WEIL, GOTSHAL& MANGES LLPWEIL, GOTSHAL& MANGES LLP Thomas Alcamo August 24, 2007 Page 2 people at...

WEIL, GOTSHAL& MANGES LLP1300 EYE STREET, NW

SUITE 900

WASHINGTON. DC 20005

(202) 682-7000

FAX: (202 ) 857-0939

EPA Region 5 Records Ctr.

237765

DAVID B. HIRDDIRECT LINE (202) 682-7175E-MAIL: david.hirdaweil.com August 24, 2007

AUSTIN

BOSTON

B R U S S E L S

BUDAPEST

DALLAS

FRANKFURT

HOUSTON

LONDON

MIAMI

MUNICH

NEW YORK

PARIS

PRAGUE

PROVIDENCE

SHANGHAI

SILICON VALLEY

SINGAPORE

WARSAW

Thomas AlcamoRegional Project ManagerU.S. Environmental Protection Agency

Region 5, Mail Code SR-6J77 West Jackson BoulevardChicago, Illinois 60604

Re: Comments on Proposed Plan for the Record of Decision Amendment,Operable Units Two and Three, Neal's Landfill Superfund Site

Dear Mr. Alcamo:

CBS Corporation ("CBS") submits these Comments on the U.S. EnvironmentalProjection Agency's ("EPA") "Proposed Plan for the Record of Decision Amendment,Operable Units Two and Three, Neat's Landfill Superfund Site" (hereinafter "ProposedPlan"). CBS requests that these Comments be included in the administrative record withrespect to any Record of Decision Amendment (hereinafter "RODA") for the Neal'sLandfill Superfund Site.1

Summary of Comments

As discussed in detail below and in the attached exhibits, no additional remedialmeasures are needed to protect human health or ecological receptors.

EPA concludes in its Proposed Plan, that its human health risk assessment shows"total risks . . . within EPA's acceptable risk range of 1 additional cancer case in 10,000

1 CBS's submission of these Comments should not be construed as admission of liabilityor responsibility for the Neal's Landfill Superfund Site. Any claim which the UnitedStates may have against CBS with respect to that Site was resolved under the terms of theConsent Decree in United States v. CBS Corp.. no. IP 83-9-C (S.D. Ind.), and the UnitedStates is barred from taking any further action against CBS by the Covenant Not to Sue inthat Consent Decree.

DC] \248804«1\5BZ801! DOO80758 0052

WEIL, GOTSHAL& MANGES LLP

Thomas AlcamoAugust 24, 2007Page 2

people at all exposure points." Proposed Plan at 19. The non-cancer risks are also foundto be below the hazard quotient of 1.0 for all routes of exposure except fish ingestion 1-mile from the Neal's Landfill site. Id. These risks are well below the threshold fortaking remedial action under the National Contingency Plan ("NCP"), see 40 C.F.R.§ 300.430(e)(2)(i)(A)(2), and under U.S. EPA OSWER Directive 9355.0-30 "Role of theBaseline Risk Assessment in Superfund Remedy Selection Decisions" (April 22, 1991).

Even these minimal risks are based on unrealistic assumptions about humanbehavior, such as that an individual would eat the equivalent of at least 10 grams of fishcaught from locations in Richland Creek 365 days a years for 30 years. If more than oneperson tried to follow this 30-year diet, the 1-mile station at Richland Creek would befished out long before the 30-year period was over. EPA's risk calculations also do nottake into consideration the fact that PCB levels in fish have been declining in recent yearsdue to the successful source control remedy and are projected to continue to do so, evenwith no additional remedial actions.

Also, EPA's calculations of ecological risks do not justify additional measures.EPA only calculated theoretical risk, using overly conservative assumptions, and withoutincorporating field data. EPA uses as its reference animals, the mink and the kingfisher,both of which are territorial. However, the affected area would only provide sufficientfood to sustain one animal in each of these species. EPA's guidance documents providethat the focus of an ecological risk assessment should be to protect a species population,not an individual animal. See U.S. EPA OSWER Directive 9285.7-28 "Issuance of FinalGuidance: Ecological Risk Assessment and Risk Management Principles for SuperfundSites" (October 7, 1999).

EPA's choice of a kingfisher as a reference animal is completely unjustifiedbecause Conard's Branch is too shallow for kingfishers to dive feed during non-stormconditions and too turbid to allow kingfishers to see their prey during storm conditions.Moreover, even though EPA identifies the kingfisher as a reference animal, it does notuse data concerning the effects of PCBs on kingfishers in its analysis. Instead, EPA useschicken and pheasant data in lieu of the missing kingfisher data, but the most recent dataindicates that kingfishers are less sensitive to PCBs than either chickens or pheasants.Moreover, EPA ignores the blue heron, even though blue herons have been observedfeeding in the upper reaches of the study area, and data exists concerning the effects ofPCBs on the blue heron. EPA has not used the blue heron as a reference animal becauseuse of this receptor would show lower risk levels.

EPA also makes errors in identifying provisions from other laws that would be"applicable or relevant and appropriate requirements" ("ARARs") for the remedial

DC I \248804W1\5BZ801! DOC\80758.0052



alternatives under consideration. In particular, EPA identifies certain requirements fromthe State of Indiana National Pollution Discharge Elimination System ("NPDES")program and certain air pollution requirements as ARARs which are neither legallyapplicable nor relevant or appropriate.

I. No Additional Remedial Measures Are Necessary To AddressAny Significant Risk To Human Health Or The Environment

Neither the risk to human health nor the risk to ecological receptors justify anyadditional remedial measures.

A. The Data Shows No Significant Risk To Human Health

In its Proposed Plan, EPA sets forth its analysis of the risk to human healthpresented by the springs, soils, and sediment affected by PCB-contamination originatingfrom Neal's Landfill. These calculations are based on extremely conservative - andunrealistic - assumptions about exposure. But even if EPA' assumptions are accepted,the risk levels calculated by EPA are minimal and would not be sufficient to requirefurther remedial action. Moreover, if EPA had used more realistic assumptions in itsanalysis, the calculated risk would be substantially lower.

The principal risk determined through EPA's calculations is the risk to individualswho catch and eat fish in Richland Creek within 5.5 miles of Neal's Landfill. EPAcalculates cancer and non-cancer risk levels for the ingestion of fish caught 1, 3, and 5.5miles from the landfill. The highest cancer risk for exposure to polychlorinated biphenyl("PCB") Aroclors calculated by EPA is 2.67 x 10"5 for fish caught at the 1 mile location;the risk goes down to 7.2 x 10"6 for fish caught at the 5.5 mile location. Proposed Plan at19. Dermal contact and ingestion of water in Conard's Branch and Richland Creek"produce a cancer risk of less 5 in 1,000,000" (or less than 5 x 10"6) and dermal contactand incidental ingestion of soil and sediment at Conard's Branch and Richland Creek"produce a cancer risk of 1 in 1,000,000 or less." (1 x 10"6 or less). Id. Also, accordingto EPA's calculations, the non-cancer risk for persons eating fish caught 1-mile from thelandfill is measured by a hazard quotient of 1.59. All other non-cancer risks evaluatedhave hazard quotients of less than 1. Id.

EPA notes in the Proposed Plan, "the results of [EPA's] human health riskassessment show that total risks are within EPA's acceptable risk range of 1 additionalcancer case in 10,000 people at all exposure points." Id. (emphasis supplied). In U.S.EPA OSWER Directive 9355.0-30 "Role of the Baseline Risk Assessment in SuperfundRemedy Selection Decisions" at 1 (April 22, 1991), EPA announced its policy that:

DC1:U48804\01\5BZ801!DOC\807580052



"[wjhere the cumulative carcinogenic site risk to an individual based on reasonablemaximum exposure for both current and future land use is less than 10 (-4) [less than 1 in10,000] . .. action generally is not warranted." This concept has been incorporated intothe NCP at 40 C.F.R. § 300.430(e)(2)(i)(A)(2). Therefore, no further action is necessaryto achieve the level of protection of human health from cancer risks required in CERCLAremedial actions.

Although EPA describes the 1 additional cancer case in 1,000,000 people (or 1 x10"6) as "a point of departure that EPA uses," Id., there is no requirement that CERCLAremedial actions must meet that level. This is especially true in the case of Conard'sBranch and Richland Creek where the potential for human exposure is so remote andunlikely. These privately-owned locations are neither highly accessible nor verydesirable for fishing, swimming or wading. In addition, recent data has shown that PCBlevels in fish in Conard's Branch and Richland Creek have been declining, and theselevels are projected to continue to decline, even with no further remedial action. Thus,over time, even the minimal risk calculated by EPA will diminish significantly.

In order to reach the calculated cancer risk levels for ingestion of fish set forth inthe Proposed Plan, EPA made several unrealistic assumptions. EPA assumed that anindividual would eat between 6 and 18.3 grams of pelagic fish and between 4 and 12.2grams of benthic fish caught at the designated locations 365 days a year for 30 years.Even making these assumptions, EPA only calculated a cancer risk of 2.67 x 10"5 for fishcaught at the 1 mile location and a cancer risk of 7.2 x 10"6 for fish caught at the 5.5 milelocation.

These assumptions are wholly unrealistic:

• First, it is highly unlikely that an individual would constrain his or her diet on aconstant basis every day for 30 years to pelagic or benthic fish caught at thesespecific locations. Humans like variety in their diet, and are more likely to relyprimarily on food purchased from retailers than on fish caught themselves. Thereis no evidence of individuals in the Bloomington area who rely primarily for foodon fish that they caught themselves, let alone fish caught at these specificlocations.

