HEALTH CONSULTATION
SHIAWASSEE RIVER SUPERFUND SITE
HOWELL, LIVINGSTON COUNTY, MICHIGAN
The Agency for Toxic Substances and Disease Registry (ATSDR) received a request for a public health consultation relative to the September 2001 Record of Decision (ROD) for the Shiawassee River Superfund site near Howell, Livingston County, Michigan.
Due to a lack of residential and non-residential exposure information for the South Branch of the Shiawassee River, the public health hazard is indeterminate. A need exists, however, for more current fish tissue concentration data, because the most recent data are from 1987. In addition, data on the consumption of game or livestock species from the river channel and/or floodplain is needed to fully characterize the extent of the contamination. Multiple data reports from the last few years have all confirmed the presence of residual PCB hotspots in both the river sediment and the floodplain soil of the South Branch. Given the moderately high concentrations still found at certain hotspots and the reasonable assumption that other, still undetected hotspots in soil and sediment exist, people living in the vicinity of the South Branch could also be exposed to harmful levels of PCBs through incidental ingestion of soil and sediment. To a lesser extent, exposure is possible via ingestion of river water and dermal exposure to soil and sediment hotspots.
Fish advisories should remain in effect, and health education measures should be initiated to ensure that all residents and non-resident recreational participants are informed of the potential health risks from PCB exposure. The Michigan Department of Community Health (MDCH) suggests that a demographics analysis be carried out to address data gaps for residential and non-residential recreational patterns of use for the South Branch. Finally, MDCH, the Michigan Department of Environmental Quality (MDEQ), and the regional U.S. Environmental Protection Agency (EPA) office should work together to address the lack of empirical data on natural attenuation processes for the South Branch.
The purpose of this health consultation is to determine whether this site poses
a threat to public health in response to a "Request for Department of Community
Health Review," as submitted by MDEQ. This consultation was requested to
address indirect and direct exposures to these contaminated media. More specifically,
MDEQ requested that MDCH determine whether the EPA ROD is sufficiently protective
of local residents and recreational users of the South Branch river system.
PCB contamination was detected in the river sediment and floodplain soil associated
with the South Branch of the Shiawassee River (Superfund site ID#: MID980794473 
).
MDCH conducts health assessments for ATSDR (charged with assessing public health
risks at Superfund sites) under a cooperative agreement.
MDCH and ATSDR consider environmental data and community health concerns when forming a health opinion. Health outcome data (morbidity and mortality) may be considered as well. On the basis of their conclusions, the agencies then make recommendations to ensure public health and safety. This consultation will provide the necessary overview of the past and current site conditions; discuss the available environmental, exposure, toxicological, and regulatory information; and provide specific public health action items.
Site Description and History
The South Branch of the Shiawassee River (hereafter referred to as South Branch) originates from Coon Lake near Marion Township just south of Howell, Michigan (Livingston County). It flows north for approximately 40 miles through Shiawassee County into the Shiawasseetown Reservoir (aka Shiawasseetown Pond). An aging dam is located at the reservoir (near Shiawasseetown) to regulate flow into the main stem of the Shiawassee River, which continues to flow north through Saginaw and Bay Counties. The Shiawassee converges with the Tittabawassee and Saginaw Rivers near the city of Saginaw and empties into Lake Huron via Saginaw Bay. See Figure 1 for an overview of this 40-mile stretch of river.
The South Branch is 20-45 feet in width, with a reported average width of 25 feet during normal periods of flow. Late summer (low) and peak flows in 1982 were reported to be 15 and 75 cubic feet per second (cfs), respectively (ENTACT 2003). Late summer depth of the water channel is 1-2 feet, with scour depths of up to 5 feet seen along the river banks. The adjacent 100-year floodplain is 50-300 feet wide during the first river mile (as measured from M-59; see Figure 1) and is heavily overgrown with mixed woodland and marsh-like areas. Land immediately adjacent to the South Branch is predominantly forested floodplain, agricultural, or undeveloped, with several residences located along the river channel. Presumably, river sediments contaminated by PCBs are deposited in the floodplain during periods of high flow in the river, which generally occurs seasonally (MDCH 2001).
The Cast Forge Company site (CFC, hereafter referred to as "former CFC property" or the "study area") consists of the property located at 22440 W. Highland Road (M-59), Howell, Michigan (Livingston County) and the section of the South Branch between M-59 and Steinacker Road, comprising approximately 8 miles of the river (EPA 2001). See Figure 2 for an overview of the study area. CFC manufactured aluminum wheel castings for the automobile industry starting in 1969. The former CFC property covers approximately 51 acres. It is bordered on the north and east by wetlands, on the west by the South Branch, and on the south by M-59. The developed portion of the property is fairly flat, draining predominantly to the wetlands on the eastern part of the property. An underground culvert drains these wetlands into the South Branch (ENTACT 2003).
From 1969 to 1972, CFC used oils containing PCBs as part of the on-site manufacturing
process. The company stopped using PCB-bearing oils in 1972, although some residual
PCBs may have remained in the equipment until thoroughly flushed in 1976 (Tetra
Tech 2001). The company discharged its waste process water from 1969 to 1977
to on-site lagoons that discharged to the South Branch. Spills and leaks of
PCB-containing hydraulic oils into the process water, discharge and overflow
from the lagoons and tanks, and disposal of sludge from the lagoons and tanks
resulted in PCB contamination of the former CFC property. Contamination was
subsequently found in the soils of CFC property, in floodplain soils and river
sediment of the South Branch, in sediments of nearby wetlands, and in fish taken
from the river (ERG 1980, MDNR 1977). In October 1977, the company installed
a wastewater treatment system and has been transporting wastewater off-site
for disposal ever since.
