HEALTH CONSULTATION
EXPOSURE INVESTIGATION
OREGON STATE PENITENTIARY
SALEM, MARION COUNTY, OREGON
Groundwater contamination was discovered in the drinking water supply wells at the Oregon State Penitentiary (OSP) on October 5/6, 1989. Two wells were contaminated by tetrachloroethylene (PCE), trichloroethylene (TCE), trans-1,2-dichloroethylene (DCE), and vinyl chloride (VC) For some period before the detection of the groundwater contamination, the OSP population, including inmates, staff, and six staff residences received a portion of their drinking water from the contaminated wells. This report is an evaluation of the potential doses and health effects to the prison population from exposure to the groundwater contaminants.
PCE was the primary contaminant used by the facility dry cleaning operation and is the primary contaminant detected in groundwater at the OSP wells with lesser concentrations of TCE and DCE (one VC value was also measured, but it was at the detection limit and was not replicated in later analyses). The maximum PCE concentrations were 1690 parts per billion (ppb) in Well 3 and 20.5 ppb in Well 1 and maximum TCE concentrations were ~62 ppb in Well 3 and ~8 ppb in Well 1. Thirty day and 10 year PCE and TCE exposure doses are calculated for both inmates, staff, and residents using the maximum concentration values.
Most of the short term and long term exposure doses from ingestion and inhalation of PCE are slightly greater than the health guidance values. Because of the health-protective nature of the dose calculations and uncertainty factors used to derive the guidance values, the PCE and TCE drinking water exposures to OSP inmates and workers are not likely to produce adverse health effects. However, inmates and staff working in the dry cleaning facility of the OSP laundry may have been exposed to significantly higher PCE doses via inhalation. Based on average air concentrations for other dry cleaning facilities, inhalation exposure doses may have been 100 times greater than the ingestion dose from the OSP water supply wells.
The Agency for Toxic Substances and Disease Registry (ATSDR) recommends additional evaluation of the occupational inhalation exposures that may have occurred at OSP to determine whether occupational exposures were within regulatory limits. This evaluation should also include a determination of the PCE air concentrations in the laundry building, the number and frequency of people employed at the facility, and the cleaning and air handling equipment in use at the facility.
In October 1989, drinking water supply wells at the Oregon State Penitentiary (OSP) were found to be contaminated by PCE, TCE, DCE, and VC. The OSP supply wells were taken out of service and OSP subsequently began obtaining water from the Salem municipal water system. For some period before detection of the groundwater contamination, the OSP population, including inmates, staff, and residents of six staff houses received a portion of their drinking water from the contaminated wells. This report is an evaluation of the potential doses and health effects to the prison population from exposure to the groundwater contaminants.
The prison population during the period of exposure included approximately 2,000 inmates and approximately 400 OSP facility staff members. Six staff residences that are outside the security fence, but on the OSP grounds, were also supplied with water from the contaminated wells. The families living in those houses were exposed to groundwater contaminants and are considered to be part of the OSP population. Different members of the OSP population were exposed to contaminants in different ways or different exposure scenarios. Those exposure scenarios and the resulting exposure doses are described in subsequent sections of this exposure evaluation.
Following detection of the contaminants in October 1989, the State of Oregon began an extensive study of the contaminant source(s) and proposed remedial activities. Those activities have resulted in numerous documents detailing contaminant distributions and concentrations (SECOR, 1998) as well as a Baseline Human Health Risk Assessment (SECOR, 2000). Detailed site and background information is available in those documents and will not be repeated here. ATSDR has also conducted a groundwater modeling study to evaluate rates of contaminant transport and potential down gradient contaminant concentrations (GIT, 2000). Pertinent information will be referenced as appropriate.
The source of the OSP contamination has been attributed to use of PCE at the onsite dry cleaning facility (SECOR, 1998). TCE, DCE, and VC are present in much lower concentrations than PCE and are attributed to degradation of PCE (SECOR, 1998). Dry cleaning operations have occurred at OSP since the late 1950s. Although it is unknown exactly when PCE might have been released from the dry cleaning facility and entered the underlying aquifer, the maximum duration of exposure via OSP supply wells is less than 30 years (~1989-1959 minus travel time from source to supply well).
