P43B in November 1992. The concentrations detected in test well No. 14 and the monitoring well exceed the MCL of 6 µg/L. However, the BEHP is thought to be associated with sample handling either in the field or the laboratory.

Maynard town well No. 3 is currently not used because of taste and odor problems. That well will remain off-line until environmental investigations at STA are completed (43). Maynard has two other town wells (Nos. 1 and 2) (not shown on Figure 5). Those wells are more than one-half mile upgradient from the installation boundary and STA study areas. Any groundwater contamination on STA is not expected to migrate upgradient toward those wells.

Although not part of the public water supply, there is a spring located off Track Road in Maynard that was posted several years ago by the Maynard Board of Health to discourage residents from obtaining water there (44). As part of the environmental investigation at the Annex, water samples were collected from the spring (FWGW12) in June and October 1992. Those samples did not contain chemicals at levels exceeding drinking water standards (4,15).

The nearest Hudson town well (Figure 5, TW2) is downgradient of Watershed 2 (Table 5). White Pond (Maynard's surface water source for potable water) is also in Watershed 2. Within that watershed, manganese exceeded the secondary drinking water standard (50µg/L) in samples from monitoring wells in and adjacent to study areas A11, P7, and P48 (Table 6). During routine monitoring of the nearest municipal well, the Hudson Cranberry well, manganese was detected at concentrations up to 250 µg/L (45). Iron concentrations in groundwater from a monitoring well in study area P48 also exceeds drinking water standards. However, those levels of manganese and iron in groundwater appear typical for the area rather than the result of past disposal activities at STA.

The nearest Sudbury town well (Figure 5, TW3) to STA is downgradient of Watershed 6 (Table 5). A boundary well (BW3) downgradient of study areas in Watershed 6 did not contain contaminants exceeding comparison values (4). A second monitoring well was installed downgradient of study area P2 where small fuel oil and pesticide spills occurred in the past (Figure 6). Sampling was done in September and December 1993 and, with the exception of sodium, which could be related to storage of road salt, no site-related contaminants were detected in groundwater (40).

Private Wells

The Army has not sampled groundwater from any off-site private wells. Reportedly, some residents in the Lake Boon community have had their well-water sampled. ATSDR requested the private well data from the Lake Boon Association; however, the information has not been provided (46,47).

Groundwater data were evaluated for study areas on STA that potentially could affect private wells. A discussion of study areas with groundwater contamination within watersheds that flow toward private drinking water sources in each town is presented in this section. Due to specific community concerns, private wells on Lake Boon are addressed separately as the Lake Boon Drainage Area (Watershed 5). Because surface water and groundwater flow in this wetland area are interrelated, surface water and sediment sampling results from Watershed 5 are also discussed. A more detailed description of installation-wide sampling activities and results are presented in the Surface Water and Sediment section.

Contaminant concentrations in groundwater from a number of monitoring wells in Watershed 1A exceed comparison values. Study areas A4, P44A, P44B, and P14 are closest to private wells; samples from monitoring wells exceeded comparison values for manganese and lead (Table 6). Comparison values were not exceeded for the VOCs, pesticides, or other metals. As previously discussed, manganese has been detected in the downgradient Guest House (Figure 5, OS1) and Civil Defense (Figure 5, OS4) wells; however, lead has not been detected in those wells. Private wells on Old Puffer Road and Parker Street are upgradient of Watershed 1A (Figure 5, RW1); therefore, on-site contaminants are not likely to migrate toward those private wells, but rather toward the downgradient Guest House and Civil Defense wells which are not used as potable water sources.

Private wells on Riverside Park (a private road near the Assabet River Bridge on White Pond Road [Old Lancaster Road]) (Figure 5, RW3) are downgradient of Watershed 3. The nearest STA monitoring well (DM9A) is between the private wells and study area A9/P12. Two explosives, 1,3,5-trinitrobenzene and 2,4,6-trinitrotoluene, were detected at levels exceeding comparison values in November 1992 (Table 6). Solvents and fuel oil constituents were detected in groundwater from DM9A during the June and November 1992 sampling; however the concentrations did not exceed comparison values (4). Groundwater samples from other monitoring wells in the study area contained VOCs at concentrations above comparison values. Additional monitoring wells (see Figure 6) were installed between well DM9A and the private wells to evaluate possible groundwater contaminant migration from A9/P12 (6). Sampling results for those intervening monitoring wells will be reviewed when they are available and, if needed, ATSDR will recommended further public health action at that time.

