Governor Chris Christie • Lt.Governor Kim Guadagno
  Search
new jersey department of environmental protection
NJ Home Page Services A to Z NJ FAQs NJ Departments/Agencies departments
site remediation program

SRP Home | DEP Home

SRP Community RelationsSite Information ProgramHudson Chromate Project Basis & BackgroundPart 3

 

Summary of the Basis and Background of the Soil Cleanup Criteria for Triavalent and Hexavalent Chromium (September 19988)
Previous Section Table of Contents

III. Soil Cleanup Criteria for Trivalent and Hexavalent Chromium

The Department currently issues guidance on trivalent and hexavalent remediation goals or the methodologies to determine them on a site-specific basis. Because there are no existing soil cleanup standards for chromium, Public Law 1997, Chapter 278 allows the development of an Alternative Remediation Standard (ARS) to be used as a substitute. It is the intention of the Department to first issue this guidance as soil cleanup criteria (SCC) on an interim basis, and subsequently, formalize these protocols and derived values as soil cleanup standards via rule making procedures.

III.A. Trivalent Chromium

III.A.1. Cancer Endpoint

There is none as there is no carcinogenicity assessment available for Cr+3 on IRIS.

III.A.2. Noncancer Endpoint

The Department's Site Remediation Program prepared a draft basis and background document which established a residential land use SCC for Cr+3 of 78,000 mg Cr+3/ kg DW soil and a position that Cr+3 should not be regulated in a nonresidential land use situation (NJDEP, 1995a). The Department employed standard USEPA risk assessment methodology in the derivation of the Cr+3 SCC. The exposure pathway examined was incidental soil ingestion with a noncancer endpoint. It is important to note that these SCC are only applicable to insoluble forms of Cr+3.

The Department determined that ACD is not a relevant endpoint for Cr+3 due to the insolubility of Cr+3 compounds typically found in the environment. The toxicity information used to develop the Cr+3 SCC is based on an animal study where rats were fed Cr+3 (Ivankovic and Preussman, 1975). This study indicated that no real adverse health effects were detected even at the highest levels tested. The Department supported the use of NJDEP 1995a as the basis for justifying the 78,000 mg Cr+3/ kg DW soil as a residential land use SCC for Cr+3. It is important to note that this position was entirely consistent with that of the USEPA.

On September 3, 1998 the Department proposed the use of a 78,000 mg Cr+6 /kg DW soil Cr+3 SCC. However, incorporation of the USEPA changes to the IRIS database, which came out simultaneously on September 3, 1998, result in a recalculated Cr+3 SCC of 120,000 mg Cr+3/kg DW soil for an ingestion pathway in a residential setting. The Department now proposes to use this 120,000 mg Cr+3 /kg DW soil value as the Cr+3 SCC. The Department still chooses to not regulate Cr+3 under a nonresidential land use scenario. All the constraints of the previous numbers still apply.

III.A.3. Ecological Endpoints

The information available on the potential ecological effects of Cr+3 is limited. The central issue is whether or not the Cr+3 is in a form available to ecological receptors. In the Department's opinion, Cr+3 generally is found in the environment predominantly in an insoluble or complexed form. Therefore, the potential for severe ecological effects is minimized provided the Cr+3 did not originate specifically from a recent discharge of a soluble Cr+3 compound. The Department has determined that it will assess the ecological hazard on a case-by-case basis. This has been mandated by Hazardous Discharge Site Remediation Act, N.J.S.A. 58:10B which precludes the development of formal ecological standards until after the Environment Advisory Task Force is constituted and has determined the appropriate way to resolve the ecological concerns.

III.A.4. Impact to Ground Water

The Department intends to assess the potential impact of Cr+3 on ground water and surface water on a case by case basis. However, to date the Department has not encountered cases involving significant amounts of soluble Cr+3. This is partial confirmation of the Department's position that Cr+3 is found in the environment predominantly in an insoluble or complexed form provided the source isn't a recent discharge of a soluble Cr+3 compound. Chromium contamination of ground water and surface water is mainly associated with Cr+6, which is far more water soluble than Cr+3.

