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




Site Map
Figure 1. Site Map

Demographic Statistics
Figure 2. Demographic Statistics


Quality Assurance

In preparing this report, ATSDR relied on the information provided in the referenced documents andby contacts with the Massachusetts Department of Environmental Protection (MADEP), communitymembers, and GAF. ATSDR assumes that adequate quality assurance and control measures weretaken during chain-of-custody, laboratory procedures, and data reporting. The validity of theanalyses and conclusions drawn in this document are determined by the availability and reliability ofthe information.

Human Exposure Pathway Evaluation and the use of ATSDR Comparison Values

ATSDR assesses a site by evaluating the level of exposure in potential or completed exposurepathways. An exposure pathway is the way chemicals may enter a person's body to cause a healtheffect. It includes all the steps between the release of a chemical and the population exposed: (1) achemical release source, (2) chemical movement, (3) a place where people can come into contactwith the chemical, (4) a route of human exposure, and (5) a population that could be exposed. In thisassessment, ATSDR evaluates chemicals in the soil and groundwater that people living in nearbyresidences may consume or come into contact with.

Data evaluators use comparison values (CVs), which are screening tools used to evaluateenvironmental data that is relevant to the exposure pathways. Comparison values are concentrationsof contaminants that are considered to be safe levels of exposure. Comparison values used in thisdocument include ATSDR's environmental media evaluation guide (EMEG) and cancer riskevaluation guide (CREG). Comparison values are derived from available health guidelines, such asATSDR's minimal risk levels and EPA's cancer slope factor.

The derivation of a comparison value uses conservative exposure assumptions, resulting in valuesthat are much lower than exposure concentrations observed to cause adverse health effects; thus,insuring the comparison values are protective of public health in essentially all exposure situations.That is, if the concentrations in the exposure medium are less than the CV, the exposures are not ofhealth concern and no further analysis of the pathway is required. However, while concentrationsbelow the comparison value are not expected to lead to any observable health effect, it should not beinferred that a concentration greater than the comparison value will necessarily lead to adverseeffects. Depending on site-specific environmental exposure factors (for example, duration ofexposure) and activities of people that result in exposure (time spent in area of contamination),exposure to levels above the comparison value may or may not lead to a health effect. Therefore,ATSDR's comparison values are not used to predict the occurrence of adverse health effects.

The comparison values used in this evaluation are defined as follows: The CREG is a concentrationat which excess cancer risk is not likely to exceed one case of cancer in a million persons exposedover a lifetime. The CREG is a very conservative CV that is used to estimate cancer risk. Exposureto a concentration equal to or less than the CREG is defined as an insignificant risk and is anacceptable level of exposure over a lifetime. The risk from exposure is not considered as asignificant risk unless the exposure concentration is approximately 10 times the CREG and exposureoccurs over several years. The EMEG is a concentration at which daily exposure for a lifetime isunlikely to result in adverse noncancerous effects.

Selecting Contaminants of Concern

Contaminants of concern (COCs) are the site-specific chemical substances that the health assessorselects for further evaluation of potential health effects. Identifying contaminants of concern is aprocess that requires the assessor to examine contaminant concentrations at the site, the quality ofenvironmental sampling data, and the potential for human exposure. A thorough review of each ofthese issues is required to accurately select COCs in the site-specific human exposure pathway. Thefollowing text describes the selection process.

In the first step of the COC selection process, the maximum contaminant concentrations arecompared directly to health comparison values. ATSDR considers site-specific exposure factors toensure selection of appropriate health comparison values. If the maximum concentration reported fora chemical was less than the health comparison value, ATSDR concluded that exposure to thatchemical was not of public health concern; therefore, no further data review was required for thatchemical. However, if the maximum concentration was greater than the health comparison value,the chemical was selected for additional data review. In addition, any chemicals detected that did nothave relevant health comparison values were also selected for additional data review.

Comparison values have not been developed for some contaminants, and, based on new scientificinformation other comparison values may be determined to be inappropriate for the specific type ofexposure. In those cases, the contaminants are included as contaminants of concern if currentscientific information indicates exposure to those contaminants may be of public health concern.

The next step of the process requires a more in-depth review of data for each of the contaminantsselected. Factors used in the selection of the COCs included the number of samples with detectionsabove the minimum detection limit, the number of samples with detections above an acute or chronic health comparison value, and the potential for exposure at the monitoring location.


Table 1.

On-Site Soil Sampling Results 1991-1993
Chemical Depth (feet)ConcentrationRange (ppb)Comparison Value(ppb)Source
Carbon TetrachloridePile32.75,000CREG
Total PetroleumHydrocarbons5.5-11.5ND-120,000,000No CVNo CV
*Source: Phase II: Comprehensive Site Assessment Report

CREG = Cancer Risk Evaluation Guide
EMEG = Environmental Media Evaluation Guide
EMEG-i = Environmental Media Evaluation Guide - Intermediate
J = estimated value
LTHA = EPA's Lifetime Health Advisory
ND = Not Detected
No CV = No Comparison Value Available
ppb = parts per billion
ppm = parts per million
RBC = EPA's Risk Based Comparison
RMEG = Reference Dose Media Evaluation Guide

Table 2.