Private wells on Concord Road and Parmenter Street (Figure 5, RW2) are downgradient of Watershed 2. Iron and manganese were detected in groundwater from monitoring wells in study areas A11, P7, P48 and a boundary well (BW2). Although the concentrations of the metals exceeded comparison values, most were typical of regional levels. Maximum concentrations of manganese (1,700 µg/L) and iron (5,200 µg/L) were detected in a monitoring well at study area A11 (Table 6). The area was previously used as a leach field for barracks septic systems and water purification system backwash. Leachfields produce anaerobic conditions in soils which can result in the release of manganese to groundwater (48). Groundwater from boundary monitoring wells (BW1 and BW2) contained VOCs and BNAs during single sampling rounds, but subsequent sampling of those wells were negative for the previously detected analytes (4).

Private wells around Willis Pond (Figure 5, RW6), and along Hudson Road near Crystal Lake (Figure 5, RW8) are downgradient of Watershed 6. An additional well, along Maynard Road (route 27) near Willis Hill (Figure 5, RW7), is upgradient of the STA study areas. A boundary monitoring well (BW3) is between Watershed 6 study areas and downgradient private wells (RW6 and RW8). Chloromethane was detected in groundwater from that well in 1991; but, two subsequent samples did not contain the substance. No contamination exceeding comparison values in groundwater has been detected in BW3 (4).

Private wells on Thicket Circle, Wildwood Road, and Lakewood Road (Figure 5, RW4) are downgradient of Watershed 4. Soil samples from study area A6 contained metals, polyaromatic hydrocarbons (PAHs) and PCBs (4). Metals, PAHs and nitroglycerin were detected in soil samples from study area P22 (4). However, groundwater monitoring for those study areas does not indicate that site activities have contaminated groundwater.

Lake Boon (Boons Pond) Drainage:
Private wells on Lake Boon are downgradient of Watershed 5 (Figure 5, RW5). Homes with drinking water wells nearest STA are on Northshore Drive, Hallocks Point, and Dawes Road (Figure 7). Northshore Drive and Hallocks Point are across Sudbury Road from STA. The nearest homes to STA are on Dawes Road, less than 500 feet from Study Area P58.

The direction of groundwater flow in the vicinity of study areas A5, P40, P31 and P58 (watershed 5) is southwest toward Lake Boon. There is a wetland at the STA boundary on Sudbury Road that flows into Lake Boon when the water table rises due to rainfall or snow melt. The study areas are in the recharge area for the buried valley aquifer in the town of Stow (Figure 4). The drinking water supply for Stow residents is from private wells tapping this aquifer. Groundwater data from monitoring wells in study areas within watershed 5 (Figure 7) are discussed below. Table 6 lists groundwater contaminants, from either monitoring wells or potable wells, that were detected at concentrations exceeding comparison values.

Manganese levels in groundwater from the majority of monitoring wells in Watershed 5 exceed the comparison value. As discussed previously, these findings are typical for the region. The monitoring well with the highest concentration of manganese (1,000 µg/L) also had levels of iron and sodium exceeding comparison values (OHM-40-29). These findings are thought to reflect water quality near the bedrock Marlboro Formation (4).

Tetrachloroethylene (PCE) was measured in groundwater from monitoring well EHA 1 in study area P40 in 1983 (22 µg/L) and 1984 (30 µg/L) at levels exceeding the comparison value of 5 µg/L (MCL). The PCE levels in subsequent groundwater samples collected from the same well are approximately ten-fold less (June 1992 [2.9 µg/L] and October 1992 [2.7 µg/L]). The source of the PCE contamination is not known and none of the other groundwater samples in the area contained the chemical. Because the PCE concentration in groundwater from study area P40 does not currently exceed comparison values, and the chemical has not been detected in downgradient monitoring wells, it is not considered a contaminant of concern.

