Toxic Chemicals Program, continued March 22, 1995 Chee Choy, Environmental Specialist Bureau of Environmental Services City of Portland 1120 S.W. Fifth Ave., Room 400 Portland, OR 97204-1972 Re: Screening Level Risk Assessment Report for the Columbia Slough Sediment Project, February 1995.

Dear Choy:

This letter constitutes Northwest Environmental Advocates’ (NWEA) comments on the above-referenced SLRA. For the purpose of these comments, we are treating the SLRA as a draft document which, when finalized, will incorporate comments from NWEA, the Oregon Department of Enviromental Quality (DEQ) and others.

Based on our analysis of the SLRA, we anticipate that this document will undergo major changes. We have extremely serious reservations about the methodology employed in this SLRA. The primary reasons, described in further detail below are: 1) the apparent adoption of an “acceptable” risk level of 10-4 in contradiction with federal and state law and policy; 2) the use of averaging techniques in order to create the appearance of risks which are lower than those actually measured; 3) the lack of analysis of the relationship of this RI/FS with on-going dredging of most Slough waterbodies; and 4) an overly restrictive approach to source control that is neither mindful of Clean Water Act requirements nor on-going and future Total Maximum Daily Loads (TMDL).

Role of the SLRA

• The SLRA collected and analyzed only the top 2 cm. of Slough sediments. SLRA at 2-3. As the RI/FS proceeds, the City, with the regulatory oversight of the Oregon Department of Environmental Quality (DEQ), will evaluate how to go about sediment remediation of sites identified in this document based on those top 2 cm. Meanwhile, the Multnomah County Drainage District No. 1 is dredging the Slough on an on-going basis. By failing to acknowledge this fact, the SLRA fails to analyze the implications of this dredging to the sampling protocol chosen, the results, and ultmiately the usefulness of the entire risk assessment. Worse yet, it fails to address issues related to the real world including: 1) the appropriate disposal of the dredged spoils; 2) the impact to beneficial uses of
  resuspended contaminated sediments; and 3) the utility of doing the RI/FS in the first place.
• This document purports to not be a decision document. For example, the following statement appears in numerous places in the document: “The results of the SLRA are, therefore, best used to determine relative ranking between sites, not to determine absolute site risks. Determination of absolute site risks for the riskiest sites will be conducted as part of the focused remedial investigation.” However, the SLRA proposes to identify the sites of highest risk using cut-offs allegedly based on analyzed risks. These determinations will then be used to determine future data collection, evaluation and ultimately remediation plans. In fact, as it stands now, the SLRA is a decision document.
• On the other hand, if it is a decision document, the SLRA should not use the relative approach. While, ultimately, priorities must be set, the actual risks should dictate the City’s remediation response. For example, the Hazard Score ranking is a relative system which compares scores rather than simply draws conclusions about risks as part of an absolute system.
• The document does not consistently identify the role of the SLRA. At several locations the SLRA states that it “is designed to identify chemicals that present the greatest potential risk to aquatic life [and wildlife] while eliminating from further evaluation those posing negligible risk.” SLRA at 5-1 and 6-1. Elsewhere, the document states that “the scope of this SLRA…is to focus the remedial investigations on those sediment locations presenting the greatest relative risk.” SLRA at 5-3. Which is it? The work done in this SLRA certainly does not merit the elimination of certain chemicals from consideration. The SLRA does not even mention dioxins and furans and is very slim on references to recently published works in a field which is developing rapidly. If the SLRA is proposing to eliminate certain chemicals from consideration for the risks they pose to wildlife but not humans, that is equally inappropriate.
• The SLRA states that “Other factors, including public use and the impact of source control measures, will be factored into a final sediment location hazard ranking at a later date.” SLRA at 7-2. No rationale is given for not factoring in information at this stage about existing disturbances such as dredging and other in-water activities. The SLRA should state a rationale for including “public use” as a basis for ranking given the fact that fish move through the Slough and wildlife use all areas.
• For Priority C sites, the SLRA recommends monitoring for sites related to inactive sources and only “possible follow-up monitoring” for those sites related to active sources. SLRA at 7-3. Priority D sites for inactive sources are recommended delayed but Priority D sites for active sources would be subject to a “case-by-case evaluation for possible source control and enforcement.” SLRA at 7-3. There is no rationale presented for the different treatment of sites related to active vs. inactive sources. Additionally, the consultants should be cautioned from showing their bias towards protecting the interests of operating industries.
• The discussion of source controls and monitoring does not explain how these activities fit within current and future Total Maximum Daily Loads (TMDL).