On-Site Groundwater Monitoring Well Sampling Results 1991-1997
ChemicalConcentrationRange (ppb)Comparison Value(ppb)Source
Benzo (a) AthraceneND-0.340.092RBC
Chloro-octadecaneND-115No CVNo CV
2-MethylnapthaleneND-30No CVNo CV
Ortho-terphenylND-70No CVNo CV
Total PetroleumHydrocarbonsND-580,000No CVNo CV
*Source: Phase II: Comprehensive Site Assessment Report and Short Term Measure Status

Note: Abbreviations defined on next page.

CREG = Cancer Risk Evaluation Guide
EMEG = Environmental Media Evaluation Guide
EMEG-i = Environmental Media Evaluation Guide - Intermediate
J = estimated value
LTHA = EPA's Lifetime Health Advisory
ND = Not Detected
No CV = No Comparison Value Available
ppb = parts per billion
ppm = parts per million
RBC = EPA's Risk Based Comparison
RMEG = Reference Dose Media Evaluation Guide


The text summarized below is from a publication by the National Institutes of Health called Cancer: Rates and Risks, 1996 4th Edition.

Colon and Rectum Cancer 1:

Together, the colon and rectum make up the large bowel. The colon refers to the upper five or sixfeet of the large intestine, the rectum to the last five or six inches. Because of the similarity of thetissue in the colon and rectum and the occasional difficulty determining in which region a tumor hasarisen, malignancies in these two bowel segments are often lumped together as "colorectal cancer."

Fifty-six thousand colorectal cancer deaths are estimated to have occurred in the United States in1994 (Boring et al.,1994). Colorectal cancer ranks second in cancer incidence for the combined U.S.population. Among men this is the third most commonly occurring malignancy (after prostate andlung cancers); among women it ranks third (after cancers of the breast and lung).

A number of studies have found an association between red meat consumption and colorectal cancer(Willett et al., 1990; Giovannucci et al., 1994) Cooked meats have been found to containcompounds, including a class of heterocyclic amines, which are mutagenic and carcinogenic inanimal models. They are produced during high temperature cooking such as broiling or frying(Sugimora, 1986). Countries and regions with the highest per capita dietary fiber consumption tendto have the lowest colorectal cancer rates. Recent investigations have also suggested that intake offolic acid, found in many vegetables, may reduce the risk of colorectal cancer (Glynn and Albanes,1994).

Several epidemiologic investigations have now found that regular use of aspirin is associated with areduced risk of colorectal cancer, although the evidence on this link is not wholly consistent(Garewal, 1994). Some epidemiologic and clinical studies have suggested that other nonsteroidalanti-inflammatory drugs (NSAID) may protect against colonic neoplasia. The possible protectiveeffect of aspirin and other NSIAD on colorectal carcinogenesis is an active area of research.

An association between low physical activity and large bowel malignancy has become one of themost consistent epidemiologic findings for this disease in recent years. Well over a dozen studiesusing several different methods of physical activity assessment have demonstrated this association(Lee et al., 1991). The association with colorectal cancer has been noted for both occupational andleisure-time activity.

The extent to which genetic factors, in isolation or in interaction with environmental factors, play a role in sporadic cancers is unresolved. Family history of colorectal cancer in a first-degree relative has been estimated to confer approximately a three-fold risk of this malignancy.

Large bowel cancer mortality rates have declined for white men and women in the U.S. in recentyears. Investigations have recently presented evidence linking this decline to improved earlydetection procedures (Chu et al., 1994). The increased use of sigmoidoscopy and fecal occult bloodtests (followed by colonoscopy) may have an important role in reducing mortality from large bowelcancer.

Lung and Larynx Cancer(2):

Primary lung cancer accounts for about 15 percent of all cancer cases in the U.S.; however, because of its high death rate, it accounts for 29 percent of all cancer deaths (Boring et al., 1994). Lung cancer is the leading cause of death in most countries. The worldwide incidence of lung cancer is substantially lower in females- a difference generally attributed to lighter tobacco consumption by women, although other factors may play a role.

In the U.S., lung cancer incidence has risen more sharply in females than in males in recent years,reflecting the growing popularity of cigarette smoking among females over the past several decades(Ries et al., 1994). Currently, more women die each year from lung cancer than breast cancer(Boring et al., 1994).

Cigarette smoking is the major cause of lung cancer. The link was first suspected in the 1920s and1930s, and today, after multiple case-control and cohort studies, the overwhelming evidence isdocumented in more than 20 reports of the U.S. Surgeon General. Smoking is currently estimated tocause 85 percent of all lung cancer deaths. Lung cancer mortality increases with increasing dose, asdetermined by number of cigarettes smoked daily, smoking duration, and inhalation patterns. Therisk of dying from lung cancer is 22 times higher among male smokers and 12 times higher amongfemale smokers than among people who have never smoked (U.S. DHHS, 1990).