Four Town FOCUS reported that phosphorus levels in Lake Boon are rising and nitrate levels are stable. The community group is concerned that phosphorus contamination, possibly from explosives, from STA may be contributing to the increased phosphorus levels in the lake (18). ATSDR reviewed the environmental sampling results for explosives and phosphate /phosphorus from STA study areas in the Lake Boon watershed.

Explosive-like compounds were detected in groundwater from monitoring well OHM-BW-4 in June 1992. That monitoring well is downgradient of A5 and P40 but upgradient of P31 (see Fig 7). However, the compounds could only be tentatively identified as 4-nitrotoluene (3.8 µg/L) and cyclonite (RDX)(5.5 µg/L), and could not be confirmed upon further analysis. Neither compound was detected in a subsequent (October 1992) groundwater sample from the same monitoring well. Furthermore, these chemicals were not measured in any other monitoring wells in the area during either round of sampling.

Groundwater from study areas A5 and P40 was not analyzed for phosphates. However, phosphates were detected in filtered groundwater samples from a monitoring well (BW4) downgradient of P40 (40 µg/L [June 1992] and 15 µg/L [October 1992]) (14). Groundwater was also sampled from monitoring wells on a 31 acre tract of STA land which borders Sudbury Road west of North Shore Drive (Figure 8)(49). The area was a proposed construction site for the Massachusetts Air National Guard (MAANG) Electronics Communications Training Complex. Although total phosphorus was detected in MW2, MW11 and MW17 (20 - 1300 µg/L), soluble phosphorus was only detected in MW2 (30 µg/L). No phosphorus was measured in groundwater from MW14 (FWGW4) and MW 13 (FWGS3), which are on the STA boundary and downgradient of MWs 2, 11 and 17 (14,49). Therefore, it is unlikely that the phosphorus levels in groundwater from STA would affect Lake Boon water quality.

Municipal wastewater treatment plants in the area surrounding STA are permitted to discharge into the Assabet and Sudbury River Basins. The effluent (treated water discharged to the receiving body of water) from these plants has been reported to contain total phosphorus concentrations ranging from 750-16,000 µg/L (or 0.75-16 mg/L) (50). Phosphorus is present in water in dissolved, colloidal (a suspension of finely divided particles), or particulate states and originates primarily from agricultural runoff and wastewaters containing detergents. It may exist as orthophosphate, polyphosphate, or in organic compounds. Phosphorus, which is an essential plant nutrient, occurs in natural waters in smaller amounts than nitrogen.

Although there were few contaminants in groundwater from Watershed 5 at levels exceeding comparison values, no monitoring wells were hydraulically downgradient of study areas P31 or P58. Three new monitoring wells were installed (see Figure 6) 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 (41 ppb over the maximum contaminant level of 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 (DDE and BHC) and the explosive 1,3-dinitrobenzene were also detected. No volatile organic compounds were found (40). Dawes Road is about 500 feet from study areas P31/P58 and is in the direction of groundwater flow. Sampling of private wells at the six homes on Dawes Road is planned (100).

In this area surface water and groundwater are connected, and sediment and surface water has been sampled (51-53). Sample analyses included metals, VOCs and BNAs, pesticides/PCBs, explosives, and phosphates (14). The only contaminants measured in sediment and surface water that exceed comparison values were metals (Table 8). Test pits were also sampled in P31 but none of the contaminants detected exceeded comparison values.

The concentration of iron detected in sediment samples from study areas P31 and P58, and Lake Boon exceed regional background levels (Table 8). Corresponding surface water samples contained iron concentrations exceeding both the background ranges and the secondary drinking water standard. These findings are not unexpected; elevated iron concentrations commonly occur in surface water draining from wetlands (54). Further, the secondary drinking water standard for iron (300 µg/L) is based on taste rather than health criteria. The surface water is not used for drinking water, and exposures from incidental ingestion are not likely to cause adverse health effects (see additional discussion in the Toxicological Evaluation section of this document).