• Second, even if someone were to restrict his or her diet to fish caught at a specificlocation, it is much more likely that that person would use one of the several otherfishing locations in the Bloomington area that are less remote and contain moreplentiful and desirable fish than Richland Creek.

DCl \248804O1\5BZ801I.DOC\80758.0052

WEIL, GOTSHAL& MANGES UP


• Third, it is highly unlikely that any one would eat as much of the benthic fish asEPA supposes, since benthic fish are generally considered less suitable forconsumption, and greater quantities of the more desirable pelagic fish can befound in fishing locations elsewhere in Bloomington.

• Fourth, EPA's calculations are based on the assumption that anglers would eatfish as small as four inches long. Studies, however, have shown that anglersprefer to eat fish at least six inches long. Limiting the ingestion analysis to fishsix inches or longer would greatly reduce the amount of affected fish available forconsumption at these locations.

Moreover, EPA assumes that PCB levels in fish will remain constant. EPA doesnot take into account in its 30-year analysis the fact that PCB levels in these fish havebeen declining, and are projected to continue to decline at a rate of 4% per year. Thus,over a 30-year period, the levels of PCBs in the fish are expected to go downdramatically without additional remedial measures.

But even if EPA's assumptions were correct, only one person could be exposed tothis risk. The fish population of Richland Creek is so limited that if more than one personwere to try to fish it at 1-mile station to consume the amount assumed by EPA, the Creekwould be fished out many years before the 30 year period assumed by EPA. This wasdemonstrated in the analysis performed by EPA's own contractor, Tetra Tech whodetermined that, at the 1-mile location, Richland Creek can only produce 6 grams ofpelagic fish and 4 gram of benthic fish to feed one consumer. Thus, at the 1-milelocation, the risks calculated by EPA would only affect a single individual assuming thatthat person acted in a way that defied human experience and common sense.

Even further away from the site, it is unlikely that EPA's hypothetical diet ofpelagic and benthic fish could feed more than a single consumer on an every day basis for30 years. EPA assumes that there is a 10 percent increase in the consumption of fish foreach creek mile, but there is no scientific basis for that assumption.

CBS has attempted to recalculate the risk, correcting some of EPA's errors. CBSused more plausible fish consumption rates of 1 gram per day at the 1-mile location, 4grams per day at the 3-mile location and 6.4 grams per day at 5.5-mile location, butcontinued to use EPA's unrealistic assumption that person would eat fish caught at theseunsuitable fishing locations 365 days a year for 30 years. CBS then factored into itsanalysis the data showing the 4% year declining trend of PCB levels in fish. CBSderived fish ingestion cancer risk levels of 1.7 x 10"6 at the 1-mile location, 2 x 10"6 at the3-mile location, and 3.6 x 10 6 at the 5.5-mile location. But even these risk levels are

DCI \248804«l\5BZ801!.DOC\80758.0052



extremely overstated because they assume an angler eating fish from these locationsevery day for 30-years, which defies human experience and common sense.

EPA also attempts to calculate risks and hazards using PCB toxicity equivalents("TEQs") for certain PCB congeners. But EPA's TEQ analysis is based on methods thathave been strongly challenged during peer review by the National Academy of Sciences,and is currently under reconsideration. Therefore, CBS the TEQ based analysis shouldbe eliminated from the risk assessment analysis.

EPA also calculated non-cancer risks associated with exposure to PCBs iningesting fish and dermal contact and ingestion with surface water, soil, and sediment. Inmaking this type of analysis, EPA's threshold for potential risks is a hazard quotient("HQ") greater than 1. Proposed Plan at 18. With respect to fish ingestion, EPA onlycalculated an HQ greater than 1 at the 1-mile location, where it came up with a value of1.59. The HQ's calculated for fish ingestion at the 3-mile and 5.5 mile locations weresubstantially below 1 (0.61 and 0.42 respectively). The HQ's calculated for dermalcontact and ingestion of surface water, soil, and sediment were also well below 1. Id. at19.

EPA's HQ calculation of 1.59 for fish ingestion at the 1-mile location isunrealistically high. As noted above, EPA assumes a daily fish ingestion rate at thatlocation of 4 grams of benthic fish and 6 grams of pelagic fish. Given the small size offish at that location and data collected about the fish consumption by anglers, a morerealistic fish ingestion rate at that location would be one gram. Using the more realisticone gram per day ingestion rate, the HQ is more accurately calculated as 0.18 (withoutany decline over 30 years in PCBs levels in fish) or 0.1 (factoring in the projected 4%rate of decline).

In sum, even if EPA's calculations are accepted as accurate, the Neal's Landfillsite in its current posture does not create a risk to humans that approaches EPA's actionthreshold of an additional cancer case for 10,000 and only exceeds EPA's HQ thresholdof 1 for a single exposure. But even these calculations are based on unrealisticassumptions concerning the amount of fish available for ingestion and the frequency ofexposure. If more realistic assumptions are used, the calculated risk would be evenlower. Therefore, there is no basis for remedial action.

A more detailed discussion of the errors in EPA's human health risk assessment iscontained in Exhibit 1 and the tables attached thereto.

DC 1 : \248804\01\5BZ801 !.DOC\80758.0052


Thomas AlcamoAugust 24, 2007Page?

B. Risks To Ecological Receptors Do Not Justify Further Action

The risks to ecological receptors are equally insubstantial. Even EPA's analysisshows that ecological risks are limited to a small geographic area, if corrections are madeto EPA's overly conservative assumptions, the projected risk levels are even lower.

EPA's approach to ecological risk assessment is to calculate hazardous quotientsfor just two animals, the mink and kingfisher. According to EPA guidance, this approachis very conservative and should only be used as a screening device to rebut untested riskassumptions. But EPA should not rely on the theoretical projections of risk to selectadditional remedial actions, without field verification.

EPA has failed to use the most recent data about PCB levels in fish (collected inNovember 2003 and November 2005) in its calculations and has used novel and undulyconservative means of calculating toxicity reference values ('TRV's) for both animals.EPA also ignores that there has been a downward trend in PCB levels in fish in the studyarea over recent years. Therefore, its conclusions overestimate risk.

For both the mink and the kingfisher, EPA calculates hazard quotients greaterthan 1 in the upper reaches of the study area. But these numbers are driven by EPA'serroneous assumptions about the amount of time these animals forage from Conard'sBranch. This is a small area which could only support the foraging of one animal foreach of the two reference species on a part time basis. Therefore, EPA's calculation ofthese hazard quotients is unjustified.

Both mink and kingfishers are territorial animals, and the upper reaches of thestudy area is so small that it could only be occupied by a single animal of each species.Thus, by focusing on these animals, EPA is inappropriately calculating the risks only toan individual member of each species, and not to the species population. According toEPA's guidance, the focus of ecological risk assessment is to protect the population of aspecies, not individual animals. See U.S. EPA OSWER Directive 9285.7-28 "Issuanceof Final Guidance: Ecological Risk Assessment and Risk Management Principles forSuperfund Sites" (October 7, 1999).

EPA's selection of toxic reference values ("TRV's") for the theoretical minkreceptor is based on methods that are not used on other mink Superfund risk assessmentsand for which the scientific basis has not been externally peer reviewed. EPA disregards,for unjustified reasons, more recent TRVs for mink developed in the literature. Thisresults in an overly conservative assessment of the risk to mink. Even with such anoverly conservative approach, the EPA calculated risk to mink is limited to a small area

DCl \248804«1\3BZ80I!.DOC\80758.0052


Thomas AlcamoAugust 24, 2007PageS

that could only impact one animal and EPA has not shown how this limited area couldimpact the overall population of mink in the Richland Creek corridor. To correct EPA'serrors and provide a more realistic theoretical calculation of risk to mink, CBS hasperformed its own calculations of theoretical projection of risk to mink, which arecontained in Exhibit 2, Appendix B, attached hereto.

EPA also relies on the analysis of a dead mink purportedly found on Route 48 andRichland Creek bridge and kept in a freezer by the person who found it. Data from thisroad kill does not meet any of EPA's established chain of custody or other quality controlor quality assurance standards. Despite the location in which the animal was found, it isunknown whether or how much the animal was exposed to the study area or to othersources of PCBs. In making its analysis, EPA assumed certain lipid values which cannotbe verified. EPA also relied on some questionable studies in making its analysis. Evenso, the levels found by EPA in the liver of the dead mink are below the levels determinedin other studies to have potential population level effects in mink. In sum, this dead minkis an unreliable source of data that ultimately has no consequence.

EPA's choice of a kingfisher as a reference animal is completely unjustified. It isunlikely that kingfishers will use the Conard's Branch reach extensively because it is tooshallow. Kingfishers are dive feeding birds who require water of sufficient depth tocapture fish. EPA has acknowledged that in non-storm conditions Conard's Branch is tooshallow for kingfishers to dive feed, but contends that during storm conditions Conard'sBranch would be deep enough for kingfishers. But kingfishers need clear water to locatetheir prey and therefore would not dive feed in Conard's Branch during storm conditionswhen the water would be turbid. So there is no basis to select the kingfisher as areference animal in either condition for the upper reaches of the study area.

The selection of the kingfisher as a reference animal is especially incongruousbecause there is no significant data in the literature concerning the effects of PCBs onkingfishers. Indeed, EPA acknowledges that insufficient data for kingfishers exists, andinstead relies on data concerning the effects of PCBs on chickens and pheasants.

