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

PUBLIC HEALTH ASSESSMENT

MID-ATLANTIC WOOD PRESERVERS
HARMANS, ANNE ARUNDEL COUNTY, MARYLAND


ENVIRONMENTAL CONTAMINATION AND OTHER HAZARDS

The public health assessment will evaluate potential health effects associated with thecontaminants of concern that are identified. The selection of contaminants of concern is basedon a number of factors, including the toxicity of the substances and the potential that peoplecould be exposed to them. Other factors considered in identifying contaminants of concern are:

  1. Concentrations of contaminants on and off the site. Quality of data collected in the fieldand laboratory, and the adequacy of methods employed to sample environmental mediasuch as air, water and soil.
  2. Comparison of site-related concentrations of contaminants with backgroundconcentrations, if available.
  3. Comparison of on-site and off-site concentrations with health assessment comparisonvalues for (a) non-cancer adverse health effects and (b) cancer.
  4. Community health concerns.

In the data tables that follow in the On-Site and Off-Site Contamination subsections, thecontaminants listed do not necessarily cause adverse health effects at the exposure levels foundat or near the site. Instead, the contaminants listed are those which will be evaluated further inthe public health assessment.

The observed levels of contaminants are evaluated further if they exceed comparison values(CVs). Human exposures to concentrations greater than comparison values may, but do notnecessarily, cause adverse health effects. Comparison values for cancer-causing chemicalsinclude Cancer Risk Evaluation Guides (CREGs). A CREG is the contaminant concentrationestimated to produce one excess cancer case in a population of 1,000,000 given lifetimeexposure to the contaminant. CREGs are calculated from USEPA's cancer slope (i.e. potency)factors.

Comparison values for health effects other than cancer include Environmental Media EvaluationGuides (EMEGs), Reference Dose Media Evaluation Guides (RMEGs), Cancer Risk EvaluationGuides (CREGs), Lifetime Drinking Water Health Advisories (LDWHA), MaximumContaminant Levels (MCLs), Maximum Contaminant Level Goals (MCLGs), ProposedMaximum Contaminant Levels (PMCLs), Proposed Maximum Contaminant Level Goals(PMCLGs), and World Health Organization values (WHO). Occupational guidelines are usedwhen other comparison values are not available.

EMEGs are concentrations of contaminants in air, water, and soil below which are not expectedto produce human exposures which may cause adverse, non-cancer health effects, assumingstandard intake rates of contaminated media and body weights. EMEGs are generally based onthe ATSDR chronic exposure Minimum Risk Levels (MRLs) values. RMEGs are similar toEMEGs; however, they are based on USEPA's Reference Doses (RfDs). CREGs are based onUSEPA's Cancer Slope Factors and provide an estimate, based on one in a million chance, of aperson's risk of developing cancer upon long-term exposure to a contaminant. MCLs, MCLGs,PMCLs and PMCLGs were developed by EPA for drinking water. MCLs are enforceablestandards representing contaminant concentrations that EPA deems protective of public health(considering the availability and economics of water treatment technology) assuming a lifetime(70 years) consumption rate of two liters of water per day. MCLGs are non-enforceable goalsmore stringent than the MCL standards. MCLGs are contaminant concentrations which mayresult in life-time exposure levels that EPA believes will result in no known or anticipatedadverse health effect, allowing for an adequate margin of safety. PMCLs and PMCLGs areMCLs and MCLGs, respectively, that are being proposed.

Several conventions were adopted in the on-site/off-site contamination tables presented below. Comparison values for on-site concentrations were computed for adults because only workersare expected to have access to the site. Comparison values used for potential off-site exposureswere calculated for children because children could potentially gain access to areas surroundingthe MAWP and because children are generally considered more sensitive than adults to the toxiceffects of chemicals.

Comparison values for total chromium in soil, air, and water, presented in the following tables,are based on those for the more toxic hexavalent form of chromium. This conservative choice,protective of human health, was made because many samples were analyzed only for totalchromium rather than its component forms (trivalent and hexavalent chromium).

No air sampling of fugitive dust was actually conducted at the MAWP site. Thus, the airconcentrations that appear in tables were estimated from shallow soil sample measurementsassuming that soil produces 1.0 X 10-7 kilogram (kg) of dust per cubic meter (m3) of air (or,equivalently, 0.1 milligram (mg) dust/m3 air).(2) The estimated air concentrations are also basedon the assumption that all of the dust in air comes from the relevant contaminated source (e.g.on-site or off-site soil). The estimated air levels are compared to inhalation CVs in the tablesbelow. Because the site-specific air levels are estimates rather than measured values,conclusions drawn from their use should be interpreted with caution.

A. On-Site Contamination

Soil
On-site soils have been monitored for chromium and arsenic since 1978 when the CCA spilloccurred (see Site Description and History section). Both total chromium and hexavalentchromium levels were measured. Monitoring for copper began in 1980. Sampling for thesethree metals has been done periodically through 1990, although the data for hexavalentchromium and copper is more limited. Because copper was not found to exceed any CV in soilor any other medium, it is not a contaminant of concern. Thus, copper will not be discussedfurther in this public health assessment.

Table 1 presents the range of concentrations found for chromium and arsenic in on-site soil overcalendar time as well as CVs. The data are presented separately for shallow and deep sampledepths. No samples restricted only to surface soil (defined as 0.0 - 3 inches deep) were taken atthis site. The shallow samples in Table 1 range in depth from 0.0 - 0.5 foot to 0.0 - 1.0 foot,depending on the sampling date, and thus include both surface and deeper soil. Therefore, theshallow samples presented may underestimate the true concentrations in surface soil. Deepsamples are defined here to be those samples taken at depths greater than or equal to 2.0 feet.

Table 2 presents the estimated air concentrations and CVs for on-site dust.

In comparing soil samples taken on different dates, one should keep in mind that apparentincreases or decreases in contaminant levels reflect not only changes in time but differences in sampling locations and depths. Nonetheless, some general trends over time are apparent.

Initial measurements taken in 1978 after the spill indicated a maximum total chromium level of 3,110 mg/kg in soil near the overflow pipe, although only 4.0 mg/kg of the more toxichexavalent form were identified (see Table 1). Concentrations generally declined over time butremained within the same order of magnitude until 1990, when the maximum detectedhexavalent chromium level had fallen to 0.70 mg/kg.

The most comprehensive sampling was done as part of the RI in 1989. The RI analyses foundlevels of total chromium in shallow soil samples ranging from 10 mg/kg west of ShipleyAvenue, to 865 mg/kg near the drip pad. No RI samples were analyzed for hexavalentchromium. Because hexavalent chromium is significantly more toxic than the trivalent form,USEPA undertook additional soil sampling in May 1990, to determine the relativeconcentrations of both types of chromium. The six post-RI samples contained 0.70 mg/kg orless of hexavalent chromium. Total chromium results ranged from 23.2 mg/kg in the easternportion of the Storage Yard to 570 mg/kg in the Treatment Yard near the drip pad.

A review of Tables 1 and 2 reveals that chromium is a contaminant of concern. Estimates oftotal chromium in air exceeded the inhalation CV for cancer (CREG) on at least one samplingdate, whereas estimates of airborne hexavalent chromium were slightly less than its CREG (seeTable 2). However, neither total nor hexavalent chromium levels surpassed the soil ingestionCVs -- based on non-cancer adverse effects -- at any time or sampling depth (see Table 1).

Like chromium, arsenic levels have also declined since the original sampling in 1978 when the maximum detected arsenic concentration was 5,660 mg/kg. By 1984, this level was reduced to2,100 mg/kg, and by 1989 the maximum detected concentration was 1,200 mg/kg. The mostrecent sampling, in 1990, found a maximum concentration of 633 mg/kg in the treatment yardnear the drip pad. Despite this decline, arsenic levels in 1990 were still higher than the cancer-based CVs for the soil ingestion and inhalation exposure routes (see Table 1). Thus, arsenic is acontaminant of concern.

A search of the Toxic Chemical Release Inventory (TRI)(3) for the years 1987 through 1989revealed that the MAWP is the only industrial source of chromium and arsenic in the areasurrounding the site. The RI sampling results also clearly implicate the MAWP as the source ofcontamination. The metals were detected above background levels in shallow soil (0.0 - 0.5feet) and deeper soil (3.0 - 3.5 feet and at the groundwater table) in samples collected fromthirteen on-site locations in the Treatment Yard (see Table 1). As illustrated in Figures 2 - 3, thedistribution patterns among these metals are nearly identical, clearly implicating the CCAsolution as the source of contamination. The highest concentrations of contaminants were foundin samples taken near the drip pad.

