HEALTH CONSULTATION

BENDIX CORPORATION/ALLIED AUTOMOTIVE
ST. JOSEPH, BERRIEN COUNTY, MICHIGAN

FIGURES

Figure 1. Bendix Corporation

Figure 2. Residential Well and Monitor Well

TABLES

Table 1. Concentrations of Volatile Organic Chemicals found in groundwater samples from monitoring wells on and near the Bendix site, 1981-1994.

ND - Not Detected (for medians, not detected in 50% or more of the samples.)

NR -Reported data not sufficient to determine medians.

— =Not Sampled or Not Analyzed for.

Note: References 3 and 5 only list data for 1,1-dichloroethane, 1,2-dichloroethane, trans-1,2-dichloroethylene, trichloroethylene, and vinyl chloride for samples collected between 1984 and 1988. Only the 1981 data gives separate concentrations for cis-1,2-dichloroethylene.

Table 2. Volatile Organic Chemicals found in sludge and soil samples collected from the Bendix site, 1982, 1984, 1994.

ND - Not Detected

NA -Not Analyzed for

Table 3. Volatile Organic Chemicals found in groundwater samples from temporary monitoring wells in the source areas at the Bendix site, 1994.

Reference: 1

Table 4. Volatile Organic Chemicals found in groundwater samples from residential wells 1984-1997 

ND - Not Detected

NA -Not Applicable

BQL -Below Quantification Limit

RESPONSIVENESS STATEMENT

The MDCH released a draft of this Health Consultation for public comment on April 8, 1998. The comment period lasted until May 8, 1998. A representative of Bosch Braking Systems wrote to MDCH with comments on the draft, presented below with responses from MDCH. The pages and paragraphs cited are those in the draft Consultation reviewed and may not agree with the current draft of the Consultation.

General Comment

(1) "Potential human health evaluations in the Draft Health Consultation are based on groundwater monitoring well data instead of residential well data, and on estimates that residents had probably been exposed for at least 10 years. There are no data to support these overly conservative and unlikely exposure scenarios."

Response: When there are no data available on water from a residential well, MDCH believes that data on water from nearby monitoring wells are acceptable as a substitute. The results from the monitoring well may be different from those from the residential well, for example, the two wells may tap different aquifers or different layers in a strongly stratified aquifer, but in the absence of positive evidence of such a situation, it is an appropriately conservative assumption to use the monitoring well data. We do agree there is no data on the development of the plume, since the earliest available data, from 1985, showed it fully developed to Lake Michigan, as we stated in the Consultation. A slowly developing plume, taking ten years or more to extend from the area of the residential wells to the lake, is consistent with the available information on the subsurface soil and modeling of the transport of contaminants through the groundwater. In the absence of contrary data or information, we believe it is likely that residents of the area may well have been exposed to groundwater contaminants for 10 years or longer.

(2) "In addition, there should be further discussion of the uncertainties and limitations of the data and human health exposure information used in this report. This information should be included in the conclusion section of the Health Consultation to allow the reader to assess the inherent uncertainties associated with these assumptions."

Response: MDCH respectfully disagrees. The Consultation adequately, for its purposes, discusses the uncertainties and limitations of the data and human health exposure information used within it. The discussion of such matters does not properly belong in the conclusion section.

Specific Comments

Comment 1 - Page 1 para 1: The phrase "and several other residential wells.." should be deleted because there are no data to indicate that other wells may have been contaminated. In addition, "some years" should be deleted because there are no data available to indicate that there has been an exposure or the duration of the presumed exposure.

Response: MDCH agrees that there is no data stating that other residential wells were contaminated or that documents any human exposure to any great extent. However, as laid out in our response to the first part of the "General Comment" above, we believe that the data and information presented in this Consultation supports our conclusion that other residential wells may have been contaminated and that the contamination may have been present for some years before it was discovered.

Comment 2 - Page 2, para 1: Please clarify the text: "to several closed lagoons in the southeast corner of the property." The RI/FS investigations identified three source areas of groundwater contamination: the former south lagoon (not a major source of VOCs to the groundwater), the loading dock area, and the north parking lot. Please correct the dates the lagoons were in use to; "between 1965 and 1975."

Response: The section has been modified as suggested.

