Woodward-Clyde wEngineering & sciences applied to the earth & its environment

December 17, 1996 Submitted Via Federal Expressand Certified U.S. Mail

Ms. Ginny NarseteCommunity Involvement CoordinatorOffice of Public Affairs (P-19J)U.S. EPA77 West Jackson Blvd.Chicago, EL 60604

Subject: Review of Proposed Plan Summary: Floodplain/Wetland AreaFields Brook Superfund SiteAshtabula, OhioProject No. 86C3609S

Dear Ms. Narsete:

On behalf of the Fields Brook Action Group (FBAG), Woodward-Clyde Consultants (WCC)is submitting written comments to the Proposed Plan and other documents provided in theInformation Repository. The comments provided in this submittal have been prepared and/orreviewed by legal and technical representatives of the participating companies and consultantswho currently are retained by FBAG (Woodward-Clyde Consultants, Gradient Corporation,PTI Environmental Services, Bechtel, and de maximis, inc.).

If you have any questions regarding this submittal, please do not hesitate to contactMr. Joseph Heimbuch, de maximis, inc. at (941) 365-8444 or me at (216) 349-2708.

Sincerely,

Martin L. Schmidt, Ph.D.Senior Project Manager

MLS:kmc

cc: Joseph Heimbuch, de maximis, inc.FBAG Technical CommitteeFBAG Steering CommitteeDavid Ludwig, PTIChuck McConnel, BechtelTony Wolfskill, WCC

Woodward-Clyde Consultants • A subsidiary of Woodward-Clyde Group, Inc.30775 Bainbridge Road, Suite 200 • Solon, Ohio 44139216-349-2708 • Fax 216-349-1514

L:\337FB\86C3609S\NARS1217.DOai7-Dec-96\SOL

C O M M E N T I E S P O H S E I I P O I T

USEPA ADMINISTRATIVERECORD REPOSITORY FILESOU-1 AND OU-4

Prepared forFields Brook Action GroupAshtabula, Ohio

December 17,1996

Woodward-Clyde ^

30775 Bainbridge RoadSuite 200Solon, Ohio 44139216/349/270886C3609S

Proposed Plan \

Comments On OSEPA Proposed Plan

GENERAL COMMENTThe Proposed Plan provided by USEPA accurately summarizes the alternatives described in FWAFS prepared by FBAG and submitted to USEPA in October 1996. The Proposed Planrecommends Alternative VII as the preferred remedy for the FWA. Since this alternative is theFBAG preferred alternative, there appears to be complete agreement as to remedy selection forthe FWA.

The FBAG is not in agreement with the FWA delineation sampling program as described in theProposed Plan. FBAG described a delineation program to USEPA that involved furtherdelineating existing response areas developed in the FS. FBAG did not propose to collectadditional samples at fifty-foot intervals in the entire FWA.

SPECIFIC COMMENTS

Page 5 Summary of Alternatives • bottom of 2nd paragraph and

Page 8 D) Remedial Activity Locations • middle of second paragraphThe text provided in the Proposed Plan Summary of the Fields Brook Superfund SiteFloodplain/Wetland Area indicates that soil sampling data will be collected at fifty-foot intervals inthe floodplain during delineation sampling programs. This could be interpreted to mean acrossthe entire Floodplain/Wetland Area. FBAG has concerns that USEPA has misinterpreted theintent of the delineation sampling program discussed during the September 26, 1996 meeting andas described in Section 4.5 of the FWA FS provided to USEPA in October 1996. As a generalcomment, FBAG does not intend to collect soil samples in the entire FWA at fifty-foot intervalsacross the entire FWA. FBAG would like to specifically point out the following items related tothe proposed design delineation sampling.

• During earlier phases of the investigation more than 250 floodplain samples have beencollected in the five FEUs. These sampling programs were approved by USEPA; in order tocharacterize the FWA, develop remedial alternatives and satisfy Remedial Investigation (RI)objectives regarding the nature and extent of the 11 COCs. The previous sampling programswere based on systematic sampling along transects and spatially distributed so as to satisfyUSEPA concerns regarding definition of locations of COC concentrations on both sides ofFields Brook. Grid sampling across the entire FWA is not necessary. It is a waste of financialresources.

