PETITIONED PUBLIC HEALTH ASSESSMENT
SOUTHERN MARYLAND WOOD TREATING
NATIONAL PRIORITIES LIST (NPL) SITE
HOLLYWOOD, ST. MARY'S COUNTY, MARYLAND
The tables in Appendix 2 list contaminants in each medium. Thosecontaminants are evaluated in subsequent sections of the publichealth assessment update to determine if exposure to them haspublic health significance. ATSDR selects and discussescontaminants based upon several factors. They includeconcentrations on and off site, the quality of the field andlaboratory data, sample design, comparison of on- and off-siteconcentrations to background concentrations (if available),comparison of on- and off-site concentrations to public healthassessment comparison values for noncarcinogenic and carcinogenicendpoints, and community health concerns.
The listing of a contaminant in the tables does not mean that itwill cause adverse health effects if exposure occurs at thespecified concentrations. Contaminants included in the tablesare further evaluated in this public health assessment update. The potential for adverse health effects resulting from exposureto contaminants of health concern is discussed in the PublicHealth Implications section.
Comparison values for ATSDR public health assessments arecontaminant concentrations in specific media that are used toselect contaminants for further evaluation. ATSDR and otheragencies developed those values to provide guidelines forestimating the media concentrations of a contaminant that areunlikely to cause adverse health effects, given a standard dailyingestion rate and standard body weight. See Appendix 3 for adescription of the comparison values used in this public healthassessment update.
EPA completed a more extensive investigation of the contaminationin all media at the SMWT site since the original ATSDR publichealth assessment was prepared. Figure 7 shows the location ofthe monitoring wells and soil samples. The public healthassessment presents the maximum concentrations of thecontaminants of health concern found at the site. No distinctionis made between RI/FS data previously reviewed and the morerecently collected remediation predesign data. The quality ofboth data sets is comparable (see the Quality Assurance andQuality Control section for further discussion); therefore,combining the values does not present a quality assuranceproblem.
Analytical data obtained from EPA show the presence ofpolynuclear aromatic hydrocarbons (PAHs), the main constituentsof creosote; phenolic compounds, a class of chemicals associatedwith pentachlorophenol (PCP) wood-treatment processes; andvolatile organic compounds (VOCs), most likely from fuel oil orcarrier oils used to dilute the creosote and PCP for use in woodtreatment. PAHs found at the SMWT site were grouped ascarcinogenic and noncarcinogenic; the health effects of thechemicals within the respective groups are similar. Theconcentrations of specific PAHs in each group are totalled foreach medium, on and off site, and appear in the tables (seeAppendix 2)(1,10).
The highest concentrations of contaminants at the SMWT site arein the soils within the piling containment wall (the processarea, also called the containment area) and in the land treatmentarea. In only a few samples were metals detected atconcentrations greater than those normally found in Marylandsoils (1).
In one small area on the northeastern side of the SMWT property, EPA's sampling indicated slight soil contamination outside thefence. That area is near a former railroad which may be a sourceof the low-level contamination. The contamination is limited tosurface soils, which are covered by grasses, locust trees, andbriars. The area is not very accessible, and EPA plans to removethe soil during the remediation. Table 4 lists the contaminantsfound in the area, and Figure 8 shows the areas where EPA plansto remediate the surface soils throughout the SMWT site.
Many contaminants are more concentrated in subsurface soils(greater than 2 feet deep) than in surface soils (0-2 feet deep). It is not unusual to find higher contaminant concentrations insubsurface soils at inactive sites, where sources of continuedcontamination have been eliminated because the more volatileorganic compounds evaporate and because many contaminants migratedownward over time. Table 1 (all tables are in Appendix 2) showsthe maximum concentrations of specific contaminants found insoils on site.
The only surface water on site is a freshwater pond in thesouthwestern corner of the site. The surface water contains lowconcentrations of a few contaminants. See Table 1 for themaximum concentrations of contaminants found in on-sitesediments. Table 2 shows the maximum concentrations ofcontaminants in on-site surface water.
PAHs, VOCs, and PCP-related phenolic compounds are found only inthe shallow water table aquifer under the SMWT site. There aretwo organic layers in the shallow aquifer, one floating on top ofthe water table and one settled at the bottom. EPA installed apiling wall around the process area that contains most of thecontaminated shallow groundwater. There is some shallowgroundwater contamination outside the piling wall (on the southside) in the vicinity of monitoring well 12 (MW12)(see Figure 7).Table 2 lists the maximum concentrations of the contaminantsfound in the groundwater on site (1).
According to data reviewed, no on-site analyses of water from theNanjemoy or Aquia aquifers had been conducted. EPA tried to sample the on-site production well, but the well was obstructedand samples could not be obtained. Water samples from fourmonitoring wells in the Upper Chesapeake aquifer do not havedetectable contamination.