One of the activities completed as part of the remedial investigation is the Well Survey Status Summary Report (Appendix D; SECOR, 1998). This report, which evaluates the potential contamination of offsite private water wells, found that past, current and future exposure in private water wells down gradient of the OSP is unlikely. Also, because the affected OSP supply wells were taken out of service in 1989, current and future direct exposure to the OSP population is unlikely (indirect current and future exposures are evaluated in the Baseline Human Health Risk Assessment; SECOR, 2000). Consequently, this report will focus on past direct ingestion and inhalation by the OSP population for the period prior to 1989.
A preliminary draft of this exposure dose evaluation was presented to community members, OSP facility and state representatives on September 19, 2000. ATSDR representatives toured the OSP facility and interviewed staff and inmates on September 20, 2000. This report includes revised exposure dose evaluations based on information presented to ATSDR at the community meeting and site tour.
CONTAMINANT CONCENTRATIONS AND DISTRIBUTIONS
PCE was the primary contaminant used by the facility dry cleaning operation and is the primary contaminant detected in groundwater at the OSP wells with lesser concentrations of TCE and DCE. VC, toluene, and bromoform have been intermittently detected, but at concentrations very close to the respective detection limits for each chemical and below levels of health concern. Subsequent analyses have not detected VC, toluene, or bromoform. DCE concentrations are below levels of health concern. Exposure doses will be calculated for PCE and TCE.
Following detection of the groundwater contaminants in OSP supply wells 1 and
3, the wells were taken offline and were destroyed the following year. Only
two sets of samples were collected from the affected supply wells. Although
a number of monitor wells were developed to define the distribution of contaminants,
the monitor wells are screened at discrete depths and are not representative
of the contaminant concentrations that were present in the supply wells. Maximum
contaminant concentrations measured in supply wells 1 and 3 are used to estimate
past exposure doses.
Table 1. VOC Analyses at the OSP Water Supply Wells in ppb (from Grimm, M.W., 1989). Bolded values are used in subsequent exposure calculations.
| 10/5/89 Results | Well 1 ppb |
Well 2 ppb |
Well 3 ppb |
Well 5 ppb |
| TCE | 7.6 | 0.2 | 61.4 | ND |
| PCE | 20.5 | 0.8 | 1,690 | 0.2 |
| 10/6/89 Results | Well 1 | Well 2 | Well 3 | Well 5 |
| TCE | 5.0 | ND | 20.0 | ND |
| PCE | 10.0 | ND | 77.0 | ND |
| DCE (trans-1,2) | ND | ND | 21.0 | ND |
| Vinyl chloride | ND | ND | 1.0 | ND |
| Toluene | 1.2 | ND | ND | ND |
| Bromoform | 4.0 | ND | ND | ND |
The maximum contaminant concentrations presented in Table 1 are used for subsequent exposure calculations. Standard water supply procedures at OSP were to pump each well for one month and then switch to an alternate well (SECOR, 1994). If additional water was needed, the primary well was augmented with water from one of the other wells. The following exposure calculations will assume that all water came from the primary well and that the affected wells 1 and 3 were each used for 92 days per year.
The maximum PCE concentrations were 1,690 ppb in Well 3 and 20.5 ppb in Well 1 (Grimm, MW., 1989). Subsequent samples had much lower concentrations (77 and 10 ppb, respectively). PCE concentrations in adjacent monitor wells varied with depth from 40 to 2,400 ppb. However, these depth-specific samples are not good indicators of the total PCE concentration present in Supply Wells 1 and 3 which were reportedly screened from 30-200 ft below ground surface (SECOR, 1998). Consequently, maximum PCE concentrations of 1,700 (1,690) and 21(20.5) ppb are used to calculate exposure doses from Wells 3 and 1, respectively. Maximum TCE concentrations were ~62 ppb in Well 3 and ~8 ppb in Well 1. Although both PCE and TCE concentrations in Well 1 were much lower than in Well 3, the lower Well 1 values are time-integrated with Well 3 values to calculate long term or annual exposure doses.
PCE and TCE are both volatile organic compounds (VOCs). VOCs dissolved in water tend to leave the water and become airborne when exposed to the atmosphere. Studies have shown that inhalation exposure to airborne VOCs from water might be as large or larger than the exposure from ingestion of the water. One study indicated that VOC exposure due to inhalation and dermal contact during a 10 minute shower is approximately equal to the same dose received by ingesting 1.3 liters of the same water (Wan, et.al., 1990). Other studies have shown that, depending on the specific organic compound, 50 to 90% of VOCs in water will volatilize during showering or laundering and become airborne (Andelman, et.al., 1989).