Lead was detected in 1991, exceeding comparison values and regional background levels, in a single sediment sample from a wash in the vicinity of study areas P31 and P58 draining toward Lake Boon. However, those levels are not likely to adversely affect health because people who may occasionally be in the swampy area would be expected to ingest contaminated sediment only incidentally or not at all. Lead concentrations in five sediment samples collected in 1992 from the area, did not exceed comparison values.

The potential for lead in sediment to mobilize to surface water and groundwater was also considered. Most lead is retained strongly in soil, and very little is transported into surface water or groundwater. However, at pH of 4-6, the organic lead complexes become soluble and leach out (55). Low pH values are typical of wetland drainage (54). Although such conditions exist in the sampling area -- pH values of 4.3 -- elevated concentrations of lead were not detected in the surface water (51,54). The Army has not sampled groundwater downgradient of P31 and P58. Reportedly, some wells in the Lake Boon community have been sampled. ATSDR requested the sampling results from the Lake Boon Association but that information has not been provided (46,47).

C. Surface Water and Sediment Contamination

Surface water and sediment data were reviewed to determine if contaminants detected at study areas within STA watersheds could migrate toward surface waters and, in the case of White Pond, a potable water supply. Contamination in on-site media is discussed in terms of those surface waters on and adjacent to STA and the individual watersheds that flow toward them. Figure 5 relates study areas within each watershed (groundwater flow) to surface waters and potable water supplies. Contaminants detected that exceed comparison values in surface water and sediment samples are listed in Tables 8 and 9. EPA's Ambient Water Quality Criteria (AWQC) were also used as comparison criteria to evaluate the potential for bioaccumulation of contaminants in fish. Ambient Water Quality Criteria are developed for the protection of aquatic organisms and their uses (56).

To characterize the extent of surface water contamination within and adjacent to the installation, 21 surface water and 20 sediment samples were collected from Taylor Brook, Honey Brook, Willis Pond, Puffer Pond, the Assabet River, and an unnamed stream near study area P26 in May 1992 (4). Previous surface water and sediment sampling of Lake Boon was conducted by Four Town FOCUS in 1990, and USATHAMA in 1991. USATHAMA also collected surface water and sediment samples from White Pond in 1991. A discussion of the data from those studies follows.

Iron and manganese were detected in the majority of surface water and sediment samples collected on and off STA (Table 8 and 9). None of those surface waters are used for drinking water. The concentrations of iron and manganese detected in the surface waters are generally typical for the region. White Pond, which is used as a public water supply, did not contain contaminants at levels exceeding comparison values (52).

Assabet River

Three off-site surface water samples were collected from the Assabet River in 1992 (4). Bis(2-ethylhexyl)phthalate was detected in all three samples at concentrations above the comparison value (see Table 9). The high of 200 µg/L was detected in a surface water sample collected upstream (upgradient) of STA and is apparently not site-related. That concentration was the highest measured in either groundwater or surface water samples from STA. The Assabet River is not used as a potable water source. Bioaccumulation of BEHP in aquatic organisms has been documented for several aquatic species. Fish and other seafood are frequently contaminated, especially in industrialized areas. However, rapid metabolism of BEHP in higher organisms seems to prevent biomagnification in the food chain (57).

Arsenic was detected in sediment samples from the Assabet River but not in surface water (4). Sediments in aquatic systems often have higher arsenic concentrations than does the water. Most sediment arsenic concentrations reported for U.S. rivers, lakes, and streams range from 0.1 to 4,000 mg/kg (58). The highest concentration of arsenic (140 mg/kg) in the Assabet River sediment was detected near the outfall of study area P9, an unnamed stream between study area A7 and A9. Arsenic concentrations in sediment samples from P9 ranged from 2 to 11 mg/kg, an order of magnitude less than the concentration detected in Assabet River sediment (4). Numerous discharge points along the river such as stormwater outfalls, wastewater and industrial process treatment plant discharges, and private septic systems can contribute to such contamination in the environment.