Public Input

• The SLRA Figure 1-1 shows only one point of input from citizens, regulators, and a “technical action committee.” The technical action committee should be identified and information provided about how the public can observe its deliberations. The SLRA should explain why, if this is not a decision document, this is the only point in the RI/FS where the City intends to take public input. The SLRA should clearly identify how the City intends to factor in public and regulatory input received on this document.
• Given the importance of this stage of the RI/FS, the fact that it appears to be the only stage at which public input is being solicited, and the extremely informal manner in which public input has been taken, the SLRA should note which agencies have been provided with an opportunity to comment on it.

Pollutants of Concern

• The Slough is currently deemed “water quality limited” (WQL) for bacteria, chlorophyll a, pH, copper, lead (RM 0-6), dioxin (RM 0-7), lead (N. Slough), PCBs (N. Slough). It can be anticipated, based on the tissue residue information presented in Table 4.1-2b of the Oregon DEQ’s 1994 305(b) report, that the Slough will also be determined to be WQL for PCBs, lead, zinc, mercury, arsenic, dioxin, copper, cadmium, and chromium. DEQ may also agree, or may be required to agree as a result of litigation in Northwest Environmental Advocates v. Browner (Oregon TMDL program), that the Slough is also WQL for lead, chromium, PCBs and benzidine on the basis of sediment contaminant levels. Additionally, if the Slough is deemed WQL for toxic pollutants, such as dioxin, found in fish tissue, it is irrelevant whether the contaminant has actually been measured in the sediment in order to deem the levels in the sediment to be presenting an unacceptable risk.
• Dioxin is very toxic and is not factored into any of the risk assessment presented in the SLRA nor even mentioned once in the document. This oversight is not explained anywhere although the document claims to build upon the historical data on the Slough and not only have dioxins have been detected in the Slough but these data are currently the basis for a DEQ determination that RM 0-7 is WQL for dioxin. Dioxin, according to current water quality standards, poses a risk at levels not detectible in water and sediments. While the SLRA estimates the levels and risks of benzidine, also widely not detected, the analysis does not even contemplate doing estimates for dioxins.

Treatment of Pollutants

• Trichloroethylene (TCE) is stated to be of “low toxicity” to humans and mammals. SLRA at 4-17. The document also notes that it is a possible human carcinogen. This pollutant is not included in the risk assessments on the basis of insufficient information on the risks that it poses. If TCE is of such little concern in this RI/FS, why is it the pollutant of concern at a nearby Superfund site?
• The SLRA contradicts itself. The document notes that there are chemicals for which no ecological effects data could be found. SLRA at 6-9. On the other hand, it states that additional data searches will be done during the detailed risk assessments. Either the materials exist or they do not. Based on our reading of the SLRA we believe that an inadequate data search has been conducted. No rationale has been provided for putting off the incorporation of significant advances in science at the SLRA stage of the RI/FS.
• The SLRA makes no reference to EPA’s Dioxin Reassessment. Health Assessment Document for 2,3,7,8-Tetrachlorodibenzo-p-Dioxin (TCDD) and Related Compounds, August 1994, External Review Draft. The Dioxin Reassessment not only makes relevant observations about the risks associated with dioxin which should have been analyzed in the Slough, but also about the body burdens to which Slough risks add. Equally important, the Dioxin Reassessment draws conclusions about the risks posed by dioxin-like co-planar PCBs.
• References on risk assessments do not include the work done by the National Marine Fisheries Service (NMFS) on the sublethal effects of brief exposures to toxins on salmonids. Salmonid populations have been reported to be present in the Lower Columbia Slough. Regardless of the species present, the analysis of sublethal effects on fish is relevant to the evaluation presented in the SLRA.
• The SLRA notes that PAHs were not detected when measured in surface water. SLRA at 3-5. The fact that they were not detected and are not expected to bioaccumulate does not mean PAHs are not a problem.