Air samples were taken near workers and at site boundaries duringan 11-day period when an EPA contractor was installing monitoringwells and collecting samples on site. All but three of the 79samples were taken at four air monitoring stations at the fenceline near the pond, near the processing area, at the mainentrance, and near the tanks on the east side of the site. Thosemonitoring stations were designed to measure the concentrationsof particulates and coal tar pitch volatiles (CTPVs), whichinclude PAHs, in the ambient air at the perimeter of the site. The other samples were obtained from monitors in the drillingzone while employees were drilling the wells. Only three sampleshad detectable CTPVs (0.17, 0.13, and 0.068 milligrams per cubicmeter (mg/m3); all are within the permissible exposure limit(PEL) of 0.2 mg/m3 set by the Occupational Safety and HealthAdministration (OSHA) (11). That PEL applies to workers only. ATSDR is unable to evaluate nearby resident's past exposures toair contaminants due to limitations of the data. However, ATSDRwill review air emissions information associated with siteremediation design and implementation. For more detail, seeATSDR's Plan of Action to Evaluate the Proposed RemedialAlternative for the Southern Maryland Wood Treating Site. Particulate concentrations were extremely low, ranging from adetection limit of 0.0099 mg/m3 to 0.577 mg/m3. The AmericanConference of Governmental Industrial Hygienists (ACGIH) standardfor particulates is 10 mg/m3 (12). The EPA particulate standardsfor ambient air are 0.150 mg/m3 (24-hour average), and 0.050mg/m3 (annual median) for particulates smaller than 10 microns indiameter. Because EPA samples were collected to determinecompliance with the OSHA standards during remedial activities,the particulate analyses were for total nuisance particulates. Therefore, the particulate analytical results cannot be comparedto the ambient air standard, which is set at levels to protectthe general population.
There are 13 above-ground tanks, one underground fuel tank, apropane tank, a boiler treatment-water makeup tank, and tworetorts on site. Three tanks are empty; one contains water fromRI/FS activities, and two contain 26 gallons or less of liquid. None of the tanks sampled contained 2,3,7,8-tetrachloro-di-benzo-p-dioxin (2,3,7,8-TCDD). However, approximately 11,960 gallonsof wastes contain other dioxin isomers. There are approximately2,140 gallons of wastes from the original process plant in tanksthat do not contain any dioxin isomers. Table 3 lists themaximum concentrations of the contaminants found in the liquidsand sediments in the tanks on site (1).
The tanks seen during the site visit were constructed of single-wall steel with a coating of asphalt. The materials used at thefacility could have damaged that type of coating during spillsand overfills while the facility was in operation. The tankswere not equipped with spill and overfill protection. Theasphalt coating, even if in good condition, offers poorprotection from corrosion, and does not meet current standardsfor corrosion protection. Also, the high soil-moisture contentand high groundwater level at the site could speed up thecorrosion process.
In addition to the possibility of a release as a result ofcorrosion, there is a potential for tank puncture as a result ofhunting in the area. During the site visit, ATSDR staff observedthree deer on site, and local sources said that hunting is commonin the area. A stray bullet from a rifle could pierce the wallof the tank and cause a sudden release into the environment. Because there are no alarms or secondary containment systems, theliquid contents could be completely released before such anaccident was discovered.
Although the tanks did not appear to be leaking during the sitevisit, they are old and unmaintained. EPA has not conductedintegrity testing on any tanks. Several tanks are outside of thecontainment wall area.
Several waste drums off site have been used during the remedialand sampling activities. Access to the drums is unrestricted atvarious locations in the woods and fields around the site. Theoff-site drums are not labeled and could be vandalized. EPA hasadvised ATSDR that the drums contain groundwater from bailing themonitoring wells when samples are collected. Empty shotgunshells were found near some of the off-site drums (13,14).
Several buildings and sheds remain on the property. Thebuildings are constructed of cinder blocks; the sheds are made oftreated lumber poles and tin roofing and siding. The locationsof those structures are shown in Figure 3. ATSDR's previouspublic health assessment identified the structures as pathwaysfor potential exposure to site-related contaminants. The wood-scrape tests on which that determination was based are notrepresentative of contamination that occurred during thefacility's operation because the samples were collected fromtreated lumber. However, no information was available regardingsurface contamination of the cinder blocks and concrete in thebuildings. During plant operations, it is likely that wood-treating chemicals were spilled in the process building.
Two streams are on the SMWT property - Old Tom's Run, whichbegins at the discharge of the pond on site, and a tributary onthe eastern side of the site that begins outside of the fencedarea (see Figures 2 and 9). The two streams join approximately1950 feet south of the site. Questionable surface-water resultswere reported for chromium at one sampling location. Theduplicate samples showed markedly different concentrations ofchromium; one showed 323 ppb, the other was a nondetect. Because other samples from the eastern tributary did not haveelevated chromium levels, and the duplicate sample showednondetectable levels of the metal, that single high-chromiumanalytical result is considered inaccurate and is not included inTable 5.
The upper section of Old Tom's Run has contamination from discharges associated with the SMWT site. The surface watercontains phenolic compounds (14). Samples in Old Tom's Run alsohad elevated lead concentrations, both in the surface water andsediments. See Table 5 for the maximum concentrations ofcontaminants in the surface water of Old Tom's Run. Because mostof the contaminants at the SMWT site are not very water soluble,the highest concentrations of contaminants are found in the OldTom's Run sediments (see Table 4). EPA took 17 samples from OldTom's Run, above the confluence with the eastern tributary, and20 samples below that point (see Figure 9). However, EPA has notsampled the banks along the creek to adequately determine thehorizontal extent of contamination that may have occurred duringflooding along the creek bed.
Off-site groundwater sampling has been conducted in monitoringwells and private drinking-water wells in the area. Sampling hasbeen done by both EPA and the state of Maryland (through the St.Mary's County Health Department [SMCHD]). Thus far, nocontaminants from sources on the SMWT site have been measured ineither the shallow or the deep groundwater residential wells offsite (see Figure 10). However, some elevated metalconcentrations and low PAH concentrations have been detected inshallow monitoring well 15 (MW 15)(see Figure 7) outside the fenced area near the SMWT site.