Arsenic does not appear to biomagnify in the food chain, although marine algae and shellfish tend to bioconcentrate the metal. Much of the arsenic in fish and shellfish is an organic form that is essentially nontoxic (58).

Lake Boon

Lake Boon is not used for drinking water, and exposures from incidental ingestion of sediment or surface water containing iron and manganese are not likely to cause adverse health effects (see additional discussion in the Toxicological Evaluation section of this document). Because the potential for biomagnification of iron and manganese from lower trophic levels to higher ones is low, uptake of iron and manganese in fish is not a concern (59).

On-Site Streams

On-site streams within Watersheds 1A and 1B, drain toward Watershed 5, which flows toward Lake Boon. On-site streams where surface water samples were collected in 1992 included Honey Brook (one sample), a small unnamed tributary near study area P26 (one sample), Taylor Brook (two samples), the northern end of the Puffer Pond wetland (one sample), the southern end of the Puffer Pond wetland (one sample), and an unnamed tributary west of White Pond Road near study area P26 (one sample) (Table 9). Manganese and iron levels, although typical for the region (4), were detected at concentrations exceeding drinking water comparison values.

White Pond

Surface water and sediment samples were collected from White Pond (a Maynard potable water source) in April 1991 (52). Samples were analyzed for VOCs, BNAs, pesticides, PCB and explosives. Metals were detected within background ranges for the region. Organic contamination was not detected and has not been detected during subsequent routine water quality monitoring.

Watershed 2 drains toward White Pond. The only study area upgradient of White Pond is P7. Manganese levels in groundwater from monitoring wells at P7 exceed comparison values; however, those levels appear typical for the region. Manganese was also detected in crossgradient groundwater at study areas A11 and P48. Manganese concentrations in White Pond are not elevated.

Puffer Pond

Watershed 1A drains toward Puffer Pond. Surface water, sediment and fish sampling has been done in the pond (4,40,60,61). Endrin, a pesticide, and metals were detected in surface water or sediment at levels exceeding comparison values. However, Puffer Pond is not used for drinking water. Fish from the pond were contaminated with metals and pesticides; but, only mercury in one chain pickerel was at levels of health concern. Additional fish sampling is planned. Further sampling is planned during Phase II investigations (5). Fishing in Puffer Pond is currently limited to "catch and release."

Willis Pond and Crystal Lake

Surface water samples were collected from Willis Pond in 1992 (4). Bis(2-ethylhexyl)phthalate (BEHP) was detected in one of three samples at a concentration of 5.3 µg/L. The concentration of BEHP detected is less than the EPA AWQC of 5.92 µg/L or the Massachusetts drinking water standard. Willis Pond is not used for drinking water.

Although Watershed 6 drains toward Willis Pond, none of the study areas in that watershed are upgradient of the pond. The nearest study areas to Willis Pond are in Watershed 1A, which is crossgradient. Potential contamination from study areas in either Watershed 1A or 6 is, therefore, unlikely to affect Willis Pond.

Crystal Lake is also in Watershed 6. Study areas within that watershed are upgradient of the lake. A boundary monitoring well located downgradient of those study areas contained chloromethane in 1991. However, two subsequent groundwater samples from the same well did not contain the substance. Crystal Lake is not used for drinking water and chloromethane will not concentrate significantly in aquatic organisms (62).

Vose Pond and Cutting Pond

No sampling related to the environmental investigations at STA have been done at either pond. Both ponds are upgradient of study areas within Watershed 1A and therefore, would not be affected by potential contamination from the installation.

Possible cumulative watershed impacts on the surrounding environment were further evaluated in the Phase II investigations (40). Elevated arsenic levels were detected in sediments and surface water from watersheds 1A (Puffer Pond and Taylor Brook), 1B (Honey Brook, Taylor Brook and a tributary), 2 (Marlboro Brook), 5 (at sites P31 and P58). For the most part, the elevated levels of metals in surface water and sediment do not extend offpost. Data on sediments and surface water leaving the Annex do show an impact on Marlboro Brook in watershed 2. However, the brook first passes through a small pond just south of Marlboro Road (Moore Road) that acts as a trap for sediments. Further investigations are planned in areas where site-related sources of contamination are suspected (40).