Airborne concentrations of volatilized PCE and TCE have been estimated using two models (Foster and Chrostowski, 1988; Life Systems, Inc., 1989). These models use the physical and chemical properties of the respective VOCs, water concentrations, water flow rates, water temperatures, and volume of whole house or shower area to generate estimated air concentrations. Air concentrations were estimated for individual cells based on continuous running of water in sinks and for the main shower area. Cell and shower area volumes and water temperatures were based on observations obtained during the 9/20/00 site visit. Maximum water concentrations for Wells 1 and 3 (Table 1) were used as model input. The Life Systems, Inc. (1989) equation and associated parameters are listed in Appendix 1.
Air concentration results from the Life Systems, Inc. (1989) model are presented in Table 2. Results from the Foster and Chrotowski (1988) model, which are very similar, were used to confirm the Life Systems model results and are not specifically listed in Table 2. These air concentrations are calculated using water concentrations from each contaminated supply well (Wells 1 and 3) and for each potential exposure area (the OSP shower areas and cells, and staff residence whole house and shower areas).
In addition to the water (Table 1) and air concentration values (Table 2) used
in calculating total exposure doses (in the following section), airborne PCE
was also present in the laundry facility due to direct releases from the dry
cleaning operation. ATSDR currently has no information about the type of dry
cleaning equipment or the volume of PCE used. In order to understand the potential
magnitude of the inhalation exposure in the laundry facility, ATSDR has assumed
that the air concentration in the laundry building was 2/3 of the average value
for dry cleaning operators (Andrasik and Cloutet 1990; from ATSDR, 1997). The
resulting estimate of OSP PCE air concentration of 80 mg/m3 is reduced
from the industry average due to the large volume of the laundry building relative
to a standard dry cleaning operation. It was also very likely that there was
a distinct gradient in PCE air concentrations in the laundry building with much
higher concentrations around the dry cleaning equipment and greatly reduced
concentrations away from the equipment.
Table 2. Estimated VOC air concentrations in OSP exposure areas based on water concentrations from contaminated OSP wells. Air concentrations are estimated using the Life Systems, Inc., 1989 model.
| Water Source | Exposure area | PCE mg/m3 | TCE mg/m3 |
| Well 1 | Cell (with running water) | 3.1e-03 | 5.7e-03 |
| OSP shower area | 1.5e-02 | 1.2e-03 | |
| Well 3 | Cell (with running water) | 2.5e-01 | 8.9e-03 |
| OSP shower area | 1.2e+00 | 4.4e-02 | |
| Well 1 | Staff residence whole house | 1.7e-03 | 1.7e-04 |
| Staff residence shower | 9.3e-02 | 3.6e-02 | |
| Well 3 | Staff residence whole house | 1.4e-01 | 1.3e-03 |
| Staff residence shower | 7.6e+00 | 2.8e-01 |
PCE is classified by the International Agency for Research on Cancer as a Group 2A carcinogen which is "probably carcinogenic to humans" (ATSDR, 1997). It will be evaluated for both potential cancer and non-cancer health effects. The cancer classification of TCE has been withdrawn (Environmental Protection Agency [EPA], 2000) and TCE will only be evaluated for non-cancer health effects. The maximum concentrations of DCE, toluene, and bromoform in the supply wells were lower than appropriate health comparison values and are not likely to produce adverse health effects. Exposure doses and potential health effects from PCE and TCE will be evaluated.
Exposure doses are calculated on the basis of the following equation:
Dose = (Chemical Conc. x IR x EF x ED) / (BW x AT)
| where: Chemical Conc. | = concentration of each contaminant in mg/liter |
|
IR
|
= ingestion rate in liters per day |
|
EF
|
= exposure frequency in days/year |
|
ED
|
= exposure duration in years |
|
BW
|
= body weight in kg |
|
AT
|
= averaging time in days |
For the non-cancer health evaluation, the resulting doses in units of mg/kg/day, are compared with various contaminant specific studies that have investigated the health effects due to exposure to the respective contaminants. To determine potential cancer risk, the PCE doses are multiplied by the pathway-specific cancer slope factors which are expressed in units of inverse dose [1/(mg/kg/day)]. The PCE cancer slope factors are 0.052 for ingestion and 0.002 for inhalation (EPA, 2000).