Quitting smoking reduces the risk of death from lung cancer; after ten years the risk of lung cancerdeath from former smokers is about 50 percent of the risk of continuing smokers (U.S. DHHS,1990). Those who reduce daily usage and those who smoke filtered, low-tar cigarettes gain somebenefit, although they still have lung cancer risks much higher than nonsmokers (Lubin et al.,1984). Increased risk of lung cancer has also been associated with the smoking of pipes and cigars,but at a lower level of risk than that for cigarettes. This may be due to a less intense patternsmoking, with cigar and pipe smokers typically inhaling less deeply and less frequently.

Radon is an inert gas produced by the radioactive decay of radium and uranium. Whileconcentrations of these elements vary widely, they are found everywhere in the crustal rock of theearth. Radon itself is radioactive and may cause lung cancer. Radon may also enter homes bymigrating from the earth through the cracks in the foundation, or through a hole for a sump pump, orin rare cases, via private water wells, by dissolving in drinking water.

Exposure to airborne asbestos appears to be the largest cancer threat in the workplace, raising therisk of lung cancer and mesothelioma (a cancer that arises in the lining of the chest cavity, ormesothelium) as well as asbestosis, a lung disease (Blot and Fraumeni, 1992). The risk fordeveloping these three diseases is substantially higher for workers in a number of asbestos industries,including miners and millers, and textile, insulation, shipyard, and cement workers.

Air pollution has been suspected as a cause of lung cancer, but it has been difficult to establishdefinite links. Of special concern are the effect of the byproducts of the combustion of fossil fuels,most notably polycyclic hydrocarbons (PAHs). Although suggestive, the association of lung cancerand PAHs has not yet been conclusively demonstrated, as other components of air pollution mayalso be carcinogenic.

Finally, there is laboratory evidence of a protective effect against lung cancer with increased intakeof vitamins A, C, E, selenium, and other micronutrients. The clearest and most consistentassociations occur with the consumption of fresh fruits and vegetables.

Melanoma of the Skin(3):

Melanoma is a cancer of the cells that produce and transport the pigment melanin. In the U.S.,approximately 32,000 new cases of skin melanoma were projected for 1994 (Boring et al., 1994).Melanoma can occur on any skin surface, but in light-skinned populations, a clear excess occurs onthe trunk in men the lower extremities in women, and the head and neck regions and arms in bothsexes. In dark-skinned populations, melanomas occur most often on the palms of the hands and thesoles of the feet. Over the past several decades, the incidence of melanoma has increaseddramatically in the U.S. and several other countries, posing a threat to public health (Devesa et al.,1987; Glass and Hoover, 1989; Ries et al., 1994).

Melanoma represents only about 5 percent of all skin cancers in the U.S., but it accounts for about75 percent of all skin cancer deaths (Boring et al., 1994). Although the precise cause of melanoma isunknown, numerous clinical and epidemiological studies in the past decade have identifiedcharacteristics associated with increased risk of melanoma: a history of sunburns, fair skin, numberof moles, presence of dysplastic or other atypical moles, previous melanoma, family history of melanoma, and immunosuppression.


Blot WJ and Fraumeni JF Jr: Lung and pleura. In Cancer Epidemiology and Prevention, 2nd ed.(Schottenfeld A and Fraumeni JF Jr, eds.). Philadelphia: WB Saunders, 1992.

Boring CC, Squires TS, Tong T: Cancer Statistics 1994. CA Cancer J Clin 44:7-26, 1994.

Chu KC, Tarone RE, Chow WH, et al.: Temporal patterns in colorectal cancer incidence, survival,and mortality from 1950 through 1990. J Natl Cancer Inst 86:997-1006, 1994.

Devesa S, Silverman DT, Young JL Jr, et al.: Cancer incidence and mortality trends among Whitesin the United States, 1947-1984. J Natl Cancer Inst 79:701-770, 1987.

Garewal H: Aspirin in the prevention of colorectal cancer. Ann Inter Med 121:303-4, 1994.

Giovannucci E, Rimm EB, Stampfer MJ, et al.: Intake of fat, meat, and fiber in relation to risk ofcolon cancer in men. Cancer Res 54:2390-7, 1994.

Glass AG and Hoover RN: The emerging epidemic of melanoma and squamous cell skin cancer.JAMA 262: 2097-2100, 1989.

Glynn SA and Albanes D: Folate and cancer: A review of the literature. Nutr and Cancer 22:101-19, 1994.

Lee IM, Paffenbarger RS, and Hsich CC: Physical activity and risk of developing colorectal canceramong college alumni. J Natl Cancer Inst 83:1324-9, 1991.

Lubin JH, Blot WJ, Berrino F, et al.: Modifying risk and developing lung cancer by changing habitsof cigarette smoking. Br Med J 288:1953-1956, 1984.

Ries LAG, Miller BA, Hankey BF, et al.: SEER Cancer Statistics Review, 1979-1991: Tabes andGraphs, National Cancer Institute. NIH Publ. No. 91-2789, Bethesda, MD, 1994.