The analyses of the subsurface samples taken during the RI revealed a sharp reduction inconcentrations of these metals with increasing depth in soil. The depth to which contaminationextended was estimated to average two feet in the Treatment Yard and westernmost third of theStorage Yard. By the time depths of 3.0 to 3.5 feet were reached, the metals' concentrations haddeclined and began to approach background levels. The only exception was the sample takenadjacent to the northern edge of the drip pad, which had concentrations exceeding thebackground levels. All soil samples taken deeper than 3.5 feet contained concentrations ofchromium and arsenic representative of background levels. The total volume of degraded soilscurrently located on-site was estimated to be approximately 5,200 cubic yards in place. Table 1summarizes the concentration ranges for soil contaminants of concern. Table 2 summarizesestimated concentration ranges for air contaminants of concern.

Groundwater
Groundwater sampling of on-site monitoring wells was done initially in 1978 after the CCAspill, and periodically through 1979. A ten year gap followed before additional water sampleswere taken.

The most extensive sampling was done during the RI in 1989. As part of this effort, tengroundwater monitoring wells were installed on site and sampled (Figure 4). Seven of themonitoring wells were screened in the upper Patapsco aquifer (above the discontinuous claylenses) and three wells were screened in the lower Patapsco. The monitoring wells weresampled during February and March 1989 and the analyses were performed on unfiltered watersamples. Four of the monitoring wells (Nos. 2, 3, 4 and 8) were analyzed for the TargetCompound List (TCL) and the Target Analyte List (TAL).(4) The other six monitoring wellswere analyzed for arsenic, copper, and chromium only.

None of the TCL/TAL substances were identified above background concentrations (norselected comparison values) in any of the RI samples. Thus, these substances will not beconsidered further here. Table 3 presents the results of the groundwater sampling of on-sitemonitoring wells for chromium and arsenic. Drinking water CVs are also shown.

Hexavalent and total chromium were found at levels exceeding the non-cancer CVs for samplestaken through 1979 (see Table 3). By 1989, levels of total chromium had declined below theCV. While hexavalent chromium was not sampled in 1989, it probably was also below its CVsince hexavalent chromium is a subcomponent of total chromium. Thus, chromium, alreadyidentified as a contaminant of concern, was found to exceed its CV for on-site groundwater inthe past, but not in the present.

Arsenic levels in the early years of sampling were higher than the cancer-based CV (CREG). By1989, arsenic was not detected in samples from on-site groundwater monitoring wells. However, the detection limit of the lab method used to analyze the samples for arsenic was 500times higher than the CREG (0.010 mg/Liter (L) vs. 2.0 X 10-5 mg/L, respectively). Thus, it isnot possible to say whether or not 1989 levels were below the CV. However, based on the levelsfound in samples taken on the earlier dates, arsenic is considered a contaminant of concern.(5)

B. Off-Site Contamination

Soils near Storm Sewer Outfall
Surface water runoff from the Treatment Yard flows to a storm drain that runs northwardbeneath Shipley Avenue. The storm water is released from an outfall into a flood plainapproximately 400 feet east of Stony Run Creek. Only a few off-site soil samples have everbeen analyzed for contamination. The results and CVs are shown in Tables 4 and 5.

The most comprehensive sampling was done as part of a post-Remedial Investigation (Damesand Moore, 1990a). Four soil samples were collected from two locations near the storm seweroutfall on March 12, 1990. At each location, a sample was collected at shallow (0.0 - 0.5 feet)and at deeper (3.0 - 3.5 feet) levels. The sampling locations were at the mouth of the outfall andat a spot approximately 33 feet downgradient of the outfall in the flood plain. The samples wereanalyzed for arsenic, total chromium and copper.

Comparing these results to the local mean background concentrations, chromium appeared to bepresent at background concentrations. The local background concentration for chromium isabout 63 mg/kg. The 1990 observed shallow soil concentrations for chromium ranged from 40.0- 59.7 mg/kg. Even though the observed concentrations were within the range of expectedbackground levels, they still exceeded the comparison values for both ingestion and inhalation.

Arsenic concentrations in 3 of the 4 samples collected appeared elevated over the expected localbackground concentration of 6.1 mg/kg. The observed levels -- with a peak of 37.1 mg/kg --alsoexceeded both the inhalation and ingestion CREG values for arsenic. Thus, arsenic is acontaminant of concern. Tables 4 and 5 summarize soil contaminant and estimated air contaminant data.

Surface Water and Sediment
Surface waters from the Stony Run Creek were sampled periodically, beginning in 1978 after theCCA spill. Samples downstream from the site taken prior to the RI, had either very low or nodetectable amounts of total chromium, hexavalent chromium, and arsenic. Three surface watersamples and five sediment samples were collected during December 1988 as part of the RI. Allof the collected samples were analyzed for chromium and arsenic. These metals were notdetected in any surface water samples.

Stony Run Creek sediments were sampled twice in 1983 and again in 1988 during the RI. Chromium was detected at concentrations within the normal range expected for this area. Arsenic levels were below the detection limit of 0.01 mg/L in the RI. Thus, it does not appearthat Stony Run Creek is currently being adversely affected by the MAWP site.

Groundwater
Off-site groundwater monitoring wells were tested for contaminants several times in the yearfollowing the spill (see Table 6). Total chromium levels ranged from not detectable to amountswhich exceeded the drinking water non-cancer CV for children. The same pattern was observedfor hexavalent chromium which was found at only slightly lower values than those observed fortotal chromium. This indicates that most of the chromium found in these wells in the year afterthe spill was in the hexavalent form. Arsenic was detected at levels above the cancer-based CVfor drinking water in 1978 and 1979.

Off-site private wells were tested more frequently than monitoring wells (see Table 7). A wellsurvey was conducted as part of the RI/FS to identify the wells that may have beencontaminated. Total chromium in private wells downgradient from the site peaked at 19.5 mg/Lshortly after the 1978 CCA spill, and declined with time until 1984, when none was detected.

Testing for hexavalent chromium in private wells was conducted only in the year after the spill. From Table 7 one can see that most of the chromium in private wells, like the monitoring wells,was in the hexavalent form at levels that exceeded the non-cancer CV. Thus, once again,chromium meets the criteria for being a contaminant of concern.

Few samples from private wells were monitored for arsenic. The arsenic concentrationsmeasured were generally low. By 1989, no arsenic was observed above the detection limit forthe lab method used to analyze the samples (0.01 mg/L). However, the detection limit exceededthe CV based on cancer risk. Thus, it is not possible to say whether the actual concentrations in water were higher or lower than the CV.


TABLE 1.

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN ON-SITE SOILa
CONTAMINANTSSAMPLE DEPTHbRANGE OF CONCENTRATIONS (MG/KG)
DATES SAMPLED
11/1/786/1/792/19/802/17/8410/11/8410/25/84
Total
Chromium
Shallow525.034.2113.1 -1908.36.9 - 585.918 - 1882,700
Deep4.5 -3,11018.4 - 130.0NANA
Hexavalent
Chromium
Shallow<.50NTNTNTNTNT
Deep<.80 - 4.0NTNANA
rsenic Shallow400.0106.92.1 - 11.30.7 - 531.517 - 120NT
Deep2.4 - 5,66069.0 - 804.0NANA
NUMBER OF
LOCATIONS
SAMPLED
Shallow
1
1
2
16
6
1
Deep
3
1
NA
NA
NUMBER OF
SAMPLES
TAKEN
Shallow
1
1
3
16
6
1
Deep
4
8
NA
NA

aSources: Table 1-2, pp. 1-11 and 1-12, and Appendix J, Dames and Moore, 1990b, for 11/1/78 - 10/25/84sampling dates. For 1/4 - 2/10/89, Table 4-1, p. 4-2, Dames and Moore, 1990b. Table 1-4, p.1-40, Dames and Moore, 1991a, for 5/24/90 sampling date.

bShallow samples were taken at depths ranging from 0.0 - 0.5 to 0.0 - 1.0 foot except for samples taken on 2/19/80 whichwere taken at a depth of 1.5 feet. Deep samples taken at depths 2 feet. Not enough information was provided todetermine sampling depths for samples taken on 2/17/84 and 10/25/84.


TABLE 1 (Con't.)

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN ON-SITE SOIL
CONTAMINANTSSAMPLE DEPTHRANGE OF CONCENTRATIONS (MG/KG)
DATES SAMPLEDSOIL INGESTIONc
10/25/84 1/4 - 2/10/895/24/90Soil
Concentration
(mg/kg)
Source
Total
Chromium
Shallow58 - 1,640 10 - 865 23.2 -570.0
3,500
RMEG
Deep BDL - 82 NA
Hexavalent
Chromium
Shallow<0.10 NT 0.40 -0.70
3,500
RMEG
Deep NT NA
ArsenicShallow41 - 2,100 BDL -1,200 10.8 -633.3
0.40
CREG
Deep BDL - 133 NA
NUMBER OF
LOCATIONS
SAMPLED
Shallow
4
13
6

Deep
13
NA
NUMBER OF
SAMPLES
TAKEN
Shallow
4
13
6
Deep
18
NA

cSoil ingestion non-cancer comparison values and CREGs were computed for adults using the assumptions that adults incidentally ingest 100 mg soil/day, and weigh 70 kg. The total chromium and hexavalent chromium non-cancer comparison values for soil ingestion were calculated using an oral RfD of 0.005 mg/kg/day. The arsenic CREG for soil ingestion was calculated using an oral cancer potency factor of 1.75 risk/(1 mg arsenic/kg bodyweight/day).