Cancer Risk = Dose (mg/kg/day) x Cancer Slope Factor (mg/kg/day) -1
Estimation of exposure doses requires delineation of specific exposure scenarios that define who is exposed, the type of exposure, and the frequency and duration of exposure (Table 2). At OSP, the exposed population consists of male inmates who are present 24 hours/day for 365 days per year. Although ATSDR did not have access to records for the inmate population prior to 1989, SECOR (2000) indicated that only one inmate was in intermittent residence for more than 9 years before May 1989. The Baseline Human Health Risk Assessment used an exposure duration of 9.25 years based on the residence time of that inmate. This evaluation will protectively assume that exposure durations of 10 or 20 years could occur for both inmates and workers.
Inmates are assumed to ingest 2 liters of water per day, which is the average daily water intake of an adult (EPA, 1990). Workers are assumed to ingest only 1 liter of water per day from OSP sources because they are only present during working hours. Adult staff residents ingest 2 liters of water per day, while children ingest 1 liter per day. Ingestion exposure frequency is 30 days per month and/or 365 days per year for both inmates and staff residents, but only 25 days per month and/or 250 days per year based on 5 working days per week for OSP staff.
Inhalation exposure frequencies vary with each activity (Table 3). The general inmate population showers 2 times per week for an assumed 15 minutes per shower for a total of 129 minutes per month. OSP staff assigned to the shower area were exposed for 8 hours per day, 2 days per week for a total of 4128 minutes per month. Inmates who continuously run water in their cell sinks are exposed for an assumed 12 hours per day and 7 days per week (or 21600 minutes per month). Both inmates and staff working in the laundry building were exposed for 8 hours per day for 5 days per week (or 12000 minutes per month). Staff residents were exposed to shower areas for 15 minutes per day for 7 days per week, or 450 minutes per month. Residential whole house exposures are assumed to have been 18 hours per day for 30 days per month (or 32400 minutes per month). These exposure assumptions along with the exposure concentrations (Tables 1 and 2) are used in calculating the exposure doses in the following section.
Table 3. Exposure assumptions used in calculation of exposure doses.
| Dose Parameters | Inmates | OSP Staff | Staff Residents |
| Ingestion Rate Adults Children |
2 liters/day NA |
1 liter/day [2] NA |
2 liters/ day 1 liter/ day |
| Inhalation Rate Adult Resting Light Activity Child Resting Light Activity |
7.5 liters/minute 20 liters/minute NA NA |
NA 20 liters/minute NA |
7.5 liters/ minute 20 liters/ minute 4.8 liters/ minute |
| Exposure Frequency Ingestion [1] Intermediate Chronic |
30 days/month 365 days/year |
25 days/month 250 days/yr |
30 days/month 365 days/year |
| Inhalation Showering Sink in cell Whole house Laundry facility |
129 min./ month 21600 min./ month NA 12000 min/month |
4128 min./ mo. [3] NA NA 12000 min/month |
450 min/month NA 32400 min/month NA |
| Exposure Duration Intermediate Chronic |
1 month 10 years |
|
1 month 10 years |
| Body Weight Adult Children |
70 kg NA |
70 kg NA |
70 kg 10 kg |
| Averaging Time Intermediate Chronic Cancer |
30 days 10 years 70 years |
30 days 10 years 70 years |
30 days 10 years 70 years |
| [1] The intermediate exposure assumes all water is from Well
3. The chronic exposure averages well supply over a year with 92 days/year
supplied from Well 3 and 92 days/year from Well 1. [2] Fifty percent of average daily water intake for staff is from OSP wells. [3] Shower area staff only. |
|||
Table 4 presents the calculated 30 day and long term or chronic (10 year) exposure doses to OSP inmates and staff and residents of the OSP houses that were hooked up to the OSP water wells. These exposure doses are summed across all exposure routes for the respective exposure scenarios. For example, the general inmate exposure dose includes doses from inhalation during showering, ingestion of drinking water and inhalation from running water in cell sinks. Exposure doses for individual exposure routes (ingestion or inhalation) are included in Appendix 2. All exposure doses are presented in units of mg- contaminant/kg- body weight/day). The 30 day exposures represent maximum doses because all water is supplied from the highly contaminated Well 3. The chronic or 10 year exposure dose is lower because the water source is time-integrated to include supply from much lower concentrations in Well 1 as well as non-contaminated Wells 2 and 5. Table 3 also includes the PCE lifetime cancer risk for exposure durations of 10 and 20 years.