Comment 3 - Page 5, para 3: "Water collected from a monitoring well on property between RW-1 and theirs (OW-29) in March 1985 contained as much as 9,950 ppb of five VOC contaminants of concern (1,1-dichloroethane, 1,2-dichloroethane, trans-1,2-dichloroethylene, trichloroethylene, and vinyl chloride) ..." is misleading and should be deleted from the text or additional discussion provided clarifying the difference between actual residential well data and separate monitoring well data. As written, it implies that exposure to 9,950 ppb VOCs was possible. This is not accurate. A comparison of the concentrations detected in RW-1 to OW-29 show that the vinyl chloride and TCE in RW-1 are two orders of magnitude lower than those observed in OW-29. This health evaluation should be based on realistic exposures.

Response: As stated in our response to the General Comment (1), in the absence of data from a residential well, MDCH considers it perfectly reasonable to use data from nearby monitoring wells. We do recognize that, as pointed out here, the concentrations in residential wells may be significantly different from those in nearby monitoring wells. Indeed, adjacent monitoring wells may contain vastly different concentrations of contaminants if they are at different depths. However, there is no information available to indicate that the factors responsible for the differences between the concentrations in RW-1 and OW-29 are also in effect between RW-6 and OW-29. We believe that it is a reasonable, conservative assumption to base our health evaluation on the worst plausible case.

Comment 4, Page 5, para. 4: Please correct the dates the lagoons were in use to; "between 1965 and 1975", and remove the last sentence of the paragraph. It is not possible to accurately assess the year that the plume left the plant property.

Response: The dates the lagoons were in use has been corrected. We agree that it is not possible to accurately date when the plume left the plant property. We believe that our estimate of the minimum time that people might have been exposed to the contaminants is reasonable.

Comment 5, Page 6, para 2: The use of the maximum contaminant concentrations detected in the northwest plume monitoring wells as the exposure point concentration is unrealistic and overly conservative. Such an exposure scenario is not probable. This is supported by the results of the RI that show the concentrations detected in OW-29 to be two orders of magnitude greater than those detected in RW-1. Assuming exposure to the concentrations detected in OW-29 greatly exaggerates the potential health hazards. Potential exposure should be evaluated based on the concentrations detected in the RW-series.

Response: The assumption is conservative but is not unrealistic or improbable. We have no sampling data from RW-6, the private well closest to the center of the plume, or well construction information to determine whether the well is screened within the contaminant plume. There is no information available on which to base a decision on whether the relative concentrations between RW-6 and OW-29 would be similar to those between RW-1 and OW-29

Comment 6, Page 7, para 1: The term "moderately increased" is used repeatedly to discuss the risks associated with potential exposure to VOCs. The use of "moderately increased" should be defined and discussed in terms of the uncertainties associated with the evaluation. Because the health evaluations presented in this report are based on the extremely conservative and unlikely exposure assumption, any conclusion based on this evaluation are likely to greatly overestimate potential risks.

Response: ATSDR practice is to describe the increase of cancer risk attributable to an exposure to a carcinogen in the environment on a qualitative scale that includes "moderately increased" as one level. The assumptions of exposure involved in each health statement in this Consultation are either explicitly or implicitly included within the statement.

Comment 7, Page 8, para 3: The discussion of potential risks associated with inhalation of vinyl chloride during showering should be modified. The evaluation can not be made based on the concentrations of vinyl chloride detected in the groundwater alone. Some type of air modeling is required to estimate potential exposure and risk because there is no actual indoor air sampling data.

Response: J.C. Little, in Reference 21 cited in the referenced paragraph, presents just such a model. We have raised the visibility of the reference to Little in the paragraph.

Comment 8, Page 9, para 1: The discussion of potential inhalation exposure from the northeast plume should be deleted. Based on available site data including chemical data (e.g., soil, soil gas, and groundwater), geologic information, and exposure modeling, Bosch does not believe that inhalation of VOCs migrating from the groundwater through subsurface soils is a likely route of exposure.

Response: MDCH and MDEQ respectfully disagree, pending additional data and information from the forthcoming study.

Comment 9, Page 9, para 2: Plume contaminants have not been detected in the lake water or beach areas, therefore, there is no exposure to recreational users. This should be included in the Conclusion and Summary section of this report.

Response: Statements to that effect have been added.

Table of Contents