Abbreviations: NT = Sample Not Tested for that Compound; BDL = Below Detection Limit; NA = Not Applicable (no samples taken at that depth); RMEG = Reference Dose Media Evaluation Guide; CREG = Cancer Risk Evaluation Guide; MG = milligram; KG = kilogram.


TABLE 2.

RANGE OF CONCENTRATIONS OF CONTAMINANTS ESTIMATED IN ON-SITE AIRa
CONTAMINANTSSAMPLE DEPTHbRANGE OF CONCENTRATIONS (MG/M3)
DATES SAMPLED
11/1/786/1/792/19/802/17/8410/11/8410/25/84
Total
Chromium
Shallow5.25 X 10-53.42 X 10-61.13 X 10-5 -

1.91 X 10-4

6.9 X 10-7 -

5.86 X 10-5

1.8 X 10-6 -

1.88 X 10-5

2.70 X 10-4
Hexavalent
Chromium
Shallow< 5.0 X 10-8NTNTNTNTNT

Arsenic
Shallow4.00 X 10-51.07 X 10-52.1 X 10-7 -

1.13 X 10-6

7 X 10-8 -

5.31 X 10-5

1.7 X 10-6 -

1.20 X 10-5

NT
NUMBER OF
LOCATIONS
SAMPLED
Shallow
1
1
2
16
6
1
NUMBER OF
SAMPLES
TAKEN
Shallow
1
1
3
16
6
1

aSources: Table 1-2, pp. 1-11 and 1-12, and Appendix J, Dames and Moore, 1990b, for 11/1/78 - 10/25/84sampling dates. For 1/4 - 2/10/89, Table 4-1, p. 4-2, Dames and Moore, 1990b. Table 1-4, p.1-40, Dames and Moore, 1991a, for 5/24/90 sampling date.

bShallow samples were taken at depths ranging from 0.0 - 0.5 to 0.0 to 1.0 foot except for samples taken on 2/19/80 whichwere taken at a depth of 1.5 feet. Not enough information was provided to determine sampling depths for samples takenon 2/17/84 and 10/25/84.


TABLE 2 (Con't.)

RANGE OF CONCENTRATIONS OF CONTAMINANTS ESTIMATED IN ON-SITE AIR
CONTAMINANTS SAMPLE DEPTH RANGE OF CONCENTRATIONS (MG/M3)COMPARISON VALUES
DATES SAMPLEDDUST INHALATIONc
10/25/841/4 - 2/10/895/24/90Air
Concentration
(mg/m3)
Source
Total
Chromium
Shallow5.8 X 10-6 -

1.64 X 10-4

1.0 X 10-6 -

8.65 X 10-5

2.32 X 10-6 -
5.70 X 10-5

8.3 X 10-8

CREG
Hexavalent
Chromium
Shallow< 1.0 X10-8NT4.0 X 10-8 -

7.0 X 10-8


8.3 X 10-8

CREG

Arsenic
Shallow4.1 X 10-6 -

2.1 X 10-4

BDL -

1.2 X 10-4

1.08 X 10-6 -
6.33 X 10-5

2.3 X 10-7

CREG
NUMBER OF
LOCATIONS
SAMPLED
Shallow
4
13
6

NUMBER OF
SAMPLES
TAKEN
Shallow
4
13
6

cAir concentrations for dust were calculated using the assumption that soil produces 1.0 X 10-7 kg of dust per cubic meter of air (or, equivalently, 0.1 mg dust/m3 air). The total chromium and hexavalent chromium CREGs for dust were calculated using an inhalation cancer potency factor of 4.1 X 101 risk/(1 mg hexavalent chromium/kg body weight/day). The arsenic CREG for dust was calculated using an inhalation cancer potency factor of 15 risk/(1 mg arsenic/kg body weight/day).

Abbreviations:NT = Sample Not Tested for that Compound; BDL = Below Detection Limit; NA = Not Applicable (no samples taken at that depth); CREG = Cancer Risk Evaluation Guide; MG = milligram; M3 = cubic meters.



Table 3.

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN ON-SITE GROUNDWATER MONITORING WELLSa,b

CONTAMINANTS
SAMPLE
LOCATION
RANGE OF CONCENTRATIONS (MG/L)
COMPARISON VALUESc
DATES SAMPLED
11/1/781/8/796/797/798/792/27 -
5/4/89
Concentra-
tion
(MG/L)
Source
Total
Chromium
Upgradient
NA
NA
NA
NA
NA
BDL - 0.088
0.175
RMEG
Downgradient
2.25
0.27
0.10 - 33.0
0.15 - 10.6
0.10 - 3.0
BDL - 0.151
Hexavalent
Chromium
Upgradient
NA
NA
NA
NA
NA
NT

0.175

RMEG
Downgradient
ND
ND
ND - 24.0
ND - 8.10
ND - 3.0
NT

Arsenic
Upgradient
NA
NA
NA
NA
NA
BDL

2.0 X 10-5

CREG
Downgradient
.0158
0.155
0.019 - 20.4
0.031 - 6.7
0.027 - 3.4
BDL
NUMBER OF
LOCATIONS
SAMPLED
Upgradient
NA
NA
NA
NA
NA
3

Downgradient
1
1
2
2
2
9
NUMBER OF
SAMPLES
TAKEN
Upgradient
NA
NA
NA
NA
NA
2
Downgradient
1
1
2
2
2
8

aSources: Table 1-2, p. 1-16, Dames and Moore, 1990b for 11/1/78 - 8/79 sampling dates. Table 4-1, p. 4-3, Dames and Moore, 1990b, for 2/27 - 5/4/89 sampling dates.

bAll samples were unfiltered water for analysis.

cComparison values computed for adults assuming that adults on average consume 2 L water/day and weigh 70 kg. The totalchromium and hexavalent chromium non-cancer comparison values for ingestion were calculated using an oral RfD of 0.005mg/kg/day. The arsenic CREG for ingestion was calculated using an oral cancer potency factor of 1.75 risk/(1 mgarsenic/kg body weight/day).

Abbreviations:NT = Sample Not Tested for That Compound; BDL = Below Detection Limit; NA = Not Applicable (Nosamples taken); ND = Not Detected ; RMEG = Reference Dose Media Evaluation Guide; CREG = CancerRisk Evaluation Guide; MG = milligram; L= liter; KG = kilogram.



TABLE 4

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN OFF-SITE SOILa
CONTAMINANTS SAMPLE DEPTHb RANGE OF CONCENTRATIONS (MG/KG)
DATES SAMPLEDSOIL INGESTIONc
1/5/8310/25/8410/25/841 -2/89
3/12/9
0
Soil Concentration(mg/kg)Source
Total
Chromium
Shallow
4.70
18.0
9.0
9.9
40.0- 59.7
10
RMEG
Deep
30 - 48
15.2- 54.2
Hexavalent
Chromium
Shallow
NT
ND
NT
NT
NT
10
RMEG
Deep
NT
NT
ArsenicShallow
BDL
5.0
NT
BDL
15.6 - 27.5
0.001
CREG
Deep
3.6 - 9.6
BDL - 37.1
NUMBER OF
LOCATIONS
SAMPLED
Shallow
1
1
1
1
2

Deep
2
2
NUMBER OF
SAMPLES
TAKEN
Shallow
1
1
1
1
2
Deep
2
2


aSources:Table 1-2, pp. 1-11 and 1-12, and Appendix J, Dames and Moore, 1990b, for 1/5/83 -10/15/84 sampling dates. For 1 - 2/89, Table 4-1, p. 4-2, Dames and Moore, 1990b. Table 1-4, p. 1-40, Dames and Moore, 1991a, for 3/12/90 sampling date.

bShallow samples were taken at depths ranging from 0.0 - 0.5 to 0.0 - 1.0 foot. Deep samples taken at depths > 2 feet. Not enough information was provided to determine sampling depths for samples taken on 1/5/83 and 10/25/84.

cIngestion non-cancer comparison values were computed for pica children using the assumptions that pica children incidentally ingest 5,000 mg soil/day, and weigh 10 kg. The chromium and hexavalent chromium non-cancer comparison values for soil ingestion were calculated using an oral RfD of 0.005 mg/kg/day. The arsenic CREG for soil ingestion was calculated using an oral cancer potency factor of 1.75 risk/(1 mg arsenic/kg body weight/day).

Abbreviations:NT = Sample Not Tested for that Compound; BDL = Below Detection Limit; NA = Not Applicable (no samples taken at that depth); RMEG = Reference Dose Media Evaluation Guide; CREG = Cancer Risk Evaluation Guide; MG = milligram; KG = kilogram.




TABLE 5.