Sugimura T: Past, present, and future of mutagens in cooked foods. Environ Health Perspect 67:5-10, 1986.

U.S. Department of Health and Human Services: A Report of the Surgeon General: The HealthBenefits of Smoking Cessation. U.S. Department of Health and Human Services, Centers forDisease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smokingand Health. DHHS Publ. No. (CDC) 90-8416, 1990.

Weillett WC, Stampfer MJ, Colditz GA et al.: Relation of meat, fat, and fiber intake to the risk ofcolon cancer in a prospective study among women. New Engl J Med 323:1664-72, 1990.


This appendix contains the comments received during the public comment period for GAFand ATSDR's response to those comments. The comments have been numbered and are initalic with the response directly below each comment.

  1. The prior owner [of the GAF plant] Ruberoid bought the plant in 1920 and after World WarI, asphalt shingles and roll roofing products were manufactured. The manufacturing ofasphalt roofing products long predates the purchase of the plant by GAF in 1967.

As noted in the summary of this public health assessment, the previous industrial activities at thissite are mentioned including roofing and insulation materials manufacturing. ATSDR only detailed the history of the site as it pertains to the industrial activities of GAF.

  1. As an assessment of potential exposures to particulate matter and VOCs you use estimationsof these materials based on data provided to you from GAF from 1990. Conclusionsconcerning exposure are made based on this data. First, as you indicate, this does notrepresent an actual measure of emissions, secondly it does not represent characteristicemission levels prior to 1990. Another document from the Millis Board of Public Health filesdated 4/2/83 discusses a report from David Gordon Associates, Inc. dated 1/11/83 and titled"Addendum to Dispersion Study of Asphalt Saturator for GAF Corporation, 60 Curve Street,Millis, Mass." Table II of that report identifies chemicals found in three separate emissiontests done by David Gordon Associates. Ambient air quality modeling data concerningspecific chemicals are included. Please see Tables I and II of that report for results. Youstate, on page 6 of your report, that "ambient air data are not available on or near the GAFproperty." Please reassess your conclusions based on this data which is now "available."

Tables I and 2 (different numeration methods) of the Addendum to Dispersion Study of AsphaltSaturator for GAF Corporation dated January 11, 1983 (mentioned above) are also based onmodeling data, not ambient air sampling data. Stack emission samples were taken (listed in Table II:volatile organic compound results) and the chemical concentrations in the emissions were runthrough air dispersion models approved by the Environmental Protection Agency. This is a similarmethod used in the 1990 study presented in the public health assessment.

ATSDR analyzed the modeling results from the 1983 report that was sent during the publiccomment period determine the past health impact from GAF emissions. All volatile organiccompounds in the worst-case scenario are below federal ambient air guidelines or within backgroundconcentration range for residential areas. Based on the modeling, there are no health effectsexpected based on the worst-case scenario of emission concentrations listed in the 1983 reportAddendum to Dispersion Study of Asphalt Saturator for GAF Corporation prepared by David Gordon Associates, Inc.

  1. If data from 1990 can be "modeled" from plant production numbers and known informationconcerning typical emissions from this type of manufacturing plant, why can't the same be done for assessment of exposures in the decades prior to 1990?

In the public health assessment on page 6, ambient air was modeled based on GAF using fuel oiland natural gas (GAF converted to natural gas in 1991). Particulate matter in plant emissions wasbelow the air standard before GAF converted to natural gas in 1991. Therefore, the modeling resultsbased on GAF burning fuel oil can be applied to conditions prior to 1990 assuming that the samemanufacturing practices existed in the past. Also, see response to #2.

  1. You report that Millis has elevated rates of lung and skin cancers. This raises questionsconcerning exposure to PAHs in general. Since it is known that these chemicals are the mostcommon to be found in air emissions from these types of manufacturing plants [roofing], whywasn't the issue discussed more fully? Have increased rates of lung and skin cancers beennoted in other towns with long term manufacturing of asphalt products, especially plantslocated in more rural areas as in Millis?

The lung and skin cancer issue was not discussed more fully because the air emissions from GAFwere not high enough to further research a cause and effect relationship. Based on modeling results,no adverse health effects are expected. According to the Massachusetts Department of Public Health,Bureau of Environmental Health Assessment, if there is concern about a significant elevation fromthe Massachusetts Cancer Registry, the individual(s) is (are) instructed to submit a written request tothe chief of the Community Assessment Unit for a Phase II investigation to further evaluate the health outcome data.

The levels of PAHs emitted from GAF were not high enough to warrant a comparison analysis withother roofing plants. Therefore, ATSDR did not evaluate health outcome data from other rural areas in the United States with roofing manufacturing plants.

  1. Could the withdrawal of large volumes of water to run a major bottling plant have impactedthe groundwater flow in the decades prior to the groundwater testing done in the 90's in response to the oil spills on the GAF site? Does including that information changeassessments of typical patterns of movement of contaminants on groundwater at the site?