D. Biota

The Army conducted fish studies in Puffer Pond in 1991 and 1992. The initial study (60) was requested by the Fort Devens Preventive Medicine Activity because a remedial investigation at STA (63) indicated environmental contamination in and around Puffer Pond. People ate fish caught from Puffer Pond and there was concern that fish might be contaminated. The second study (61) was conducted to confirm the results of the primary screening-level study (60).

In the 1991 study, three fish samples (a chain pickerel; a crappie, two perch, and a sunfish, combined; and six shiners, combined) were analyzed for metals and pesticides. A fourth sample -- six brown bullheads, combined -- was analyzed for pesticides only. The samples were fish fillets with the scales removed but the skin left on. The pickerel was the only sample that contained contaminant concentrations exceeding comparison values. The level of total mercury (1.2 ppm) in the large (2 lbs. 12 oz.) pickerel exceeded the Food and Drug Administration's (FDA) action level of 1 ppm (methylmercury) (60). Between 50% and 100% of total mercury concentrations in biota are methylmercury (64, 99). There are significant differences in the toxicity between the various forms of mercury (65) (see Appendix 9 for additional discussion).

After the initial finding of mercury contamination in pickerel from Puffer Pond, fishing was limited to "catch and release." The policy was implemented in April 1992 by order of the Commander of Fort Devens. Signs are currently posted along the shore of the pond (18).

In a subsequent fish study at Puffer Pond, individual skinless fillets (six brown bullheads and six black crappie) were analyzed for metals and pesticides (61). Metals (cadmium, chromium, copper, lead, mercury, and zinc) and pesticides (diazinon, ronnel, and chlorpyrifos) were detected in the fillets but at concentrations below comparison values; assuming chronic daily intakes (CDIs) calculated in accordance with EPA Region I guidance. The CDI are based on estimates of 48 fish meals per year and 0.284 kg (10 ounces) of fish per meal (61). However, actual fish consumption rates for the area are not known.

Of the fish species sampled in Puffer Pond, only the chain pickerel had tissue contaminant concentrations exceeding FDA action levels (60). Additional pickerel were not sampled in the subsequent Puffer Pond fish study (61). The elevated concentration of mercury found in the pickerel is not surprising. It is not uncommon to observe elevated levels of mercury in older specimens of long-lived predatory fish (66). Findings from studies of toxics in fish from "clean water" in the state showed similar levels of mercury in a chain pickerel (1.1 ppm) (64, 99) and a smallmouth bass (1.3 ppm) (66). "Clean water" was defined as those water bodies having little or no historical or present sources of pollution (66).

ATSDR also reviewed the work plan for additional fish sampling in Puffer Pond (5). The proposed sampling and analytical methods should address the data gaps identified in the previous fish studies. Sampling of edible fish species will include top predators (large mouth bass or chain pickerel), foragers (yellow perch or black crappie), and bottom feeders (brown bullhead). Fish sampling will also been done at an off-site background pond. Background levels of mercury may be high because acidic surface water can cause leaching of metals (67). Minister's Pond in Stow will be used for comparison sampling. Sediment and surface water samples will be collected in conjunction with the fish sampling in both Puffer and Minister's ponds (5,68).

E. Other Sources

Additional sources of off-base contamination were identified using EPA's Toxic Release Inventory (TRI). That database is an annual compilation of self reports from chemical manufacturers of the quantity of toxic chemicals released into each environmental medium (e.g., air, water, land); the manufacturing facilities employ more than 10 people and are in Standard Industrial Classification Codes 20 through 39. STA is not a manufacturing facility and, therefore, is not subject to reporting releases to the TRI; however, they must comply with all other state and federal reporting requirements for actual releases.

Environmental releases reported from 1987 to 1991 by industries in Maynard, Hudson, Sudbury, and Stow were reviewed. No releases were reported for industries in Maynard, Sudbury or Stow. Five industries in Hudson reported air releases of varying amounts. Digital Equipment Corporation released acids, 1,1,1-trichloroethane, freon, and hydrogen fluoride; Ditric Optics released 1,1,1-trichloroethane; Middlesex Research Manufacturing Company, Inc. released toluene; Lance Corporation released styrene; and Hudson Lock released zinc. No releases to other environmental media were recorded in the TRI.