Risk Assessment

• The methodology used by the SLRA to determine the relative risks of sites, the Total Hazard Scores, is flawed. SLRA at 7-1. Instead of averaging the human health and ecological scores, the SLRA should identify which areas pose the highest risks for cancer, non-carcinogenic toxicity, and ecological values. There is no rationale given for averaging, a technique which results in avoiding the identification of risks where they exist. Averaging also occurs in the development of the human scores and the ecological scores with the result of smoothing peaks and creating the appearance of less than calculated risks. Simply put, there is no rationale for a process that can take a showing that nearly 100% of sites pose a risk of cancer greater than 1×10-6 and turn it into a conclusion that, at most, 34 sites should continue to be subject to the RI/FS.
• The SLRA recommends the delay of all action for 185 Priority D sites and the delay of most action for 81 Priority C sites on the basis of Total Hazard scores and presumably these sites’ presenting risk levels within the 1×10-4 to 1×10-6 range. SLRA at xvi. However, no sites which represent a cancer risk greater than 1×10-6 — which is nearly 100% of all sites — should be eliminated from study under the RI/FS. SLRA at 4-10, Figure 4-2.The SLRA fundamentally misunderstands federal and state law and guidance on this issue. For example, the SLRA states that most cancer risk levels “fall within the range of risks identified as ‘acceptable’ by EPA under Superfund guidance (e.g., 10-4 to 10-6).” SLRA at 4-10. This is not an accurate reflection of federal or state law. EPA requires clean-up to the 1×10-6 level unless it is shown to be impracticable. In the event that meeting the 1×10-6 level is impracticable, the clean-up levels are set at a level that is practicable but in no event presenting a greater than 1×10-4 level of risk. This is refered to as the “point of departure concept.” See, e.g., 55 Fed. Reg. 8718, March 8, 1990. Oregon requires a clean-up to background levels unless impracticable. Even bills introduced by industry for consideration in the current legislative session identify practicable as 1×10-6, not the 1×10-4 level advanced by the SLRA. The SLRA has presented no evidence on the question of practicability.
• The SLRA establishes a scheme for four Priority Classifications but provides no basis for the cut-offs of hazard scores. SLRA at 7-3. For example, it states “Based on the final magnitude and distribution of the hazard scores, a hazard score greater than 1,000 was used as the basis for assigning Priority Classification A.” SLRA at 7-3. Since the ranking should be done on the basis of the risk assessment, the failure to address the connection makes the choice of hazard scores appear arbitrary. The SLRA provides no insight into what it means to base the cut-offs on the “final magnitude and distribution” of the scores.
• For humans, otter and heron values the SLRA notes several times that the results should be used to identify sites of higher relative risks rather than interpreted as absolute risks. Only for the otter and heron values, however, does the SLRA suggest obtaining corroborating evidence of impacts to populations. The City appears to be taking the position that wildlife must be treated as laboratory guinea pigs. Moreover, the suggestion that only a very gross measurement of risk, namely population declines and poisoned individuals, is an appropriate indicator of whether predicted risks were legitimate is objectionable. The SLRA does not suggest who will conduct such studies. Nor does it suggest how long such studies would take given the multigenerational impairment caused by contaminants producing sublethal effects. The SLRA does not state, but may imply, that remediation should be put on hold while these studies take place. If not, then the studies would merely be academic.If the SLRA wanted to better identify the actual risks to wildlife species, it should not have homogenized all the “other” fish, not all of which are necessarily in the food chain of the otter or the heron.
• The use of the most highly exposed mammal and bird species does not necessarily ensure the protection of all other species. SLRA at 6-1. Some species are more sensitive than others to stressors including contaminants. For example, otter was chosen because it is believed to have a diet that is 100% fish/crayfish. However, mink, also known to have a diet very high (90%) in fish are known to be the most sensitive species to contaminants such as PCBs. Relatively little information is known about the sensitivity of otter. In the last five years, numerous studies have attempted to identify the role of contaminants in both species, none of which are cited as references. Some species could be more a risk based on their ratio of contaminant intake to body weight. The justification should be presented as to why the representative species were chosen and an explanation given on why they are representative of the most sensitive wildlife beneficial use both present and affected by the contaminated sediments.