Yes, withdrawing a large volume of water can alter the groundwater flow in any area. However,past records of groundwater withdrawal rates, aquifer recharge, flow system dynamics, otherpumping wells, and the other associated data necessary for an accurate evaluation of pumpingdrawdown effects do not appear to be available in this particular area of Millis. All of these pastgroundwater data elements would have to be accounted for to develop a reliable estimate of past effects.

  1. In assessing the past impact of the facility on municipal wells, would one expect to findconsistent positive ground water test well results on a site 1/3 mile away from a well more than fifteen years after those wells tested positive for VOCs?

To establish a pathway from a source to exposure, one would expect to find a consistent link ofdetected contamination. The historical lack of such a link, or plume, from the GAF site to the Milliswells suggests the possibility of multiple sources of contamination. In fact, Millis has a history ofgroundwater contamination from several sources; MADEP has documented at least three other sitesnear the Millis municipal wells with a confirmed hazardous material release that contaminatedgroundwater within the last ten years. Also, historical groundwater monitoring results from GAF tenyears ago also show contamination within the interior area of GAF property and little to none in thewells located at the boundary of GAF. It is also ATSDR's understanding that the municipal watersystem in Millis closely monitors water quality to prevent potential exposure to drinking water thatdoes not meet federal and state water quality standards (for example, the closing of wells #1 and 2).

  1. It is unknown how much material [from GAF] is buried at local unlined landfills in Millis. Itspotential impact on municipal wells #1 and #2 is not addressed at all in the report. Whetheror not private wells could be potentially impacted is not discussed either.

There were no disposal records pertaining specifically to GAF or groundwater monitoring dataavailable for these landfills in Millis. As mentioned in comment #6, there are several groundwatercontamination sources near the Millis municipal wells. However, municipal water is a regulateddrinking water source which is required to be monitored. Private wells throughout the United Statesare generally not regulated (other than for bacteria) and can be potentially impacted from any sourceof environmental contamination. The private wells located nearest to GAF were hydrologically upgradient and would not be affected by GAF.

  1. A potential future problem source of groundwater contamination that is not discussed is theburied railroad car containing "a solid asphalt shingle/salt mixture" that was found whileremoving the leaking underground storage tank at the site in 1992. Because the plant isupgradient and close to the municipal wells, wouldn't some recommendations concerningappropriate continued monitoring of the buried railroad car be reasonable at this time?

The groundwater monitoring taking place at GAF would detect any contamination that potentiallythreatens the municipal wells.

  1. Were any other asbestos containing products [besides asbestos containing concrete] storedon-site? A document from the town of Millis concerning taxes indicates that in 1965Ruberoid paid $500.00 to the town for an "asbestos storage shed". Why would a plant thatdidn't make asbestos products need an asbestos storage shed?

GAF responded in writing to ATSDR's inquiry regarding the asbestos manufacturing and storageissues. As quoted from their response: "Finally, with respect to the Seventy-Fifth Anniversary of theIncorporation of the Town of Millis, it is apparent that the asbestos-containing products mentionedtherein were not and are not matters of concern. This excerpt shows that Millis was notmanufacturing these asbestos-containing products, but rather distributing products manufacturedelsewhere. As the excerpt states, the products manufactured at Millis were various asphalt materials,waterproof papers, and reflective insulation. The other products -- asbestos-cement sidings, asphaltand vinyl floor tile, gypsum products, and certain insulation products -- were handled only fordistribution into New England."

The storage shed appears to have been used for storage and distribution of asbestos containingmanufactured products that were manufactured elsewhere.

  1. Given the lack of independent air quality testing results and lack of ambient air qualitysample data, the question arises as to why ATSDR would draw [the] conclusion that there isno threat to human health from hydrocarbon particulate matter or from VOC emissions asstated in the report.

The Clean Air Act of 1990 under Title V, Permits and Title III, Air Toxics gives allowance to drawconclusions based on the projected ambient air concentration from approved (by the EnvironmentalProtection Agency) air models.

ATSDR has worked on projects where ambient air sampling data were available and modeling wasdone to validate the accuracy of models for long-term air quality evaluations. In most cases, the airmodels are more conservative (they over-predict concentrations) than actual ambient airconcentrations. In this case of Millis, we made assumptions which would make the models tend to over-predict ambient air concentrations.

  1. Relative to the incidence of lung cancer, Appendix D, Summary of Cancer Risks states "Ofspecial concern are the effects of the by-products of the combustion of fossil fuels, mostnotably polycyclic hydrocarbons (PAHs)". With those two pieces of information, the aboveconclusion that there is "not a threat to human health from.... VOCs in manufacturingemissions at GAF" is at least arguable. Also, of note was the level of VOC emissions stated inthe report of 0.01mg/m3. This level was based on an estimate, yet ATSDR was able toconclude from an estimate, which was based on an indirect measurement, that VOC levelswere "100-1000 times (this is a range of a magnitude of 10 times) less than the EPA's health based comparison values..."