III.B. Hexavalent Chromium

III.B.1. Cancer Endpoint

For Cr+6 relative to a cancer human health endpoint, the critical exposure route is the inhalation pathway. To calculate a cleanup criterion for Cr+6, it is necessary to assess the Cr+6 soil concentrations present. Calculation of a particulate emission rate and a dispersion pattern are next. Finally, the Cr+6 soil concentration that would be associated with a given risk using standard USEPA cancer dose-response relationships and equations would need to be determined. Currently this is done on a site by site basis as an ARS.

The Department currently provides the guidance that USEPA sanctioned models can be used to predict particulate emissions as well as particulate dispersion. This is as per Public Law 1997, Chapter 278. The selection of given emission and dispersion models is preferably based on a comparison of predicted model results and actual field data as a means of verifying the appropriateness of the selections. The Department has already used this mechanism to approve the use of AP-42 (Cowherd et al. 1989; USEPA 1995a) in combination with the Fugitive Dust Model (USEPA 1991) to generate ARS data at several chromate processing waste sites in Hudson County.

The Department is allowing site-specific information to be used in the models. The considered inputs include particle size distribution, local weather parameters, traffic patterns, lot size, and configuration as well as other factors.

Residential and nonresidential exposure scenarios are being employed where appropriate. These situations are being differentiated via the inputs to the standard USEPA risk assessment equations. Based on analysis by the Department of model outputs, the major difference between these two scenarios is not the human lifetime exposure assumptions, but rather the impact of vehicular traffic. Vehicular traffic when present dominates the emission levels. Consequently, the calculated remediation goals for nonresidential sites are lower than comparable residential sites. Note that the risk level employed in both scenarios is an excess cancer risk of 1 x 10-6 specified by Public Law 1997, Chapter 278.

Once a site-specific number is generated, compliance is determined by comparing the calculated value to the 95th % upper confidence limit of the mean of the observations. With respect to the inhalation pathway, the Department is considering moving away from a "bright line" approach. This is in recognition that a single exceedance at a single location should not necessarily require a remedial action for the entire site, particularly when the phenomenon being evaluated is likely derived from an area source and not a point source.

For a nonresidential exposure scenario, the Department proposes to use the logic employed in the above-described ARS to develop a Cr+6 SCC for a cancer endpoint. The intention is to comply with the requirements of N.J.A.C. 1:30 by employing USEPA approved air models in combination with USEPA exposure assessment equations and USEPA standard assumptions. The Department is currently proposing to use AP-42 as the particulate emissions model and the Industrial Source Complex - - Short Term, Version 3 (ISCST3) as the particulate air dispersion model. The Department's selection of ISCST3 over FDM was influenced by the fact that the defaults for the ISCST3 are already known. In addition, FDM requires much more site-specific data input than ISCST3. Consequently, an ARS can be much more readily determined at much less expense if ISCST3 is used instead of FDM. Furthermore, initial evaluation indicates that this pairing of models produced more protective results than the AP-42/FDM pairing.

To generate a generic nonresidential land use Cr+6 SCC for a cancer endpoint, the Department intends to input protective assumptions into the models (AP-42 and ISCST3) and calculate a value. Preliminary calculations indicate that such a SCC is 20 mg Cr+6/kg DW soil if a site size of 2 acres is presumed. Site size is a major factor impacting the dispersion model. The Department is currently in the process of determining what a reasonable nonresidential site size should be for New Jersey. The Department projects that the site size default value will be 2 acres. Be advised that the larger the site, the less stringent the SCC. The 20 mg Cr+6/kg DW soil value also incorporates default values for other parameters, which are specific for New Jersey, as well as USEPA defaults. The fraction of the total amount of particles present represented by respirable particles (particles 10 microns in diameter or less) and the percent silt present are the major factors impacting the emissions model. The USEPA defaults employed for these parameters were taken from AP-42 Compilation of Air Pollution Emission Factors.