• Assumptions on the consumption rates and sensitivities of amphibians and reptiles is not supported by any scientific evidence. SLRA at 6-3.
• Risks to wildlife should not be averaged. Instead, the worst risk should be evaluated.
• The SLRA claims to be conservative because at various stages in the analysis conservative assumptions are made. The document also identifies non-conservative assumptions, for example where certain pollutants and risks are not included in calculations. However, the SLRA does not identify all the non-conservative assumptions or factors. Notwithstanding this mixture of largely non-quantifiable conservative and non-conservative assumptions, the SLRA concludes that the analysis has been conservative. This is a fallacy.
• What the SLRA gives, namely noting certain conservative assumptions, it takes away by stating that the results may indicate unrealistically high “paper risks.” SLRA at 2-2. To be truly conservative, the SLRA should embrace the conservative assumptions based on the current inadequate state of our knowledge and the lessons of the past. Going through the motions and then condemning the result is not the hallmark of a conservative approach.
• The SLRA notes that it is “not possible to accurately predict whether the effects of a particular combination will be antagonistic, additive or synergistic.” SLRA at 6-12. The most conservative analysis would consider the synergistic effects of multiple pollutants, not just the additive effects. While the SLRA dismisses synergism as just one possible interaction among multiple chemicals, it is in fact a documented effect. This lack of conservatism should be noted more clearly and accounted for where possible.
• Use of one-half of detection limits to evaluate non-detections is appropriate. SLRA at 3-2. As changes in technology advance the level of detection, the analysis of risks and complex biotic responses to contaminants at ever lower levels of exposure keeps pace. Even the SLRA acknowledges that if the Slough had been evaluated even ten years ago, the levels of non-detection would have been considerably greater. SLRA at 4-12. It is therefore likely that as technology improves, today’s non-detections will give rise to detections. The SLRA’s statement that this practice gives rise to inflated “paper risks” is inappropriate. These paper risks are balanced against very real but unquantified risks present in the Slough environment.
• The SLRA proposes to only utilize predicted levels of contaminants in tissue. SLRA at 6-5. Where actual chemical concentrations measured in tissue are higher than those which are predicted, the actual levels should be used. They represent the risks presented by contaminants in the Slough and should be used to determine where risks pose too great a danger and remediation is necessary. Predicting the risks from site specific tissue levels should only be used for establishing the relative priorities between sites as remediation proceeds. The beneficial uses dependent upon the Slough are affected by the total — real — risks.
• Risks from other contamination must be factored in to adequately assess the additional risks posed by the contaminated sites. The decision not to include the risks from surface water contamination because “chemicals in surface water may originate from many sources” simply ignores the fact that all the species evaluated suffer from the risks of the sites in addition to whatever other risks are present. SLRA at 3-5. The SLRA states that its model does not include so-called “short-term contributions of chemicals” such as CSOs, stormwater. SLRA at 3-7. During the rainy season, however, CSOs and stormwater are not short term contributors to the Slough waters; they provide a substantial part of the volume of the Lower Slough. Clean-up levels should consider all sources of risk. The decision to clean up sites should consider the total risks because the end point is protection of the beneficial uses.
• Carcinogenic risks posed by the contaminated sediments to humans similarly avoid reference to body burdens. SLRA at 4-10. The real risks presented to those who consume Slough fish are greater than just the risks from the contaminated sediments.
• The SLRA notes that health advisories are done in various states on the basis of 1×10-5 to 1×10-6 risk level and that clean-up standards are established on the basis of a 1×10-4 to 1×10-6 risk level. SLRA at 4-12. Besides grossly misstating legal requirements on clean-up standards, this observation points out that clean-up standards set at the 1×10-5 level are equivalent to health advisories in other locations. Clearly such levels of clean-up would not be protective of the beneficial uses.