As stated in Appendix D under the lung and larynx cancer section: "Although suggestive, theassociation of lung cancer and PAHs has not been conclusively demonstrated, as other componentsof air pollution may also be carcinogenic." There are several chemicals released during fossil fuelcombustion. Based on the information provided to ATSDR, the levels of VOCs reported beingemitted during natural gas combustion (during worst-case scenarios) at GAF appear to be belowlevels associated with health effects.

Page 6 of the public health assessment lists the predominant volatile organic compounds (VOCs) asanthracene, pyrene, and fluoranthene. The 100-1000 times less than EPA's comparison refers to 100less for anthracene, and 1000 less for pyrene and fluoranthene. EPA's health risk comparison values(or Risk-Based Concentrations) are chemical concentrations corresponding to fixed levels of risk (alifetime cancer risk of 1 person in 1,000,000 getting cancer). This is a conservative risk calculationdesigned to show chemical concentrations in an environmental medium (i.e., air, water, soil) thathave the potential to increase one's "risk" of getting cancer. The levels of VOCs emitted from GAFare well below (ranging from 100 to 1000 times below) this conservative risk-based concentration.

  1. The first paragraph on page 10 in section D states that "The MADPH reported an increase(that was statistically significant) in the rate of lung... cancer in the town of Millis whencompared to rates for the state. Yet the last paragraph in section D states that "ATSDRcannot conclude any association between contaminants from any source (including GAF)and the incidence or mortality of cancer in Millis, Massachusetts" Given the above, would the antithesis of this conclusion also be correct? One cannot conclude any disassociation between contaminants from any source and the incidence of cancer in Millis.

Based on the reported stack emissions and air modeling data, ATSDR was not able to conclude that the contaminant concentrations were at levels of health concern. Also see comment #4.

  1. Page 3: The storage tank listed in the report as currently closed, is still going through theremediation process supervised by the MADEP.

According to GAF staff and records, the underground storage tank (UST) was closed in place inJuly 1992. The Chief of the Millis Fire Department and a representative of MADEP were presentduring the closure. On November 19, 1992, GAF submitted groundwater sampling data from thevicinity of the UST. These data indicate that the UST had not affected the groundwater.

  1. Page 5 Air: Ambient air quality should be required as a full study and evaluation samplingis needed at this site. Using previous records (modeling) from studies nine years ago using particulate matter estimates vs. actual ambient air testing is futile.

Please see comment #10.

  1. Page 19 Quality Assurance: "ATSDR assumes that adequate quality assurance and controlmeasures were taken during chain of custody, laboratory procedures, and data reporting.The validity or analysis and conclusions drawn in this document are determined by thatavailability and reliability of the information." Perhaps one government agency should takethe time to prove a theory. A theory in this case could have the potential to prove the futurewith regard to health concerns. As I quote you in your conclusion on the following page 8-9:"ATSDR concludes there is no current threat to the town of Millis municipal water qualityfrom GAF groundwater contamination, etc. However, private well owners should test theirwell water for VOCs since TCE was detected and Ann & Hope Company well water from anunknown source. If said wells (private) are "upgradient" then why the concern?

Private wells are generally not required to be tested for chemical contamination. As noted byMADEP, there have been several documented hazardous substance releases in Millis that have (orcould potentially) contaminated the groundwater. As a precaution to prevent exposure to thesepotential groundwater contaminants, it is recommended that private well owners who use well wateras a potable water source have it tested for possible chemical contamination.

  1. I ask your staff take an additional evaluation on actual findings, rather than baseinformation on this case, for all the residents of Millis, now and in the future.

ATSDR reviewed all available information including, community concerns, health outcome data,environmental sampling data, safety studies, and field technician evaluations. If additionalinformation becomes available in the future, ATSDR will evaluate the provided information andupdate this public health assessment.

  1. Has the site done a [groundwater] sampling of the bottom of the buried rail car? Has therebeen any testing of the "saturator" where solvents were used to clean asphalt on a regular basis?

On November 19, 1992, GAF submitted to MADEP groundwater samples from the vicinity of theburied rail car. These data indicate that the rail car had not affected the groundwater. Twenty-threegroundwater monitoring wells are also located throughout the GAF property boundary and aretested frequently. The groundwater sampling data do not indicate that the saturator is a source ofgroundwater contamination. According to GAF, the only solvent used to clean the saturator waskerosene.

  1. Is there an explanation why the test well sampling was interrupted by obstructions placed bythe owner? Is there an explanation why the plume of contaminants is traveling northwestfrom the release site as opposed to the northeast travel of groundwater?

ATSDR is not aware of owner-related obstructions to monitoring well testing. There are severalpublicly available reports dating from 1990 that show the groundwater test results of the on-sitemonitoring wells. Each map that ATSDR reviewed showed the groundwater contamination plumegenerally flowing in a northeast direction. Note that some variability in groundwater flow is possibledepending on weather conditions (i.e., rainfall levels).