• If grid sampling were to be considered, it should have occurred in 1993-1994 when the RIactivities were in progress and not at the design phase of the project and would have replacedthe systematic sampling program (which was approved by USEPA and implemented byFBAG).

L:\337FB\86O3609S\124CMRSP.DOC\17-O6C-9WSCM. 1

Comments On USEPA Proposed Plan

• Since the FS has been completed based on the approved systematic sampling program, andresponse areas identified, it is appropriate at this time to conduct a delineation samplingprogram to support the final FWA design. The purpose of the delineation sampling programis to properly define the response areas that have been developed in the FWA FS for coverand excavation and substantially reduce the uncertainty that any significant contaminated areashave been missed. The following text was contained in the FWA FS to describe thedelineation sampling program.

"The isoconcentration lines that have been drawn based on currently available datahave been used to develop the potential remedial response areas presented in this FS.Figures 2-3a through 2-12 illustrate the isoconcentration mapping for PCBs andhexachlorobenzene. In order to establish cutlines that can be used during integration ofthe FWA remediation with the SOU activities, a design delineation program is plannedto be implemented. The delineation sampling will involve collection of additional soilsamples in both residential and industrial FEUs to further delineate current responsearea boundaries and sample in areas where data points are widely spread. A samplinggrid in these areas consisting of approximately 50 foot grid lines will be used toestablish sampling locations. Samples will be selected to refine existing response areasand fill in data needs. Soil samples will be analyzed for PCBs in all five FEUs andhexachlorobenzene in selected response areas in FEUs 4 and 8. "

In reviewing the information provided during the September 26, 1996 meeting with USEPA, it isFBAG's position that an agreement was reached regarding delineation sampling and a consistentlevel of post-remediation long-term monitoring requirements. The delineation sampling programproposed by FBAG at that time consisted of approximately 200 samples located within the FWA.A fifty-foot grid would be used to identify possible sampling locations, in order to better defineresponse boundaries, fill in data gaps, and to develop cutlines for use in preparing designdrawings. USEPA requested that sampling maps be submitted prior to field work in order toreview proposed sampling locations. As described in the sampling plan submitted to USEPA onDecember 2, 1996, the number of delineation samples was increased from 200 to 248 soilsamples. The delineation program and O&M sampling proposed by FBAG, supplement existingdata and are sufficient and protective for the remedy selected.

L:\337FB\86C3609S\124CMRSPDOC\17-Dec-96\SOL

___________________Comments On USEPA Proposed Plan

Page 7 B) Landfill RequirementsThe description of the on-site landfill should be referred to as the on-site consolidation area, to beconsistent with the FS text and language used to describe this area by FBAG. During theSeptember 26, 1996 meeting with FBAG and USEPA, the on-site consolidation area wasdescribed as a storage area that would contain materials from the following areas of the FieldsBrook site:

• Soil excavated from FWA response areas.

• Sediment excavated from Fields Brook as part of Sediment Operable Unit (SOU)remediation

• Excavated haul road material from construction of FWA and SOU remedial areas

MfciutliBMil Clyle L\337FB\B6C3609S\124CMRSP.DOC\17-Dec-96\SOL 3

RI Report

Comments On Remedial Investigation Report For TheHolds Rrooknoodplaln/Wedind Area EPA

GENERAL COMMENTThe Remedial Investigation Report for the Floodplain/Wetland Area (FWA) provides a generaloverview of the site chronology, activities completed and analytical data collected. There aresome minor inconsistencies that should be noted regarding the following: use of the termoperable unit for the FWA, description of FEUs (Floodplain Exposure Units) vs. EUs (SedimentExposure Units), and analytical parameters obtained during FWA studies. Overall, the reportreferences and summaries information provided in deliverables prepared by Gradient, EAEngineering and Woodward-Clyde.