In other words, what EPA is actually doing is using chickens and pheasants as areference animal, even though there is no evidence of chickens and pheasants in Conard'sBranch. How EPA gets to the chicken and pheasant is by making a specious argumentthat kingfishers will dive feed in Conard's Branch during storm events even though thewater would not be sufficiently clear for them to do so. EPA then announces that, in theabsence of sufficient data about PCB effects on kingfisher, it will use chicken andpheasant data. EPA takes this last step by assuming that PCBs will have the same effect

DC 1 :\248804\OI\5BZ801 !.DOO80758.0052



on kingfishers as on chickens and pheasants, even though more recent field studiesindicate that kingfisher are less sensitive to PCBs than chickens or pheasants.

At the same time EPA is making its unjustified assumptions about kingfisherbehavior and the supposed similarities between kingfishers and chickens or pheasants,EPA ignores the blue heron. Unlike the kingfisher, the blue heron is well suited to feedin Conard's Branch and has been observed at the confluence of Conard's Branch andRichland Creek. Moreover, data exists in the literature about the effects of PCBs on theblue heron. Thus, the blue heron is a much more suitable reference animal for this sitethan the kingfisher. CBS has calculated the theoretical risk to the blue heron. Thisanalysis is found in Exhibit 2, Appendix A. When the current situation is analyzed usingthe blue heron as a reference animal, the risks presented are well within an acceptablerange.

A more detailed description of the errors in EPA's analysis of the ecological riskis contained in Exhibit 2 (and its two appendices) attached hereto. When properlyanalyzed, the minimal ecological risk provides no justification for further remedialmeasures at the Neal's Landfill site.

II. The Proposed Plan Sets Forth Requirements As "ARARs" That Do NotQualify As "ARARs" Under CERCLA

Even if EPA were to select one of Alternatives 2-7 as a remedy, certain of theNPDES2 requirements and air emission requirements identified as ARARs in theProposed Plan would not qualify as ARARs3 under CERCLA.

2 4 'NPDES" is acronym for "National Pollutant Discharge Elimination System," which isthe federal and state system implementing the permit requirements of the Clean WaterAct.

3 An acronym for "applicable or relevant and appropriate requirements." See 43 U.S.C.§9621(d)(2).

DCl \248804\01\5BZ80P.DOC\80758.0052



A. Identified NPDES Provisions Are Not ARARs For This Site

1. In General, NPDES Provisions Should Not Be ConsideredARARs At This Site

In the Proposed Plan, EPA asserts that "certain regulations enacted by the State ofIndiana under its federally approved NPDES program are relevant and appropriate todischarges from the plant." Proposed Plan at 33. It is important to note that even EPArecognizes that these requirements are not legally "applicable," but only attempts tojustify them under the "relevant and appropriate" prong of the term "ARAR," as thatconcept is used in section 121(d) of CERCLA, 42 U.S.C. § 9621(d) and the NCP.However, these NPDES requirements are not "relevant and appropriate" to the watertreatment plant.

The NCP defines "relevant and appropriate requirements" to mean substantiverequirements of federal or state law, which "while not 'applicable' to a hazardoussubstance, pollutant, contaminant, remedial action, location or other circumstance at aCERCLA site, address problems or situations sufficiently similar to those encountered atthe CERCLA site that their use is well suited to the particular site." 40 C.F.R § 300.5.This provision requires EPA to look at the specific circumstances of a site to determinewhether the proposed requirement is "well suited."

In this case, NPDES requirements are very ill-suited to the Neal's LandfillSuperfund Site, which is very different from the situation in which they would normallyapply. The NPDES requirements are intended to apply to the "discharge of pollutants,"which is a defined term with a specific meaning under that statute. 33 U.S.C. §§ 1311,1362(12). The term "discharge of a pollutant" is defined to mean "any addition of anypollutant to navigable waters from any point source." 33 U.S.C. § 1362(12). Thisdefinition does not apply to the treatment plant at the Neal's Landfill Site because there isno "addition" of a pollutant.4

A treatment facility at Neal's Landfill would not add any pollutant to navigablewaters. Rather, the plant intercepts the already existing flow of groundwater as it mergesinto the surface water and removes pollutants before the water flows further. Instead ofadding pollutants, rather it would subtract pollutants from the water. The treated water is

4 Although the Neal's Landfill facility has had a NPDES permit because it was requiredunder the terms of the 1985 Consent Decree. Such a permit requirement has madeunnecessary under the 1986 amendments to CERCLA. 42 U.S.C. § 9621(e)(l).

DCl \248804«I\5BZ801!.DOC\80758.0052



released back into the natural system of flow where it would have flowed if it had notbeen intercepted.

The alternatives under consideration create a very different situation from theordinary circumstances in which the NPDES program applies: an industrial discharger orpublic sewer system adding both water and pollutants to a body of navigable water. Inthose circumstances, the discharger is adding materials which would not otherwise flowinto these bodies of water. The treatment facility is also different from most Superfundremedies involving the pumping and treating of contaminated groundwater. At thosesites, groundwater is extracted and treated, and the treated water is then discharged intosome other body of water into which it would not naturally flow if not intercepted,resulting in the addition of pollutants to a new body of water. Accordingly, NPDESrequirements may be ARARs at those sites. But this treatment facility does not involveadding anything to Conard's Branch or Richland Creek which would not have flowedthere if no action were taken. Thus, NPDES requirements should not be ARARs.5

2. 327 IAC 2-1-6, Table 1 As A Whole Is Not An ARAR For ThisSite

EPA and the State of Indiana have identified 327 IAC 2-1-6, Table 1 as anARAR. This table contains water quality limits for a wide variety of pollutants identifiedby the State of Indiana from any source regulated by the State under its NPDES program.No effort is made to identify any specific pollutant that may be discharged from thetreatment facility.

This approach is totally inconsistent with the CERCLA process of identifyingspecific ARARs that apply to specific CERCLA cleanups. Under CERCLA, an ARAR issupposed to be a specific "standard, requirement, criteria or limitation" that is relevantand appropriate to the specific hazardous substance, pollutant or contaminant at issue. 42U.S.C. § 9621(d)(2)(A). Rather than identify limitations that would be applicable to theeffluent from a treatment facility, EPA and the State simply identify the standardsapplicable to all possible pollutants that might be regulated under the program. Indeed,this approach is inconsistent with the way EPA and the State administer the NPDESprogram. When issuing an NPDES permit, the Indiana Department of Environmental

5 Instead, the treatment facility at Neal's Landfill is analogous to Superfund remedies inwhich treated groundwater is reinjected into the same groundwater system in which itwas flowing before being treated. In those circumstances, NPDES requirements aregenerally not identified as ARARs.

DCl \248804«1\5BZ80I!.DOC\80758.0052



Management (IDEM) does not simply list all the standards on Table 1; rather, it identifiesthe pollutants in the discharger's wastewater and only includes the limits for thepollutants identified.

As noted above, the NCP defines "relevant and appropriate requirements" tomean substantive requirements of federal or state law, which "while not 'applicable' to ahazardous substance, pollutant, contaminant, remedial action, location or othercircumstance at a CERCLA site, address problems or situations sufficiently similar tothose encountered at the CERCLA site that their use is well suited to the particular site."40 C.F.R § 300.5. The process of identifying ARARs then involves identifying specificstandards which have specific relationship to the hazardous substances found at the site.Instead of doing that, the EPA and the State simply refer to a generic table listing allsubstances regulated (327 IAC 2-1-6, Table 1). A general reference to the table is notsufficient. EPA and the State must identify the specific substances of concern that arepresent at this site and require limitations. Both CERCLA and the NCP provide that theonly State standards which may qualify as ARARs are those which have been identified"in a timely manner." 42 U.S.C. § 9621(d)(2)(A)(ii); 40 C.F.R. § 300.5. The NCPfurther provides that a remedy must only meet those ARARs which are identified at thetime a Record of Decision is signed. 40 C.F.R. § 300.430(f)(l)(ii)(B).

At Neal's Landfill, the only hazardous substance, pollutant or contaminant ofconcern that has been identified are PCBs. Accordingly, effluent limitations with respectto other substances may not be ARARs because there is no information to show that theymay be relevant or appropriate for this Site. Therefore, the general reference to 327 IAC2-1-6, Table 1 as an ARAR is not appropriate.

3. The Purported Identification Of A PCB Discharge Limit IsErroneous And Inapplicable To CBS

Even with respect to PCBs, neither EPA nor the State has identified a legallyenforceable effluent limitation which would qualify as an ARAR. The State of Indianahas admitted that it is impossible to calculate a viable effluent limitation for PCBs usingthe methodology in 327 IAC 5-2-11 and 327 IAC 5-2-11.2, because the resulting numberwould be a level that could not be reliably confirmed using existing measurementtechnology. Instead, the State claims to use the effluent limitation of 0.3 parts per billion("ppb") in permits because that is the limit of quantification using existing technology.This 0.3 ppb effluent limitation for PCBs cannot qualify as an ARAR because it has not

DCl \248804«1\5BZ801!DOC\80758.0052

WEIL. GOTSHAL& MANGES UP


been "promulgated" by the State as a regulatory standard, as is required under 42 U.S.C.§9621(d)(2)(A)(ii).6

But the level which would be calculated applying 327 I AC 5-2-11 and 327 IAC 5-2-11.2 is not an ARAR, either. The State admits that it does not include this standard inits NPDES permits and that this standard is not verifiable using existing technology. InNRDC v. Costle, 568 F.2d 1369, 1380 (D.C. Cir. 1977), the D.C. Circuit held that "whennumerical effluent limitations are infeasible, EPA may issue permits with conditionsdesigned to reduce the level of effluent discharges to acceptable levels." EPA haspromulgated 40 C.F.R. § 122.44(k)(3), which provides that best management practices("BMPs") may be used "to control or abate the discharge of pollutants when .. .[n]umeric effluent limitations are infeasible." This provision is incorporated into theIDEM regulations. See 327 IAC 5-2-1.5.