RANGE OF CONCENTRATIONS OF CONTAMINANTS ESTIMATED IN OFF-SITE AIRa
CONTAMINANTS SAMPLE DEPTHb RANGE OF CONCENTRATIONS (MG/M3) COMPARISON VALUES
DATES SAMPLEDDUST INHALATIONc
1/5/8310/25/8410/25/841 - 2/893/12/90Soil Con-
centration
(mg/m3)
Source
Total
Chromium
Shallow 4.70 X 10-7 1.8 X 10-69.0 X 10-79.9 X 10-74.0 X 10-6 -

5.97 X 10-6


8.3 X 10-8

CREG
Hexavalent
Chromium
Shallow
NT
ND
NT
NT
NT

8.3 X 10-8

CREG
ArsenicShallow
BDL
5.0 X 10-7
NT
BDL
1.56 X 10-6 -

2.75 X 10-6


2.3 X 10-7

CREG
NUMBER OF
LOCATIONS
SAMPLED
Shallow
1
1
1
1
2

NUMBER OF
SAMPLES
TAKEN
Shallow
1
1
1
1
2


aSources:Table 1-2, pp. 1-11 and 1-12, and Appendix J, Dames and Moore, 1990b, for 1/5/83 -10/15/84sampling dates. For 1 - 2/89, Table 4-1, p. 4-2, Dames and Moore, 1990b. Table 1-4, p. 1-40, Dames and Moore, 1991a, for 3/12/90 sampling date.

bShallow samples were taken at depths ranging from 0.0 - 0.5 to 0.0 - 1.0 foot.

cDust concentrations were calculated using the assumption that soil produces 1.0 X 10-7 kg of dust per cubic meter of air(or, equivalently, 0.1 mg dust/m3 air). The total chromium and hexavalent chromium CREGs for dust were calculated using an inhalation cancer potency factor of 4.1 X 101 risk/(1 mg hexavalent chromium/kg body weight/day). The arsenic CREG for dust was calculated using an inhalation cancer potency factor of 15 risk/(1 mg arsenic/kg body weight/day).

Abbreviations:NT = Sample Not Tested for that Compound; BDL = Below Detection Limit; NA = Not Applicable (no samples taken at that depth); MG= milligram; M3 = cubic meters.



Table 6.

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN OFF-SITE GROUNDWATER MONITORING WELLSa,b,c
CONTAMINANTS RANGE OF CONCENTRATIONS (MG/L) COMPARISON VALUESd
DATES SAMPLED Concentration
(mg/L)
Source
11/1/78 1/8/79 6//79 7/79 8/79
Total
Chromium
ND - 0.05
0.05 - 0.15
ND - 10.75
ND - 11.70
ND - 10.0
0.05
RMEG
Hexavalent
Chromium
ND
ND
ND - 9.00
ND - 9.10
ND - 8.3
0.05
RMEG
Arsenic
0.013
0.016 - 0.048
ND - 0.019
0.011 - 0.066
ND - 0.027
5.7 X 10-6
CREG
NUMBER OF
LOCATIONS
SAMPLED
2
3
6
6
6

NUMBER OF
SAMPLES
TAKEN
2
3
6
6
6

aSource: Table 1-2, pp. 1-15 and 1-16, Dames and Moore, 1990b.

bAll samples were unfiltered water for analysis.

cAll sampling sites were downgradient from the MAWP site.

dComparison values computed for children assuming that children on average consume 1 L water/day and weigh 10 kg. Thetotal chromium and hexavalent chromium non-cancer comparison values for ingestion were calculated using an oral MRL of 0.005mg/kg/day.

The arsenic CREG for ingestion was calculated using an oral cancer potency factor of 1.75 risk/(1 mg arsenic/kg bodyweight/day).

Abbreviations:NT = Sample Not Tested for That Compound; ND = Not Detected ; RMEG = Reference Dose Media Evaluation Guide; CREG = Cancer Risk Evaluation Guide; MG = milligram; L= liter; KG = kilogram.



Table 7.

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN OFF-SITE PRIVATE GROUNDWATER WELLSa,b
CONTAMINANTS SAMPLE
LOCATION
RANGE OF CONCENTRATIONS (MG/L)
DATES SAMPLED
8/15/789/25/7810/2/7810/30/781/8/796/797/79
Total
Chromium
Upgradient
NA
NA
NA
NA
ND
NA
NA
Downgradient
ND
19.5
ND
5.20
ND - 8.40
ND - 8.50
ND - 8.70
Hexavalent
Chromium
Upgradient
NA
NA
NA
NA
ND
NA
NA
Downgradient
7.7
ND
ND
0.01
ND - 5.00
ND - 7.80
ND - 7.0
ArsenicUpgradient
NA
NA
NA
NA
0.014
NA
NA
Downgradient
ND
ND
ND - 0.002
ND
0.006
ND
ND
NUMBER OF
LOCATIONS
SAMPLED
Upgradient
NA
NA
NA
NA
1
NA
NA
Downgradient
1
1
2
1
2
2
2
NUMBER OF
SAMPLES
TAKEN
Upgradient
NA
NA
NA
NA
1
NA
NA
Downgradient
1
1
2
1
2
2
2

aSources: Table 1-2, pp. 1-15 and 1-16, Dames and Moore, 1990b for sampling dates 8/15/78 - 8/30/84. Table 4-1, p. 4-3, Dames and Moore, 1991b, for sampling date 3/1/89.

bAll samples were of unfiltered water.

Abbreviations:NT = Sample Not Tested for That Compound; NA = Not Applicable (no sample taken); ND = NotDetected; BDL = Below Detection Limit; RMEG = Reference Dose Media Evaluation Guide; CREG =Cancer Risk Evaluation Guide; MG = milligram; L= liter; KG = kilogram; MRL = Minimum RiskLevel.



TABLE 7 (Con't.).

RANGE OF CONCENTRATIONS OF CONTAMINANTS DETECTED IN OFF-SITE PRIVATE GROUNDWATER WELLS
CONTAMINANTS SAMPLE
LOCATIONS
RANGE OF CONCENTRATIONS (MG/L) COMPARISON VALUESc
DATES SAMPLEDConcentrations (mg/L)Source
8/791/5/832/9/838/30/843/1/89
Total
Chromium
Upgradient
NA
ND
0.01
NA
BDL
0.05
RMEG
Downgradient
ND - 11.0
NA
NA
ND
NA
Hexavalent
Chromium
Upgradient
NA
NT
NT
NA
NT
0.05
RMEG
Downgradient
ND - 11.0
NA
NA
NT
NA
ArsenicUpgradient
NA
ND
0.01 - 0.03
NA
BDL
5.7 X 10-6
CREG
Downgradient
ND
NA
NA
ND
NA
NUMBER OF
LOCATIONS
SAMPLED
Upgradient
NA
1
2
NA
2

Downgradient
2
NA
NA
2
NA
NUMBER OF
SAMPLES
TAKEN
Upgradient
NA
1
1
NA
2
Downgradient
2
NA
NA
2
NA

cComparison values computed for children assuming that children on average consume 1 L water/day and weigh 10 kg. Thetotal chromium and hexavalent chromium non-cancer comparison value for ingestion were calculated using an oral RfD of0.005 mg/kg/day. The arsenic CREG for ingestion was calculated using an oral cancer potency factor of 1.75 risk/(1 mgarsenic/kg body weight/day).

C. Quality Assurance And Quality Control

Much of sampling done prior to the RI was apparently done in absence of regulatory oversight. For most dates, the total number of samples for any medium was generally very small. Whilethe quality of these samples appeared sufficient to signal that contamination had occurred, they were generally inadequate to characterize the full extent of contamination at the site.

In contrast, the data collected during and after the RI underwent strict scrutiny for qualityassurance/quality control (QA/QC) requirements. Among the parameters verified were: holdingtimes, matrix spike recoveries, surrogate compound recoveries, instrument calibration andduplicate sample recoveries. All RI samples were analyzed according to USEPA's ContractLaboratory Program (CLP) specifications, and were judged by MDE to meet QA/QC standards. The post-RI soil analyses were conducted using standard laboratory methods. However, CLPmethods were not used for total chromium and arsenic analyses. Nonetheless, the post-RIsampling appeared to meet QA/QC standards.

Two monitoring wells (Nos. 1 and 8) that were sampled during the RI did have elevatedchromium levels of 62 and 151 µg/L, respectively. However, the source of the contaminationshown in these latter results is questionable because a high concentration of aluminum was alsodetected in well No. 8. The high concentration of aluminum indicates that the well may havebeen improperly constructed prior to sample collection. Aluminum is relatively insoluble inwater and is frequently used as an indicator of unusually high levels of particulates in unfilteredsamples. The presence of particulates in a water sample extracted from a newly installed wellcauses the laboratory test results for metals to appear to be higher than the true values are. Thewater sample extracted from monitoring well No. 1 was not analyzed for aluminum. Thus, it isit is not possible to assess whether the apparently high levels of chromium detected in well No. 1 were accurate or not.