The State of Michigan filed a lawsuit against CFC in 1977 for contaminating the environment with PCBs. The final consent judgment in the case required the company to remove contaminated materials from the former CFC property for proper off-site disposal and to finance remediation of part of the South Branch. The company carried out the required clean-up of its property in 1981 and initiated clean-up of the river channel in 1982. The original extent of remediation was planned to cover the entire 8-mile study area (see Figure 2); however, cost over-runs and greater-than-expected sediment contamination allowed for remediation only of roughly 1.5 miles. In addition, the site was placed on the National Priorities List (NPL, a registry of "Superfund" sites) at this time because of concerns about possible exposure to PCB contamination.
MDPH (now MDCH) issued an advisory against consumption of fish taken from the South Branch of the Shiawassee River in 1979.1 The current 2003 MDCH Michigan Family Fish Consumption Guide advises against eating any species or size of fish caught from the South Branch between M-59 and Byron Road–a stretch that includes the study area (MDCH 2003). The Guide also advises against consumption of all carp and larger northern pike and smallmouth bass from the stretch of the Shiawassee River that lies between the towns of Byron and Owosso–outside the study area but within the 40-mile reach of the South Branch.
The Michigan Department of Natural Resources (MDNR) contracted for a Remedial Investigation/Feasibility Study (RI/FS) of the site in 1986.2 The RI identified areas of residual PCB contamination both on- and off-site and indicated that the 1.5-mile stretch of river immediately downstream of the former CFC property contained the highest amounts of residual contamination (Warzyn 1992). This is the same 1.5-mile stretch that was remediated in 1982.
EPA conducted additional sampling in 1999 to delineate the area of contamination with greater certainty; however, the river sediment and floodplain soil sampling were focused within the 8-mile study area. Five floodplain locations (Transects 9, 16, 26, 28 and 37) and six sediment (Transects 4, 7, 11, 12, 25, and 27) locations were reported to be PCB "hotspots," according to this sampling event (Tetra Tech 2001). Transects from the study area can be found in Figure 2. The 1999 data formed the basis of the ROD issued by the EPA in September 2001 (EPA 2001). The ROD selected "Excavation and Off-Site Landfill Disposal of Contaminated Soil" as the chosen remedial action to clean up river sediments to 5 parts per million (ppm) and floodplain soil to 10 ppm, followed by post-remediation monitoring. This remediation decision assumed that, along with physical removal of sediment with PCB concentrations above 5 ppm, natural attenuation would further reduce PCB concentrations in the river to a level at which fish advisories would no longer be needed. Furthermore, the selected remedial action level (RAL) was based on a "projected, post-remediation Surface Weighted Average Concentration (SWAC) goal of 1 mg/kg [milligram of PCB per kilogram of sediment] along the first river mile downstream of the facility" (EPA 2001).
Currently, MDEQ is investigating the sediment along the entire 40-mile stretch of the South Branch (Gannett Fleming 2003). The final report from this sampling is still being created as of the writing of this health consultation; however, preliminary data confirm residual PCB contamination within the original 8-mile study area. The data from Gannett Fleming (2003) are of limited usefulness for this analysis because the data collected did not follow the transect system followed in Tetra Tech (2001) and ENTACT (2003); however, Gannett Fleming (2003) provides the only recent sediment data taken outside the original study area.
A pre-design investigation (PDI) was performed during March, April and July 2003, prior to execution of any remedial actions included in the EPA ROD. Results of this most recent sampling event confirmed residual "hotspot" contamination in floodplain soils (from Transects 9, 26, 28, and 37) and river sediment (from Transects 4 and 7 as well as two supplemental transects) (ENTACT 2003).
The former MDPH, working under a cooperative agreement with the Agency for
Toxic Substances and Disease Registry (ATSDR), prepared a Health Assessment
(HA) for the South Branch in March 1989. The HA concluded that the site was
a potential public health concern because of the possibility of exposure to
PCBs in the soils, sediment, and biota (ATSDR 1989). The HA recommended further
sampling of wetlands along the river and suggested that the advisory against
consumption of fish remain in effect. The HA did not recommend any follow-up
health studies.
ATSDR issued a Site Review and Update for the Shiawassee River in 1993. It documented visits to the site by MDPH staff in 1991 and 1993 (ATSDR 1993). Circumstantial and anecdotal evidence indicated that both fishing and canoeing along the river were still occurring, despite public fish consumption advisories. The document also noted that development along the river continued, and although not many points of public access to the river existed at that time, continuing development might create new public points of access in the future. During these site visits, local residents informed MDPH staff of infrequent harvests of snapping turtles from the river channel (reportedly to be sold to restaurants in the metropolitan Detroit area). Finally, the Site Review reported two cattle farms located near the river channel, on properties where the cattle grazed on seasonally flooded fields within the floodplain. To further investigate, MDPH and Michigan Department of Agriculture (MDA) staff collected and analyzed a composite milk sample from the farms. The MDA laboratory reported no detectable PCB contamination at a method detection limit (MDL) of 1 part per billion (ppb) [ATSDR 1993]. No new assessment or consultation actions were suggested for this site in the Site Review document.
The 1994 revision of the Site Review addressed the findings of the Health Activities Recommendation Panel, convened by ATSDR and MDPH to evaluate the data and information developed for the Shiawassee River Superfund site. This panel determined that occasional incidental exposures to sediments, surface water, and fish contaminated by PCBs probably occur (ATSDR 1994). No additional health actions were recommended because MDPH was providing community health education at the time, in addition to maintaining the fish consumption advisory for the river. Finally, the panel supported additional wildlife and fish sampling, as recommended by MDPH.