Assuming arguendo that the 0.3 ppb NPDES discharge limits for PCBs wouldqualify as an ARAR for a treatment plant at the Neal's Landfill site, CBS notes that EPAand IDEM are constrained by the terms of the 1985 Consent Decree and the decision ofthe Supreme Court of Indiana in Canard v. IDEM, 614 N.E.2d 916 (Ind. 1993) fromimposing on CBS a PCB discharge limit more stringent than 1.0 ppb without CBS'sconsent.

4. EPA Is Taking An Inconsistent Approach To Sampling

One of the NPDES provisions identified as a potential ARAR is 327 IAC 5-2-13(f). This provision calls for composite sampling. This is not a standard or limitation.In operating the interim treatment plant required under Paragraph 59(a) of the ConsentDecree, CBS has historically taken grab samples. CBS contends that sampling

6 EPA has adopted its own effluent limitation for PCBs, however, one which is higherthan the unpromulgated limitation of 0.3 ppb proposed by the State. On June 29, 1998,EPA adopted a comprehensive rule concerning the handling of PCBs. Disposal ofPolychlorinated Biphenyls (PCBs), 63 Fed. Reg. 35384 (June 29, 1998). In the rule, EPAestablished a presumptive discharge standard of 3 ppb for discharges to navigable water.40 C.F.R. § 761.79(b)(l)(ii), 63 Fed. Reg. at 35458. In the preamble to the rule, EPAdescribed this 3 ppb limit as the "level historically used to regulate discharges of PCBs."63 Fed. Reg. at 35418. EPA also concluded in its PCB rule that water with a PCBconcentration of 0.5 ppb was suitable for any use, including drinking water. 40 C.F.R.§ 761.79(b)(l)(iii), 63 Fed. Reg. at 35458. Therefore, there is no basis for adopting aneffluent limitation of 0.3 pbb for this Site.

DCl \248804\OI\5BZ8011DOO80758.0052



requirements should not be designated as ARARs because they are not standards, criteria,or limitations. But if EPA chooses to select a water treatment remedy, it should considerwhether EPA should mandate composite sampling where grab sampling has been reliedupon historically.

5. The Technical Impracticability Waiver For By-Passes IsJustified

To the extent that NPDES provisions are identified as ARARs, CBS agrees withEPA that a technical impracticability waiver is appropriate with respect to the applicationof these requirements to water that is not treated in the water treatment plant.

One provision in particular which should not be applied as an ARAR is 327 IAC5-2-8(11), which concerns by-passes. Most of the remedial alternatives underconsideration involve using the existing plant to capture and treat flows of similarvolume. Although, in some of these alternatives, modifications would be made in thecollection system, not all the flows from springs around Neal's Landfill would bediverted to the plant. Moreover, in periods of high flow from the collected sources, theplant would not be able to capture flows in excess of 1 cubic feet per second (or 450gallons per minute) on a consistent basis. Although the existing system is often capableof handling 500 or even 550 gallons per minute in certain situations, it cannot handlevolumes greater than 450 gallons per minute in all situations. Therefore, if this provisionis designated as an ARAR, there must be a technical impracticability waiver with respectto flows that are not collected by the system and with respect to flows collected by thesystem but which are greater than 1.0 cubic feet per second or 450 gallons per minute.

6. ARARs Must Be Established At The Time Of Agency Decision-Making

Section 327 1AC 5-2-11.1 (h) should not be considered an ARAR because it doesnot provide for any standard, requirement or limitation, but rather describes a processthrough which IDEM may impose more stringent standards than those currently in effect.Section 121(d) of CERCLA, 42 U.S.C. § 9621(d), requires EPA to determine ARARsbased on the legal requirements that are in effect at the time the remedy is selected, not onthe basis of requirements which may come into effect in the future. 42 U.S.C.§ 9621(d)(2)(A)(ii), (4); 40 C.F.R. § 300.430(0(1 )(ii)(B). Application of this provisionwould reserve a right for IDEM to impose more stringent standards later. Listing thisprovision as an ARAR is inconsistent with CERCLA's requirement that ARARs must beidentified at the time of remedy selection.

DCl \248804« 1 \5BZ801! DOC\80758.0052



7. IDEM Internal Administrative Requirements Should Not BeEstablished As ARARs

Many of the NPDES provisions identified as ARARs do not set forth limits orrequirements that IDEM imposes on dischargers, but rather describe the procedures thatIDEM is supposed to follow in establishing such limits. These include: 327 IAC 5-2-1 l(d), (e), (f), (g) and (h) and 327 IAC 5-2-11.1 (a), (b), (d), (0, (g), and (h) as ARARs.These provisions should not be designated as ARARs. The term ARAR in 42 U.S.C.§ 9621(d)(2)(A) refers to the specific "standard, requirement, criteri[on], or limitation"itself, not the means by which it was determined.

Section 327 IAC 5-2-1 l.l(d) is particularly incongruous because it relates to theestablishment of WQBELs for discharges of metals, and is neither applicable or relevantand appropriate since metals are not contaminants of concern at this site.

The Proposed Plan also cited various other administrative provisions of the IDEMNPDES program as ARARs. CBS does not believe that these provisions should beconsidered ARARs for the reasons explained in the following table:

IAC Code Provision Requirement Reason(s) Why Not AnARAR

327 IAC 5-2-8(3) Permittee shall take allreasonable steps tominimize or correctnoncompliance with thepermit

(1) There will be no permit;and (2) this is not a"standard, requirement,criteria, or limitation" under42 U.S.C. §9621 (d)(2).

327 IAC 5-2-8(7) Permittee shall provide aright of access to IDEM andits contractors

(1) There will be no permit;and (2) This is not a"standard, requirement,criteria, or limitation" under42 U.S.C. §9621(d)(2)(A).

327 IAC 5-2-8(8) Permittee shall maintain thefacility in good order

(1) There will be no permit;and (2) this is not a"standard, requirement,criteria, or limitation" under42 U.S.C. § 9621(d)(2)(A).

DCI \248804«I\5BZ80I!.DOC\80758.0052



Permittee's tampering withmonitoring device is subjectto criminal sanctions

327 IAC 5-2-8(9) (1) There will be no permit;(2) This is not a "standard,requirement, criteria, orlimitation" under 42 U.S.C.§9621(d)(2);and(3)enforcement provisions arenot ARARs.

327 IAC 5-2-8(10) Establishes reportingrequirements

Reporting requirements forCERCLA remedies are setforth in consent decrees orother enforcementdocuments

327 IAC 5-2-8(12) Requirements for upset Paragraph 59(a) of theConsent Decree supersedesthis provision because itdoes not require CBS totreat flows greater than 1.0cubic foot per second.

327 IAC 5-2-8(14) Provisions relating tosigning of reports andcriminal sanctions for falsereports

(1) There will be no permit;(2) This is not a "standard,requirement, criteria, orlimitation" under 42 U.S.C.§ 9621(d)(2)(A); (3)procedural and enforcementprovisions are not ARARs;and (4) these issues aregenerally addressedseparately in a consentdecree or other enforcementdocument.

327IAC5-2-ll(a)(l)-(4) Definitions of "AverageMonthly Discharge,""Average WeeklyDischarge," "ContinuousDischarge," and "DailyDischarge," used in

Definitions are not in and ofthemselves "standard[s]requirement[s], criteria, orlimitation[s]" under 42U.S.C. §9621(d)(2), andtherefore not ARARs.

DCl \248804« I \5BZ801! DOO80758.0052



calculating effluentlimitations.

327IAC5-2-ll(a)(5)(C) Method of calculation fordaily discharges

This methodology is not inand of itself a "standard,requirement, criterion], orlimitation" under 42 U.S.C.§ 962 l(d)(2), and thereforenot an ARAR.

327 IAC 5-2-1 l(g) Provisions relating to non-continuous dischargers

The Neal's Landfill plantoperates on a continuousbasis.

327 IAC 5-9-2(a), (c), (d),(e)

Refers to best managementpractice programs for"ancillary manufacturingoperations"

There are no "ancillarymanufacturing operations"in this situation. Thisprovision is not an ARAR,because it is not applicableor relevant and appropriateunder CERCLA. It is ill-suited for this situation.

327 IAC 5-9-2(i) Discharger will amend thebest management practiceswhen there is a change tothe system that affects thedischarge

(1) This is not a substantiverequirement of a permit, buta procedural provisionrelating to IDEM'sadministration of theprogram; and (2) underCERCLA, 42 U.S.C.§9621(d)(2)(A),(4),andthe NCP, 40 C.F.R.§300.430(f)(l)(ii)(B),ARARs must be identifiedand are fixed at the time ofremedy selection.

(1) This is not a substantiverequirement of a permit, buta procedural provision

327 IAC 5-9-2(3) Provision relating to theamendment of bestmanagement practices plan

EX • I :\248804VO 1 \5 BZ801! r>OC\80758 005 2

WEIL, GOTSHAL& MANGES LJLP


I if it proves ineffective

I

relating to IDEM'sadministration of theprogram; and (2) undi.vCERCLA, 42 U.S.C.§9621(d)(2)(A),(4),andthe NCP, 40 C.F.R.§ 300.430(0(1)01X3),ARARs must be identifiedand are fixed at the time ofremedy selection.

327 I AC 5-2-13(a). (c), (d), I Monitoring provisions(e). (0 i

These provisions are notARARs in and ofthemselves because they donot establish standards, butrather set forth monitoringrequirements for assessingcompliance with standards.CBS would presume thatmonitoring requirementswould be set forth in aStatement of Work or otherdocument developed byEPA, as opposed to thisregulation. See thediscussion of 327 IAC 5-2-13(f) in the text of thisletter.