• The SLRA assumes that both children and adults “can” swim in the Slough. SLRA at 4-3. It is worth noting that children and adults do, in fact, swim in the Slough.
• The SLRA concludes that “the use of U.S. EPA-derived reference doses likely will result in overestimates of potential risk.” SLRA at 4-7. This is not known to be true because EPA reference doses are out of date. See, e.g., the Dioxin Reassessment. See also, Reproductive and Developmental Toxicants: Regulatory Actions Provide Uncertain Protection, U.S. General Accounting Office, October 1991, which finds that two thirds of decisions for study chemicals were not based on reproductive and developmental toxicities.
• The SLRA relies on EPA criteria which are designed to protect 95% of species in an aquatic community. SLRA at 5-4. The SLRA should evaluate whether these criteria are still up-to-date. For example, the water quality standard for copper has widely been called into question. In evaluating the strengths and weaknesses of the ecological effects the SLRA states that the sensitive species these criteria intend to protect may not inhabit the Slough. SLRA at 5-5. Either the species do or they do not. The SLRA should not take credit for conservatism on a whim.
• The SLRA states that an HQ target value equivalent to 0.3 was chosen conservatively based on the possibility of additive or synergistic interaction. SLRA at 5-6. It is not possible for one number to do both. In fact, the SLRA appears to acknowledge that when in the same paragraph it states that there is evidence that toxicity is additive. SLRA at 5-6. It further cites a study for the proposition that additive effects usually stop after more than three or four chemicals. SLRA at 5-7. The SLRA makes no mention, however, of any studies on synergistic effects, notwithstanding their documentation. For example, synergistic effects of mercury and PCBs have been measured in mink.
• The risk assessment does not include the role of other non-contaminant stressors such as temperature, low dissolved oxygen, high pH, disease or habitat pressures. SLRA at 5-3. Such stressors can be significant. For example, the combination of cold temperatures with contaminants are known to cause mortality in mink. At the very least, the contribution of these other stressors to the risks of the species evaluated should be sufficient cause to eliminate the frequent references to over-conservative “paper risks.”
• Comments on bald eagles regarding the size of their range, their potential for exposures, and the role of non-aquatic prey in their diets may be true for those frequenting the Slough but not for all bald eagle populations. SLRA at 6-2 and 6-4. In fact, evidence exists to the contrary for all of the above. Supposition should not take the place of fact.
• The SLRA makes much of the uncertainty factors which range up to 2,000. However, the analysis does not include dioxin, the Dioxin Reassessment, chemicals for which toxicological data is lacking and many other factors. Discussion of large uncertainty factors should be balanced with discussions regarding lack of conservatism. Use of statements regarding maximums such as “range up to 2,000” can be deceiving and should be augmented with other information as they are when the SLRA reports on maximum contaminant loads and maximum risks.
• In reference to the lack of dermal toxicity data for wildlife, the SLRA should investigate research on turtle shells as contaminant pathways.
• Toxicity data for wildlife is based on the Contaminant Hazard Review series. SLRA at 6-6. These are an excellent source but they are dated. There are numerous more recent studies not cited. Use of older studies is a basis for adjusting the uncertainty factors.
• The SLRA does not discuss consistency of the consumption assumptions underlying human estimates with results of other fish consumption surveys. SLRA at 4-5. For example, the Columbia River Intertribal Fish Commission fish consumption survey and the Dioxin Reassessment both note likely rates of fish consumption for subsistence and recreational diets respectively.
• Results from the risk assessments for otter and heron indicate high levels of risk to these species. In response, notwithstanding the SLRA’s position that these are modeled risks from specific areas of contamination which will not be borne out in, for example, actual fish tissue levels, the SLRA states that “Corroborating data will be sought for the detailed risk assessments — for example, declines in resident otter [and heron] populations, or sick or dying animals — which could directly support the conclusions of significant risk to wildlife species.” The SLRA does not explain why such corroborating data, which would not represent information on the risks posed by any specific site(s), would be either useful or appropriate.