For a residential land use scenario, the Department proposes to use the USEPA soil screening value of 270 mg Cr+6/kg DW soil. However, the USEPA value was determined using loam soil as a default in the USEPA Soil Screening Level (SSL) equations. The Department is currently evaluating the feasibility of substituting values for a New Jersey soil type for the loam default into the USEPA SSL equations to generate a more New Jersey oriented residential air value. Doing so would still be consistent with USEPA guidance and the requirements of N.J.A.C. 1:30. However, until such time that it is decided to use New Jersey soil data, USEPA defaults will remain in place. It is notable that USEPA personnel have indicated that the USEPA SSL equations are not designed to generate a nonresidential land use value.

While the USEPA on September 3, 1998 modified the IRIS database regarding the toxicity of Cr+6 , an evaluation by the Department of the new data indicates the September 3, 1998 proposed value of 20 mg Cr+6 /kg DW soil would still be the most protective for a nonresidential land use scenario. Consequently, the Department is not altering its September 3, 1998 position.

III.B.2. Noncancer Endpoint

III.B.2.A. Ingestion Pathway

The Department proposes to use the USEPA SSL value for the residential ingestion pathway. This value is 390 mg Cr+6/kg DW soil in the May 1996 technical background document (USEPA 1996). However, with the September 3, 1998 change in the IRIS database, this number becomes 240 mg Cr+6 /kg DW soil. For the nonresidential ingestion pathway, the use of the standard USEPA risk assessment equations and defaults previously yielded a value of 10,000 mg Cr+6 /kg DW soil. Using the new IRIS information in combination with these same standard USEPA equations and assumptions yields a Cr+6 SCC of 6,100 mg Cr+6/kg DW soil.

III.B.2.B. Dermal Pathway

The Department recognized the need to develop a procedure that would relate the ACD endpoint to Cr+6 soil concentrations. This is because soil concentration data are traditionally the type of data generated during remedial investigations and is how all the other soil cleanup criteria are expressed. Consequently, an ARS procedure was developed which allows for the use of site-specific conditions in the development of a site-specific ACD-based Cr+6 cleanup level. This procedure is considered applicable to all types of Cr+6 contaminated sites within the State.

The Department intends to utilize an agitated extraction test to reflect, but not duplicate, the assumed exposure scenario. The Department readily admits that ephemeral phenomena such as a puddle or a wet soil are dynamic and complex. Variable in configuration and duration of existence, these systems are complicated by other factors. Disturbance, turbulent flow, diffusion, adsorption, ion exchange, percolation, and evapotranspiration are some of the factors influencing the concentration of dissolved ions present. Certainly, there are no standardized analytical methods currently available that are designed to simulate these events. However, the use of a standard method is desirable in order to assure the availability of the test procedure in commercial laboratories. Therefore, the Department determined that the only viable approach was to obtain a protective measure of the extractability of the hexavalent chromium contaminated material using an approach that employed a standardized method and also reflected, but made no claim to duplicate, the assumed exposure scenario. This would then be used to estimate the hexavalent chromium concentration in soil that would produce the MET.

The Department in consultation with Larry P. Jackson, Ph.D. of Environmental Quality Management developed such an approach based on the concept of soil-water partition coefficients (Kd) which utilizes experimentally derived data from American Society for Testing and Materials Method D3987-85 (Method 3987). Jackson (1998) is the result of this collaboration. Jackson has extensive experience with the validation of Method 3987 and has won numerous awards from the American Society for Testing and Materials for his work with hazardous waste characterization and management including leaching methods and chemistry. He is currently the liason between American Society for Testing and Materials Committee D34 and the USEPA on this topic and also serves as a consultant to USEPA in this area as well. Jackson believes Method 3987 is suitable for such a use.