SPECIFIC COMMENTS

Page 1-1:The text notes that USEPA has divided the site into four areas of concern, three of which areconsidered "operable units" associated with the Fields Brook Superfund Site. The text incorrectlydescribes the operable units. The 1986 SOW originally described two operable units, theSediment Operable Unit (SOU) and the Source Control Operable Unit (SCOU). The Floodplainand Wetland Area (FWA) has never been officially designated or referred to as an operable unitby FBAG. The FBAG voluntarily proceeded with FWA sampling, risk assessments and feasibilitystudy alternatives beginning in 1994. Figure 1 does not accurately depict the SOU and FWAinvestigation areas. The Source Control Operable Unit is not shown.

Page 2-3:Section 2.1.2 discusses SCOU activities and indicates that all field work was completed in the Fallof 1995. The text does not reference the completion of a Feasibility Study that identified remedialalternatives that were developed for the remaining source control areas that represented potentialrecontamination sources to Fields Brook.

Page 2-4:Section 2.1.4 refers to the Floodplain/Wetlands Area Operable Unit. This reference is incorrect.The text does correctly indicate the FBAG voluntarily conducted the investigation.

Page 3-2:The text describes groupings of FEUs for residential and industrial land use areas. There appearsto be some confusion regarding actual designation of FEUs (floodplain exposure units) and EUs(sediment exposure units). The listing provided describes exposure units (EUs) for the sediment

WhNMlWMll CljUle W L:\337FB\B6C3609S\124CMRSP.OOCM7-Dec-ge\SOL 1

Comments On Remedial Investigation Report for TheHolds Brook Hoodplaln/Wetland Area EPA

operable unit and incorrectly describes the residential and industrial EUs. The designation breakfor residential and industrial FEUs in the FWA was Route 11 and in the SOU was State Road. Inthe SOU, residential EUs included: Reach 1, 2-1, 2-2, 3, 4, 11-1, 11-2, 5-1, and 5-2. and inindustrial EUs included 11-3,1-4,6,7-1, 8-1,8-3, 8A, 13-1, 13-2, and 13A.

It should be noted that the FWA project only considered response areas in residential areas (FEU2 and 3) and industrial areas (FEW, 6, and 8).

Page 3, section 3.3, Figures:The figures provided in the RI (Figure 3 through 7) that identify Phase I, II, and III FloodplainSample Locations depict the proposed locations as shown in the Work Plan. They do not showactual surveyed locations and minor sample numbering modifications that occurred during fieldsampling. Figures showing final sampling locations that were surveyed to exact locations wereprovided in the October 1996 Feasibility Study.

Page 3-4:In the second full paragraph, the text indicates that all samples were analyzed for the 130chemicals. This is not correct. The 137 samples collected during Phase III were only analyzedfor the 11 COCs identified for the FWA (1,1,2,2 tetrachloroethane, tetrachloroethene,trichloroethene, vinyl chloride, benzo(a)pyrene, hexachloroethane, hexachlorobenzene,hexachlorobutadiene, arsenic, beryllium, and PCBs).

L:\337FB\B6C3609SM24CMRSP.DOC\17-Dec-96\SOL 2

Appendix AComments On Remedial Investigation Report For The

Fields Brook Floodplaln/WeUand Area EPA

The source of the summary analytical data tables included in this appendix was not referenced.FBAG provided USEPA with electronic files. It is assumed that data provided in these summarytables is consistent with the FBAG database.

L:\337FB\86C3609S\124CMRSP.DOC\17-Dec-96\SOL 1

Baseline Risk Assessment

Comraeits On Baseline Hnman Health Risk Assessment Report________________fields Brook Floodplaln/Wedand Arei EPA

GENERAL COMMENTSThe FWA risk assessment follows a previous risk assessment done for the Fields Brook SOU,where PCB cancer risks were assessed with a cancer slope factor of 7.7, and PCB non-cancerrisks were assessed with an RfD of 1 x 10"4. At the time of the SOU risk assessment, IRIS listeda CSF of 7.7, while no RfD for PCBs was given. The RfD used in the Fields Brook SOU riskassessment was recommended by EPA Region V, and was based on a study of aroclor 1016 inrhesus monkeys. Many PCB studies have been completed since the SOU risk assessment, andIRIS now lists a revised CSF of 2.0 (high end of a recommended range). There remains no RfDfor PCBs, however ATSDR recommends a minimum risk level (MRL) of 2 x 10"5, based onaroclor 1254. The change in the CSF has the effect of reducing estimated cancer risks, while thechange in the RfD has the effect of increasing estimated noncancer risks. EPA does not considerthe revised CSF in calculating cleanup goals, which would have decreased risks and raised thecleanup goal, but they do consider the revised RfD, which increases risks. EPA does this despitethe fact that the new CSF is listed on IRIS, while significant doubt remains about the RfD. As aresult, the FWA risk assessment contains an unfair bias because EPA has failed to accept a validscientific basis for a change in the toxicological factor if it leads to a decreased risk.

