• You are here: EPA Home
• About Region 5
• Superfund
• Ecological Risk Assessment & the Ecological Technical Center
• Region 5 Case Studies
• Indiana Harbor Site

Indiana Harbor Site Proposed Sediment Clean-up Goals

Executive Summary

The purpose of this report is to derive sediment clean-up numbers for appropriate ecological resources using existing data. The ecological resource that was deemed appropriate was defined as the assessment endpoint: functioning benthic invertebrate communities. Proposed clean up goals (PCUGs) were derived from a large database of sediment data that included the Indiana Harbor Area. The ranges of PCUGs are presented in Table 2.

The PCUGs are values that can be used as target goals for clean up activities. The lower number of a given PCUG range is the concentration below which no adverse effects would be expected. Therefore, cleaning up to below that number should result in no adverse effects. This is shown with the "less than" sign. The higher number of the PCUG range is the concentration above which there is a 50% chance that adverse effects would occur. The concentrations between the lower and higher values are unknown areas, where the occurrence or lack of occurrence of adverse effects in this range cannot be predicted.

There are uncertainties associated with the derivation and use of these PCUGs. While these uncertainties can not be quantified at this time using available data, this does not mean that the PCUGs cannot be used as targets for removal actions at this site.

Background

Routinely, in the Superfund program, we perform an ecological risk assessment to determine what, if any, contaminants at a given site are having adverse impacts on ecological receptors. However, on this site, the Natural Resource Trustees performed considerable work in the area and produced An Assessment of Sediment Injury in the Grand Calumet River, Indiana Harbor Canal, Indiana Harbor, and the Nearshore Areas of Lake Michigan (USFWS 2000). A significant data set was used to support this report that included sediment contaminant concentrations, toxicity tests, and benthic community surveys collected from the Grand Calumet River and the Indiana Harbor Canal. While the data are spread out over the entire area with only a few locations in the area we are specifically addressing in this action, these data can be used to generate sediment PCUGs at this time.

The PGUGs were derived using this data set in a manner that is consistent with Ecological Risk Assessment Guidance for Superfund: Process for Designing and Conducting Ecological Risk Assessments (U.S. EPA 1997). The Guidance details an eight-step process (described on this website: Eight-step Ecological Risk Assessment) for determining risk, making risk management decisions and developing preliminary remedial goals. To maintain consistency with the guidance, an assessment endpoint was first selected. An assessment endpoint is the "explicit expression of the environmental value that is to be protected" (U.S. EPA 1992). An assessment endpoint (i.e. something to protect) is necessary to derive concentrations that are protective. The risk to that assessment endpoint was then qualitiatively evaluated. Finally, the sediment PCUGs were calculated using the methods detailed below.

For purposes of this evaluation, "functioning benthic invertebrate communities" was used as the assessment endpoint.

Rationale for selection:

Benthic invertebrate communities are comprised of a mixed group of organisms that inhabit bottom substrates. Benthic invertebrates are susceptible to exposure to contaminants because they live and feed directly in the sediments where most contaminants are concentrated.

Benthic invertebrates play a vital role in the aquatic community, including mineralization, recycling of organic matter, nutrient and energy cycling which support the productivity of the entire ecosystem. Benthic invertebrates usually comprise a significant amount of biomass (the sum total of all living biological matter in a particular area) in aquatic systems and as such, are considered an important energy source for the system. Many species of fish, mammals, birds and other wildlife rely primarily on benthic invertebrates as a food source and for some species, benthic invertebrtes may the only food consumed.

Risk Evaluation

U.S. Fish and Wildlife Service (USFWS 2000) performed a detailed analysis of the sediment contamination in the Grand Calumet River and Indiana Harbor Canal. Their analysis included sediment chemistry, benthic toxicity testing, and benthic community surveys. They concluded that two lines of evidence – sediment chemistry and sediment toxicity – indicated the presence of conditions sufficient to injure sediments and sediment-dwelling organisms in areas that included the Indiana Harbor Canal. The evidence suggests that there may be risk to the selected assessment endpoint: functioning benthic invertebrate communities.

Derivation of Sediment Preliminary Clean Up Goals

In U.S. EPA 2000, a database was developed that contained 92 published reports and 1657 sample locations. This database included the data from the Indiana Harbor area. One purpose of that report was to evaluate the ability of consensus-based sediment quality guidelines to predict toxicity in freshwater sediments from the Great Lakes Region. These consensus-based sediment quality guidelines are taken from MacDonald et al. (2000), also discussed in USFWS (2000). These values are presented in Table 1.

In MacDonald et al. 2000, the purpose of their evaluation was to derive sediment concentrations that could be used to predict toxicity, or lack of toxicity. The threshold effect concentration (TEC) is the concentration below which adverse effects are not expected to occur. The probable effect concentration (PEC) is the concentration above which adverse effects are likely to occur more often than not. MacDonald evaluated the usefulness of these sediment effect concentrations (SECs) and found that 21 TECs, including four trace metals, eight individual PAHs, total PAHs, total PCBs and seven pesticides were able to accurately predict the absence of toxicity. Of the PECs, they found a total of 16 were able to predict toxicity within their evaluation criteria. Concensus-based PECs for arsenic, cadmium, chromium, copper, lead, nickel, zinc, napthalene, phenanthrene, benzo[a]anthracene, benzo[a]pyrene, chrysene, pyrene, total PAHs, total PCBs, and sum DDE provide an accurate basis for predicting toxicity. Insufficient data were found to evaluate PECs for mercury, anthracene, fluoranthene, fluorene, dieldrin, sum DDD, sum DDT, total DDT, endrin, heptaclor epoxide and lindane. The TECs and PECs that were found to accurately predict toxicity (or lack thereof) are indicated with an asterisks in Table 1. Note chemicals that were not found to be predictive are still useful as guides. This is because although the chemicals did not meet the predictive ability criteria set forth in MacDonald et al. (2000) for different reasons, since the Proposed Clean-Up Goals (PCUGs) are targets or objectives, they are still useful for general monitoring purposes.