F. Quality Assurance and Quality Control

In preparing this Public Health Assessment, ATSDR relies on the information provided in the referenced documents. The Agency assumes that adequate quality assurance and quality control measures were followed with regard to chain-of-custody, laboratory procedures, and data reporting. The validity of the analyses and the conclusions drawn in this document are determined by the availability and reliability of the referenced information.

The majority of the environmental data presented in this public health assessment is from the Site/Remedial Investigation (RI) report (4). Generally, the methodology used in the RI and that proposed in Phase II studies (5)(6) is appropriate for characterizing contamination at STA. Data inadequacies for the OHM Remedial Investigation (4) are discussed below.

Environmental samples were analyzed for specific chemicals (or known breakdown products) suspected at the study areas being investigated. However, study area P41 (Bunker 303) was used in the past to store pesticides and herbicides (4). One of the herbicides stored there was Ureabor; several formulations of that weed killer exist (70). The exact identity of the active ingredients in the Ureabor stored in Bunker 303 is uncertain; however, monuron (chemical name: 3-(p-chlorophenyl)-1,1-dimethylurea) is thought to be one of the components. Monuron, which is nearly water insoluble, is moderately toxic if ingested. Contact with skin, eyes and clothing should also be avoided (69,70). Past and proposed (5) analysis of soil samples from the area will not detect constituents of any formulations of Ureabor, and no analytical method has been located that is suitable for detecting Ureabor in soil. Soil samples collected at Bunker 303 were analyzed for organochlorine pesticides and chlorinated herbicides with the expectation that spills and releases of pesticides and herbicides could be detected with these established methods. Pesticides were detected but herbicides were not. The extent of Ureabor contamination of surface soil at P41 cannot be determined because no analytical method is currently available.

Methylene chloride, a common laboratory contaminant, was detected in several groundwater samples; however, it was also detected in method blanks which is indicative of laboratory contamination.

N,N-Diethyl-3-methylbenzamide (DEET) was detected in some of the groundwater samples collected in June 1992, it was not present in the October samples. DEET, which is the active ingredient in some insect repellents, is not believed to be a site-related contaminant. Field workers, using insect repellent during the June sampling, probably inadvertently contaminated some of the groundwater samples with DEET.

Several unknown BNAs (Base/Neutral/Acid Extractables are compounds, such as phenol, classified by the analytical techniques used to identify them) and VOCs were detected in groundwater from monitoring wells located throughout STA. The reported concentrations generally ranged from 10 to 20 µg/L. Some of the unknowns are dispersed throughout STA and are, most likely, naturally occurring. Others were detected in a single round of sampling; they were not reported in previous or subsequent groundwater samples from the same monitoring well. Unknowns are compounds that can not be identified using available analytical methods. Some are possibly breakdown products of known chemicals.

ATSDR also reviewed data reported as Laboratory Test Results -Fort Devens Annex, Lake Boone Drinking Water Recharge Area, Stow, Massachusetts, sampled December 4, 1990 (51). No information was given regarding the specific sampling locations, the field or analytical methods employed, or QA/QC procedures. The sampling was reportedly done by Four Town FOCUS (71). ATSDR requested a confirmatory copy of that data in a letter to a member of Four Town FOCUS (47); those data have not been provided.

G. Physical and Other Hazards

Physical hazards are associated with study areas throughout the installation. Some of the physical hazards are construction debris and general refuse at STA landfills, abandoned buildings, and structures at Demolition Area II (A6) and the training confidence course (P10). Study area P26 continues to be used as an air drop zone. However, STA personnel maintain access restrictions to the installation. Other than permitted activities, the general public should not be exposed to these hazards unless they trespass. Various physical hazards at STA study areas are discussed in greater detail in the Pathways Analyses section and Appendices 4 and 8 of this document.

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

A-Z Index

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