Since 1978, the SMCHD has sampled seven residential wells nearthe SMWT plant. The depth of five of the wells is 300-350 feet; the other two wells are less than 50 feet deep. Initially, allthe wells were sampled monthly; sampling frequency was decreasedto quarterly beginning in 1986. The sampling is ongoing, and nocontaminants have been detected in those wells (15). However,all but one of those wells are upgradient of the SMWT site;therefore, no contamination is expected in those areas.
EPA has several monitoring wells in the shallow water-tableaquifer and four wells in the Upper Chesapeake aquifer outsidethe plume of contamination. The shallow wells can be used tomonitor for plume movement in all directions in the shallowaquifer. Samples taken from the four Upper Chesapeake aquifermonitoring wells do not show any contamination; however, none ofthem are downgradient from the plume of contamination in theshallow aquifer. See Table 5 for the maximum concentrations ofcontaminants of concern in the off-site groundwater aquifers.
Based on data presented in the on-site air contamination section,no PAHs or particulate contaminants at levels of concern arecurrently migrating off site. However, in the past, SMCHDofficials received complaints from area residents about odors andblack residue coating their houses when the plant was inoperation (15). No air sampling was conducted during that time to determine what chemicals, if any, were migrating off site.
The conclusions in this public health assessment update are basedlargely upon data developed by EPA contractors and SMCHD. Whendescriptions were provided, the QA/QC measures appearedconsistent with measures normally taken with environmentalsampling and analysis. The data are assumed to be accurateunless specifically qualified.
A brief discussion of data quality is in Appendix F of the RI/FS. The RI/FS was prepared by CDM Federal Programs Corporation. Theonly major data problem identified resulted from incorrect tuningof the Gas Chromatograph/Mass Spectrophotometer (GC/MS), whichaffected the Phase II volatile organic analyses. Those resultsare designated by EPA to be used for engineering purposes onlybecause effects of the improper instrument calibration on thedata quality could not be determined (10).
The analytical data developed by Dames and Moore and presented inthe "Hazardous Waste Remedial Action Predesign Report" appear tobe of high quality for the contaminants listed in the ATSDRtables (See Appendix 2). There were some analytical difficultieswith laboratory control samples for certain organic compounds,which resulted in a few PAH data points for environmental mediabeing listed as estimated (J). The laboratory reports statedthat the PAHs were present in the samples, but, due to thelaboratory difficulties associated with the determination ofPAHs, the quantities were estimated. However, because of thelaboratory difficulties, those values are estimated higher thanthe true values are believed to be (1). To be conservative, theJ values from the PAHs have been included in all ATSDR tables inAppendix 2.
In tank samples, there were matrix effects on some contaminantsresulting from high concentrations of organic compounds. Whensamples were diluted to overcome the matrix effects, Dames andMoore qualified the diluted data points as estimated (J). Whensumming the carcinogenic and noncarcinogenic PAHs, if any valuehad a J qualifier, the total is qualified with a J in the ATSDRtables (1).
Two QA/QC issues were associated with the Dames and Mooreanalytical data for the metals of concern. Analysis of duplicatewater samples from monitoring well 22 showed conflicting results. One sample showed 87.5 ppb lead; but, in the other sample, leadwas not detected at 2 ppb. Because the 87.5 ppb was notconfirmed in the duplicate sample, ATSDR has not included thatdata point in the tables. Analysis of duplicate surface-watersamples from one sampling location in the eastern tributary alsoshowed conflicting results. One sample showed 323 ppb chromium;chromium was not detected in the other sample. Because theduplicate sample showed nondetectable levels of the metal, thatsingle high-chromium analytical result is considered inaccurateand is not included in Table 5 (1).
Some of the water samples analyzed for metals were filtered, aprocedure which removes any metals adhering to particulates andalso removes colloidal chemicals. The filtered samples reflectonly dissolved metals concentrations instead of total metals;therefore, the metals concentrations reported may be low. Metalsdata from filtered samples do not represent the true dose tohumans if the unfiltered water is ingested. ATSDR does notconsider filtered samples adequate for comparison to EPA DrinkingWater Standards, which are based on unfiltered samples. Thesamples taken of residential wells by St. Mary's County HealthDepartment were unfiltered (16). All water samples taken byDames and Moore were filtered before metals analysis (1,17). TheRI/FS does not state whether the contractor filtered the watersamples before analysis for metals, but it does say that theorganic compound samples were unfiltered before analysis, whichcould imply that the metals samples were filtered (10). ATSDRhas assumed that all water samples taken by EPA contractors formetals analysis were filtered. The metals data in Tables 2 and 5are all from EPA contractors and are assumed to representdissolved metals concentrations.
There are no apparent physical hazards to the general publicassociated with the site because it is secured by a high fence.
TOXIC CHEMICAL RELEASE INVENTORY
ATSDR conducted a search of the EPA Toxic Chemical ReleaseInventory for St. Mary's County for 1987, 1988, and 1989. Forany of 300-plus toxic chemicals, EPA requires that themanufacturing industry report annual estimated releases into theenvironment. No facility chemical releases were listed for thecounty during those three years.
To determine if people are exposed to contaminants released fromthe SMWT site, ATSDR evaluated the environmental and humancomponents that lead to human exposure. This pathways analysisconsists of five elements: 1) source of contamination, 2) environmental medium in which the contaminants may be presentor may migrate, 3) points of human exposure, 4) routes of humanexposure such as ingestion, inhalation, or dermal absorption, and5) receptor population. ATSDR classifies exposure pathways ascompleted, potential, or eliminated. A completed exposurepathway existed in the past, and may exist in the present orfuture if all five elements of an exposure pathway link thecontaminant source to a receptor population. Potential pathways,however, are defined as situations in which at least one of thefive elements is missing, but could exist. Potential pathwaysindicate that exposure to a contaminant could have occurred inthe past, could be occurring now, or could occur in the future. Pathways are eliminated when at least one of the five elements ismissing and will never be present. Completed and potentialpathways may also be eliminated when they are unlikely to exist. A list of all completed, potential, and eliminated exposurepathways at SMWT appear in Table 6 of Appendix 2.