D. Physical And Other Hazards

Physical and other hazards associated with the MAWP site were primarily associated withoccupational activities at the site. While it was a working industrial facility, certain hazardsexisted that were to be expected with any environment that contains heavy machinery and rawmaterials. Such machinery and materials at the site included forklifts, trucks, and large stacks ofpressure-treated wood. Because the chain-linked fence prevents access to the site and the facilityis no longer in operation, physical hazards that may still exist on site should not pose a threat toneighboring residents.

PATHWAYS ANALYSIS

Exposure Pathways evaluated in this public health assessment are listed in Table 8. Todetermine whether nearby residents are exposed to contaminants migrating from the site, theenvironmental and human components that lead to human exposure must be evaluated. Thispathways analysis consists of five elements:

  1. a source of contamination,
  2. transport through an environmental media,
  3. a point of exposure,
  4. a route of human exposure, and
  5. an exposed population.

Exposure pathways are classified as completed, potential, or eliminated. Completed pathwaysexist when the five elements are present and indicate that exposure to a contaminant hasoccurred in the past and/or is occurring now. Potential pathways are those that may haveoccurred in the past or present, or could occur in the future. In eliminated pathways, at least oneof the five elements is and was missing, and will never be present. Completed and potentialpathways, however, may be eliminated when they are unlikely to be significant. Table 8 and thediscussion that follows identify the completed, potential, and eliminated pathways at this site.

A. Completed Pathways

Surface Soil Pathways

    Inhalation of dusts

Much of the outdoor surface of the MAWP facility is unpaved dirt, gravel, and asphalt. Trucksfrequently travel across the unpaved storage yard and the area surrounding the drip pad andstorage shed. Dust may be emitted from vehicle travel over unpaved surfaces or by wind erosionover these surfaces. As described in the Remedial Investigation, some surface soil samples arecontaminated with chromium and arsenic. This contaminated soil may become airborne and beinhaled by persons in the immediate vicinity of the dust cloud.

Contaminated dust may also be transported off site by wind. The prevailing wind direction forthe entire year, measured at the National Weather Bureau Station at Baltimore WashingtonInternational Airport, is west-northwest. During the summer months, when dry conditions maylead to the generation of wind blown particulates, the prevailing wind direction is more westerly. The Feasibility Study (1991) called for interim measures to reduce the risk from inhalation of fugitive dust, including placing varying amountsof gravel over contaminated areas to limit exposure by this pathway.

The persons who may be exposed to airborne dust originating from the MAWP will be limited tothose in the immediate vicinity of the site. These people include on site and neighboring siteworkers and neighboring residents. As distances from the site increase, dispersion and falloutwill cause a decrease in the concentrations of the contaminants of concern in the dust. Also, thefrequency of exposure will depend upon the direction of the population from the site and thefrequency that the wind blows in that particular direction.

Surface soil at the sewer outfall (i.e., off-site soil) was also analyzed and found to contain lowlevels of arsenic. Routine exposure of people to these soils is unlikely due to the proximity ofthe outfall to the Gunther Transport Trucking Company and the difficult access route to the area. However, some children may occasionally play around the outfall. Because the outfall area iscovered with vegetation, it is unlikely that fugitive dust is produced. Thus, children playing inthe area are probably not at risk of inhaling contaminated dust from this area.

    Incidental Ingestion and Dermal Contact

Ingestion of, and dermal contact with soil is often considered an exposure route that applies onlyto young children. However, adults are known to inadvertently ingest small amounts of soilthrough direct contact with soil and subsequent inadvertent ingestion through associated hand-to-mouth contact. Opportunities for inadvertent ingestion are somewhat reduced by the interimmeasure of placing gravel over portions of the site. As in the case of dust inhalation describedabove, these exposure pathways are of greater concern for on-site workers than neighboringresidents. Older children may be capable of approaching and gaining access to the site, althoughsite access is greatly restricted by a chain link fence that is locked when the facility is closed.

B. Potential And Eliminated Pathways

Surface Soil Pathways

    Incidental Ingestion and Dermal Contact

Children might incidentally ingest surface soil containing arsenic when playing near the seweroutfall. As noted above, the likelihood of such exposures is low, though not impossible, andtherefore, off-site soil ingestion is considered a potential exposure pathway. Although thispathway was evaluated considering the unlikely, worst-case scenario involving pica children(age 2-5) ingesting 5,000 mg of soil from this area per day, it is more likely that only olderchildren, age 6 to 12 years, would frequently access this area. Incidental soil intake for this group is estimated at 200 mg per day.

TABLE 8.

EXPOSURE PATHWAYS
PATHWAYTIMEEXPOSURE
POTENTIAL
SOURCETRANSPORT MEDIAEXPOSURE POINTEXPOSURE PATHEXPOSED POPULATION
On-site
Surface Soil
Past
-------
Present
Completed
---------
Completed
Spills &
Drips at
MAWP
Surface SoilOn-site

Nearby Residences

Inhalation

Ingestion

Workers

Neighboring
Residents*

Surface Soil atSewer OutfallPast
-------
Present
Potential
----------
Potential
Spills &
Drips at
MAWP
Surface Soil
&
Surface WaterRunoff
Sewer OutfallIngestionChildren
GroundwaterPast
-------
Present
Potential
----------
Eliminated
Spill at
MAWP
Infiltration through SoilHome DrinkingWater WellsIngestionUsers of
Affected Wells

*Neighboring residents would be exposed to on-site soil through the inhalation route only.

Groundwater Pathways

Exposure to the contaminants of concern in drinking water is presently an eliminated pathwaybecause there are currently no known exposed populations. While water samples from on-sitemonitoring wells were highly contaminated in the past, these wells were not used for drinkingwater. Thus, the on-site wells are part of an eliminated pathway.

Of the 17 domestic wells and 11 public water supply wells identified off-site and within a 3-mileradius of the MAWP, only 6 were listed as downgradient of the site. While the nearest is in thepathway of expected groundwater flow from the site, it is not used for drinking water. This wellwas tested by the State of Maryland and found to have concentrations of chromium, cobalt, andarsenic below USEPA Maximum Contaminant Levels (MCL).

At the time contamination was discovered (1978), a brief exposure to chromium (the highestlevel found was 19.5 mg/L) may have occurred at a neighboring residence. Because residenceswere quickly connected to the public water supply system, this pathway no longer exists. Theabsolute maximum period of potential community exposure to hexavalent chromium ingroundwater was limited to 4 years, from time the MAWP began operating (1974) until residentswere connected to a municipal water supply (1978); however, because no samples were takenprior to 1978, it is not known whether or not exposure to hexavalent chromium before 1978actually occurred.

The potential for future exposure to contaminants in groundwater depends on the likelihood thatexisting or future contaminants migrate from the site into groundwater and reach susceptiblepopulations. For groundwater, the potential routes of exposure would include dermal contactand ingestion. Inhalation while showering is not a pertinent pathway because contaminants ofconcern are not volatile (i.e. do not evaporate). They tend not to be soluble and are not likelyavailable to people as aerosols while showering.

The contaminants of concern are unlikely to contaminate groundwater in significantconcentrations in the future for several reasons. First, it is unlikely that significant newcontamination will occur because of the remediation effort underway. Thus, any futuregroundwater contamination is likely to be the result of existing environmental levels, which have already significantly declined.

Second, these metals have a limited ability to leach through soil primarily due to their physicaland chemical properties. Chromium is adsorbed by (i.e., becomes attached to) all soil materials,while arsenic becomes adsorbed preferentially by clay and humic acids (i.e. acids found in theorganic matter of soil). Although hexavalent chromium is much more mobile in soil thantrivalent chromium, the former is converted to the latter by natural processes. Although thesampling data are limited, it appears that the concentrations of hexavalent chromium in soil havedeclined from the maximum detected after the 1978 spill to the low levels detected in the 1990post-RI sampling. This decline of hexavalent chromium in soil is likely due to havingpreviously leached into groundwater shortly after the 1978 spill and/or conversion into trivalentchromium. Vertical migration of the residual trivalent chromium and arsenic has been impededby their adsorption to soils underlying the site. These adsorptive properties limit the potentialfor additional leaching in the future.

Third, the concentrations of these metals once they reach groundwater are likely to quicklydecrease. A computerized model that predicts the movement of chemicals in groundwater wasgenerated in an effort to predict the ultimate concentrations of total chromium in thegroundwater that would result from the soil contamination levels and hydrogeological conditionsobserved at the site (Dames and Moore, 1990a). Results from this model suggest that levelsshould recede to below the current federal or state drinking water standard of 50 micrograms(µg)/L (or equivalently, 0.050 mg/L) within three months, assuming that no additional chromiummigrated into the aquifer.