The Department proposes to use the results of Method 3987 to generate a Kd relationship; select a specific Kd; and use it to derive a hexavalent chromium cleanup criterion that corresponds to the 25 mg Cr+6/l MET. This is of course envisioned initially as a site-specific protocol. Duplicate samples of the highest hexavalent chromium concentration representative of the type of contamination found on a given site would be extracted at 4 different extraction solution to soil sample ratios. The samples to be extracted will be homogenized, but otherwise unmodified (not screened). This is to directly reflect the particle size structure and soil characteristics present at the site. The proposed ratios would be 4:1, 10:1, 20:1, and 40:1. Replication at each ratio is recommended, but not required. Extraction ratio (LSR) is defined as the ratio of the volume (liters) of extraction liquid or solution to the weight (kilograms) of solid sample to be extracted. The calculated Kd for each point would be plotted against its corresponding extraction ratio. A best-fit determination would be conducted and the actual type of relationship established. This relationship would then be used to derive the Kd at an LSR of 2:1 to be used in combination with the 25 mg Cr+6/l MET to calculate the ACD criterion. The intent of the Department is to follow the principles utilized by the EPA in developing their SSLs that are also based on soil-water partition concepts.

The question arises whether or not the methodology is protective or not. One aspect of Method 3987 that is conservative is the 18 hour agitation period. Clearly, the 30 rotations per minute for this time period produce more mixing than would normally occur at a site. However, Method 3987 also traditionally calls for the use of a 20:1 LSR. The Department is taking the position that while this ratio maximizes the absolute amount of hexavalent chromium extracted and better reflects equilibrium conditions, the concentration of the extracted ions is artificially decreased to an inappropriate level. Therefore the Department will require the use of a 2:1 LSR. The Department recognizes that in doing so, the assessed situation may not be at equilibrium. The Department is comfortable with this position because the conditions of the assumed exposure scenario are also likely to be non-equilibrium in nature. Again this decision is protective in nature.

Uncertainty regarding the methodology is an issue that needs to be discussed. Jackson (1998) indicates that Method 3987 can be employed in this manner; but that there are increasing concerns about reproducibility as lower LSRs such as 1:1 are used. The use of a range of LSRs (including higher LSRs) to determine the relationship between Kd and extracted soil concentrations is an effort to minimize this effect. Previously the Department would select the experimental LSR; employ extensive replication to estimate the relationship between the concentration of the extraction solution and the extracted soil concentration; and then back-calculate the soil concentration that would produce the assumed MET. The problem was that the selected LSR was on a portion of the regression subject to potentially the largest variability. By incorporating a range of LSRs, the derived Kd relationship is much more robust than the relationship (between extraction solution concentration and extracted soil concentration) produced by the Department's previous approach.

Various parties have questioned the selection of the 2:1 LSR by the Department. The initial selection of a 2:1 LSR was in part an evaluation of the results achievable using Method 3987; the theoretical conditions at which free water is available in a soil type; and a risk management decision. The EPA uses values of 2.65 grams per cubic centimeter (g/cm3) and 1.5 g/cm3 for the typical soil density and bulk density, respectively. Inserting these values into the EPA equation for soil porosity yields a value of 434 milliliters (ml) of pore space per liter of soil.

Soil porosity = 1 - (bulk density)
-----------------
(soil density)

Alternatively if 1,000 cubic centimeters (cm3), the equivalent of 1 liter, of this soil were totally saturated, there would be 0.434 liters (l) of water present. Conversely, this would mean that there are typically 566 cubic centimeters (cm3) of actual soil present for this specific case. Therefore every 0.434 l of excess water at saturation corresponds to approximately 1.5 kilograms (kg) of soil. This represents an LSR of 2:7, which is much lower than 2:1. Consequently, the 2:1 LSR is a nonconservative assumption, but far less so than 20:1. The 2:1 ratio also represents in the Department's opinion the practical lower limit at which Method 3987 can be conducted for generalized use. Previously, the Department was considering using a 1:1 LSR because laboratory testing of a selection of New Jersey soil types indicated this was the point at which free water would become available. However, to use an LSR of 2:7 or 1:1 would require the utilization of an extrapolated portion of the Kd/LSR relationship prone to very high variability. The Department determined the use of a 2:7 or 1:1 LSR to be inappropriate and "redundantly conservative."