B. Certain Indiana Air Pollution Requirements Should Not Be IdentifiedAs ARARs

Finally, CBS disputes the identification of three Indiana air pollutionrequirements relating to major sources of hazardous air pollutants ("HAPs") as ARARs.These requirements are plainly inapplicable and not "relevant and appropriate" to thecontemplated measures at Neal's Landfill One requirement, 326 IAC 2-4.1, simplyprovides that PCBs are designated as a HAP. This is not an ARAR, because in and ofitself this designation is not a "standard, requirement, criterion], or limitation" within the

DC I \248804tt1\5BZ801! DOO80758.0052



meaning of 42 U.S.C. § 9621(d)(2)(A). The second provision. 326 IAC 2-5.1-3(a)(l)(D).applies to sources of HAPs which have the potential to emit ten (10) tons or more of aHAP in a year, and the third provision, 326 IAC 2-. t5.1-?.(a)(l)(A), apples to sources ofHAPs which also have the potential to emit five (5) 'ons or more of paniculate matter.But none of the proposed alternative; involve constructing a source with the potential ofemitting ten (10) tons of a HAP or fi /e (5) tons of paniculate matter. So these provisionsare neither applicable noi relevant and appropriate. They are simply irrelevant andshould not he considered as ARARs. Also, 326 IAC 2-5.1-2(a)(l)(A) would not be anARAR because it only imposes a registration obligation, not a substantive limit onemissions.

Sincettly,

David B.

Enclosure?

cc: Dorothy M. Alke (w/ enclosures)Russell Cepko (w/ enclosures)David R. Berz (w/ enclosures)

DCl \248804«1 \5BZ801! DOO80758 0052

Exhibit 1

CBS Corporation Comments on the Human Health RiskAssessment for Neal's Landfill

Exhibit 1

Comments on the U.S. EPA'sHuman Health Risk Assessment for the Neal's Landfill Site

On behalf of the U.S. Environmental Protection Agency (EPA), Tetra Tech EM Inc.prepared a human health risk assessment (HHRA) report to assess the potential humanexposures to contaminants measured in fish tissue, surface water, bank soil, floodplainsoil, and sediment samples from Conard's Branch and Richland Creek, which aredownstream of the Neal's Landfill Site (Site) in Bloomington, Indiana (Tetra Tech,2006a). That report was a revision of EPA's previous draft HHRA (Tetra Tech. 2003a)for the Site and added refinements that substantially improved its scientific quality andsite-specificity. The resulting revised HHRA is considerably more site-specific and hastaken some important site characteristics into consideration.

CBS continues to believe, however, that the risk estimates that are provided in therevised HHRA still contain many unnecessary precautionary default assumptions thatcontribute greatly to the uncertainty of the risk and hazard estimates. It appears that, inan effort to ensure that risks at the Site are not underestimated, a number of highlyconservative default assumptions, which are not supported by available data or sitecharacteristics, have been made and have resulted in substantially overestimated Siterisks.

The key issues that have contributed to this overestimation and uncertainty include thefollowing:

• Risks due to fish consumption are still overestimated due to the fact that therevised fish ingestion rates do not reflect realistic use of the resource. Inaddition, the exposure point concentrations (EPCs) used for some locations arebased on an extremely limited number of samples. Further, the toxicityequivalent quotient (TEQ) concentrations that have been used do not reflect thecurrent toxic equivalency factors (TEF) that have been developed by the WorldHealth Organization (van den Berg et al., 2006). Finally, EPA has used amethodology that was proposed in the Agencies' Dioxin Reassessment (EPA,2000) to evaluate potential carcinogenic risks and noncarcinogenic hazardsassociated with the fish consumption pathway, despite the fact that this was aproposed draft approach that was strongly challenged during the peer reviewconducted by the National Academy of Sciences (NAS) and is currently beingreconsidered by EPA.

• Risks due to other media at the Site, including surface water, sediment and soils,are also overestimated. This is due to the use of EPCs, in many locations, thatare based on the maximum detected values and to the use of highly conservativeexposure assumptions including high ingestion and dermal contact rates, andunrealistically high exposure frequencies.

• The total combined risk and hazard estimates, which include multiple pathwaysof exposure, are inappropriate due to the fact that risks and hazards have been"mixed and matched" from various portions of the Site to derive reach-specifictotal risk estimates. As a result, in some of the areas evaluated, the estimatedrisks are not representative of the potential exposure that could be attributed tothose specific areas.

Exhibit 1

CBS has recommended alternative values based on their scientific merit and onadditional consideration of site-specific characteristics and has combined these intoalternative risk estimates. These revised risk estimates indicate that risks at this Site arelikely to be lower than have been estimated in the HHRA.

Fish Consumption

Fish Ingestion Rate

EPA has substantially reduced its fish consumption rates for the 1-Mile Site of RichlandCreek from the rates used in the previous HHRA because the fish consumption ratesused previously could not be supported by the productivity of that location. CBScommends EPA for considering the productivity and sustainable harvest from the 1-MileSite in revising its fish consumption rate. However, there are still two major issues withthe revised consumption rates estimated by EPA for this site.

First, the calculations of sustainable yield still assume that anglers would eat a four inchand above fish. Studies have shown that anglers prefer to keep and eat fish six inchesand above . Beyond assuming that a four inch fish is desired by anglers, assuming ablanket minimum size disregards the different shapes and edible yield of fish by species.For example, sunfish and rock bass have body shapes that are considerably more stoutthan white suckers. Thus while studies show that anglers prefer pan fish (includingsunfish and rock bass) of at least six inches in length, it is conceivable that some ediblefood could be obtained from a panfish as small as 4 inches. Whereas the more slenderbody of a white sucker of only 4 inches in length would yield considerably less ediblemass. CBS continues to recommend that six inches be considered the minimum ediblefish size.

Second, the Tetra Tech analysis indicates that the productivity of this reach can onlysupport the estimated fish consumption rates for a single fish consumer withoutimpacting the sustainability of the fishery. While as noted above, CBS still feels thatconsumption rate for even this single consumer is overestimated it is not typical riskassessment practice to base risk estimates on maximum exposures to a singleindividual. While Appendix A acknowledges this approach, this important limitation is notpresented in the main body of the risk assessment. It is important that, if EPA continuesto use this approach, this limitation be presented more directly so that risk managers willbe able to consider this important factor in their decision-making concerning potentialremedial activities.

Consideration of the quality of the fishery and information about recreational fishconsumption rates from rivers and streams indicate that the estimated fish consumptionrates that are being applied in the revised HHRA are not appropriate for this area. Inaddition, if more than one individual were to harvest and consume fish from this reach atthe assumed rates, the result would be the depletion of the resource so that no fishingcould occur there.

As discussed in previous comments on the HHRA for Meal's Landfill (CBS, 2004; 2005),the available fish consumption literature indicates that, for small streams, fishconsumption rates in the range of 1 to 4 g/day would likely be appropriate, while in largerrivers and streams, the range of consumption rates is likely to be 1 to 12 g/day. At the 1-

Exhibit 1

Mile Site, the Tetra Tech derived productivity can only support a maximum rate of 6g/day for pelagic fish and 4 g/day for benthic fish for a single individual. While it isquestionable whether any substantial amount of fishing occurs in this area, it is likely thatif it does occur, more than one individual would be involved. If a total of 10 individualsfished this area each year, then the maximum sustainable rate per person would be 0.6g/day for pelagic fish and 0.4 g/day for benthic fish, or a total of 1 g/day for eachindividual. It is recommended that this maximum sustainable rate be used to evaluatethe 1-Mile Site. This rate is consistent with the median fish consumption rates reportedfor rivers and streams in the Maine angler survey (Ebert et al., 1993). It is important tonote, however, that the median rate of 1 g/day that was reported in the Maine surveywas based on fishing that occurred in considerably larger and more productive riversand streams. Thus, even this rate is likely to overestimate consumption from the 1-MileSite.

In attempting to derive more realistic fish consumption rates for the downstreamreaches, in the absence of specific productivity information for them, EPA has assumedthat there is a 10 percent increase in the fish consumption rate per river mile. There isno scientific justification for making this assumption as distance downstream will not bethe determining factor in rates of fish consumption. Many factors, including accessibility,habitat, availability of desirable species, and availability of alternative fisheries have asubstantial impact on the likelihood that individuals will use and consume fish from aparticular fishery.

Rather than making unfounded assumptions about downstream productivity, it isrecommended that EPA select fish consumption rates from the published literature forsimilar types of waterbodies. As discussed above, Ebert et al. (1993) providedrecreational fish consumption rates for fish obtained from rivers and streams in Maine.The median rate was 1 g/day, the arithmetic mean rate was 6.4 g/day, and the 95th

percentile rate was 12 g/day. Given that these rates are based on fish obtained byindividual anglers who obtained fish from multiple, larger, and more desirable fisheriesthan are being evaluated for Richland Creek, it is unlikely that these rates wouldunderestimate consumption from the Richland Creek reaches being evaluated. Inaddition, in its National Sludge Rule, EPA recommended the use of rates of 1 to 4 g/dayto evaluate smaller waterbodies (EPA, 2003). Thus, it is recommended that the rate of 1g/day be used as a maximum rate for the 1-Mile Site, a rate of 4 g/day be used for the 3-Mile site, a rate of 6.4 g/day be used for the 5.5-Mile Site, and a rate of 12 g/day be usedto evaluate the 12.7-Mile Site.