The model results are supported by actual sampling data. Analysis of the current metalsconcentrations compared to past results indicate that the concentrations of arsenic and chromiumhave been steadily decreasing in the shallow aquifer below, and downgradient of the MAWPsite. Sampling done during the RI indicates that chromium concentrations did not exceedcurrent federal or state drinking water standards at eight of the monitoring well locations. Twomonitoring wells (Nos. 1 and 8) that were sampled during the RI did have elevated chromiumlevels of 62 and 151 µg/L, respectively. However, these results are questionable because a highconcentration of aluminum was also detected in well No. 8. That indicates that the well mayhave been improperly constructed prior to sample collection or the sample collector did notproperly flush the well prior to collecting the sample. Aluminum is relatively insoluble in waterand is frequently used as an indicator of unusually high levels of particulates in unfilteredsamples. The presence of particulates in a water sample extracted from a newly installed wellcauses the laboratory test results for metals to appear to be higher than the true values are. Thewater sample extracted from monitoring well No. 1 was not analyzed for aluminum. Thus, it isit is not possible to assess whether the apparently high levels of chromium detected in well No. 1were accurate or not.

Further sampling of these wells, done to clarify the results, was inconclusive. A comparison offiltered versus unfiltered samples at well Nos. 1 and 8 showed dissimilar results. At well No. 1,the chromium concentration in the filtered sample was below detection, while the unfilteredsample was 88 µg/L chromium. This result indicates that the chromium present was likelyadsorbed onto particulates and not dissolved in the water. However, at well No. 8, the filteredand unfiltered analyses were virtually identical, with 68 and 69 µg/L of chromium detected,respectively. Therefore, at well No. 8 the chromium present was dissolved in the groundwater.

Additional studies characterizing the shallow and deeper portion of the aquifer showed that claylenses may be restricting the vertical migration of the groundwater.

In conclusion, the environmental fate and transport data suggest that future exposure to thecontaminants of concern in groundwater is likely to be an eliminated pathway. Furthermore,even if groundwater does become excessively contaminated in the future, there is no exposedpopulation since area residents are connected to a public water supply. Future residentialdevelopment to the northwest of the site -- in the pathway of groundwater flow -- couldpotentially expose residents through newly drilled residential wells. However, sourceremediation should prevent further groundwater contamination, and levels of contaminantsalready in groundwater are expected to decrease with time and distance from the site.

Surface Water and Sediment Pathways

Surface water and sediment sampling results did not reveal any detectable levels of site-relatedcontaminants. Therefore, no present exposure pathways could be identified. Although someconstituents may have migrated off-site through these media, the quantity is probably minimalbecause if there had been significant contamination, it would likely have been observed due tothe fact that these metals generally settle out of the water, bind to sediment particles, and remaindetectable. Because the sources are being remediated, future off-site migration and subsequentcontamination of surface water and sediment is not likely. For these reasons, exposure tocontaminants in surface water and sediments is considered an eliminated pathway.

PUBLIC HEALTH IMPLICATIONS

A. Toxicological Evaluation

In this section, the potential health effects in exposed persons affected by site contaminants arediscussed. To evaluate health effects, ATSDR has developed Minimal Risk Levels (MRL) forcontaminants commonly found at hazardous waste sites. The MRL is an estimate of dailyhuman exposure to a contaminant below which non-cancer, adverse health effects are unlikely tooccur. MRLs are developed for each route of exposure, such as through ingestion andinhalation, and for the length of exposure, such as acute (less than 14 days), intermediate (15 to364 days), and chronic (greater than 365 days). When MRL's are not available, RfD's and RfC'sdeveloped by USEPA are used to characterize safe levels for non-cancer health effects. Cancerpotencies are used to evaluate chemicals which may pose a significant increased lifetime risk ofcancer at levels observed in the environment. In this PHA, a conservative definition ofsignificant excess risk is adopted: one excess cancer case in a population of 1,000,000 givenlifetime exposure to the contaminant. For comparison, USEPA defines a range of acceptablerisk, one excess cancer case per 10,000 people to one excess cancer case per 1,000,000 people;risks greater than one excess case per 10,000 population are considered significant andunacceptable. ATSDR also develops Toxicological Profiles on chemicals commonly found athazardous waste sites. These chemical-specific profiles provide information on health effects,environmental transport, human exposure, and regulatory status.

Chromium

Chromium is a naturally occurring element of the earth's crust, present in at least two forms: hexavalent (Chromium VI) and trivalent chromium (Chromium III). All people are exposed tolow levels of chromium in air and water, but by far the most significant exposure route isthrough ingestion of food. Trivalent chromium is considered to be a nutrient, essential for goodhealth, at low levels (i.e., 0.05 - 0.2 mg/day for adults, or equivalently 0.0007 - 0.003 mgchromium/kg body weight/day) (ATSDR, 1992).

Chromium VI is more readily absorbed by the human body than chromium III and is thereforemore toxic. Both forms are associated with non-cancer health effects. Chromium VI is anirritant, and short-term, high level exposure can result in adverse health effects at the contactlocation (for example, ulcers of the skin, irritation and perforation of nasal mucosa, and irritationof the gastrointestinal tract). Chromium VI may also cause adverse health effects to the kidneyand liver. The respiratory tract in humans is a major target for inhalation of chromiumcompounds. However, no adverse non-cancer respiratory effects were observed in humansfollowing intermediate length exposures to air concentrations of hexavalent chromium at orbelow 0.001 mg chromium VI/m3. ATSDR used this human respiratory NOAEL to calculate aninhalation MRL for hexavalent chromium of 0.00002 mg Chromium VI/m3 for intermediatelength exposures (ATSDR, 1991b). The oral RfD for non-cancer health effects followingingestion of chromium VI is 0.005 mg Chromium VI/kg body weight/day.

Chromium VI is associated with the production of adverse developmental or reproductiveoutcomes, such as birth defects or fetal deaths. However, information on the specific toxicityand the dose levels required to produce such effects in humans is very limited, and most of theknowledge about chromium's adverse reproductive effects comes from animal studies (ATSDR,1991b). The doses needed to produce such effects in rodents given chromium VI in food orwater are relatively high (ranging from 3.5 mg/kg/day for decreased sperm production to 57mg/kg/day for fetal deaths and birth defects) (ATSDR, 1991b). These levels exceed themaximum estimated oral dose to humans from drinking water and soil associated with MAWPsite (see the following discussion).

Chromium III is generally much less toxic than Chromium VI(6) and is in the form thought to be anutrient at low levels. Chromium in food is primarily in the trivalent form. The oral RfD forthe ingestion of trivalent chromium is 1.0 mg Chromium III/kg body weight/day, 200 timesgreater than the oral RfD for the hexavalent form. There is currently no inhalation MRL fornon-cancer effects for trivalent chromium (ATSDR, 1991b).

Long-term inhalation during occupational exposure to low levels of a variety of hexavalentchromium compounds has been shown to increase the risk of lung cancer. Based on thisevidence, EPA has classified hexavalent chromium as a group A (known human) carcinogen. Studies of workers exposed only to trivalent chromium compounds have consistently found noexcess cancer risks. Therefore, trivalent chromium compounds are not considered to becarcinogens through the inhalation route. Neither trivalent nor hexavalent chromium compoundsare considered to cause cancer through the oral route (ATSDR, 1991b).

The carcinogenic potency factor for chromium derived by USEPA is based on the occupationalepidemiology investigations that linked inhalation exposure to lung cancer, with Mancuso(1975) being the principal study.

At the MAWP site, the occupational and residential exposure routes of concern include past andcurrent inhalation and incidental ingestion of soil, and past ingestion of groundwater. Exposureto chromium through soil-related routes (inhalation and ingestion) is predominantly to thetrivalent form, although levels of chromium VI in soil were elevated in the year following theinitial CCA spill.(7)In contrast, the high levels of chromium in groundwater observed after thespill were mostly in the hexavalent form.(8)

The potential health risks posed to on-site persons and neighboring residents by levels ofchromium found on and off-site were evaluated by Dames and Moore in the RI. MDE reviewedthe Dames and Moore assessment and conducted its own evaluation, finding general agreementbetween the two. Results show that past and current total chromium levels in soil exceed thelocal background concentrations and pose an elevated cancer risk by the inhalation pathway,based on worst case assumptions:

  • that model-based estimates of airborne chromium are accurate;
  • that people were exposed to the maximum concentration of chromium ever detected at thesite;
  • that exposed individuals would be inhaling these levels of airborne particulates over alifetime; and
  • that all chromium in shallow soil was in hexavalent form.

Using these assumptions, the maximum estimated chromium levels in on-site air were 1.91 X 10-4 mg chromium per m3 air (see Table 2), significantly higher than the cancer-based inhalationcomparison value of 8.3 X 10-8 mg of hexavalent chromium per m3 air (ATSDR, 1991b). Inevaluating this potential cancer risk, it is important to keep in mind that the assumptionsemployed are highly conservative in order to protect public health, and that the true cancer riskassociated with site is probably much lower. For example, if one uses measured levels ofhexavalent instead of total chromium in shallow soil to estimate the inhalation cancer risk, thenthe maximum estimated airborne chromium VI level never exceeded the cancer-based inhalationcomparison value (see Table 2). However, very few measurements of hexavalent chromiumrelative to total chromium were made at this site, and thus conclusions based on the hexavalentchromium measurements contain a great deal of uncertainty.