Sources of Pollutants

• The work to identify sources should begin immediately and is long overdue. The public (tax and rate-payers) should not have to pick up the tab for costs associated with the RI/FS and remediation actions that stem from the practices of currently operating industries which have and/or continue to contribute to the contamination. Source control work will not begin until sources of toxic contaminants are identified therefore this identification should proceed as expeditiously as possible.
• The SLRA makes no reference to the Total Maximum Daily Loads (TMDLs) that are being and will be developed in the future. Between current and future TMDLs, source controls for many of the toxic contaminants discussed in this report will or should be addressed. For the sake of the environment and bureaucratic efficiency, these two programs should work together and the SLRA should reflect that.
• In no event should priorities for source control efforts be established on the basis of calculated risks in the SLRA or any other remediation document. First, acceptable risk levels for remediation are not the same as those used for establishing water quality standards. Second, the SLRA presents no analysis of the fate of pollutants from various sources. While some high priority sites, as ranked by the SLRA, are likely to be related to specific sources, the inverse is not necessarily true. Some high priority sources may not be related to high priority sites. High water volumes associated with the 13th street CSO, for example, could drive contaminants in its effluent farther into the Slough than a lower volume source. In order to both meet TMDLs and protect the environment, source control must be based on more than the contamination of specific sites in need of remediation.
  The SLRA notes at least one reason why there is no rationale for setting priorities for source control efforts based on priorities for site remediation. The SLRA notes that historical data collected prior to 1989 were deemed unusable because of “age and therefore chemical fate (e.g., degradation, transportation, etc.)…” SLRA at 2-1. The fact that contaminants move in aquatic systems is the very reason why source controls should be implemented across-the-board.Source control efforts should proceed in accordance with the water quality-based approach embodied in section 303 of the Clean Water Act which requires the protection of the most sensitive beneficial uses. The statement that source controls “will be an on-going activity” is highly questionable in the absence of any evidence. To date no source controls, with the exception of those stimulated by litigation (e.g. CSO removal), have been implemented. In fact, the SLRA itself diminishes the importance of source control by ignoring information on water column, sediment transport and source information. See, e.g. SLRA at 3-5.
• Statements regarding source controls made by the consultant in this and future documents should be consistent with the Clean Water Act and should identify the governmental entity responsible for or taking responsibility for implementing the law. The SLRA appears to suggest that the City, rather than the DEQ, will be responsible for any enforcement or implementation of source controls. That is not accurate. The SLRA should identify the role of the “Source Control Division of BES” and explain why it hasn’t done anything and the nature of its relationship with the DEQ.
• The SLRA states that “When appropriate, Priority A sites are also recommended for focused RI/FS evaluations.” What does this mean? The SLRA should state what factors will be taken into account to deem “appropriateness.”
• Given the length of time the City has been working on Slough issues it is remarkable that it apparently has not yet entered source data on a GIS format in order to be able to compare known sources with the sites. This is a strong indication of the City’s commitment to environmental clean-up and to serious source control in particular.
• The SLRA does not provide any justification whatsoever for designating sites based on their proximity to active and inactive sources. On its face it appears to be a convenient way to put off doing remediation for those sites related to operational facilities. Doing source control at sites near active sources is not mutually exclusive with remediation any more than doing source control for all sources in the Slough across-the-board is mutually exclusive with conducting remediation activities at sites near inactive facilities.