Recognizing there would be continuing objections to the Department's selection of a 2:1 LSR, the Department then sought the opinion of an outside reviewer, who could act as an objective third party, to determine the appropriateness of the selection of a 2:1 LSR. The chosen outside reviewer was Jon Chorover, Ph.D., Assistant Professor of Environmental Soil Chemistry at Penn State University. Chorover (1998) is the result of this effort. Chorover has published in the area of metals behavior and has served as a workshop organizer for the Soil Science Society of America on the topic of transport modeling.

Chorover used Fick's Law to evaluate the assumed exposure scenario. The main parameters examined were water column depth, time, soil porosity, and soil bulk density. After making a number of simplifying assumptions, he established a mechanism for calculating the depth of soil contributing to the diffusive flux. The ratio of water column depth to depth of soil contributing to the diffusion is directly relatable to the LSR issue. Certainly, there are additional factors that are involved and were not considered, such as evapotranspiration, turbulence, configuration, adsorption behavior, multi-directional movement, etc. However, to include all factors would make the analysis too complex. The Department opted to define the situation to a reasonable extent and then establish the equivalent of a protective Tier I screening level. The use of Method 3987 provides this measure of conservativism because of the long duration and nature of the mixing involved with the extraction. Chorover (1998) provides support that it is appropriate for the Department to require Method 3987 be run at an LSR of 2:1. In fact, the use of 2:1 LSR is now considered by the Department to be actually somewhat nonconservative in nature.

If the Department wished to develop a specific statewide ACD-based SCC for Cr+6, this could be accomplished through the analysis of laboratory data from a study which simulated a recent spill (having occurred approximately 24 hours prior) of a soluble Cr+6 solution (sodium dichromate). Using a native New Jersey sandy soil containing a relatively small amount of organic matter would serve the purpose of making the value more pertinent to the locality as well as being protective. Such a soil would be treated with a Cr+6 solution and extracted at multiple LSRs using Method 3987. The relationship between Kd and LSR would then be calculated and used to generate the Kd at a LSR of 2:1. This Kd value would then be used to back-calculate the Cr+6 soil concentration that would generate a 25 mg Cr+6/l solution. By determining the Cr+6 soil concentration which would produce the equivalent of a 25 mg Cr+6/l solution under fully saturated soil conditions, a SCC could be established. While this has not been formally done, such a calculated ACD-based SCC would be applicable for both residential and nonresidential land use conditions. The initial phases of developing such a criterion have been initiated by the Department with a current scheduled project duration of 6 months barring unforeseen developments.

III.B.3. Ecological Endpoints

The assessment of the ecological impacts of Cr+6 will be handled on a case by case basis. Again the development of ecological criteria is precluded by Hazardous Discharge Site Remediation Act, N.J.S.A. 58:10B.

III.B.4. Impact to Ground Water

Impact to ground water (and potentially surface water) by Cr+6 is being assessed by employing one of two options. Both use 100 micrograms of Cr+6 per liter (ug Cr+6/l) as a compliance value. This value is derived from the maximum contaminant level allowed in drinking water which is 100 ug Total Cr/l.

The first option is the use of USEPA Method 1312 (Synthetic Precipitation Leaching Procedure) to assess the Cr+6 availability. This is an agitated extraction method that utilizes unbuffered water at a pH of 4.2 as the extraction solution. The 100 grams of the sample to be evaluated is placed in a container with 2 liters of the extraction solution. The contents of the container are mixed for 18 hours at room temperature. The distribution of the Cr+6 between the extraction solution and the extracted sample are determined. A site-specific Kd is then calculated using a soil partitioning equation. This Kd is then used to back calculate the Cr+6 concentration in soil that would produce a level of 100 ug Cr+6/l.

The second option is the direct assessment of the ground water through laboratory analysis. Ground water is sampled and analyzed for Cr+6. This value is then compare to the appropriate compliance value, which in this case is 100 ug Cr+6/l. Under this second option, there are additional requirements that must be met. First, area of the highest concentration of Cr+6 contamination must be evaluated. Secondly, the Cr+6 contamination must be in contact with the ground water.

Next Section