Ingestion of on-site soil contaminated with chromium does not pose a risk to adults, based onpast and current chromium levels and worst case ingestion scenarios. Based on the maximumlevel of total chromium(9) detected in on-site soil, the highest daily doses of total chromium towhich persons on-site (primarily adult workers) could be exposed are estimated to be 0.0039 mgchromium/kg body weight/day from incidental ingestion of soil, and 0.000077 mg chromium/kgbody weight/day for inhalation of dust (corresponding to an estimated air concentration of0.00027 mg chromium/m3 air). This soil ingestion dose is lower than the oral RfDs for bothhexavalent and trivalent chromium, and thus no non-cancer adverse effects to on-site workerswould be expected to occur from the oral route. There is no Chromium III MRL or RfD tocompare to the estimated dust levels of chromium at the site. The estimated dose from dust doesexceed the Chromium VI inhalation MRL, but this comparison is probably not appropriatebecause most of the dust is likely to be in the trivalent form (ATSDR, 1991b).

Off-site exposure to chromium through ingestion of soil does not pose a non-cancer hazard toadults. However, in the worst case analysis, children exhibiting pica behavior(10) are assumed toconsume up to 50 times more soil per day than adults and weigh, on average, seven times lessthan adults (ATSDR, 1988). Thus, even though off-site soils contain significantly lesschromium than on-site soils, in the unlikely event that children regularly play near the seweroutfall over a long period of time, doses could reach as high as 0.0298 mg total chromium/kgbody weight/day. While this dose is lower than the oral RfD for chromium III, it exceeds theoral RfD for chromium VI. Since chromium in soil is likely to be in the trivalent form, theobserved levels are unlikely to be hazardous to pica children. However, under a worst caseanalysis, if all of chromium is in the hexavalent form --a highly unlikely scenario -- then picachildren could potentially be exposed to levels which cause kidney or liver damage (ATSDR,1991b). Again, these worst case analyses are performed to be maximally protective of publichealth, and almost certainly overstate the true risk.

People breathing chromium in areas surrounding the site would not be expected to experiencenon-cancer adverse effects since off-site exposure estimates were 10 times lower than on-siteexposures, the latter which were considered safe for non-cancer health effects.

Levels of chromium in private wells downgradient from the site exceeded concentrationsconsidered safe for long-term exposure. The estimated maximum dose of hexavalent chromiumto adults and children drinking water from contaminated wells was 0.56 and 1.98 mg chromiumVI/kg body weight/day. These doses exceed the oral RfD for chromium VI of 0.005 mg/kg/day(ATSDR, 1991b). However, because the potential exposure period was short (area residentswere connected to a public water supply after contamination was discovered) possible exposureto chromium in the year after the CCA spill is not thought to have posed a significant risk tohuman health. Measurements made from 1983 to the present have shown that chromium in off-site wells is well below the maximum water concentration thought to be protective of humanhealth.

There is no health risk associated with on-site groundwater wells since they have never beenused for drinking water.

Arsenic

Arsenic was detected in surface soil and groundwater at the MAWP site. Although arsenic is a naturally occurring element in the earth's crust, it is used in large quantities to pressure-treatwood. Pure arsenic is a gray-colored metal, but this form is not common in the environment. Rather, arsenic is usually found combined with one or more other elements such as oxygen,carbon, chlorine, and sulfur, which determine its form as inorganic or organic. Arseniccombined with inorganic elements is referred to as inorganic arsenic, whereas arsenic combinedwith carbon and hydrogen is referred to as organic arsenic. The inorganic forms of arsenic areusually more toxic than the organic forms (ATSDR, 1991a).

Data gathered at the MAWP site do not give arsenic concentrations by specific form. To bemost protective of human health, the health discussion below assumes that all site-related arsenicis inorganic.

Exposure can occur at MAWP by incidental ingestion of contaminated surface soil, inhalation of fugitive dust, and possibly by ingestion of groundwater (although affected wells are no longerused for drinking water). Absorption through the skin is not usually an important pathway. Depending on its chemical form, arsenic quickly enters the bloodstream as it is absorbed by thestomach, intestines and lungs. At low to moderate levels, arsenic does not accumulate in thebody since it can be converted by the liver to a less toxic form that is easily excreted in the urine(ATSDR, 1991a).

Inorganic arsenic has been recognized as a human poison since ancient times. Humans aregenerally more sensitive to arsenic toxicity than test animal species for both non-cancer andcancer health effects. The exact human lethal dose is unknown, but is estimated at 1 to 3 mg/kgbody weight from case studies of accidental poisonings. Potential oral exposures to humansestimated for the MAWP site are well below this lethal threshold (see the following discussion). Studies with rats and mice have found the acutely lethal doses that will kill 50% of the testanimals -- the LD50s -- range from 15 to 110 mg/kg body weight (ATSDR, 1991a).

Short-term oral exposure to lower levels of arsenic -- from 0.02 to 0.06 mg/kg/day -- mayproduce injury in a number of different body tissues. Doses ranging from 0.02 mg/kg/day to0.15 mg/kg/day can cause stomach irritation, pain, nausea, vomiting and diarrhea, depending onindividual susceptibility. Prolonged exposure may cause decreased production of red and whiteblood cells, abnormal heart function, liver or kidney damage and impaired nerve function. Themost characteristic effect of arsenic exposure is a pattern of skin abnormalities, including theappearance of dark and light spots on the skin, and small corns on the palms, soles and trunk. Some of these abnormalities may lead to skin cancer. Developmental effects in humans fromexposure to arsenic are not considered a plausible outcome for exposures that are not toxic to themother (ATSDR, 1989).

Based on the aforementioned and other adverse non-cancer outcomes, the oral chronic exposureRfD for arsenic has been set at 0.0003 mg arsenic/kg body weight/day. This dose is set at a leveldeemed highly protective of public health with a large margin of safety. This RfD assumeslong-term exposures and is 67 times lower than the lowest level found to actually be toxic tohumans. There is presently no inhalation MRL or RfD for non-cancer effects associated witharsenic (ATSDR, 1991a).

The estimated maximum oral doses of arsenic to workers and children from contaminated soiland groundwater exceed arsenic's chronic oral RfD, but are below the levels documented to betoxic in actual studies of human exposure (ATSDR, 1991a). Thus, while a theoretical risk ofadverse, non-cancer health effects exists to workers and children who ingest soil from the site, these effects are unlikely based on actual experience.

It is well established that arsenic can cause cancer in humans exposed through the oral orinhalation routes. Inhalation of arsenic is known to cause an increase in human cancers of thelung (ATSDR, 1991a). Air concentrations of arsenic greater than or equal to 2.3 X 10-7 mg/m3pose a significant excess lifetime risk of lung cancer. This concentration corresponds to aninhalation dose of 6.7 X 10-8 mg arsenic/kg body weight/day.

A dose-response relationship between arsenic in drinking water and skin cancer has beenobserved (Tseng et al., 1968). Epidemiology studies have found that individuals who drankwater which contained arsenic at or above 0.3 mg/L had an increased frequency of skin and/orbladder, kidney, lung and liver cancers (ATSDR, 1989). Based on these epidemiology studies,the oral dose associated with a significant excess lifetime risk of cancer was estimated to be 5.7X 10-7 mg arsenic/kg body weight/day.(11)

The potential cancer risk from ingestion of groundwater is currently eliminated because residentsnearby the site do not use private wells for drinking water. Nearby residents were connected toa community water supply shortly after the contamination of residential wells occurred and wasdiscovered. Past exposures to arsenic in groundwater, if they occurred at all, were very brief andtherefore, would not be expected to pose a significant excess cancer risk.

Of concern, however, is the potential cancer hazard posed by arsenic exposures that could resultfrom downgradient wells that may be installed and used as a drinking water source. Currentarsenic concentrations in off-site groundwater have declined with time to below the detectionlimit. As discussed earlier, the detection limit for arsenic is higher than the level which poses nosignificant cancer risk (ATSDR, 1991a). Therefore, it is not possible to evaluate whether or notarsenic levels in water have declined below that which poses a significant cancer risk.

Cancer risk was evaluated for the populations most vulnerable to arsenic exposure by inhalationof fugitive dust and incidental ingestion of surface soil. These populations are: on-site workersand children who play in areas near the site where elevated arsenic levels were found. Inevaluating inhalation cancer risks, it is important to remember that they are based on modelestimates rather than actual samples of air concentrations. Thus, the estimates of airconcentrations and the cancer risk assessments which depend on them contain a good deal ofuncertainty. Therefore, the inhalation cancer risks estimated for this site must be consideredprovisional.