  1. Why are there no test wells along the southern boundary of the site?

There are five monitoring wells along the southern perimeter of GAF property; GT-10, GT-8, GT-7,GT-3, and GT-1. However, most of the monitoring wells are located downgradient of thegroundwater flow path and around the groundwater contaminant plume.

  1. What are the risks of PAHs associated with the contamination?

ATSDR is assuming this comment is referring to health risks associated with the ingestion of PAHcontaminated groundwater at GAF. There is no risk to PAH contaminated groundwater related toGAF because there is no known human exposure to this groundwater. Monitoring wells are not used as a potable water source.

  1. Air stripping of municipal wells 1 & 2: Does the water merely go through a stack, mix with air and release the VOCs back into the ecosystem?

Air stripping (also called in-well vapor stripping) involves the creation of a groundwater circulation pattern and simultaneous aeration within the stripping well to volatilize VOCs from the circulating groundwater. Air-lift pumping is used to lift groundwater and strip it of contaminants. Contaminated vapors may be drawn off for aboveground treatment or released in to the vadose zone (area between land surface and water table) for bioremediation. Partially treated groundwater is forced out of the well into the vadose zone where it reinfilterates to the water table. Untreated groundwater enters the well at its base, replacing the water lifted through pumping. Eventually, the partially treated water is cycled back through the well through this process until contaminant concentration goals are met. Reported advantages of air stripping include lower capital and operating costs due to use of a single well for extraction of vapors and remediation of groundwater and lack of need to pump, handle, and treat groundwater at the surface (EPA, 1997 Groundwater Remediation Technologies Analysis Center, )

  1. The [air stripping] process exposes the water to the open atmosphere, thus requiringchlorination. What are the short and long term effects of chlorine? What are the effects whenit is heated and becomes a vapor (such as when showering, bathing or cooking)?

All municipal water (public water supply) in the United States is required by the Safe DrinkingWater Act to be protective against harmful levels of bacteria. This is generally done throughchlorination, as is the case in Millis. Chlorine levels are monitored daily in municipal systems. Achlorine residual is not harmful to your health. Chlorine protects against several gastrointestinaldiseases and other infections. It is also safe to cook and bathe with chlorinated municipal water.

  1. Can we assume the treatment on municipal wells 1 & 2 is risk free?

All municipal water systems are regulated under the Safe Drinking Water Act (SDWA) and itsamendments. This entails frequent water quality monitoring. The finished product (water that goesto your tap) from municipal wells 1 & 2 would be within EPA regulations and safe to drink.

  1. Municipal well #3 is not shown on your map; it is a mile from the site.

Thank you.

  1. Citizens complain about malfunctions at GAF [with regards to air emissions, sand,limestone, oil, asphalt fumes], however, it is unclear whether GAF documents thesemalfunctions.

According to GAF, records are kept that document malfunctions and these records are submitted toMADEP to the extent required by applicable laws and regulations. The records sent to MADEP are available to the public.

  1. The Millis Board of Health responded to complaints of airborne sand being deposited on thepremises. The Board of Health took samples, returned and took more samples. Were thereany citations from this event? Was the volume of discharge established? How do these types of events affect the Gradient reports?

ATSDR is not aware of any citations because of airborne sand, however, GAF made severalmanufacturing controls in regards to fugitive emissions (listed on page 13 of the public healthassessment). These control measures are detailed in the Gradient reports.

  1. What are the health effects of ingestion of sand, limestone, and oil coated granules on thestomach, liver, kidney, and colorectal area of humans [from fugitive emissions]? What arethe possible effects on the skin.

At this time, ATSDR did not find any toxicological studies involving short-term exposure tolimestone dust, sand, or oil coated granules in humans or experimental animals via ingestion.Limestone dust is predicted to be an irritant of the eyes, skin, and respiratory tract due to frictionalaction. However, such fugitive emissions pertaining to GAF are not expected to cause healthproblems. Dust concentration, exposure duration, amount of skin exposed, and amount of dustingested are not expected to be significant for nearby residents. Clothing would be enough to protectnearby residents' skin from abrasion.

  1. This report does not mention the TRC Environmental Corporation report of June 1994,which describes the production stacks as being too short, and not installed using the bestengineering practice. The stacks were raised in 1994 to TRC specifications.

Neither ATSDR or GAF are familiar with the "1994 TRC report." Please provide more informationregarding where ATSDR can obtain a copy of this report. However, on page 13 of the public healthassessment the stack extensions are mentioned as one of the completed emission control measures.

  1. Has there been a review of the RES permit for an alteration to the process of adding odorousgas or vapors released from the asphalt coating operation which according to the 1994 TRCreport were discharged to the atmosphere through four stacks in the roof of the processbuilding?

Restricted Emission Status (RES) permits are permits that restrict emissions below federal operatingpermit thresholds. GAF submitted a RES application dated September 1994 that was prepared byGAF's consultant Earth Tech. The MADEP reviewed and approved the application by letter datedApril 25, 1995. This permit does not relate to "an alteration to the process of adding odorous gas orvapors released from the asphalt coating operation."