Christina Bush and Erik R. Janus (MDCH), as well as Janet (Sunny) Krajcovic (MDEQ), conducted a site visit on May 29, 2003. The three inspected the former CFC property, including wetland areas on the eastern side of property, culverts, and drainage ditches connecting the various wetland areas found on site. The river channel adjacent to the site, both on and off the former CFC property, was examined. Various signs of both human recreation and wildlife use along the river banks adjacent to the former CFC property were noted during the site visit. Several instances of human recreation were observed near Transect 5, including a hunting blind and a pile of empty beer cans. In addition, several occurrences of dark orange-colored bank staining were observed along the first mile or so of river sediment. In particular, this orange coloration of the sediment was noted at Transect 4, along with "oily" patches of sediment.
Following examination of the former CFC property and adjacent river channel, the team visited several public access points to the river, as seen from road crossings. Residences were noted along the river channel at Bowen Road and Chase Lake Road (both of which are within the original 8-mile study area, see Figure 2), and signs of human recreation within the floodplain were seen in many instances, some of which involved children. (A site visit conducted by MDPH staff in 1991 also had noted signs of recent fishing activity near the Bowen Road bridge crossing.) Although much of the privately owned property along the river channel may be undeveloped, several residences were observed along the river channel, with well maintained lawns extending to the river channel.
To estimate expected exposure point concentrations for both river sediment and floodplain soil hotspots, a meta-analysis was conducted through use of the combined data of Tetra Tech (2001) and ENTACT (2003). The data of Gannett Fleming (2003) were not included in this meta-analysis because this study did not use the same transect system; however, the Gannett Fleming data were considered in the overall assessment. These three data reports represent the most recent and most extensive characterization of the PCB contamination found in river sediment and floodplain soil of the 40-mile stretch of the South Branch.
Estimates for "average" and "reasonable maximum exposure" (RME, analogous to the 95th percentile upper confidence limit [UCL]) were calculated after combining the raw data from Tetra Tech (2001) and ENTACT (2003). This UCL is interpreted as follows: It is 95% certain that a future sample (i.e., PCB media concentration) taken from a particular hotspot should fall below the 95th UCL value for that hotspot. These estimated exposure concentrations were compared to any existing and relevant health-based regulatory levels. Raw data and results of the meta-analysis can be found in Appendix A.
The Shiawassee River Superfund site has a long and varied history when it comes to environmental contamination data. This report will give attention only to the most recent studies in order to characterize current site conditions. Consult Appendix B for an extensive review of site history.
EPA conducted a final feasibility study (FS) in 1999 to assess the most current extent of PCB contamination in on-site soils, in South Branch river sediment, and in adjacent floodplain soils (Tetra Tech 2001). On-site sampling showed remaining residual on-site soil contamination in low-lying wetlands east and northeast of the former facility buildings. Sediment and soil samples were collected from the 61 transects set up during the 1986 RI; five floodplain (Transects 9, 16, 26, 28 and 37) and six sediment (Transects 4, 7, 11, 12, 25, and 27) locations were shown to be PCB hotspots. EPA used this final FS data in the ROD for the site. The final FS followed a fish-shocking protocol in the final FS to examine stunned fish for lesions; no lesions were found through this sampling effort.
Gradient Corporation (2000) performed a statistical evaluation of the 1999 data for those samples taken within 3.5 miles downstream of the former CFC facility. Gradient reported a "sample [arithmetic] mean concentration of PCBs in river sediments downstream of Cast Forge property" of 5.43 ppm; however, Gradient considered the geometric mean a "more appropriate and more stable estimate." The reported corresponding geometric mean was 2.14 ppm, with a 95th upper confidence limit (UCL) of 3.37 ppm. This 95th UCL (3.37 ppm) is one full order of magnitude higher than the current Michigan "sediment standard" of 0.33 ppm for PCBs in sediment that is protective of aquatic organisms. This standard is based on a site-specific, back-calculated sediment concentration (considered protective of aquatic organisms by MDEQ) and on the MDL for SW-846 Test Method 8082 for the more highly chlorinated Aroclor mixtures (1242, 1248, 1254, and 1260).3
MDEQ recently prepared a work plan outlining the steps necessary for delineation of river sediment contamination within the South Branch (Gannett Fleming 2003). This study sampled sediment along the entire 40-mile stretch from the former CFC facility to the Shiawasseetown Reservoir (see Figure 1). The 40-mile stretch was split into three smaller study areas: Study Area 1 consisted of the area between the former CFC facility and the Chase Lake Road crossing (about 7 miles downriver); Study Area 2 consisted of the 33-mile stretch between Chase Lake Road and the Shiawasseetown Reservoir; and Study Area 3 was the reservoir itself. The final report from the sampling is still pending as of the writing of this health consultation; however, preliminary data results confirm residual PCB contamination within Study Area 1.
The results from Gannett Fleming (2003) were not included in the meta-analysis because this sampling event did not use the transect system used in most other recent reports. However, the unique aspects of this data collection were that sediment samples were taken from outside the original 8-mile study area and that some samples were screened for polychlorinated dioxin (PCDD) and furan (PCDF) congeners. All PCB samples taken outside the original 8-mile study were non-detect at an MDL of 0.33 ppm. Two of 13 samples screened for PCDDs/PCDFs showed detectable concentrations but were found to be consistent with Michigan background concentrations. The MDL for the PCDD/PCDF screen was 4 parts per trillion.
The EPA pre-design investigation (PDI) sampled river sediment, on-site soil, and floodplain soil during March to July 2003 (ENTACT 2003). Sampling was limited to the first river mile downstream from the former CFC facility and the "hotspots" identified in both sediment and soil in the final FS report. The PDI confirmed residual PCB contamination in on-site soils, river sediment, and floodplain soils.