Development of EPCs

The fish tissue EPCs used for some locations of Richland Creek are highly uncertaindue to the fact that they are based on an extremely limited number of samples. Forexample, in the 3-Mile reach, the total PCB EPC for pelagic fish is based on only threefish samples and the total PCB EPC for benthic fish is based on only five fish tissuesamples. The EPCs for TEQ are based on even more limited data with only one sampleavailable for pelagic fish and one sample available for benthic fish collected from the 3-Mile reach. At the 5.5-Mile reach, there are no fish tissue data available for PCBs orTEQ in pelagic fish, which are the fish that are most likely to be consumed. The onlydata available are for eight benthic fish that were collected.

Exhibit 1

In addition, Tetra Tech has calculated TEQ concentrations for fish tissue samples thatwere analyzed for PCB congeners. However, these calculated TEQ concentrations donot incorporate the most current TEF approach recommended by the WHO (van denBerg et al., 2006). In its revised recommendations for TEF, WHO has modified itsrecommended TEF for a number of PCB congeners. Based on the revised approach,CBS has recalculated the TEQ concentrations for all of the fish tissue scenariosevaluated. The revised TEQ concentrations are shown below and compared with theoriginal values used by Tetra Tech.

Exhibit 1

Reach1-Mile1-Mile3-Mile3-Mile5.5-Mile5.5-Mile12.7-Mile12.7-Mile

Type ofFish

PelagicBenthicPelagicBenthicPelagicBenthicPelagicBenthic

EPA ReportedTEQ (ng/kg)

3.58

1.42.3NE3.3NENE

CBS RecalculatedTEQ (ng/kg)

3.59.31.01.4NE2.3NENE

NE = Not evaluated by Tetra Tech.

Finally, Tetra Tech has based its benthicfish tissue EPCs for the 1-Mile Site on the fishtissue data for white suckers. As was discussed in previous comments on the draftHHRA that were submitted by CBS (2004; 2005), it is extremely unlikely that whitesucker will be consumed. In fact, the productivity study conducted by CBS at the 1-MileSite found that sucker constituted less than twenty percent of the fish biomass (52.5kg/hectare) for fish that might be harvested for consumption (e.g., rock bass, yellowbullhead, white sucker and green sunfish). Thus, it is very unlikely that sucker would beexclusively harvested and consumed from this reach (Christian, 2006).

In its HHRA conducted for the Lemon Lane Landfill Site, EPA acknowledged that whitesucker are "generally less desirable for consumption" than sunfish and that "white suckerare generally expected to make up a small portion of a typical angler's diet" (Tetra Tech,2006b). In that report, however, EPA stated that some anglers may be moreopportunistic and thus may ingest a higher proportion of sucker. While this might be trueif sucker were the only fish available in Richland Creek, this is not the case.Opportunistic anglers catch and consume the fish that they can catch most easily. In thecase of Richland Creek, an opportunistic angler may catch and consume both pelagicfish and benthic fish but it is highly unlikely that that individual will preferentially consumesucker, given the higher available biomass of other, more desirable species. Instead ifthere are individuals who do consume any sucker, they also likely consume otheravailable species. This information needs to be discussed in the risk characterization toprovide perspective on the risk estimates that are presented based on the exclusiveconsumption of sucker.

Toxicitv Criteria

As discussed in previous comments to EPA on the HHRA for the Site (CBS, 2004;2005), there is considerable uncertainty associated with the use of dioxin TEQ toevaluate PCB congeners, including the uncertainties associated with the application ofindividual toxic equivalence factors, as well as the uncertainty associated with theselection of a cancer slope factor for dioxin. These uncertainties and the scientificvalidity of the entire approach have been the subjects of considerable debate amongmembers of the scientific community and were a primary focus of the review of EPA'sDioxin Reassessment, which was conducted by the National Academy of Sciences(NAS).

Exhibit 1

A recent report by the MAS, subsequent to its review of the draft Dioxin Reassessment,noted that only four of the 12 PCB congeners for which WHO has developed TEFs wereincluded in the Dioxin Reassessment and thus were evaluated by the NAS (MAS, 2006,p. 53). The NAS report did not consider the other eight PCB congeners "because ofconcerns about the accuracy of previous in vivo and in vitro toxicological (relativepotency) results." The NAS also concluded that the use of TEQ to estimate toxicity inhumans may be inaccurate or incorrect due to the fact that the TEFs have beendeveloped to evaluate the relative toxic potency of a mixture to which an animal isdirectly exposed by dietary intake but may not be an appropriate measure for internalTEQ concentrations and potential toxic effects (NAS, 2006, p. 61).

The NAS report also cast substantial doubt on the use of the current cancer slope factor(CSF) for dioxin to establish potential cancer risks due to TEQ. The NAS committeeunanimously concluded that the current weight of scientific evidence favors the use of anon-linear model for extrapolating carcinogenic effects of dioxin-like compounds at lowdoses (p. 135). However, in accordance with current EPA policy, the CSF of 150,000(mg/kg-day)"1, which was used to evaluate potential risks due to the presence of thesecongeners in fish from Richland Creek, is based on a linear non-threshold model. Thisissue makes the use of the alternative CSF of 1.5E-06 (mg/kg-day)'1 presented in therisk results even more uncertain.

Finally, the HHRA has reported the results of an evaluation of the noncarcinogenichazards posed by the "dioxin-like" PCB congeners, using a reference dose of 1 pg/kg-day, based on the value it states was proposed in EPA's draft Dioxin Reassessment.However, the draft Dioxin Reassessment does not recommend or propose a referencedose for dioxin or dioxin-like compounds and EPA has never published a reference dosefor dioxin in either its Integrated Risk Information System (IRIS) or Health EffectsAssessment Summary Tables (HEAST). While EPA used an allowable daily intake of 1pg/kg-day to derive its ambient water quality criterion in 1984, and other agencies suchas the Agency for Toxic Substance and Disease Registry (ATSDR) have derivedallowable concentrations based on non-carcinogenic effects, EPA does not discussnoncarcinogenic potential in IRIS or in HEAST. In addition, the screening Risk-BasedConcentrations derived by EPA Region III and Preliminary Remediation Goals derivedby EPA Region IX, which are commonly used as screening values, do not consider thenoncarcinogenic potential for dioxins and provide no reference dose (even a provisionalreference dose from NCEA). Consequently, CBS recommends that the evaluation ofnoncarcinogenic effects due to TEQs be removed from the HHRA.

Due to the uncertainties associated with the evaluation of TEQ, both for the cancer andnoncancer endpoints, CBS recommends that all TEQ analyses be eliminated from theHHRA. Such an analysis is unnecessary and contributes greatly to the overalluncertainty associated with the reported results of the HHRA.

Exposure Duration

EPA has used an exposure duration of 30 years for the fish consumption pathway.While it is conceivable that an individual could fish Richland Creek every year for 30years, it is highly unlikely that this would occur, even if the individual were to reside inthe area for that period of time. Instead, an individual who regularly fishes from year toyear is likely to be an avid sport angler who will visit higher quality fisheries duringregular fishing trips.

Exhibit 1

Although it is highly unlikely that an individual will fish the Site for this extended period oftime, it is possible that it could occur. However, it appears that the EPCs for fish tissues,which are based on recent sampling data, are not accurate predictors of the PCSconcentrations to which fish consumers would be exposed over a 30-year period. Thetrend analysis conducted by QEA on behalf of CBS (QEA, 2006) indicates that PCBconcentrations in fish tissues have been declining at a rate of approximately four percentper year. At this rate of decline, the 30-year average fish tissue concentrations would besubstantially lower than the current concentrations, which have been used as the basisfor estimating potential risks due to the fish consumption pathway. Thus, the HHRAshould clearly note that fish tissue concentrations are declining and that the use ofcurrent fish tissue concentrations in combination with an assumed exposure duration of30 years is a highly conservative approach and likely overstates the potential for cancerrisk.

Surface Water Contact

The HHRA has concluded that direct contact with surface water provides the greatestpotential for exposure due to the direct contact pathways. However, these estimates ofexposure are exaggerated due to the assumptions that have been used to derive them.The overestimated exposure assumptions include the use of the maximum detectedconcentration at each location, inflated assumptions about exposure frequency,exposure time, and frequency of water ingestion, and an unlikely dermal surface area.

EPC for Surface Water

The surface water data upon which the EPCs have been based exclude a number ofsurface water samples that have been collected at RCVP and RC43. While Table B-12in Appendix B reports a total of 3 samples at RC43 and 4 samples at RCVP, there areover 60 surface water sample results available for each of these stations. CBSacknowledges that the majority of those surface water sample results were taken duringstorm conditions and thus are not likely to be representative of the levels of PCBs towhich individuals may be exposed since it is not likely that people will visit these sitesduring significant rain events. However, there have been a few non-storm eventsamples taken at these locations since the 1999 source control remedy. These includethe following:

Station IDRCVPRCVPRCVPRCVPRCVPRC43RC43

Sample IDNL 1893RC0019NL4410NL 4491NL4492NLH 0756NL4437

Date11/14/025/27/033/26/044/16/044/16/048/1/013/26/04

Result0.0140.013

Exhibit 1

Tetra Tech (2006) has based its surface water EPCs for the 1-Mile and 3-Mile Sites onthe maximum surface water concentration at each location, despite the fact that therewere a number of non-detect concentrations at each location and the long-term averageconcentrations were considerably lower. The maximum concentration was usedbecause there were too few samples to calculate an upper confidence limit on thearithmetic mean (UCL). However, it is important to note that the use of a maximumconcentration is not representative of the long-term surface water concentration to whichrecreationalists might be exposed. Instead, when one is considering long-termexposures, the average concentrations measured over time are most relevant, as hasbeen acknowledged by EPA.