Several exposure pathways are likely to have been completed inthe past at the SMWT site. Plant workers and nearby residentscould have been exposed to wood-treating chemicals predominantlythrough air releases during plant operation. However,concentrations of hazardous materials reaching receptorpopulations are unknown because no environmental sampling wasdone during active facility operations. Those pathways aredescribed here.
Plant employees were probably exposed to wood-treating chemicalsduring plant operations via inhalation, skin contact, andinadvertent ingestion of compounds. SMCHD staff said that theskin of plant workers appeared yellow. The Chief Medical Officerfor SMCHD said that a study of occupational exposure of plantworkers, in collaboration with Johns Hopkins University, wasproposed, but never undertaken (15).
Staff from SMCHD also reported that nearby residents wereconcerned about plant operations emitting odors and depositingblack residue on their houses. However, specific informationabout those releases is unavailable (15).
Although plant workers and nearby residents were probably exposedto wood-treating chemicals in the past, ATSDR is unable toevaluate such exposures because the concentrations of chemicalsthat reached the people working, living, or playing near the siteare unknown.
Trespassers may also have been exposed to environmentalcontaminants on site before the fence was put up around theproperty. The fence appears to have virtually eliminatedunauthorized entry. The potential for short-term exposures tocontaminated soils, surface water, sediments, and ambient airexists for any persons on site. Details regarding possiblecontact of trespassers with on-site contaminants cannot bedetermined because specific activities that would lead toexposure are unknown.
A potential off-site exposure pathway exists because PAHs,phenolic compounds, and metals were detected in sediment andsurface water in Old Tom's Run. The creek flows through aforested area with sparse undergrowth. A path along the westernside of the SMWT site property leads to the creek. Access to thearea is unrestricted, and it is possible that children play inthe stream. Residences are less than half a mile northwest ofthe area. Skin contact with and ingestion of sediments andsurface water by children during play could occur, however, thefrequency and duration of exposure to contaminants is difficultto evaluate. Possible exposure scenarios will be presented inthe Public Health Implications section of this public healthassessment update.
Groundwater contaminated with VOCs, phenolic compounds, PAHs, andmetals is confined at present to the shallow aquifer at the SMWTsite. There are no on-site drinking water wells, and off-siteresidential wells in the vicinity showed no site-relatedcontaminants when analyzed by EPA in 1986. SMCHD has beenmonitoring residential wells quarterly since 1978, and continuesto do so. However, the analytes measured do not include allsite-related contaminants. Although the plume appears to belargely confined to the site, samples from monitoring well 15west of the SMWT boundary (see Figure 7) showed contamination. The shallow groundwater follows local topography, which shouldprevent contaminants from affecting the shallow wells of twohomes less than half a mile west of the boundary. In addition,periodic sampling of EPA's shallow monitoring wells should detectplume movement before local, shallow residential wells areaffected.
In addition to the concern about lateral movement of contaminantsin the shallow aquifer, vertical migration of chemicals to deepergroundwater is possible. An abandoned supply well believed tohave been built in 1965 transects the shallow groundwatercontamination plume. Deterioration of the well casing couldallow contaminants to reach the underlying Aquia and Piney Point-Nanjemoy aquifers. Groundwater flow in those aquifers is largelytoward the southeast. Drinking water for area residents is drawnfrom both aquifers. Until the abandoned supply well on site isproperly sealed, the potential exists for contamination of thedeeper aquifers.
Lead was detected in on- and off-site monitoring wells. However,no residential wells had elevated levels of lead, based onresults reported by SMCHD.
Past, present, and future exposure to contaminants in a varietyof environmental media is possible for on-site remedial workers.It is unlikely that such exposures would be at levels of concernhowever, provided appropriate work practices, as defined by thestate or federal regulatory or permitting authorities, such asthe Occupational Safety and Health Administration (OSHA), arefollowed. Those include worker education, certification,supervision and training, and use of personal protectiveequipment.
ATSDR's previous public health assessment identified thebuildings on the SMWT site as a potential exposure pathway forsite-related contaminants. However, the data show that buildingsamples were actually collected from treated timbers. Thepresence of PAHs and PCPs would be expected because thestructural timbers are treated with those chemicals to preventdecay. Timbers impregnated with wood-treating chemicals arewidely used as construction materials.
It is possible that surface contamination of the cinder blocksand concrete in the process building may have occurred duringplant operations. However, no additional sampling of thebuildings has been done other than that previously described. The buildings, shown in Figure 8, will be demolished ifremediation of contaminated soils takes place at the SMWT site(14). Therefore, based on available information, the on-sitebuildings represent an eliminated pathway rather than a potentialpathway.
Surface soils are also contaminated in a small area off site,immediately northeast of the property boundary (see Figure 8). Residential areas to the north and west are within one half mile, and it is possible that children may occasionally be in thevicinity. Exposure to contaminants is unlikely, however, becausethe contaminated soils are covered by grasses, and the area isovergrown with briars and thorny locust saplings.
Food-chain exposure pathways include the consumption of plants,animals, or other food products raised in contaminated soils, incontaminated water, or irrigated with contaminated groundwater orsurface water. Some chemicals may accumulate in biota. Chemicals such as mercury and other metals can concentrate intissues of certain organisms at levels higher than the chemicals'concentration in water or in vegetation. No biota sampling hasbeen conducted at the SMWT site.