Sampling

• A substantial percentage of the sediments identified as presenting the highest carcinogenic risks (26 out of 39 of those with risk levels greater than 1×10-4) came from the random samples. SLRA at 4-10. The only samples with HIs greater than 1.0 for 1994 data were random samples. SLRA at 4-11. The SLRA should evaluate the implication of these results that more random samples should be taken.
• The SLRA should discuss the likelihood that any sample, but particularly those chosen on a random basis, could have been collected at the edge of a toxic area.
• The SLRA notes that a weakness of its tissue analysis was the compositing of all fish species other than carp and crayfish. SLRA at 2-5. This is a very questionable practice and no scientific rationale is provided nor are the problems identified with such a practice thoroughly analyzed. The SLRA notes one problem, namely that species accumulate chemicals to different degrees. It provides no evidence that potential such averaging or smoothing of data would seriously compromise the results. The SLRA does not note, however, that by compositing the “all other species” category, this data cannot be used to analyze the food chain connections to higher levels. Not all higher food chain mammals and birds consume all types of fish. The potential for underestimating the accumulation of contaminants exists by averaging all fish species and not knowing which are eaten by whom.
• Fish tissue generally represents a larger geographic area than sediment data. Tissue is a great indicator of the contamination levels of the entire waterbody including but not limited to the areas of highest contamination. Tissue is a good indicator of why source controls should be implemented in a very restrictive fashion from all sources. The SLRA should discuss these issues.
• Certain pollutants such as aromatic hydrocarbons (AH) do not bioaccumulate in fish. Studies in the Puget Sound have employed evaluation of stomach contents as well as hepatic and biliary levels of pollutants such as the AHs in order to more able to evaluate their impacts. The SLRA does not discuss these methods and the shortcomings of their limited tissue sampling approach.
• The SLRA notes that a weakness in its tissue data is that the fish were caught in areas up to three miles long. SLRA at 2-5. The fact that fish move is a problem encountered in any collection of field tissue. The SLRA does not use this opportunity to discuss the results of the biomonitoring experiments conducted by Fishman in 1989 and 1993 that may have used caged clams and crayfish. Conversely, if this data were used in the SLRA, the document should discuss the appropriateness of doing so.
• The failure to assess any surface water contaminant loads in the SLRA is the loss of another sampling season to begin the work of source control. Source control has yet to begin notwithstanding the many regulatory mechanisms including the RI/FS in which it could take place. Source control costs less to taxpayers and to the environment than remediation of already contaminated sediments. Therefore such work should begin immediately. The SLRA should note and evaluate the relationship of the RI/FS with current and future TMDLs.
• Sediment sampling was performed for only the top 2 cm. SLRA at 2-3. The SLRA presents no rationale for the choice of 2 cm. This sampling protocol was used in the Bi-State Lower Columbia River Water Quality Program based on an alleged intention to sample only those sediments deposited within a year. (NWEA did not agree with this decision at the time.) This is noted as a weakness but no substantive discussion evaluates the ramifications of what could be an extremely serious flaw in the study design.
• The SLRA notes as a weakness in its sediment sampling that it was done at high-flow conditions. SLRA at 2-3. The dates of sampling are not noted. Most important, the SLRA does not discuss the ramifications of this weakness to the study’s results. Nor does it state what future actions will be taken to assure that the entire 1994 data set is not flawed. The SLRA does not describe how focused remediation investigations of high-risk locations identified through this study will establish whether or not the weaknesses in study design (e.g., 2 cm and high-flow conditions) were fatal. There is no evaluation of the ramifications of the combination of the two weaknesses and dredging activities.