Limitations of Environmental Data under Consideration
This uncertainty discussion focuses on the three reports that contained the most recent environmental media data (Tetra Tech 2001, ENTACT 2003, and Gannett Fleming 2003); the uncertainties fall into 5 major categories:
Because of the discrepancies between the reported results of ECCS and GLA, a second fixed facility laboratory (ESC) was retained to re-examine split samples taken during March and April 2003. ESC used the same extraction and analysis methods as GLA. ENTACT (2003) performed statistical analyses on the three sets of data and stated that a better overall correlation existed between the ECCS and ESC data sets (as opposed to either the ECCS/GLA comparison or the ECS/GLA comparison). In all instances where PCBs were detected, the results reported by ECCS were higher than the results reported by either GLA or ESC. This health consultation used the ECCS data within the meta-analysis because these higher results are considered more conservative and thus more protective of human health.
The Gannett Fleming data were not incorporated into the meta-analysis that included the ENTACT (2003) and Tetra Tech (2001) data, because the Gannett Fleming sampling effort did not use the same transect scheme as the two other reports. It is recognized that the use of the 95th upper confidence limit as a "reasonable maximum exposure" is dependent on sample size as well as on sample mean and standard deviation and that deriving these values from smaller data sets includes a higher amount of uncertainty. [If Gannett Fleming (2003) data matched up with the existing transect system, these data could be included in the meta-analysis, thus potentially reducing uncertainty as a result of the presence of more data.]
To determine whether nearby residents are, have been, or are likely to be exposed to contaminants associated with a property, ATSDR and MDCH evaluate the environmental and human components that could lead to human exposure. Actual exposure to a contaminant occurs through an exposure pathway. An exposure pathway contains five elements: (1) a source of contamination, (2) contaminant transport through an environmental medium, (3) a point of exposure, (4) a route of human exposure, and (5) a receptor population. An exposure pathway is considered complete if there is evidence that all five of these elements are, have been, or will be present at the property. Alternatively, an exposure pathway is considered complete if there is a high probability of exposure. It is considered a potential exposure pathway if at least once of the elements is missing but could be found present at some point. An incomplete pathway exists if at least one element is missing and will never be present. Table 1 shows the PCB exposure pathways expected for the South Branch.
Air: Insufficient site-specific data exist to permit a determination of whether adverse health effects would occur from PCB exposure via the inhalation route. No ambient air samples or studies for the site were identified. The ease with which PCBs can cycle back and forth between air, water, and soil is dependent on the amount of chlorination of congeners found at the site. Lighter PCBs (i.e., those with fewer chlorine atoms) are more likely than heavier PCBs to evaporate into the air, where they can be quickly dispersed by ambient wind conditions. Heavier PCBs are more likely to sorb to sediments when released into aquatic environments (although there can be some release to the water column from sediments). The PCB mixtures found at the site are primarily Aroclor 1242 and 1248 (heavier PCBs), and they are more likely to settle into the sediment rather than to evaporate into the air. Therefore, inhalation of PCBs from the ambient air is considered an incomplete exposure route.
Surface Water/Groundwater: No water samples from the South Branch were identified; however, groundwater samples collected by MDNR in 1988 from on-site monitoring wells contained no detectable PCB concentrations. It is expected that small amounts of PCBs will partition from the sediment into the water column. Once in the water column, these PCBs are likely either to partition to air via volatilization or to be transported downstream (where they could volatilize to air or be re-deposited to sediment). The Shiawassee River is not used as a public water supply. Therefore, exposures to PCBs in surface water would likely be limited to negligible amounts during human recreation in the river channel.
Table 1. Exposure Pathways of Concern at the South Branch of the Shiawassee River.
| Source | Environmental Medium and Transport | Exposure Point(s) | Exposure Route(s) | Exposed Population | Time Frame | Status |
| Release of PCBs from the former CFC facility | Air | Ambient Air | Inhalation | Residents, swimmers, boaters, anglers | Past | Incomplete |
| Present | Incomplete | |||||
| Future | Incomplete | |||||
| Water | South Branch, Shaw Lake, Shiawasseetown Reservoir | Ingestion, dermal absorption, inhalation |
Residents, swimmers, boaters, anglers | Past | Potential | |
| Present | Potential | |||||
| Future | Potential | |||||
| River Sediment | South Branch | Ingestion, dermal absorption, inhalation |
Residents, swimmers, anglers | Past | Potential | |
| Present | Potential | |||||
| Future | Potential | |||||
| Floodplain Soil | 100-year floodplain of South Branch | Ingestion, dermal absorption, inhalation |
Residents, others recreating in floodplain | Past | Potential | |
| Present | Potential | |||||
| Future | Potential | |||||
| Biota | Fish, turtles caught from South Branch; game species grazing in the floodplain | Ingestion | Local anglers and hunters | Past | Potential | |
| Present | Potential | |||||
| Future | Potential |
River Sediment: Any individual engaging in water-based recreation (swimming, boating, canoeing, angling, etc.) within the South Branch may come into contact with PCB-bearing sediments. In addition, it is likely that the sediments closer to the former CFC property are migrating downstream over time and with scour events within the river channel. The implications are such that individuals much further downstream could be exposed in the future to transported PCB-bearing sediment that is not detected and remediated. This could the case near the town of Byron, Michigan, which has a millpond fed by the South Branch that supports water-based recreation by local residents. (Personal communication with J. Krajcovic [MDEQ], May 2003)
Floodplain soil: It is a reasonable assumption that those properties along the river are occasionally contaminated with PCB-bearing sediment during flooding events in which the water overflows the river banks. In addition, significant (but unquantified) numbers of residential and otherwise private properties are within the 100-year floodplain; therefore, it is assumed that potential exposure to PCBs in soils exists.