Mercury was found in sediments and surface water of streamdischarges from the SMWT site. The most contaminated surface-water sample was 0.3 ppb, which exceeds the EPA Ambient WaterQuality Criterion of 0.146 ppb. The water criteria are intendedto protect humans who eat fish and other aquatic species. Inorganic mercury can be converted into methylmercury in theenvironment. That organic form of mercury can have higherconcentrations in certain fish than in environmental media. TheEPA water criteria consider the uptake of contaminants and bioconcentration by aquatic species. The criteria use averagefish consumption values of 6.5 grams per day for an individual. The streams and surface-water runoff immediately downstream fromthe SMWT site are too small to support an edible fish population. The first point at which a small edible fish population mightexist is at the Morgan Road Pond, which is approximately 2 milesdownstream from the SMWT site. Sediment samples and surfacewater samples for mercury at that point were below detectionlimits. Fish consumption is an unlikely source of exposure tocontaminants at the SMWT site (19).
Because the site is fenced, it is unlikely that humans wouldcontact plants or aquatic organisms on site. Terrestrial animalssuch as deer, rabbits, and squirrels could be exposed to on-sitecontaminants in the soil, surface water, or food chain. Thereare insufficient data and protocols to determine if gameconsumption would be a point of potential human contact withsite-related contaminants. It is unlikely that game would obtaina most of their food on site because of the limited vegetation,and the fence limits access. Therefore, it is not likely thatgame consumption or other food chain pathways would result inhuman exposures to SMWT contaminants.
Health effects resulting from the exposure of an individual to ahazardous substance in the environment depend on several factors. One is the route of exposure: that is, whether the chemical isbreathed, consumed with food or water, or contacts the skin.Other factors include the dose to which a person is exposed andthe amount actually absorbed. Mechanisms by which chemicals arealtered in the environment, or inside the body once absorbed, arealso important. There is much variation in those mechanismsamong individuals.
ATSDR has prepared toxicological profiles for many substancesfound at hazardous waste sites. Those documents present data andinterpret information on the substances. Preparers of thispublic health assessment update have reviewed the profiles forthe contaminants of concern at SMWT. Background information onincineration has also been reviewed.
There do not currently appear to be completed exposure pathwaysof public health concern. Remedial workers are required tofollow OSHA health and safety guidelines. Brief discussions ofthe toxicity of contaminants at the SMWT site are in Appendix 4. Under present conditions, the contaminants associated with theSMWT site would not be expected to cause illness or diseaseeither in the local population or in properly protected remedialworkers. The health effects described in Appendix 4 would becaused by higher doses than those associated with possibleexposure at the SMWT site. Because access to the site isrestricted, even exposure at those lower doses is unlikely.
Worker exposure to site contaminants during previous facilityoperations is likely. Data are insufficient to determine thedoses to which on-site workers may have been exposed in the past. Workers and nearby residents may have been exposed tocontaminants by way of completed air pathways that existed in thepast. However, because no air monitoring data are available forpast exposures, the amount or duration of exposure is not known,and a dose cannot be reconstructed. Therefore, it is not possibleto determine likely health outcomes from past exposures tocontaminated air.
A number of site-related chemicals were detected in off-sitesediments and surface water in Old Tom's Run. Access to thestream is unrestricted, and the nearest residence is within one-half mile. It is possible that children play in Old Tom's Runand contact contaminated surface water and sediments. The mostlikely route of exposure would be ingestion of or dermal contactwith contaminated media. However, the stream is not used as adrinking-water source, and only inadvertent consumption of smallamounts of water or sediments might occur if children wereplaying in the area. Limited dermal contact with contaminatedmedia may also occur during play. Such exposures would likely beintermittent if they occurred at all.
Although measurable quantities of both carcinogenic andnoncarcinogenic compounds were found in sediments and surfacewater, only pentachlorophenol (PCP) and cPAHs exceeded health-based comparison values. The health guideline used for cPAHs isderived from an interim cancer slope factor. The comparisonvalue used for cPAHs assumes a daily exposure to soils containing100 percent benzo(a)pyrene.
The health guidelines currently available for 2,3,7,8-TCDD toxicequivalents only consider effects other than cancer (see Tables 4and 5 in Appendix 2, and Appendix 3). The PCP health guidelineconsiders the carcinogenic potential of that compound over alifetime of exposure. PCP, carcinogenic PAHs (the grouprepresented by benzo[a]pyrene), and 2,3,7,8-TCDD and similarcompounds are considered probable carcinogens (category B2) byEPA. That is, they are believed to have the potential to causecancer in humans. Animal data strongly support that likelihood,even though human data are not sufficient to conclude that thosecompounds cause cancer in humans (20,21,22,23).
Consequently, the small amount of contaminated surface water orsediments to which children might occasionally be exposed whileplaying in the stream is unlikely to adversely affect theirhealth.
Evaluation of health outcome data can give a general picture ofthe health of a community. Those data can confirm the presenceof excess disease or illness. However, elevated rates of aparticular disease are not always associated with hazardoussubstances in the environment. And, even if elevated rates arenot found, a contaminant may still have caused illness ordisease.
ATSDR must depend on previously gathered data to perform a publichealth assessment. Pre-existing health outcome data are usuallyreported for population units, such as counties. Groups of peoplelikely to have been affected by the contaminants associated witha particular site are usually much smaller. Any evidence ofexcess illness or disease in the smaller group may be hiddenwithin the larger groups' rates. Also, when populations aresmall, the number of people who have a particular illness ordisease is also small. Small changes in the number of affectedpeople from year to year can cause a large change in the rate; that is, the rate is considered "unstable." For those reasons,health outcome data must be evaluated with caution.