The estimated maximum dose to on-site workers from incidental soil ingestion and inhalation of dust is 7.3 X 10-4 and 7.7 X 10-5 mg arsenic/kg body weight/day, respectively. These doses arecalculated assuming that workers are exposed to the site 250 days per year for 25 years. Boththe ingestion and inhalation dose estimates are associated with significant excess lifetime cancerrisks (ATSDR, 1992).

The estimated maximum dose to children from inhalation of airborne particles is 4.12 X 10-6 arsenic/kg body weight/day. Under the worst case scenario, the maximum average daily intakeof arsenic from soil by children who exhibit pica behavior is 4.2 X 10-4 mg arsenic/kg bodyweight/day. This worst case scenario assumes that children play near the sewer outfall every dayfor 27 years. It also assumes that they ingest 5,000, 200 and 100 mg soil/day from ages 3 to 5, 6to 11, and 12 to 29 years old, respectively. Body weights are assumed to be 15, 30, 50 and 70kgs for ages 3 to 5, 6 to 11, 12 to 17, and 18 to 29 years old, respectively. The excess lifetimecancer risk level associated with the worst-case dose is significant at 7.3 excess cancer cases per10,000 population (ATSDR, 1992).

Under a more plausible but still conservative, lower exposure scenario, the cancer risk is reducedsubstantially, although it is still exceeds a level of significance. This scenario assumes thatexposed children do not exhibit pica behavior, and play near the sewer outfall 100 days per yearfrom ages 5 to 17 years. Daily soil ingestion is assumed to be 200 and 100 mg, and bodyweights are estimated at 30 and 50 kg for children 5 to 11 and 12 to 17, respectively. Theaverage daily intake of arsenic under these conditions is estimated to be 6.3 X 10-6 mg/kg/daycorresponding to 1.1 excess cancer cases per 100,000 people (ATSDR, 1992).

Because exposure opportunities are limited, especially for the inhalation route, the cancer riskspresented above are more theoretical than real: children are unlikely to play in areassurrounding the site where elevated arsenic levels were found, and the off-site area is highlyvegetated, minimizing the chance that the soil will produce fugitive dust available for breathing.

B. Health Outcome Data Evaluation

Birth Defects

Birth defects rates are not available for the census tracts or zipcodes in the immediate vicinity ofthe MAWP site. The smallest geographic unit for which such data are available is Anne ArundelCounty. The County population is 25 times larger than that which lives in the two census tractsadjacent to the site (427,238 versus 17,092 in 1990, respectively) (Source: Anne ArundelCounty Office of Planning and Zoning). The County also spans a much larger geographic areathan the two census tracts of interest. Thus, it is very difficult to assess whether or not an excessrate of birth defects exists near the MAWP site by using County statistics. Even if localstatistics were available, they are of limited value to establish a link between birth defects andexposure to the contaminants of concern: no such exposure information is collected from theindividual parents who give birth to infants with malformations. Nonetheless, a high rate ofbirth defects in the County would suggest that further studies be done to determine where withinthe County the defects are occurring, and whether the parents who gave birth to infants withdefects were exposed to significantly higher levels of the contaminants of concern than parentswho had offspring without such defects.

The preliminary data available from the Birth Defects registry shows that from 1984-1988, theaverage annual birth defect rate for all sentinel birth defects combined in Anne Arundel County(75.4 per 10,000 live births) was about the same as the overall Maryland State rate (73.1 per10,000 live births). However, two types of birth defects were significantly higher in the Countythan in the State. Esophageal atresia was 2.5 times higher in the County than in all of Maryland(4.6 versus 1.8 per 10,000 live births, respectively). Congenital hip dislocation was 2.2 timeshigher in the County than in the State (13.8 compared to 6.3 per 10,000 live births). However,DHMH scientists who maintain the Birth Defects Registry indicate that the apparent excess ofcongenital hip dislocation in Anne Arundel County may be artificially high and is possibly dueto some doctor(s) in the County who use a more inclusive definition of this defect than othermedical practitioners to classify newborns. This possibility may be explored further if resources become available.

The causes of esophageal atresia and congenital hip dislocation are unknown. Furthermore, thehuman exposures that may have occurred as a result of site related contamination in 1984-1988(i.e. when the birth defects were reported) are much lower than those which produceddevelopmental abnormalities in animal studies (see the Toxicological Implications sectionabove). Finally, there are many difficulties, as previously discussed, in trying to use data such asthat available to establish cause and effect. Thus, it is unlikely that any excess reported defectswere the result of exposures to contaminants of concern that emanated from the site.

Cancer Mortality

Like the birth defects data, cancer statistics are not available for the census tracts or zip codesnear the site. The smallest geographic unit for which cancer mortality (death) data is available isAnne Arundel County. The same caution used to interpret birth defects data applies to usingAnne Arundel County cancer data to draw conclusions about residents near the site: The Countypopulation is 25 times larger than the population of concern, and the County is much largergeographically than the area around the site. Furthermore, the available statistics do not includeinformation on whether the individuals who died of cancer were exposed to unusually highlevels of the contaminants of concern. An additional caveat is that cancer generally takes 10 to20 years to develop after exposure to a cancer-causing agent. Thus, those who die of cancer inAnne Arundel County (and become part of the cancer mortality statistics) may not have lived inAnne Arundel County when they were exposed to any causative agent(s). Furthermore, any newcancers occurring immediately after exposure (CCA use started in 1974 and the major spilloccurred in 1978) would not even, in theory, be caused by site-related exposure sinceinsufficient time would have passed for site-related cancer to develop. All of theseconsiderations make it difficult to draw solid conclusions about the cancer risk to residentsaround the MAWP site. Nonetheless, a high rate of cancer deaths in the Anne Arundel Countywould suggest that further studies be done to evaluate whether or not the MAWP is posing anexcess cancer risk to residents near the site.

Table 9 presents age-adjusted, average cancer death rates per 100,000 population over the years1983-1987 for selected types of cancer in Anne Arundel County, and the State of Maryland forcomparison. The relative risk -- the ratio of cancer mortality rates in Anne Arundel County tothose in all of Maryland -- permits a comparison of former to the latter. From Table 9, one seesthat the overall rate of cancer in Anne Arundel County was about the same as that for the State(i.e. the relative risk is 1.07).

The rate of bronchus and lung cancer was elevated by about 16% in Anne Arundel Countycompared to Maryland (i.e. the relative risk was 1.16). This is of interest because twocontaminants of concern, chromium and arsenic, are linked to the production of lung cancer. Itis important to note that these rates have not been adjusted for smoking, which is known to causelung cancer. Thus, for example, such a high relative risk could be observed if Anne ArundelCounty residents smoked significantly more than those in Maryland. The smoking status ofthose who died from cancer remains to be determined.

The relative risk for bronchus and lung cancer was almost twice as high in men than women(1.17 vs. 1.10, respectively). Higher relative risks in men than women suggest an occupationalrather than environmental cause as more men than women have traditionally worked inindustrial settings in which exposure to cancer-causing chemicals occur. If environmentalexposures are the culprit, one would expect to see approximately the same relative risks in menand women, other important risk factors held constant. Anne Arundel County is more industrialthan many other parts of Maryland. The possibility that work-related exposures are responsiblefor some of the excess risk observed in men remains to be explored.

In summary, the available cancer data are insufficient to evaluate whether or not contaminationfrom the MAWP has been or will be associated with an excess risk of cancer to residents. Whilethe higher lung and bronchus cancer rates are provocative, more extensive analyticalepidemiology studies with a longer follow-up period (i.e. years beyond 1987) would be requiredto provide more definitive answers. Even rigorous, analytical epidemiology studies would be oflimited value for the community surrounding the MAWP because they would be unlikely toreveal any excess cancer cases attributable to the site. This is because only a small number ofpersons have potentially been exposed to the site and its contaminants, and the potential cancerrisks posed by the contamination levels at the site are low.

C. COMMUNITY HEALTH CONCERNS EVALUATION

When the public health assessment process began, no specific health concerns were identified inthe community. During the public comment period, some concerns were expressed. Thoseconcerns and a response to the concerns are in the Attachment of this public health assessment. MDE and USEPA will address any new concerns if and when they arise.

Table 9.

Average Cancer Death Rates for 1983-1987 per 100,000 Population for Selected Cancers in Anne Arundel County

Maryland

Anne Arundel
County

Relative Risk*

ALL CANCERS, all
        Males
        Females

192.8
250.2
155.5
205.4
265.3
164.2
1.07
1.06
1.06
Bronchus/Lung,all
        Males
        Females

53.4
31.6
84.3
61.8
35.3
98.6
1.16
1.17
1.10
Esophagus, all
        Males
        Females

4.9
8.5
2.2
4.7
7.6
2.7
0.96
0.89
1.23
Prostate, male 27.2 29.9 1.10
    *Relative Risk is the ratio of Anne Arundel County rates to
    Maryland Rates for each category of cancer


Next Section          Table of Contents

  
 
USA.gov: 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. #