Biota: Although fish advisories have been and are warning against consumption of fish taken from the South Branch, several anecdotal reports imply that fishing (and possibly other water-based recreation) still occurs along the river. Therefore, it is assumed that exposure to PCBs via ingestion of fish and/or turtles taken from the South Branch could occur, as it could with the ingestion of any game (i.e., white-tailed deer, cottontail rabbit) or livestock (i.e., cattle, chickens, pigs) species that graze within the 100-year floodplain.
MDNR collected and analyzed fish from the South Branch several times from 1981 to 1987 as part of the Michigan Fish Contaminant Monitoring Program (MDEQ 2003). The most recent MDNR sampling event was in July 1987 near the Marr Road bridge crossing, where fillet samples of 10 rock bass and 10 white suckers were collected and analyzed by the MDPH laboratory (ATSDR 1993). Rock bass tissue averaged 7.95 ppm, while the white sucker tissue averaged 17.52 ppm. Because of the lack of fish length data, it is impossible to derive any reliable relationship between fish body length and PCB tissue concentration from this data (ATSDR 1993). MDNR also collected fish from the South Branch at New Lothrop Road near Byron, Michigan (Shiawassee County) in July 1987, comprising 10 carp, 3 northern pike, and 7 smallmouth bass. Maximum concentrations in fillet samples were 7.45 ppm, 1.21 ppm, and 0.31 ppm, respectively. Most recently, FRG (1999) reported fish tissue and sediment data from 1994 caged-fish studies involving white suckers within the South Branch. Near the Bowen Road crossing, sediment PCB concentration was 0.72 ppm and fish concentration was 2.56 ppm; near the Marr Road crossing, sediment concentration was 0.59 ppm and fish concentration was 1.7 ppm. See Table 2 below for a summary of the most recent fish tissue concentration data.
The EPA ROD states that fishing is limited in the South Branch because of a "lack of significant game fish"; however, MDNR has sampled northern pike, crappies, white suckers, sunfish, rock bass, and bluegills near the site (MDEQ 2003). These fish are likely to be caught and possibly eaten by recreational anglers as well as by subsistence fishers; however, it should be noted that the feasibility of subsistence fishing at this site has not been demonstrated. Also, Tetra Tech (2001) conducted a fish-shocking study as part of the final FS, which reported the existence of rock bass, pumpkinseed, shiners, and minnows (which could serve as prey for larger game fish found within the river).
Table 2. Fish tissue concentrations from the "study area" of the South Branch of the Shiawassee River.
| Sampling Date | Sampling Location | Concentration (ppm) | Type of Fish | Reference |
| July 1987 | Marr Rd | 7.95 (average) | Rock bass | ATSDR 1993 |
| 17.52 (average) | White sucker | |||
| New Lothrop Rd | 7.45 (max) | Carp | ||
| 1.21 (max) | Pike | |||
| 0.31 (max) | SM Bass | |||
| May 1994† | Bowen Rd | 0.72 (sediment) | White sucker | FRG 1999 |
| 2.56 (fish tissue) | ||||
| Marr Rd | 0.59 (sediment) | White sucker | ||
| 1.7 (fish tissue) |
† Note that the sediment and fish tissue concentrations from this sampling event were taken at the same time and location.
Results of the meta-analysis: See Table 3 below for results of the recent floodplain soil and river sediment meta-analysis. Discussion of the health implications of the reported results can be found in the next section of this report.
Results from Table 3 were compared against the RALs (5 ppm for sediment and 10 ppm for floodplain soil) identified in the EPA ROD. Sediment transects 11 and 21 are below this 5 ppm RAL at central tendency ("average") exposure concentrations. All estimated RME sediment concentrations exceed the 10 ppm action level. Floodplain soil results show that only transect 7 does not exceed the 10 ppm RAL, on the basis of central tendency exposures. In contrast, all estimated soil concentrations exceed 10 ppm on the basis of RME exposures.
Table 3. Estimated total PCB concentrations in South Branch river sediment and floodplain soil "hotspots."
| Transect | Medium | Exposure Point Concentration (ppm) | |
| Central Tendency | RME (95th UCL) | ||
| 4 | sediment | 6.3 | 14.3 |
| 7 | sediment | 28.4 | 55.2 |
| 11 | sediment | 3.2 | 7.8 |
| 12 | sediment | 8.4 | 16.2 |
| 21 | sediment | 2.9 | 5.4 |
| 25 | sediment | 15.7 | 43.3 |
| 9 | floodplain soil | 12.8 | 23.7 |
| 28 | floodplain soil | 7.4 | 10.6 |
| 37 | floodplain soil | 11.6 | 35.3 |
All identified soil hotspots (Transects 9, 26, 28, and 37) had detectable levels of PCBs in the subsurface soil as well. These data were not included in the meta-analysis performed for surficial floodplain soils; however, the existence of these hotspots may be significant for the design of health education initiatives (see the Recommendations section for more information).
Health Effects: PCBs are complex mixtures of synthetic organic chemicals with no known natural source. They exist as colorless to light yellow, oily liquids or solids. They have no known smell or taste. Some PCBs are volatile and may exist as a vapor in air. Because they do not burn easily and are good insulating materials, PCBs were used widely as coolants and lubricants in transformers, capacitors, and other electrical equipment. The manufacture of PCBs stopped in the United States in 1977 because of evidence that the chemicals build up in the environment and may cause harmful effects. Products that may contain PCBs include old fluorescent lighting fixtures, electrical devices or appliances containing PCB capacitors made before PCB use was stopped, old microscope oil, and old hydraulic oil (ATSDR 2000). The former CFC facility used oils containing PCBs as part of the on-site manufacturing process, and it was reported that there was use of significant quantities of Monsanto Pydraul 312 and 312A, the primary component of which is Aroclor 1242 (Tetra Tech 2001).