Using the analytical results for all of the relevant surface water samples collected at the1-Mile and 3-Mile Sites, and assuming that concentrations listed as non-detect arepresent at half of their reported detection limits, the resulting arithmetic mean EPCswould be 0.046 ug/L for the 1 -Mile Site and 0.09 ug/L for the 3-Mile Site. It isrecommended that these values be used as the EPCs for the surface water pathways inthese areas.

Exposure Frequency

The HHRA uses extremely high exposure frequencies of 34 days/year (adults) and 51days/year (children and youths) to evaluate the surface water exposure pathways for theNorth and South Springs and all portions of Conard's Branch. For all reaches ofRichland Creek, the exposure frequency is assumed to be 68 days/year. This frequencyis based on the assumption that adult and youth recreationalists will spend 4 days perweek for 13 weeks, from June through August, and 4 days per month at each Site duringApril, May, September, and October.

While it is plausible that individuals could occasionally visit the North and South Springsor Conard's Branch, it is not likely that they will visit them, even at the reduced frequencythat has been assumed by EPA. There is no indication that this intensity of recreationalactivity is currently occurring in these areas or would be likely to occur in the future,given the characteristics of the waterbodies. EPA has acknowledged that individualsare not likely to visit these areas, due to their physical characteristics and accessibility.In fact, given these characteristics, it is very unlikely that individuals will spend anyrecreational time in these areas. The springs are shallow and are located at somedistance from public roads, making regular access unlikely. In addition, Tetra Tech(2003b)reports that Conard's Branch has very limited access, the 3 to 5 foot banks aresteeply incised and, with the exception of one large pool, the remainder of the reachranges in width from 1 to 3 feet and is less than 6 to 12 inches deep. As a result, it isvery unlikely that these areas are used for recreational purposes with any regularity.

Similarly, the 1-Mile Site is less than 2 feet deep and the reach ranges from 10 to 30 feetin width. This area does, however, have some side, channel and point gravel bars, andcan be accessed more easily from the road, making it a somewhat more desirablerecreational location. While there are very few houses nearby, it is possible thatresidents could occasionally access this area for some type of activity. Thus, whileEPA's exposure frequency of 68 days/year is very conservative, given that surface watercontact is only likely to occur during the summer months, it is still plausible.

Exhibit 1

As shown in the Data Validation Review of Neal's Landfill Fish Samples November 2003Validation (Tetra Tech, 2004), the banks of the 3-Mile Site are steep in most locationsand heavily overgrown with brush and woody debris, making access difficult. As aresult, it is unlikely that individuals regularly visit this area and have water contact there,as has been assumed in the HHRA. Similarly, the 5-Mile Site of Richland Creekconsists of an upper reach that has deep pools containing woody debris with heavilyovergrown banks, making this type of exposure less likely in this area. Thus, while it isnot impossible for individuals to have contact with surface water in these areas, it ishighly unlikely that young children will participate in these activities in such areas, andvery likely that these activities will occur with much less frequency than 68 days/year.

CBS recommends that reduced exposure frequencies be evaluated for the surface waterpathways in most areas. For the North and South Springs and the reaches of Conard'sBranch, it is recommended that no more than 12 days of exposure be evaluated. Thisequates to a visit to each area occurring one day each week during the warmest monthsof the year.

The higher exposure frequency of 68 days/year for surface water contact at the 1-MileSite is conservative but plausible, given its proximity to a few residences and its greateraccessibility. However, at the 3-Mile and 5.5-Mile Sites, it is not likely that surface watercontact will occur more than 1 or 2 days/week during the summer months. Thus, forthese reaches, an exposure frequency of 24 days/year is recommended.

Dermal Surface Area

EPA has assumed that wading occurs at the North and South Springs, all sections ofConard's Branch, and all of Richland Creek. In addition, EPA has assumed that youngchildren swim during 50 percent of their summer exposure events in Section F. Inderiving dermal surface areas, EPA has used variable depth information to determinewhat portions of the feet and legs might be exposed. While this approach reflects thespecific physical characteristics of the water bodies being evaluated, it does not reflectthe type of wading activity that is likely to occur on a regular basis. In fact, the depth ofwater available does not reflect the depth to which people are likely to wade. In mostcases, when individuals are wading, they will not wade to a depth greater than kneeheight due to the fact that deeper wading may result in loss of balance. Thus, while CBSagrees with using the assumed surface areas for the North Spring, South Spring, andSections Pre-A, A, and B, it believes that the feet and lower legs should be themaximally exposed dermal surface area during wading in the remainder of the reaches.Thus, for young children, older children and adults who wade, the skin surface areas inthese areas would be 1,101, 2,559, and 3,595 cm2, respectively.

It should also be noted, however, that adults and high-school aged adolescents are notlikely to wade in these areas on a regular basis. While it may be reasonable to assumethat young children and pre-adolescents between the ages of 7 and 13 years might beengaged in recreational activities that would involve regular wading, once children reachhigh-school age, regular activity of this type of activity is likely to lose its appeal. As aresult, the assumption that these behaviors occur every year for 12 years is likely tosubstantially overestimate exposure to those individuals. This additional level ofconservatism should be discussed in the risk characterization of the HHRA.

Exposure Time

Exhibit 1

EPA has used an exposure time of 2 hours/day to evaluate dermal exposure to surfacewater. This is a highly inflated estimate of the amount of time that individuals are likelyto wade at the Site, particularly when considered in light of the high exposurefrequencies used in the assessment. It is unlikely that an individual would spend thismuch time in the water during each exposure event. This is particularly true for olderyouths and adults who are not likely to spend substantial amounts of time in the water,even when visiting the creek.

Given the size of the water bodies and their limited appeals as recreational locations, itseems more realistic to assume that an individual spends no more than one hour incontact with the surface water during every exposure event. In its response to CBS'scomments on the risk assessments for Bennett's Dump and Lemon Lane Landfill sites,EPA has acknowledged that its assumption of 2 hours is conservative and they indicatedthat they planned to revise the exposure time assumption in those risk assessmentsfrom 2 hours per day to 1 hour per day. CBS recommends that this same revision bemade for the Neal's Landfill HHRA. Even this assumption would be likely tooverestimate potential for contact for this exposure pathway.

Water Ingestion Rate

EPA has selected a water ingestion rate of 0.0382 L/day for adults and 0.0765 L/day foryouths, based on swimming. They have derived these water ingestion rates by usingthe incidental ingestion of 0.05 L/hour (EPA, 1997) for an exposure time of 2 hours toderive a total of 0.1 L/day. For youths, they have assumed that individuals will onlyingest water during the summer months (13 weeks or 52 visits if there is a visit fourdays/week) and thus have adjusted the ingestion rate by a factor of 52/68 so that theycan combine this ingestion rate with a frequency of 68 days/year to derive their wateringestion rate. EPA has then assumed that adults will ingest water for approximatelyhalf of the time that youth receptors will. This approach is unnecessarily convoluted anddoes not reflect potential for exposure due to this pathway.

Surface water ingestion is only likely to occur if an individual is swimming or becomessubmerged in the water and accidentally ingests some water. The depth of water in theNorth and South Springs and in Conard's Branch is too low to permit swimming or toresult in submersion. As a result, the surface water ingestion pathways should beeliminated altogether from the HHRA for these reaches.

EPA has clearly acknowledged that swimming is not likely to occur in most portions ofRichland Creek. If this is the case, then the only opportunity for an individual to ingestwater would be if they were to fall, become submerged, and accidentally ingest surfacewater. It would not be anticipated that this accidental exposure would happen everytime an individual visits the creek. Instead, it is possible that this type of accidentalexposure might occur one or two times per year. Thus, CBS recommends that ansurface water ingestion exposure frequency of 2 events/year be combined with a wateringestion rate of 50 mL/event to estimate exposures due to this pathway. This willrepresent more realistic estimates of exposures due to the accidental ingestion ofsurface water.

10

Exhibit 1

Bank and Floodplain Soil Contact

EPA has also overestimated exposures and risks due to contact with bank andfloodplain soils due to the use of high exposure frequencies, high dermal surface areas,and an unreasonably high estimate of soil ingestion by 7-18 year old youths.

As with the surface waters of the North Spring, South Spring, and Conard's Branch, it isvery unlikely that individuals will regularly visit these areas and have contact with bank orfloodplain soil. EPA has reduced its exposure frequency for Conard's Branch to reflectits limited accessibility and lower desirability as a recreational area. However, even theexposure frequencies of 38 days/year for adults and 57 days/year for youths andchildren substantially overestimate the likely frequency with which these areas will bevisited or contacted. In addition, it does not appear that EPA has made any adjustmentsfor lower frequencies that would be expected at the North and South Springs.

CBS asserts that there is little potential for actual contact with bank and floodplain soilsat either of the springs or at any point along Conard's Branch. Thus, it recommends thata reduced exposure frequency of 12 days/year, representing one visit per month duringthe summer months, be used for these areas. For the 1-Mile Site, the assumedfrequency of 76 days/year may be plausible given its proximity to a few residences.However, in the downstream reaches, there is no reason to assume that contact withsoils will occur with more frequency than will contact with surface water. Thus a revisedexposure frequency of 24 day/year is also recommended for these exposure pathwaysin those areas.

In addition, the HHRA has used an upper bound soil ingestion rate of 200 mg/day foryouths, despite the acknowledgement in its guidance (EPA, 1997) that the moreintensive soil ingestion, which is sometimes demonstrated in young children, ceases tooccur by the age of six. Generally, risk assessments assume that an upper boundestimate of 100 mg/day is appropriate for older children and adults alike (EPA, 1997;2002). Thus, CBS recommends that this exposure assumption be reduced accordingly.