ATSDR staff determined that it is unlikely that exposures tocontaminants associated with SMWT could occur at a level whichcould cause illness or disease, including cancer. However,because local residents expressed concern about the rates ofcancer in Maryland and their county, ATSDR staff examined available health outcome data.
ATSDR staff met with the Chief Medical Officer of St. Mary'sCounty and discussed what health outcome data could be obtained. As a result of that discussion, and other discussions with statehealth department officials, ATSDR determined that the availablehealth outcome data were limited. The only registry dataavailable were from the Maryland Cancer Registry. Hospitaldischarge data were not consistently available. Vital statisticswere difficult to obtain and were not available at less than acounty level (15,16).
Because residents expressed concern about the cancer rate, ATSDRstaff evaluated the available cancer data for Maryland and St. Mary's County. The Maryland Cancer Registry was establishedin 1982; however, reporting was incomplete. ATSDR staff examineddata from that registry; those data were provided by theGovernor's Council on Cancer. Cancer registries are designed torecord diagnosed cases. Not everyone who is diagnosed withcancer dies because of its effects. If the number of diagnosedcases is accurately reported, it will exceed the number of cancerdeaths over time. For the period between 1982 and 1990, thetotal number of cancer cases reported to the Maryland CancerRegistry was less than the number of deaths from cancer in justone year of that period. Therefore, reporting could not havebeen complete.
Because reporting diagnosed cases of cancer was not required inMaryland before July, 1991, many cases have not yet been includedin the registry. All clinics, hospitals, and laboratories wererequired July 1, 1991, to begin reporting all cancer diagnoses(24). Currently, registry personnel are searching death andmedical records in an effort to find and verify past cases thatwere not been reported. Updating the registry is a major effortand has only been completed for the area of the state known asthe Eastern Shore. Data for St. Mary's County, part of theWestern Shore area, are expected to be verified and usable withina year (8). ATSDR staff believe the currently available datacannot be used to generate incidence rates (rates of newlydiagnosed cases of cancer per year). However, should morecomplete data become available, ATSDR will evaluate it.
Because the filing of death certificates was required by Marylandlaw, reporting of causes of death tended to be more complete thanreporting of cancer diagnoses. ATSDR staff reviewed the data inthe Maryland Vital Statistics Annual Report: 1986 and theMaryland Vital Statistics Annual Report: 1987. Cancer death(mortality) rates for Maryland and St. Mary's County werecompared with cancer death rates for the United States (U.S.). The U.S. rates were obtained from the Centers for DiseaseControl's Compressed Mortality database on WONDER (WiderangingONline Data for Epidemiologic Research). The rates were based onthe same ICD-9 codes (standard codes for medical diagnoses) andwere all age-adjusted to the 1940 U.S. standard population, whichis used for reporting vital statistics by Maryland's Center forHealth Statistics (6,7,9). Age-adjustment is a process thatcompensates for the effects on rates of death and disease ofdifferences in the age distribution of populations. When onepopulation has many more older people than another, the deathrate for the older group will be higher because more older peopledie. Comparing rates of the two populations is misleading,because the rates of the older population appear higher than therates in the younger population. The older population wouldappear to have an excess of disease; the younger population,however, actually may be more affected. By age-adjusting thedata, it is possible to determine the rate at which deaths wouldoccur if the populations had the same number of people in eachage group. Age-adjustment allows an accurate comparison of therates for one population with the rates for another.
When compared on a county-by-county basis, St. Mary's Countyranked in the lowest third of the counties in Maryland for cancerdeaths during the 1984-1986 period (6). When the average cancerdeath rates for the 1985-1987 period were compared (7), there wasa marked change in the average rate. St. Mary's County rankedseventh out of Maryland's 23 counties. The population of thecounty is relatively small; therefore, its cancer death rates aremore "unstable" than the rates of larger counties. For example,Prince George's County has approximately 10 times the populationof St. Mary's County; therefore, its death rates do not vary asmuch year to year.
Because of the variation in death rates for St. Mary's County,ATSDR staff believed that reviewing data for a longer periodwould provide a more valid comparison. The Compressed Mortalitydata in WONDER for the 10-year period between 1979 and 1988 werereviewed. Table 7 shows the cancer death rates reported for St. Mary's County, Maryland, and the U.S. during that period (9).
If any association existed between the operation of SMWT facilityand the cancer rates in St. Mary's County, that association wouldhave been expected to begin to be apparent during the 10-yearperiod for which data were examined. The facility operated from1965 to 1978. Most cancers develop after a latency period of 15to 30 years. That is, exposure to the carcinogen occurs 15 to 30years before a specific cancer type can be detected. If exposureto site contaminants had caused excess cancer, increased cancerdeath rates would be seen beginning in about 1980, provided allthree of the following conditions were met:
- sufficient number of people in the county were exposed;
- exposure occurred at sufficiently high levels; and
- persons exposed continued to live in the county.
|Year||St. Mary's County||Maryland||United States|
**Rates are age-adjusted to the U.S. 1940 standard, which is used by Maryland for reporting vital statistics.
As can be seen from Table 7, the cancer death rates for St. Mary's County varied more from year to year than rates forMaryland or the U.S. However, some variation was seen for allthe populations. For seven out of 10 years, the cancer deathrate was lower in St. Mary's County than in the state ofMaryland. Three out of 10 years, the cancer death rate for thecounty was lower than the rate for the U.S. It does not appearthat the rate of cancer death in St. Mary's County is eitherincreasing or decreasing in any consistent pattern. However, 10years may be too short a period for detecting a trend in such asmall population (6).