SPECIFIC COMMENTS

Page 3-5:The EPA risk assessment is overly-critical of the exposure survey conducted by the PRPs, andstates that "...the survey did not focus upon those individuals who reside along Fields Brook andthus does not accurately represent the population whose exposure to the Brook is most ofconcern." This is not the case. The survey area is bounded by portions of East 14th Street,Columbus, East 15th Street, Route 11, East 23rd Street, and East 21st Street. Of the 100households that were surveyed within this area, 40 of them are on property directly adjoiningFields Brook. The other 60 households, while not adjoining Fields Brook itself, include thoseacross the street from the Brook, and within this small defined area. EPA's assertion, that thesurvey did not focus on households likely to be exposed to the Brook, is false.

L:\337FB\86C3609S\124CMRSP.DOC\17-Dec-96\SOL 1

Comments On Baseline Human Health Risk Assessment Report________________Holds Brook nee Jplaln/Wetland Area EPAPage 4-3:Section 4.3.1 concerning toxicity values for PCBs states that the cancer slope factor is 2.0. Whilethis is a true statement, it is misleading because EPA used the old cancer slope factor of 7.7 toderive the cleanup goals.

Page 4-4:The text (pgs. 4-4 and 6-5) notes that endocrine disrupter effects "may be best modeled as linearand non-threshold," and that they "could theoretically be calculated in a manner similar to thecancer slope factor." We disagree. Although the precise mechanisms of endocrine disruption arenot well known at the present time, the dose-response relationship for PCB endocrine disruptereffects is likely to have a threshold, similar to developmental and systemic effects induced viaother mechanisms. Linear, non-threshold models are only appropriate for compounds that arethought to induce cancer by directly causing DNA mutations; endocrine disruption is not believedto involve mutagenesis.

Page 4-5:Section 4.3.8 concerning arsenic toxicity cites a cancer slope factor of 1.75. This value is out-of-date. The correct value of 1.5 is given in Table 5. It is unclear which value was used in the riskcalculations.

Page 5-3:PCB cancer risks were inappropriately calculated using the former cancer slope factor of 7.7(mg/kg-day)"1. It was acknowledged in the risk assessment that a new range of cancer slopefactors has recently been established for PCBs, but these new values were not used. No scientificjustification was provided to explain why the former cancer slope factor was chosen instead of therecently reassessed value.

In the risk assessment report, it is noted that the former cancer slope factor of 7.7 (mg/kg-day)"1

was based on studies using Aroclor 1260. The recent PCB reassessment was described as a "jointconsideration of cancer studies and environmental processes [...] supported by severalcomplimentary sources," perhaps implying that the reassessed values are less valid. In fact, theupper bound slope of 2.2 (mg/kg-day)"1 calculated in the recent PCB reassessment is based on thesame rat study (Norback and Weltman, 1985) as the former value of 7.7 (mg/kg-day)"1; thereassessment simply corrected three errors that had been made in the development of the 7.7value. Specifically, the revised value took into account a 1991 reevaluation of the tumorhistopathology data, a new EPA cross-species scaling factor, and initial instead of time-weighteddoses.