In general, PCBs are relatively insoluble in water. Sediments that contain PCBs can release the PCBs into the surrounding water, but the nature of the chemicals causes them to attach more strongly to soil particles rather than enter the water column. PCBs are taken up into the bodies of small aquatic organisms and fish, especially those fish that are bottom-feeders, and can accumulate through the food chain. They accumulate in the body fat and can enter breast milk. The most likely source of human exposure to PCBs is through the eating of contaminated fish, although PCBs also can be absorbed through the skin and via inhalation (ATSDR 2000b).
Oral Exposures: The floodplain soil and river sediment concentrations listed in Table 3 were first compared to the applicable clean-up standards. The MDEQ Direct Contact Criterion (DCC) for residential soil is 1 ppm (from Michigan's Part 201 Generic Cleanup Criteria and Screening Levels). This standard is meant to protect against the adverse health effects from direct skin contact with PCB-bearing soils. There is no analogous direct contact standard for sediment in Michigan; however, a remediation goal of 0.33 ppm was derived for this site by MDEQ (see Footnote 3). This particular standard was derived to be protective of aquatic organisms, but it defaulted to the MDL because it was just slightly below this MDL value for SW-846 Test Method 8082.
The range of "average" (central tendency) soil hotspot concentrations listed in Table 3 is 7.4 to 12.8 ppm, while the RME range is 10.6 to 35.3 ppm. The range of "average" sediment hotspot concentrations from Table 3 is 2.9 to 28.4, while the RME range is 5.4 to 55.2. In addition, a large number of individual sample points collected from river sediment exceed 0.33 ppm, both from within and outside of the original study area.
These ranges are well above any health-based benchmarks established for oral exposures, such as for incidental ingestion of river sediment and floodplain soil. The reference dose (RfD) established by EPA for chronic oral exposures to Aroclor 1254 is 0.00002 ppm per day (EPA 2003), which is identical to the chronic oral minimum risk level (MRL) established by ATSDR (ATSDR 2000). In addition, the PCB concentrations seen at the soil hotspots exceed the residential direct soil contact criterion of 0.22 ppm, as established in the EPA Region IX Preliminary Remediation Goals (PRG) table (EPA R9 2002).
The most probable and significant route of oral exposure to PCBs from the South Branch is through the consumption of contaminated fish and other game and livestock species collected from the river channel or floodplain soil. The presence of hunting blinds in the floodplain and anecdotal evidence regarding snapping turtle collection imply a larger potential exposure issue than merely fish. Terrestrial organisms grazing or living on the 100-year floodplain could be getting significant exposures at the soil hotspots.
The fish advisories issued by MDCH to the general population customarily use a trigger level of 2 ppm in fish tissue, which is assumed from the U.S. Food & Drug Administration (FDA). According to these same rules, sensitive subpopulations that eat one fish meal per week should not consume fish with more than 0.05 ppm. Furthermore, EPA guidance on risk-based fish consumption limits specifies that no fish meals per month should be consumed at certain PCB concentrations in fish tissue (EPA 2000). For non-cancer endpoints, this concentration is 0.38 ppm (wet weight). For cancer endpoints, this concentration is 0.094 ppm (wet weight).
FRG (1999) reported that white suckers collected from near the Bowen Road crossing (located near the transect 21 sediment hotspot) had over 2 ppm in their tissue in an area where PCB sediment concentration measured 0.72 ppm. It is likely that the sediment hotspot concentrations listed in Table 3 would lead to concentrations in fish much higher than 2 ppm, especially in such bottom-dwelling fish as the white sucker. An added complication for the fish pathway with regard to avoiding exposure within the river channel is the fact that a dam at the Shiawasseetown Reservoir prevents migration from the South Branch.
According to FDA and EPA fish consumption guidance, even fish concentrations of 1.7 to 2.56 ppm (as seen in the white sucker data from 1994; see Table 2) are 1 to 2 orders of magnitude above "do not eat" concentrations. Furthermore, this range of 1.7 to 2.56 ppm in fish tissue is related to sediment concentrations of 0.59 to 0.72, which again is differentiated by concentrations reported in Table 3 by 1 to 2 orders of magnitude.
Dose Estimation: On the basis of the data from FRG (1999), doses through the fish consumption exposure pathway were estimated through use of standard EPA methodology and exposure factors. Appendix C presents an overview of methods and assumptions used, in addition to presenting the doses estimates. The estimated doses were assumed to result from eating white sucker at the FRG (1999) contamination levels (i.e., from the 1994 site conditions of the Bowen Road bridge crossing on the South Branch). Adults from the general population could be expected to take in a dose ranging from 0.00022 to 0.00066 milligrams (mg) of PCBs per kilogram (kg) of body weight (BW) per day. [The former estimate is an average, or central tendency, while the latter estimate is an RME estimate, roughly analogous to the 95th percentile.] Adults from the recreational fisher population could be expected to take in a dose ranging from 0.0014 to 0.00044 mg/kg-day. Children of the general population (consumption rates for children of recreational fishers in Michigan could not be identified) could be expected to take in a dose ranging from 0.00061 to 0.000084 mg/kg-day.
Assuming a meal is 8 ounces of fish and that meal-sized portion contains 2.56 mg PCB per kg of fish tissue, this would translate to 0.00829 mg PCB per meal, or 0.000118 mg PCB per kg human body weight per meal (as consumed by the "average" adult from the general population).