If pre-1989 PCE or TCE supply well concentrations were greater than those measured in 1989 or later, the resulting doses would have been commensurately higher. Neither the groundwater transport modeling study (GIT, 2000) nor the analytical groundwater sampling (SECOR, 1998) indicate that a pulse of higher contaminant concentrations has migrated past the OSP supply wells or property boundary.
This type of high concentration pulse could result from a short term release or PCE spill and would be necessary to produce a historically higher, then declining concentration trend. However, maximum contaminant concentrations still occur in the laundry building area and decrease down gradient. Therefore, the use of the measured maximum supply well concentration is a health-protective, or conservative assumption, which probably over-estimates contaminant concentrations. Also, if TCE is a degradation product as indicated by SECOR (1998), its initial concentration was zero with a gradual increase to measured concentrations based on the rate of microbially mediated PCE degradation.
In addition to long term changes in VOC concentrations due to plume migration, it is very likely that concentrations at Well 3 would change during the period of monthly pumping. It is very common for VOC concentrations at remedial wells to decrease as contaminants are removed from the immediate capture area of the well. However, with no time series analyses of the supply wells, or control over the timing of the few samples taken, it is not possible to reasonably estimate such decreases. In light of these uncertainties, use of maximum measured concentrations is necessary.
Table 4. Total exposure doses and cancer risk from past groundwater exposure to inmates and workers at OSP. Note that these doses are summed across the ingestion and inhalation routes.
|
Exposure Scenarios |
PCE |
PCE 10 year |
PCE Cancer Risk 10 year |
PCE Cancer Risk 20 year |
TCE 30 day |
TCE 10 year |
| mg/kg/day | mg/kg/day | |||||
| Inmates General Population | 6.9e-02 | 1.8e-02 | 9.3e-05 | 1.9e-04 | 2.5e-03 | 7.0e-04 |
| Inmates Laundry Workers | 9.2e+00 | 9.2e+00 | 2.7e-03 | 5.4e-03 | 2.5e-03 | 7.0e-04 |
| Staff General Population | 2.0e-02 | 5.1e-03 | 3.8e-05 | 7.6e-05 | 7.1e-04 | 2.0e-04 |
| Staff Laundry Building | 9.2e+00 | 9.1e+00 | 2.6e-03 | 5.2e-03 | 7.1e-04 | 2.0e-04 |
| Staff Shower Area | 6.7e-02 | 1.7e-02 | 4.1e-05 | 8.3e-05 | 5.7e-03 | 1.6e-03 |
| Staff Resident- Adult | 1.2e-01 | 3.1e-02 | 9.7e-05 | 1.9e-04 | 3.3e-03 | 9.4e-04 |
| Staff Resident- Child | 5.1e-01 | 1.3e-01 | 3.5e-04 | NA | 1.3e-02 | 3.8e-03 |
Contaminant concentrations measured at the wells very likely over-estimate contaminant concentrations delivered to drinking water taps. Each supply well fed into a 3,500 gallon pressure tank before distribution (Trewartha, SECOR, personal communication [e-mail], Aug. 9, 2000). Depending on temperature, pressure, and volume of air space in this tank, some portion of the VOC contamination would volatilize into this air space and reduce the water concentration. Similar volatilization would occur at the taps and during cooking. However, inhalation exposure would occur to airborne PCE/TCE that volatilizes into living and working areas and any water concentration reductions would be very difficult to quantify. Inhalation exposure has been included in the exposure calculations (Table 2).
There is also some uncertainty related to the duration of exposure. None of the reports reviewed include any estimate of when the PCE/TCE may have been originally released from the laundry facility to the aquifer. The groundwater transport model (GIT, 2000) indicates that contamination would travel the ~600 feet from the laundry facility to supply well 3 in approximately 5 years. However, it is also likely that the initial breakthrough concentration was much lower than the 1989 measured concentration. Because there is no reliable estimate of the timing of the PCE/TCE release to the aquifer, use of the maximum duration of inmate residence time is the best estimate of exposure duration. Estimates of cancer risk are also extended to 20 year durations to account for prison workers that may have been employed for more than 10 years.
The laundry facility air PCE concentration which is the basis for estimating inhalation exposures to inmates and staff assigned to the laundry building is based on the arbitrary use of an industry average value. Additional work should be done to refine the air concentrations and exposure doses in this area. The laundry inhalation doses are included only to highlight the potential magnitude of PCE exposure for this scenario.
Next Section Table
of Contents