L:\337FB\86C3609S\124CMRSP.DOCM7-Dec-96\SOU 2

Comments On Baseline Unman Health Risk Assessment Report________________Fields Brook Fioodplaln/Wetland Area IPflThe statement "There is some evidence that mixtures containing more highly chlorinated biphenylsare more potent inducers of hepatocellular carcinomas in rats than mixtures containing lesschlorine by weight," (pg. 5-3 of the risk assessment) was an early hypothesis, but has not beensupported by more recent data. Based on information summarized in EPA's 1996 PCBreassessment report, there is little if any evidence of a consistent relationship between chlorinationlevels and cancer potency. Aroclor 1016 (41% chlorinated) is much less potent than Aroclor1242 (42% chlorinated), and analytical methods cannot reliably distinguish the two mixtures inenvironmental samples. Also, Aroclor 1254 (54% chlorinated) is more potent that Aroclor 1260(60% chlorinated) in female rats in the recent carcinogenicity study sponsored by GeneralElectric. It is likely that the cancer potency is far more dependent on the concentrations ofparticular congeners. Unfortunately, the relative importance of each congener is only beginningto be characterized.

Page 6-3:The text incorrectly states that "The FBAG proposed that the size of the FWA exposure units belarger than those used in this risk assessment; that is, they proposed that a smaller number ofexposure units be demarcated within the FWA for evaluation." This statement is false. The FBAGinitially proposed that the FWA exposure units correspond to the SOU exposure units, and EPAaccepted this proposal.

Page 6-4:There are several problems with the discussion of uncertainty associated with the PCB cancerslope factor on page 6-4 of the risk assessment. A significant problem is that only the draft PCBreassessment (January, 1996) is discussed. Since several changes were made in the finalreassessment report (September, 1996), this discussion should be updated to reflect EPA's finalrecommendations. For example, while the draft reassessment recommended a specific adjustmentto the exposure duration to reflect the persistence of PCBs in the body, the final reassessmentrecognized potential concerns about persistent PCB congeners but did not recommend aquantitative adjustment. Also, it is incorrectly noted on page 6-4 that "the new documentdescribes PCB cancer potency using two ranges ..." [emphasis added]. In fact, while the draftreassessment recommended two ranges, the new, final reassessment recommended a total of sixvalues - both a central and upper-bound slope value for three reference points (" high risk," " lowrisk," and " lowest risk" ). Lastly, it is incorrectly noted in the risk assessment report that thedraft document included guidance on the application of a cross-species scaling factor. While boththe draft and final reassessment reports incorporated a cross-species scaling factor when modelingslope estimates, it was not recommended that risk assessors apply an additional cross-speciesscaling factor when assessing PCB cancer risks.

L\337FB\B6C3609S\124CMHSP.DOCV17-Dec-96BOL 3

Comments On lisellne Human Health Risk Assessment Report________________fields Brook Floodplaln/Wetland Area EPAPage 6-5:In the risk assessment, it was stated that " It is likely that the risk calculated from either of thecancer slope factor underestimates actual risk from the site". We strongly disagree with thisstatement. While there are numerous sources of uncertainty in PCB risk estimates (includingendocrine disrupter effects), numerous conservative assumptions are used so that the resultsrepresent upper bound risk estimates. For example, it is noted in the Risk Assessment Guidancefor Superfund (RAGS) (EPA, 1989) that cancer risk estimates based on linear cancer models andan upper 95th percentile confidence limit "will generally be an upper bound estimate. This meansthat EPA is reasonably confident that the true risk will not exceed the risk estimate derivedthrough use of this model and is likely to be less than that predicted." This fact is even noted inTable 7 of the EPA Fields Brook risk assessment.

Table 3:This table contains typographical errors; it is not clear to what extent, if any, these errors affectthe risk calculations. The errors include:

• FEU2, maximum concentration of PCE should be 2500 |ig/kg• FEU2, maximum concentration of 1,1,2,2-Tetrachloroethane should be 580 Jig/kg

• FEU2, basis for exposure point concentration for PCBs should be listed as 95% UCL

• FEU8, maximum concentration of hexachlorobenzene should be 480,000 |J.g/kg• FEUS, 95% UCL on mean and exposure point concentration for hexachlorobutadiene

should be 1900u£/kg.