All the estimated doses for both adult and children consumption scenarios exceed both the intermediate (0.00003 mg/kg-day) and the chronic (0.00002 mg/kg-day) oral MRLs, as defined by ATSDR for Aroclor 1254. In addition, the estimated amount of PCBs that would be ingested per 8 oz portion of fish (assuming it is contaminated at 2.56 mg PCB per kg fish tissue) is 0.000118 mg, also in exceedance of both the chronic and intermediate MRL values. This chronic oral MRL is based on feeding studies done with Aroclor 1254 and Rhesus monkeys (ATSDR 2000). A lowest-observed adverse effect level (LOAEL) of 0.005 mg/kg-day was identified from these studies, based on an immunologic endpoint of decreased antibody response. A total uncertainty factor of 300 was applied to this LOAEL to derive the chronic MRL (and coincidently, the EPA oral reference dose as well). All the estimated doses for both adult and children consumption scenarios exceed the LOAEL identified from the critical study that defined the chronic oral MRL.
Dermal Exposures: Some absorption through the skin might occur if a person were to swim or stand in the South Branch; however, it is difficult to estimate an absorbed amount. At the lower PCB concentrations found at most sites of environmental contamination, the chemicals tend to adhere to organic materials in the soil and thus tend to migrate through the skin less easily than pure PCBs or technical-grade PCB mixtures. MDCH reviewed several scientific studies to evaluate the potential for dermal uptake of PCBs from soil or sediment. These studies indicated that people who are dermally exposed to very high soil concentrations of PCBs tend to accumulate very little of the chemicals in their bodies.
These studies suggest that dermal exposure to PCBs in the river channel is not expected, by itself, to result in adverse health effects. However, prolonged exposure to high levels of PCBs via ingestion and skin contact, collectively, could result in adverse health effects.
Children may be at greater risk than adults from exposure to hazardous substances at sites of environmental contamination. Children engage in activities such as playing outdoors and hand-to-mouth behaviors that could increase their intake of hazardous substances. They are shorter than most adults and therefore breathe dust, soil, and vapors closer to the ground. Their lower body weight and higher intake rate results in a greater dose of hazardous substance per unit of body weight. The developing body systems of children can sustain permanent damage if toxic exposures are high enough during critical growth stages. Even before birth, children are forming the body organs they need to last a lifetime. Injury during key periods of growth and development could lead to malformation of organs (teratogenesis), disruption of function, and premature death. Exposure of the mother could lead to exposure of the fetus, via the placenta, or could affect the fetus because of injury or illness sustained by the mother (ATSDR 1998). The obvious implication for environmental health is that children can experience substantially greater exposures than adults to toxicants that are present in soil, water, or air.
Children living on or near the South Branch of the Shiawassee River and its adjacent 100-year floodplain may have access to both the floodplain and the river channel. Residences are located along the river at several locations, and more public access points to the floodplain of the river likely exist. Young children were identified at one residence located on the river during the May 2003 site visit. Because the former CFC facility is an NPL hazardous waste site, children living nearby are at increased risk of exposure to PCBs in soil (ATSDR 2000). Residential housing located in the floodplain may have detectable levels of PCBs in house dust, so that children living there may be at even greater risk of exposure. Children and teenagers with exposed skin playing in the floodplain were also identified during the May 2003 site visit. Finally, infants and small children can be exposed to PCBs both prenatally and through consumption of breast milk, for PCBs have been shown to accumulate in breast milk and to be transferable to babies and breast-feeding small children.
Before the 1982 state-sponsored remedial actions, all property owners along the river were notified by mail of the removal action. The public response to this mass mailing is unknown.
MDNR held an informal public meeting in July 1991 to discuss the ongoing RI/FS study. Similarly, MDEQ hosted a public meeting in September 1998 to discuss its proposed plan for remedial action, with no significant objections raised at the time.
Most recently, EPA held a public meeting in July 2001 and a public comment period from July to September 2001 for interested parties to comment on the proposed plan (as later published in the September 2001 Record Of Decision), which encapsulated the "final and only remedial action" for the Shiawassee River Site. A Responsiveness Summary was generated (per Superfund law) and included as an Appendix in the 2001 ROD. Very few health-related concerns were identified during the comment period and meeting. In fact, most recorded comments were related to the choice and extent of remedial action, to concerns over the SWAC method, and to the amount of time it may take for the river to return to a more natural state. One commenter questioned whether the range of the desired remediation goal (which was based on the results of an ecological risk assessment examining the mink as the primary receptor) was going to be protective of children and/or human fetuses. Many participants in the review process expressed an opposition to any remedial action at all.
Due to a lack of residential and non-residential exposure information for the South Branch of the Shiawassee River, the public health hazard is indeterminate. A need exists for more current fish tissue concentration data, because the most recent data are from 1987. In addition, data on the consumption of game or livestock species from the river channel and/or floodplain is needed to fully characterize the extent of the health hazard. Multiple data reports from the last few years have all confirmed the presence of residual PCB hotspots in both the river sediment and the floodplain soil of the South Branch of the Shiawassee River. Given the moderately high concentrations still found at certain hotspots and the reasonable assumption that other undetected hotspots in soil and sediment exist, people living in the vicinity of the South Branch could also be exposed to harmful levels of PCBs through incidental ingestion of soil and sediment from hotspots. To a lesser extent, exposure is possible via ingestion of river water and dermal exposure to soil and sediment hotspots.
Any citizen with additional information or health concerns regarding this health consultation should contact the Michigan Department of Community Health, Environmental and Occupational Epidemiology Division, at 1-800-648-6942.
