Ralph A. Perona and Mark F. Tardiff, Neptune and Company

A conceptual site model that integrates attributes of both the physical environment and human exposure pathways is necessary for identifying data needs to support credible risk estimates in complex systems. Aspects of a risk assessment for contaminated soil or sediment that are related to both physical and exposure models include bioavailability, geomorphological sorting of particles, and particle size effects on chemical concentration. In a recent investigation at a copper mining operation a sampling design based on stream geomorphology was used to collect data to test a conceptual site model. Metals concentration data for multiple particle size fractions were obtained within three types of sediment deposits (channel, bars, and overbanks) over several miles of an ephemeral stream that had received mining-related discharges. Statistical analyses and risk assessment calculations were performed with these data to identify risk-driving metals and exposure pathways. The output of these calculations was then used to identify additional data needs for reducing uncertainty in risk estimates to acceptable levels. These data needs included additional information on background concentrations of key metals, likely metal concentrations in plant and animal foodstuffs raised on-site, and the variability and correlation in average concentrations of key metals with geographic location and particle size. The authors will demonstrate how addressing each of these data needs required a synthesis of information related to physical and exposure models.

Ronald J. Kotun, Ph.D., Tetra Tech NUS, Inc.

Geostatistics, kriging, and risk assessment are used together to minimize the amount of remediation required to attain target risk. A simple assumption in exposure assessment is that individuals randomly move throughout an exposure unit. Therefore, the exposure concentration contacted over a period of time can be estimated by the average concentration of the samples collected within that exposure unit. To account for uncertainty, this exposure concentration is typically represented by the upper confidence limit of the mean (UCL). Using a nonparametric technique, such as bootstrapping, to determine the UCL removes the need to establish the distribution of the data. Risks are typically estimated by using the UCL. If the UCL is greater than the cleanup level, then remedial action is necessary. However, the exceedance of a cleanup level by an individual sample does not necessarily indicate a significant exceedance of a target risk. Therefore, attainment of a cleanup level will indicate that a sample with a concentration greater than the cleanup level can remain in place, assuming that there is adequate representation within the exposure unit. The most contaminated soil samples are identified for removal until the UCL is less than the cleanup level. The samples proposed for remediation are replaced with a "post-remediation" concentration, such as one-half the detection limit, to calculate a post-remedial UCL. When this UCL is less than the cleanup level, there is attainment. The maximum concentration remaining is referred to as the "pick-up level." Geostatistics and kriging can be used to define the areas that need to be excavated. Areas targeted for remediation are based on the "pick-up level." Geostatistics, in combination with use of the "pick-up level," significantly reduce the amount of remediation necessary to attain the target risk, relative to treating all soil with concentrations greater than the cleanup level.

Jo Anne Shatkin, Ph.D. Menzie-Cura & Associates, Inc.

The National Toxicology Program (NTP) recently completed their two year bioassay of naphthalene, and concluded there was, "clear evidence of carcinogenicity" of the compound in rats. Recent work on whole mixture toxicity of coal tar also found significant cancer potency independent of benzo(a)pyrene, long thought to be the "bad actor" on which the relative potency of other DNA adduct-forming polycyclic aromatic compounds is based. These and other recent findings on coal tar component carcinogenicity are signals that the current relative potency model of toxicity for risk assessment is inaccurate for this mixture. These findings beg for a more holistic approach to mixture toxicity evaluation for coal tar, and for mixtures generally. This work reviews recent work on coal tar component toxicity, considers the implications for risk assessment, and proposes alternative approaches for assessing the toxicity of coal tar and other mixture toxicity, including those that provide a consistent framework for both human and ecological toxicity.

Michael W. Ehlebracht and Steven J. Germiat, Hart Crowser

Application of risk-based cleanup approaches to contaminated brownfield sites have significantly enhanced opportunities for converting barren pieces of land into much needed urban greenspace. Three case studies illustrate how risk-based cleanup approaches have allowed park developers to dramatically reduce cleanup costs while being protective of human health and the environment. A large former bulk fuel oil facility in downtown Seattle is being redeveloped into a sculpture park. Petroleum-impacted soil and groundwater still remain at the site, particularly under a major city arterial. Cost estimates for excavating residual impacted soils exceeding generic cleanup criteria totaled over $4,000,000. To minimize cleanup costs, a supplemental field investigation and modeling program was implemented to evaluate exposure pathways of concern under the future park land-use scenario. We demonstrated to state regulators that residual petroleum would not impact the adjacent marine environment and that emission of soil vapors would not likely pose a significant risk to park users. The cleanup action plan (CAP), which was approved as part of a consent decree, allows more resources to be used for park development rather than unnecessary environmental cleanup. Risk-based cleanup approaches were used to facilitate the purchase and redevelopment of a former bulk fuel oil site into a maritime park and education center. In order to meet an aggressive purchase deadline, an RI/FS and CAP were completed within three months. Use of fraction-specific petroleum analysis and risk-based cleanup approaches significantly reduced projected cleanup costs and allowed the project to move forward. Redevelopment of the former Dickman Mill timber mill into a waterfront park provided for significant restoration of intertidal habitat, while creating an upland lawn area for public access. The cleanup remedy, implemented cost-effectively as a component of park construction, addressed potential risks under the site’s future use as a waterfront park/intertidal habitat.

Deirdre C. Menoyo, BWSC, MA DEP, Rosemary Knox, MA DEP, Carol de Groot Bois, Bois Consulting Company, R. Duff Collins, Woodard & Curran

In 1993, Massachusetts implemented significant revisions to the Massachusetts Contingency Plan (MCP) regulations in a unique approach to expediting investigation and remediation of hazardous waste sites. These revisions included "privatizing" the majority of the cleanup work, that is, allowing private parties to have greater control of their investigation and remediation projects under the supervision of Licensed Site Professionals (LSPs), eliminating a significant amount of direct agency oversight. LSPs are qualified, experienced scientists or engineers licensed by the Commonwealth to render professional opinions regarding investigation and remediation actions at hazardous waste sites. Other components of the redesigned program included redirecting the agency’s focus to the sites of highest priority, from a human health and/or environmental risk standpoint, promulgating reporting and cleanup standards, and requiring DEP to audit sites to ensure compliance. In addition, the regulations allow site-specific, risk-based cleanups.

The purpose of this paper is to examine the progress of the redesigned program. A five-year review of the program in 1998 (finalized in 1999 as a Generic Environmental Impact Report) concluded that the program had successfully increased the number of cleanups completed by private parties and allowed the DEP to concentrate more resources on serious threats to public health and the environment (DEP, 1999). The site universe in 1994, just after the start of the program, consisted of almost 7,000 sites. Between 1983 and 1992, response actions were taken at about half of the sites where a release had been confirmed (DEP, 1993). At the close of FY 2000, over 14,400 of the 20,000 release cases reported to the agency since 1993 had been closed.

The overall understanding is that the redesigned program is working to achieve the intended endpoint of remediating releases of oil and hazardous materials. The DEP and the stakeholders in the program, however, recognize that there has not been enough focus on more accurate evaluation parameters to determine the progress of site cleanups and to evaluate how well LSPs and PRPs are meeting the requirements of the regulations. The DEP has initiated a long-term project that will evaluate the efficacy and timeliness of response actions at hazardous waste sites using more detailed and systematic methods. In the interim, this paper will present data that is currently available in order to provide the regulated community and other stakeholders with updated information on the redesigned program.