Table 7:We have several concerns with the discussion of general uncertainty factors in Table 7. It ismisleading to represent many of these uncertainty factors as "over- or underestimating risks,"implying that each possibility is equally likely, when in fact, they are far more likely tooverestimate risks. By definition, toxicity factors (slope factors and reference doses) are designedto err on the side of overestimating risks. For example, even though it is theoretically possiblethat humans could be more sensitive to PCBs than animals, an uncertainty factor is incorporatedinto toxicity values to specifically account for this possibility. Furthermore, despite large numbersof PCB-exposed workers, PCBs have not been causally associated with cancer in humans basedon epidemiology data. Another example is linear extrapolation from high doses to low doses.This approach will only underestimate risks if the dose-response relationship is superlinear (whichhas never to our knowledge been suggested for PCB carcinogenicity).

WOMlMMfll Clyde W L:\337FB\86C3609S\1Z4CMRSP.DOC\17-Dec-96\SOL 4

Appendix AFirst Document Entitled "Development Of Clean-Up Goal (CUG) Concentrations

________________fields Breoh floedmaln/Wedand Area EPAPage 3 bottom through page 4 top:This text states that CUGs were developed for the FWA by considering only the soil ingestionexposure route; that dermal exposure was not considered in developing CUGs. This statement isfalse. Both the FBAG and the EPA risk assessments considered both soil ingestion and dermalexposure, and both were used in EPA's development of the CUGs.

page 5 top:This section appears to state that CUGs were developed for the FWA in 1993. However, the riskassessment was not done until 1995 (FBAG risk assessment) and 1996 (EPA risk assessment),and this 1996 EPA risk assessment is the first document that contains CUGs.

L:\337FB\86C3609S\124CMRSP.DOai7-Oec-96aOL 1

Ecological Risk Assessment

Comments Regarding Baseline Ecological Risk Assessment ReportHolds Brook Floodnlaln/Wetland Area EPA

EXECUTIVE SUMMARY

Paragraph 1, last sentence:FBAG identified "ECUGs" only for the convenience of USEPA. FBAG believed, and continuesto believe, that such goals are inappropriate objectives for risk management. Risk reductionobjectives, rather than simplistic environmental media cleanup goals, are the most appropriate riskassessment standards for risk management.

Paragraph 2, sentence 4:The FBAG ecological risk assessment employed appropriately conservative assumption andapplied reasonable input values based on site-specific data. These assumptions and values werecarried through appropriately in the conclusion of the USEPA Region V baseline ecological riskassessment. In contrast, USEPA-Edison risk assessment applies values which are technicallyinappropriate and which are not conservative, they are simply wrong. For example, USEPA-Edison assumed that maximum contaminant levels represented exposure, that area use factorswere universally 1, and that chemicals were 100% bioavailable. Such assumptions are onlyjustifiable in a screening level risk assessment according to guidance (USEPA 1994), and are nottechnically sound for more site-specific and comprehensive risk assessment. More complete andcomprehensive risk assessment under the guidance is to be based on site-specific data andreasonable assumptions. Such data and assumptions were employed in the FBAG ecological riskassessment as reflected in the EPA Region V baseline risk assessment.

INTRODUCTION

Page 1, paragraph 1, first bullet:The FBAG risk assessment is not included in this report. The FBAG document was revised byUSEPA, and USEPA should be considered the author of the revised risk assessment.

Page 1, paragraph 1, third bullet:FBAG did not "develop" ecological clean-up goals, and does not believe that such goals areappropriate for risk management. FBAG provided ECUGs based on risk assessment findings atthe request and for the convenience of USEPA.

L:\337FB\86C3609S\124CMHSP.DOC\17-Dec-96\SOL 1

Comments Regarding Baseline Ecological Risk Assessment Report________________Holds Brook Floodplaln/Wetland area EPA

Page 2, paragraph 4, first sentence:The baseline ecological risk assessment report submitted by FBAG in October 1995 is NOT"provided as Attachment I". Attachment I is a report prepared by USEPA, incorporatingnumerous revisions to an electronic text file provided by FBAG.