Because of the variation, ATSDR staff believed that looking atcancer death rates on a yearly basis was less reliable thanlooking at the average rates over several years. Because smallnumbers of cancer deaths were recorded for individual bodyorgans, any rates generated for those cancer types would notallow determination of reliable trends over time. Therefore, thecombined death rates for all organs were examined. Over the 10-year period, Maryland had the highest average cancer death ratein the nation: 150.0 deaths per 100,000 person-years. The 10-year average cancer death rate in St. Mary's County was 142.9deaths per 100,000 person-years. St. Mary's County rankedeleventh out of 23 among Maryland's counties for that period,well behind the county with the highest rate. That county,Somerset, had an average 10-year cancer death rate of 172.5deaths per 100,000 person-years (6).
A person's risk of developing cancer has been associated withmany factors, including both gender and ethnic group. Therefore,ATSDR staff compared cancer death rates by gender and ethnicgroup for Maryland, the U.S., and St. Mary's County. Women inSt. Mary's County died from cancer at a rate similar to women inthe U.S. The rate was below the death rate for women inMaryland. On the other hand, men in St. Mary's County had acancer mortality rate close to, but still below, the average forMaryland. The rate was well above the rate for men in the U.S. When the data were examined by ethnic group (white and black(including African-American)), the difference between cancerdeath rates in men and women remained similar. As was true ofboth Maryland and the U.S., blacks (including African-Americans)of both genders had higher cancer death rates than did whites ofboth genders. Table 8 shows cancer death rates by gender andethnic group.
|St. Mary's County||171.1||252.8||107.2||125.4|
**Rates are reported only for whites and blacks (including African-Americans)because the number of deaths in other ethnic groups was too low to generatereliable rates.
***All rates are age-adjusted to the U.S. 1940 standard, which is used by Maryland for reporting vital statistics.
In summary, when cancer death rates were compared for the periodbetween 1979 and 1988, women in St. Mary's County died fromcancer at about the average rate for the U.S. The death ratesfor all women in Maryland were much higher. Men in St. Mary'sCounty died at a rate which was close to, but still below, theaverage cancer death rate for Maryland. Maryland's cancer deathrates for both sexes were higher than the national averages. Rates of incidence (based on newly diagnosed cases of cancer eachyear) could not be determined with available data. Based on theevaluation of contaminants and pathways, and of health outcomedata, ATSDR does not believe that exposure to contaminantsassociated with the SMWT site is likely to result, or haveresulted, in the development of illnesses and diseases, includingcancer.
ATSDR staff would have been able to evaluate the general healthof local residents more thoroughly if more health outcome datahad been available. Specifically, data were not available forgroups of people smaller than the total population of the county(such as census tract). Also, hospital discharge data andregistry data were either not available or not usable. ATSDRstaff evaluate available health outcome data to determine whichdata are of value in describing the health of the localcommunity.
1. The major community health concern expressed to ATSDRinvolves the potential health implications of the incinerationprocess EPA is planning to use to remediate the site. Specifically, how will the stack emissions, such as products ofincomplete combustion (PICs) and metals, affect community health? How will equipment failures, process upsets, and other emergencysituations affect air releases associated with remediation?
Emissions from the incinerator into the air represent a majorexposure pathway to be considered by ATSDR during the secondphase of this Agency's involvement in the SMWT site remediation(see ATSDR's Plan of Action to Evaluate the Proposed RemedialAlternative for the Southern Maryland Wood Treating Site forexpanded discussion of those issues).
Design and operating plans; trial burn plans and test results;air pollution control systems; and other facility provisions helpto define the air emission characteristics of an incinerationfacility. Consequently, those provisions will be carefullyreviewed by ATSDR as the design moves toward implementation. Similarly, ATSDR will review contingency plans, monitoring plans,and system safeguards to determine the potential for excessiveemissions and the effectiveness of corrective measures planned tominimize the impact of such conditions. Through carefulplanning, good design, operator training, and appropriatemonitoring, excessive stack emissions should not be released.
The community expressed concern to ATSDR that all PICs producedby incineration have not been identified. Although that is truefor incineration, it is also true for virtually all forms ofcombustion. It is likely to remain true as long as analyticchemistry methods cannot define all chemicals down to the levelof individual molecules. Regarding identification of PICs causedby various combustion processes, hazardous waste incineration hasbeen reviewed in considerable detail during research and as partof the incineration permitting process. The most prevalent PICs(by mass and occurrence) have been identified and examined forhealth implications using conservative (worst case) exposuresituations. When an incinerator performs as required forefficient hazardous waste destruction and removal, identifiedPICs have been found at very low levels, below health concern.
The community's concern about the environmental fate of metals inthe incineration process is valid and must be examined withrespect to site-specific metals contamination, and systemperformance characteristics, to see if health effects would beapparent or significant. The levels of metals in the soils andsediments to be treated at the SMWT site; the specific airpollution control systems; and actual trial burn results need tobe reviewed. Both metals and PIC data should be modelled usinglocal meteorologic data to identify pollutant air dispersionpatterns and locations of maximum expected ground-levelconcentrations in relation to area populations. That level ofsite-specific understanding is essential in order to evaluatepotential community health implications.
As part of this Agency's commitment to evaluate the public healthimpacts of the SMWT site, ATSDR plans to review all relevantinformation developed for site remediation. Although it is anticipated that most of the appropriate information will bedeveloped as a routine part of EPA's management of this project,ATSDR will identify data gaps, if any, and work with EPA to fillany information needed to determine public health implications.