Although these SECs were derived to predict toxicity (or no toxicity) to benthic invertebrates based on sediment chemical concentrations, they could be used to derive a range of PCUGs that would be protective of the assessment endpoint. This is precisely what was done. However, the values need to be examined in more detail. The TECs are straightforward, because any concentration below the TEC is not expected to cause adverse (toxic) effects in benthic invertebrates. However, the PECs are a bit more complicated because the PECs represent concentrations above which there is a greater than 50% chance that adverse effects could occur. Therefore, clean up levels would have to be below the PECs to be protective of the assessment endpoint (benthic invertebrate communities). The question is, how much lower?

In U.S. EPA 2000, the PECs and the mean PEC quotients (PEC-Q) were examined more closely. PEC-Q are calculated by dividing the concentration of each substance in a given sediment sample by its respective consensus-based PEC. The PEC-Qs are then summed for each sediment sample by adding the PEC-Qs that were determined for each substance. The PEC-Qs were then normalized to the number of PEC-Qs for each sample to give a mean PEC-Q for that sample. It was found that a 50% incidence of toxicity corresponds to a mean PEC-Q of 0.63 when survival or growth were used to classify a sample as toxic. By comparison, a 50% incidence of toxicity is predicted at a mean PEC-Q of 3.2 when survival alone was evaluated. In all, it was concluded that the longer-term tests in which growth and survival are measured tended to be more sensitive than the shorter-term tests. Since we are interested in deriving sediment PCUGs that are protective of benthic invertebrate communities in the long-term, it makes sense to derive PCUGs based on long-term tests.

The ranges of PCUGs calculated for use at this site are presented in Table 2. The lower values are the TEC concentrations. The higher values are calculated from the PEC values by multiplying each individual PEC by 0.6. The 16 TECs shown by MacDonald et. al. (2000) to be reliable at predicting toxicity are highlighted in yellow.

Preliminary Clean Up Goals

The range of PCUGs (Table 2) represent potential target values for clean up activities. The lower number represents the concentration that below which, no adverse (harmful) effects would be expected. Therefore, cleaning up to below that number should result in no adverse effects. This is shown with the "less than" sign. The higher number represents the concentration above which there is a 50% chance that adverse effects would occur. So, for example, cleaning up the site for arsenic to a concentration below 9.79 mg/kg dry weight should result in no adverse impacts to benthic invertebrates. If the site were cleaned up to 19.8 mg/kg dry weight, we would predict a 50% chance that adverse effects would occur. Based on the process outlined above, no predictions can be made about the impacts of concentrations between 9.79 and 19.8 mg/kg dry weight.

Reaching the PCUGs in sediments would result in decreasing current concentrations to acceptable risk levels. Not reaching the PCUGs does not necessarily mean that the sediments are not protective. If the latter situation arose, the site would have to be evaluated in light of other mitigating factors such as time, financial constraints, background contaminant concentration, technological constraints and the like. More site-specific evaluation would be necessary if the PCUGs can not be reached.

Uncertainty

There are some uncertainties associated with the derivation and use of the Proposed Clean-Up Goals (PCUGs) presented in Table 2. The first is that they were derived from a large database of sediment chemistry and toxicity information. Since this database does include the Indiana Harbor Canal (IHC) site information, the PCUGs are partially, but not wholly site-specific. More data would have to be collected from the canal to derive truly site-specific sediment clean up goals for the IHC site.

There is uncertainty in using the Sediment Effects Concentrations (SECs) in a manner for which they were not designed. The Threshold Effect Concentrations (TECs) and Probable Effect Concentrations (PECs) were designed to predict toxicity to benthic invertebrates in freshwater sediments. They were not derived as clean-up goals. There is not much information to determine if the PCUGs derived in this report are too protective or not protective enough. Again, to derive fully site-specific clean-up goals, more site-specific data would be necessary.

This report only evaluated sediment concentrations to protect one assessment endpoint: functioning benthic invertebrate communities. This is not an evaluation of the protectiveness of the PCUGs for any other group of organisms (e.g. fish, plants, mammals, birds, reptiles, amphibians, etc.). Along the same lines, this report does not evaluate the effectiveness of the PCUGs with respect to food chain bioaccumulation or biomagnification. More site-specific information is necessary to evaluate the risks to and the protectiveness of the PCUGs to these organisms.

Taking into account these uncertainties, the proposed PCUGs are useful as goals or targets against which to evaluate the success of the removal action. At this point, it is not known whether these values are over- or under-protective. However, they were derived using the best available information in a scientifically sound manner. Should more specific information be warranted, then more site-specific data would need to be collected.

Top of page

Top of page

Contacts for this site

• Brenda Jones, U.S. EPA. (jones.brenda@epa.gov)

Local Navigation

• Region 5 Superfund Home

• ---

• Basic Information
• Long-Term Cleanup
• Emergency Response & Removal
• Brownfields Program
• Revitalization & Redevelopment
• Oil Program
• Ecological Risk
• Field Support

• EPA Home
• Privacy and Security Notice
• Contact Us