Sites

• The SLRA failed to take into consideration the not-inconsequential ramifications of on-going dredging activities in the Slough. The SLRA should present the history of this dredging with a map, text and tables, obtaining the information from the Multnomah County Drainage District No. 1. The SLRA should also note the dredged spoil disposal sites.
• Sites were chosen on two bases: random or connected to potential sources. SLRA 2-2. The SLRA makes no mention of why no sites were chosen on the basis of being depositional areas although deposition is known to be the reason for higher concentrations of contaminants from a wide variety of sources including non-point sources (NPS) in other waterbodies. The SLRA appears to conclude, without testing the hypothesis, that the only sources of real concern are point sources rather than the accumulations from multiple sources due to deposition. The only reference to NPS is a statement on the usefulness of the random sites. SLRA at 2-2. If NPS are a problem, it makes no sense to talk about setting priorities in source control efforts on the basis of point sources only.
• The SLRA does not clearly distinguish between an active and an inactive site. An inactive facility could have substantial contamination in its (permitted or unpermitted) run-off.
• The frequency of targeted samples in the Lower Slough was four per mile. SLRA at 2-2. In the Upper Slough it was two per mile. Notwithstanding the lower frequency of random samples in the Upper Slough, this waterbody yielded a higher number of highly contaminated sites. This suggests that further random sampling should be done in this area.
• The SLRA does not mention the East Multnomah County Superfund site. There must be at least some remote connection between the work being done on the two sites. At the very least, the maps in the SLRA should extend to the full extent of the Slough area, including the East Multnomah County Superfund site and note its existence. Finally, since the SLRA seems to question the toxicity of trichloroethylene, perhaps the City should look into the East Multnomah site since TCE is the primary pollutant of concern there.
• The SLRA does not address Smith and Bybee Lakes, unattached lakes and “stranded slough reaches.” The latter phrase should be explained and the reaches identified. If these reaches are no longer connected but once were connected to the Slough and its contaminants and respective sources, the SLRA should address them. The SLRA should note the extent of the work being done by Metro and the City and the rationale for excluding Smith and Bybee Lakes, which are connected to the Slough, from the RI/FS.

Miscellaneous

• The SLRA notes that chemical results from earlier field screening were unusable because of inadequate QA/QC. SLRA at 2-1. The SLRA does not evaluate the potential that its own data, which apparently used higher than possible detection limits, will similarly be unusable. By doing so, the City might better evaluate whether the cost savings from data analysis at less than the best QA/QC and detection limits in future data collection efforts is worth the inability in the future to use the data at all.
• The SLRA should note that the multilingual posted warnings in the Slough were created and installed by Northwest Environmental Advocates. SLRA at 4-2. The SLRA should evaluate the benefit of installing additional warning signs.
• Figure 1-1 should identify the dates at which events occurred and are scheduled to occur under the Consent Decree or other processes.
• The background on page ix should state that the consent decree requiring the RI/FS was entered into because of a lawsuit threatened by Northwest Environmental Advocates.
• The juxtaposition of the consultant’s conclusion that the toxicity of TCE is questionable and certainly not quantifiable with the designation of Superfund status to the East Multnomah County site undermines the credibility of the SLRA. Consultants often are reluctant to do any significant underlying research to understand the details of the chemicals present and the specific beneficial uses at risk. The failure of the consultant in the SLRA to better explain its rationale for the choice of wildlife, the implications of homogenizing fish species for analysis, and its apparent lack of interest or knowledge about TCE and the East Multnomah site suggest that this project is getting less than the attention it deserves.

Conclusion

Due to lack of time, we were unable to include many citations to recently published studies that bear on the risks associated with contaminants present in the Columbia Slough. We can provide you with this information if you desire it. We have also not quoted EPA regulations and guidance but we strongly urge you to note the inconsistency of this SLRA with the law on remedial actions.

We are gravely concerned with the analysis and conclusions presented in this document. Please contact us soon to indicate how you will be finalizing this portion of the RI/FS. We would also appreciate your forwarding to us copies of written comments from other parties.

Sincerely,

Nina Bell
Executive Director

cc: Anne Levine, DEQ Northwest Region