2. Is the current groundwater monitoring program adequate todetect contamination of drinking water wells in the area? Thepublic is concerned that the production well may causecontamination of the deep aquifers.
Groundwater monitoring at SMWT is being performed both on and offsite and appears adequate to detect the migration of contaminantswithin the shallow water table aquifer. Quarterly monitoring ofnearby residential wells, including two wells drawing from theshallow aquifer, by the St. Mary's County Health Department hasnot detected any contamination. EPA's analysis of 10 privateresidential wells also did not show contamination by sitecontaminants. ATSDR recommends continued monitoring ofdowngradient residential wells with a focus on site contaminants. None of the on-site monitoring wells in the Upper Chesapeakeaquifer have shown any contamination.
The characterization of the Upper Chesapeake Aquifer downgradientof the site does not adequately identify potential migration ofcontaminants. The potential for contaminant migration into thedeep drinking water aquifers, Aquia and Piney Point-Nanjemoy, hasnot been evaluated beneath the site. Most of the domestic wellsused by residents who live near the site draw from those deeperaquifers.
ATSDR concurs with the public concern about potentialcontamination of the deeper drinking-water aquifers beneath thesite from possible vertical movement of contaminants along thecasing of the on-site production well. ATSDR recommended in the1989 public health assessment of this site, and again in thispublic health assessment update, that the on-site production wellbe properly closed.
Previous attempts by EPA to collect a sample from the on-siteproduction well were unsuccessful because the well is obstructed. Prior to proper closure of the well, EPA should attempt to removethe blockage and obtain samples to determine if the Aquia aquifer is contaminated.
3. What is the effect of the site's contaminants on the foodchain? Because some chemicals bioaccumulate in plants andanimals, are the fish, shellfish, game, and crops in the areacurrently affected? Will the incinerator emissions affect thefood chain?
The food chain should not be affected by contaminants at the SMWTsite. Because the site is fenced, human exposure resulting fromconsumption of plants and animals harvested on site is unlikely.
Mercury was found at elevated levels in the sediments and surfacewater of streams that originate on the SMWT site. Because thesurface-water tributaries leaving the site are too small tosupport edible fish populations, that potential food chainpathway is not a likely health concern. Morgan Road Pond is thefirst downstream body of water that could support a small ediblefish population; sediments and surface water in that area did notshow elevated levels of mercury. Therefore, fish consumptiondoes not represent a likely source of exposure. See the PathwaysAnalysis section for a more detailed discussion of food chainexposures.
There is insufficient information to evaluate effects ofincinerator emissions on the food chain. Once trial burn and airdispersion modelling data are available, ATSDR will evaluate thatpathway.
4. Does the county have a high cancer rate?
Data were not available on cancer incidence, that is, the rate ofnewly diagnosed cancer cases per year. ATSDR evaluated availabledata on cancer death rates for all Maryland counties. Whencancer death rates were compared for the period between 1979 and1988, women in St. Mary's County died from cancer at about theaverage rate for the United States. That rate was less than therate for all women in Maryland. Men died at a rate that wasclose to, but still below, the average cancer death rate forMaryland. Maryland's cancer death rates were higher for bothsexes than the national averages. Mortality rates are affectedby many factors, including access to medical care and whetherearly diagnosis and treatment is available. An evaluation ofavailable data, limitations of the data, and an interpretation ofthe data are contained in the Health Outcome Data Evaluation section.
5. How will sensitive populations be affected by this site andby remediation activities?
Currently, there are no known completed pathways of exposure. Therefore, no people are likely to be affected. Once remediationplans are final, ATSDR will review them and assess any potentialhealth effects that the remedial activities at the site mightcause for sensitive populations living nearby.
6. Are the chemicals at SMWT and the potential incinerator-stackemissions cancer promoters or initiators?
For the specific hazardous substances associated with the SMWTsite, it is not known for certain if they are initiators,promoters, or both. ATSDR evaluates all cancer-causingsubstances as "complete" carcinogens. That is, ATSDR assumesthat no interactions with other substances are needed for aparticular substance to be able to cause cancer. Exposure toeither an initiator or promoter might result in development ofcancer. Whether a person will develop cancer depends on severalfactors. Among those factors are the level of exposure andlength of time the person was exposed. Because it is uncertainhow cancers are caused, cancer risk is estimated using levels andtime of exposure, not whether the substance is an initiator orpromoter.
It is best not to think of cancer as a single disease, but as aset of diseases that develop by similar processes. Cancers maybe inherited, or may result from exposures to viruses, chemicals,or physical forces such as radiation. One of several theories ofhow cancers develop following exposure to chemicals is thatsubstances may be either initiators or promoters. Researchsuggests that the first step, initiation, changes the cell into apre-cancerous cell. This is believed to involve a change in thegenes (damaged DNA). The second step, promotion, causes a cellto express the change (25).
Substances which result in changes to a cell's genes areinitiators, and substances which alter the expression of achanged gene are promoters. Promoters are believed to increasethe rate of expression of cancers that occur spontaneously, or asa result of previous initiation (25). In general, initiators arebelieved to be more strongly associated with development ofcancers than are promoters. A substance might be both aninitiator and a promoter (25).
7. Will ATSDR do any follow-up health studies of the communityto see if the remediation activities affected the health ofresidents?
Because there are no currently completed pathways, and pastexposures cannot be defined, no health studies are indicated atthis time. However, once EPA's plans for remediation becomefinal, ATSDR will evaluate them to determine if health studiesare needed in association with those activities. If healthstudies are needed, ATSDR would undertake them in conjunctionwith the community and state and local health authorities.