Michigan Department of Community Health
Erik R. Janus, Toxicologist
Christina Bush, Toxicologist
Robin Freer, GIS Analyst
ATSDR Regional Representative
Mark Johnson
Regional Services, Region V
Office of the Assistant Administrator
ATSDR Technical Project Officer
Alan Yarbrough
Division of Health Assessment and Consultation
Superfund Site Assessment Branch
Agency for Toxic Substances and Disease Registry. 1988. Preliminary health assessment for Barrels, Inc, Lansing, Michigan. Atlanta: US Department of Health and Human Services; 1988 Sep 30.
Agency for Toxic Substances and Disease Registry. 1989. Health assessment for
Shiawassee River, CERCLIS NO. MID980794473 , Livingston and Shiawassee Counties,
Michigan. Atlanta: US Department of Health and Human Services.
Agency for Toxic Substances and Disease Registry. 1993. Site review and update:
Shiawassee River, vicinity of Howell, Livingston County, Michigan, CERCLIS NO.
MID 980794473 . Atlanta: US Department of Health and Human Services; 1993 Sep
(Revised August 1994).
Agency for Toxic Substances and Disease Registry. 1994. Site review and update:
Shiawassee River, vicinity of Howell, Livingston County, Michigan, CERCLIS NO.
MID 980794473 . Atlanta: US Department of Health and Human Services; 1994 Aug
(Revision of September 1993 report).
Agency for Toxic Substances and Disease Registry. 2000. Toxicological profile for polychlorinated biphenyls (PCBs). Atlanta: US Department of Health and Human Services.
Brown MP 1999. The role of natural attenuation/recovery processes in managing
contaminated sediments. Available through Sediment Management Work Group web
site, http://www.smwg.org/products/techpapers/paper2.pdf
ENTACT and Associates, LLC. 2003. Pre-design investigation report: Shiawassee River Superfund Site. Prepared for Johnson Controls, Inc. 2003 Aug.
ERG (Environmental Research Group, Inc.). 1980. Polychlorinated biphenyl investigation of Cast Forge Properties. 1980 Dec.
Fox River Group (FRG). 1999. Effectiveness of sediment removal: an evaluation of EPA Region 5 claims regarding twelve contaminated sediment removal projects. Submitted to the National Academy of Sciences. 1999 Sep.
Gannett Fleming of Michigan, Inc. 2003. Final work plan for the sediment survey of the South Branch of the Shiawassee River, Howell, Michigan. Project # 40229.000. 2003 Jan.
Gradient Corporation. 2000. Evaluation of recent PCB data in the Shiawassee River. 2000. 14 Apr.
Indiana State Department of Health. 1992. Final report on community exposure to polychlorinated biphenyls, Bloomington, Indiana. 1992 Jun.
Indiana State Department of Health. 1997. Draft health consultation for Shelly Ditch, Crawfordsville, Indiana. 1997 Jun 16.
Michigan Department of Community Health. 1997a. Michigan Department of Community Health files–Barrels, Inc.
Michigan Department of Community Health. 1997b. Michigan Department of Community Health files–Lakewood Study.
Michigan Department of Community Health. 2001. Memorandum dated 20 September 2001 from Dr. Linda Larsen (MDCH) to Janet (Sunny) Krajcovic (MDEQ). Subject: Shiawassee River Superfund Site.
Michigan Department of Community Health. 2003. 2003 Michigan family fish consumption guide.
Michigan Department of Environmental Quality. 2003. Michigan Fish Contaminant Monitoring Program 2002 annual report. 2003 Feb.
Michigan Department of Natural Resources. 1977. Fish and sediment survey of the South Branch of Shiawassee River. Staff report, Biology Section, Water Quality Division. August and October 1977.
Michigan Department of Public Health. 1987. Warner SC, Williams D, MDPH, Center for Environmental Health Sciences. Health effects of residential exposure to polychlorinated biphenyls: a study of two Detroit neighborhoods, final report.
Michigan Department of Public Health. 1992. Interim preliminary health assessment
for Carter Industrials, Incorporated, Detroit, Wayne County, Michigan, CERCLIS
No. MID980274170 . Report prepared for Agency for Toxic Substances and Disease
Registry. 1992 May 20.
Malcolm Pirnie Engineers. 1997. South Branch Shiawassee River Superfund Site, remedial alternatives feasibility study, volumes I and II. 1997 Dec.
Tetra Tech EM, Inc. 2001. Final supplemental feasibility study report for Shiawassee River Site, Howell, Michigan.
U.S. Environmental Protection Agency. 2003. Integrated Risk Information System
entry for Aroclor 1254 (CASRN 11097-69-1). Available at: URL: http://www.epa.gov/iris/subst/0389.htm .
U.S. Environmental Protection Agency, Region IX. 2002. Preliminary remediation
goals table for 2002. Available at: URL: http://www.epa.gov/region09/waste/sfund/prg/index.htm .
U.S. Environmental Protection Agency. 2002. Child-specific exposure factors handbook. Interim report. EPA-600-P00-002B. 2002 Sep.
U.S. Environmental Protection Agency. 2001. EPA Superfund record of decision:
Shiawassee River, Howell, MI. EPA ID: MID980794473 . EPA/ROD/R05-01/528. 2001
Sep 28.
U.S. Environmental Protection Agency. 2000. Guidance for assessing chemical contaminant data for use in fish advisories. Volume 2: Risk assessment and fish consumption limits. 3rd Edition. EPA 823-B-00-008. 2000 Nov.
U.S. Environmental Protection Agency. 1997. Exposure factors handbook (Volume I: General factors and Volume II: Food ingestion factors). EPA-600-P95-002F. 1997 Aug.
Warzyn, Inc. 1992. Remedial investigation report, South Branch of the Shiawassee River, Howell, Michigan. Prepared for MDNR; 1992 Jan.
Figure 1. South Branch Shiawassee River
Figure 2. Study Area Portion of Shiawassee River
Next Section Table
of Contents