Page 2, last paragraph, bullets ft to page 4:The second bullet identifies "probabilistic statistical approach" as an "unresolved issue". Inpractice, and according to draft agency-wide ecological risk assessment guidance (USEPA 1996),probabilistic exposure estimates are technically sound and most appropriate. For anotherexample, the fourth bullet identifies use of fish tissue from upstream locations to estimate"current" risks to mink and heron. In fact, such a use was never made of these data. The FBAGwas committed to remediation of the brook sediments prior to conducting the ecological riskassessment. Therefore, species potentially exposed via brook components of the overallecosystem in the context of the floodplain wetland would only be exposed to post remedyconcentrations. The floodplain wetland ecological risk assessment has no role in determiningcleanup criteria for brook sediments.

Page 3, first full paragraph, last sentence and elsewhere:Repeated characterization of the database employed by USEPA for ecological risk assessment as"questionable" is unnecessarily vague.

Throughout the report, USEPA should identify clearly and quantitatively the weaknesses believedto exist in the information base and state the effect of these weaknesses on the certainty withwhich risk assessment conclusions are drawn. It is important to note that, in contrast tostatements made in USEPA's baseline risk, a large data base on site-specific chemicalconcentrations including measurements in plant tissue, mammals, and invertebrate, has beenobtained to reduce assessment uncertainty.

Page 3, second full paragraph, last two sentences:Compositing of Phase II samples is not justification for employing highly uncertain and biasedconservative input assumptions. All Phase III samples were individual organisms whereappropriate, and there are no discernible differences between these samples and the Phase IIcomposites, indicating strongly that concern for potential to "obscure" elevated contaminantconcentrations is misplaced. In addition, Phase II samples were composited (to make analyticalmass threshold requirements) only within species and trophic guilds. All tissue measurement dataemployed in the risk assessment are scientifically sound, and examination of these data indicatesthat nothing is "obscured" by their application.

U L:\337FB\86C3609SM24CMRSP.DOC\17-Dec-96\SOL 2

Attachment IFields Brook Floodplaln/Wetlind Area EPA

Authorship:This attachment was prepared by USEPA, not by EA Engineering, Science and Technology.USEPA modified an electronic text file prepared by consultants to FBAG, and is the sole authorof this attachment.

Pagex, last sentence:It would be helpful to have citations for this statement, rather than the generalization that theapplication is "suggested by some authors" with no identification of the scientific sources.

Page 36, sections 5.1.9 and 5.1.10, last sentence of each:As discussed in comment for Page 3, second full paragraph, last two sentences above,compositing did not affect data quality negatively, and in fact allowed application of high massanalytical methods for Phase n data for comparison with individual samples analyzed by low massmethods in Phase HI. Because compositing clearly had no negative effect on data quality, there isno justification on this basis for highly uncertain and biased input assumptions.

WUOllMMll Clyde W L:\337FB\86C3609S\124CMRSP.DOai7-Dec-96\SOL 1

Attachment 11Fields Brook Fleodplaln/Wetland Area EPfl

Page 2, paragraph 1, second sentence and elsewhere:Statement that "...very little data on the extent and magnitude of contamination in the area wasavailable to conduct this risk assessment..." is repeated elsewhere (for example, on page 12), andis apparently used to justify the scientifically untenable assumption that maximum measuredchemical concentrations appropriately represent ecological exposure. In fact, hundreds ofsamples of sediments and biotic tissue) have been taken on the floodplain wetland, and have beenreported in appropriate detail in project reports. These data, rather than inappropriate assumptionof measured maximum, should have been employed to represent ecological exposure.

Page 13, section 9.0:Cleanup goals based on maximum site chemical concentrations will clearly lead to over-remediation, with associated needless habitat destruction and impact of remedy. Given the largedata base on the distribution of chemicals in the floodplain wetland and the detailed reporting bylocation provided in project documents (for example, in the RI report), use of maximumconcentrations in this calculation is inappropriate and unnecessary.

Page 14, second full sentence:It is unclear why statement that "No site-specific data was [sic] available for levels of COCs inplants or terrestrial invertebrates". A number of site-specific measurements of tissueconcentrations in plant tissue and earthworms were made and reported. These data should havebeen used to derive site-specific BSAFs, if needed for the risk assessment, rather than employingthe generic and inappropriate assumption that BSAFs equal 1.

WMMhMM'd Clyde 10 L:\337FB\86C3609SV124CMRSP.DOCVI7-Dec-96\SOL 1