Finally, the evaluation of exposure due to surface soil, bank soil, and floodplain soilcontact assumes that 100 percent of the soil ingested and contacted on a daily basis willbe contaminated soil. In fact, soils from the contaminated areas are likely to represent avery small fraction of the total incidental soil ingested daily, which will include a majorityof soils from non-contaminated areas (i.e., away from the banks and outside of floodplain areas). As a result, it is not reasonable to assume that 100 percent of the soilscontacted on each day of exposure will be contaminated soil. Given the relatively smallsizes of the floodplain areas and bank soil areas, it is reasonable to assume that nomore than half of the soil contacted daily will be derived from contaminated areas. ThusCBS recommends that an additional factor of 50 percent be used in calculating potentialexposures due to soil contact.

Sediment Contact

While the sediment contact pathways are not important pathways in terms of theircontribution to total risks, the risk estimates for this pathway are overstated. This isbecause the analysis uses (1) the maximum sediment concentration from all reaches asthe EPC despite the fact that many of the sediment samples were non-detect, (2) a veryhigh exposure frequency of 68 days/year (as discussed in the preceding comments for

11

Exhibit 1

the surface water pathways), (3) an inflated sediment ingestion rate that is not supportedby the EPA guidance upon which it is purportedly based, and (4) an unreasonably highdermal adherence factor for youths, for whom one would not expect dermal adherenceof sediment to be any greater than it would be for adults, given the same exposed bodysurface areas. However, the low risks that have been predicted in the HHRA forsediment contact pathways (cancer risks ranging from 3.9E-09 to 1.1E-07, and hazardindices ranging from 2.3E-04 to 6.7E-03), despite the use of these very conservativeassumptions, indicate that sediments at the Site do not pose a risk to the receptorsevaluated. While any refinement of the exposure parameters to make them more site-specific and relevant will further reduce risks, it is not necessary or fruitful, from a riskpoint of view, to provide additional comments on these pathways.

Calculation of Total Risks

In addition to the pathway-specific risk estimates that are presented in the HHRA, EPAhas gone on to calculate total risks by aggregating exposures and risks across multiplepathways. Tables 9 and 10 of the HHRA provide summaries of aggregated total risksand hazard indices, respectively. To calculate these, EPA has summed exposures dueto fish consumption with exposures due to direct contact with surface water, sediment,bank soil and floodplain soil to derive total risk estimates for most of the reachesevaluated. There are a number of problems with this approach that result inunrepresentative risk estimates and provide inaccurate information that can beextremely misleading for decision-making by risk managers.

First, these aggregate risk estimates assume that all individuals who have some contactwith water, soil, or sediment are also fish consumers. In all cases, the risks associatedwith the fish consumption pathways have been combined with the risks associated withother direct contact pathways. It is not unreasonable to assume that individuals who fishfrom Richland Creek may have some contact with surface water or with soil/sediment inthe area at which they fish. However, it is likely that the vast majority of individuals whomay have contact with sediment or soil do not also consume fish from the Creek.

Appendix A of the HHRA makes it clear that the fish ingestion rates derived for the 1-Mile Site are based on the assumption that a single individual is consuming all of theavailable fish. Thus, the total risk estimates that are provided in Tables 9 and 10 mayonly be relevant for a single individual at each reach and are not likely to berepresentative of the recreational population as a whole. Risk managers may be misledby these aggregate risk estimates into assuming that these total risks to individuals whouse the Creek for recreational activities apply to a substantially larger population than isactually the case.

In addition, EPA has combined fish ingestion with the other pathways evaluated at theNorth Spring, South Spring, and Conard's Branch, despite its own acknowledgementthat they do not expect fish to be consumed from these sections of the Site. Theinclusion of fish consumption risks in the total risk estimates for these areas isparticularly misleading because it implies that risks to individuals who visit these areasare much greater than they actually are, since the vast majority of the estimated risksare based on the consumption of fish that are not even present or likely to be consumedfrom these areas. These total risk estimates could consequently lead risk managers tomake remedial decisions for these upstream areas based on activities that do not occurthere. If those risk estimates are used as the basis for remedial activities at Conard's

12

Exhibit 1

Branch or the Springs, those activities will result in no change in residual risks becausethe risk estimates are driven by the fish consumption pathway, which is incomplete inthese areas. Instead, any remedial decisions that are made for North and South Springsor Conard's Branch should be based only on activities that could potentially occur inthose areas, which do not include the consumption of fish. Thus, the fish consumptionpathway should be eliminated from the total risk and hazard estimates that are providedfor Conard's Branch and the North and South Springs.

In addition, EPA presents total risks for pelagic fish and for benthic fish but does notprovide risk estimates for combined fish. If, as EPA asserts, individuals may beopportunistic fish consumers, then it is likely that their total fish consumption rate will notincrease but rather that the mix of fish will change depending upon availability andsuccess. Thus, in order to provide a more realistic estimate of risks due to consumptionof all types of fish, it is recommended that EPA present total risks that represent amixture of those types of fish rather than a sum. It would be reasonable andconservative, if it is assumed that pelagic fish are more desirable but benthic fish aremore available, to assume that 50 percent of the fish consumed are pelagic fish and 50percent benthic fish when attempting to provide estimates of total risk across allavailable species. Such an assumption is plausible but conservative for the 3-Mile and5.5-Mile Sites. However, as discussed previously, the productivity study conducted atthe 1-Mile Site indicated that suckers represented less than 20 percent of the availablebiomass of consumable fish species there. Thus, at this reach, the EPC for combinedfish should be weighted (e.g., 20 percent sucker, 80 percent pelagic fish) to reflect thisreach-specific availability.

Finally, EPA's approach for aggregating total risks and hazards double-counts (for the 3-Mile Site) and triple-counts actual direct contact exposures for the other reaches thatwere evaluated (except the 5.5-Mile and 12.7-Mile Sites where only exposure tosediment was evaluated). In evaluating the potential risks associated with sediments,bank soils, and floodplain soils along different reaches of the Site, EPA has consideredeach exposure pathway as a discrete exposure. Thus, for each evaluation, total dailysoil ingestion rates and dermal contact rates have been combined with media-specificEPCs to estimate potential risks. While this is already a highly conservative approachwhen considering a single medium, it is completely inappropriate to sum the estimatedrisks and hazards associated with multiple media because to do so, yields risk resultsthat are completely unrepresentative of potential exposures in those areas.

For example, the HHRA has used an upper bound soil ingestion rate for adults of 100mg/day, which is assumed to be a total daily soil/sediment ingestion rate. However, forSouth Spring, North Spring, Section Pre-A, Section A, Section B and Section C, EPAhas evaluated potential exposure due to the ingestion of sediment, bank soil, andfloodplain soil separately, each using 100 mg/day as an ingestion rate, and then hassummed these pathways to get an aggregate risk estimate. Thus, these aggregate riskestimates assume that individuals ingest 100 mg of sediment, 100 mg of bank soil, and100 mg of floodplain soil on each day of exposure, for a total of 300 mg/day. This is notcharacteristic of recreational exposures and is not consistent with EPA's recommendedsoil ingestion rate. Instead, if individuals have contact with sediment, floodplain soil, andbank soil on a given day of recreational activity, then using EPA's conservative rate ofsoil ingestion, these individuals will still only ingest a total of 100 mg/day but that total willbe comprised of a combination of sediment, floodplain soil, and bank soil. In order toaggregate exposures for these pathways combined, the HHRA should either apportion

13

Exhibit 1

the total soil ingestion rate among the different exposure pathways or should conduct aseparate calculation that uses an EPC that is based on a combination of sampling datafor all three media, weighted as is deemed appropriate. The same is true for the dermalpathway.

The same approach was used for the 3-Mile Site although only two dupllcative mediawere evaluated (sediment and bank soil). However, the same approach as wasdescribed above should have been used to aggregate risks so that the ingestion anddermal exposure risks are not double-counted, as they currently are.

The same problem is associated with the aggregation of the dermal pathways. For banksoil and floodplain soil, it is separately assumed that the head, hands, forearms andlower legs are in contact with these two media and the media are summed. However,during time spent engaged in recreational activities in these areas, the total surface areawill be constant but the soil/sediment contacting it will either be floodplain soil, bank soil,sediment or some combination of the three media. Similarly, the adherence factorsoverestimate the loading of soil or sediment to the skin when it is assumed that anindividual is in contact with more than one of these media on the same day and the risksdue to both are summed. If EPA intends to continue to combine potential risks due tothese pathways, then it is important that the EPCs be modified to reflect a combinationof media or that the adherence factors be adjusted to reflect the relative contribution ofeach medium to total exposure.

Revised Risk Estimates

In an effort to demonstrate the degree of overestimation of the risk estimates provided inthe HHRA, CBS has recalculated potential risks to these receptors using therecommended assumptions outlined and discussed above. These revised riskestimates, and the basis for each, are provided in the following sections.

Fish Consumption

To recalculate risks due to fish consumption, CBS has used fish consumption rates thatare more reflective of the published literature on consumption of fish from rivers andstreams. Thus, rates of 1 g/day, 4 g/day, 6.4 g/day and 12 g/day have been used toevaluate th

WEIL, GOTSHAL& MANGES LLPWEIL, GOTSHAL& MANGES LLP Thomas Alcamo August 24, 2007 Page 2 people at...

Documents

Transcript of WEIL, GOTSHAL& MANGES LLPWEIL, GOTSHAL& MANGES LLP Thomas Alcamo August 24, 2007 Page 2 people at...