According to GAF, the four stacks referenced in the question are extended roof vents situated toexhaust the humid air from the cooling loops of the roofing line. These stacks emit small amounts ofair from the vicinity of the coating line, because the coater is fully enclosed (top and sides) withaluminum sheets and plexiglass observation ports, and it is maintained under negative air pressurerelative to its surroundings.

ATSDR and GAF are not familiar with the "1994 TRC report." Please provide more informationregarding where ATSDR can obtain a copy of this report if you would like us to comment on it.

  1. In the Gradient reports that are cited, is there confirmation that the consumption of naturalgas is relative to the amount of vapors incinerated on-site?

In the Gradient reports, all air modeling using natural gas (and past fuel oil) combustion wasperformed using the EPA approved Industrial Source Complex Short Term (ISCST) air dispersionmodel. GAF maintains records of its emissions and provides routine inventories of these emissions to MADEP to document that they remain below permit requirements.

  1. Many manufacturers of fiberglass warn that fiberglass may be a possible cause of cancer. How are fiberglass particles on shingles handled [at GAF] since they are not flammable?

There is no consistent evidence that fiberglass causes cancer in humans (Meditext® MedicalManagement database). The Occupational Safety and Health Administration (OSHA) does not havea permissible exposure limit (PEL) for fiberglass fibers. In 1992, OSHA requested comment on aproposed PEL of 1 fiber/cubic centimeter for a Time Weighted Average (TWA). However, OSHAhas never acted on this proposal. The TWA of 1 fiber/cubic centimeter is also the value listed in the1998 TLVs and BEIs: Threshold Limit Values for Chemical and Substances and Physical AgentsBiological Exposure Indices by the American Conference of Governmental Industrial Hygienists(ACGIH).

Even though OSHA does not regulate fiberglass fibers, GAF has tested the Millis manufacturingline for airborne fibers. These tests have consistently shown that the fiber levels are either belowdetection limits or less than 1% of the proposed PEL.

  1. Emissions of diesel exhaust from trucks are a possible cause of cancer. The fugitiveemissions from nearly 100 trucks/day ... must be added to the air quality overview of this site.

Automobile and other background sources do contribute to the total amount of contaminants thatpeople are exposed to. State wide air quality (a select few contaminants) and vehicle emissions aremonitored by MADEP. The nearest air monitoring station to Millis is in Easton. According to the1997 Air Quality Report, "Overall, there is a substantial projected emission reduction for all fourpollutants (VOCs, nitrogen oxides, sulfur dioxide, and carbon monoxide) from 1990-1996, eventhough there has been significant growth in population and economic activity in Massachusetts." Fora copy of the 1997 Air Quality Report that summarizes all ambient air monitoring locations, call theDEP Infoline: (617) 338-2255 or you can access the report via the Internet:

  1. There is no mechanism in place to prevent the large plastic bags covering finished producton-site from blowing into the community. There must be a calculation of risk of suffocationby one of these large plastic bags by school children who walk past this site to and fromschool on windy days or play in their yards. These bags provide no warning noridentification of risks to keep out of reach of children.

The site is adequately fenced to prevent debris from blowing off-site under typical weatherconditions. However, these large bags should be anchored to prevent them blowing off during windyor stormy weather conditions. There is no calculation known that would estimate the risk ofchildren suffocating from one of these bags.

  1. The odor reduction devices installed on asphalt delivery trucks do not eliminate fugitiveemissions from those trucks, but merely restricts the release which of fumes which can buildup during the trucks trip and be released all at once in the area around the site.

According to GAF, the devices remain in place and in operation throughout the trip to and fromGAF, and the drivers never "release" fumes through the devices. The devices are highly effectivethough no device can eliminate 100% of all fugitive emissions. The odor reduction devices weredesigned and installed at GAF's request specifically to minimize odors associated with the asphaltunloading operations in Millis, but the devices are not required by law.

  1. The issue of whether asbestos was used in manufacturing at this site needs to be resolved.Mr. Johnson needs to clarify what materials were used in the manufacturer of reflectiveinsulation referred to in reference #4 of the report.

Please see comment #9.

  1. The conclusions are based on the inexact science of cancer reporting, failing to include anyaccount of the many other elements of human health which may be affected by this facility.The reliance of the opinions of advocates for this facility including the government of thetown of Millis, is disheartening.

ATSDR uses all pertinent and possible information when evaluating a site (technical and non-technical, see comment #16). No matter how extensive a state's cancer registry is, cancer reportingin itself has several flaws that can alter the statistical outcome of results. However, it is the best reporting methodology known to account for large numbers and types of cancer cases.

  1. The distance of some homes from the GAF site boundary is about 60 feet; not 0.25 mile asstated in the report.

The distance has been changed in the public health assessment.

1. National Institutes of Health called Cancer: Rates and Risks, 1996 4th Edition
2. National Institutes of Health called Cancer: Rates and Risks, 1996 4th Edition
3. National Institutes of Health called Cancer: Rates and Risks, 1996 4th Edition

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. #