This profile summarizes an expedited site characterization and remediation project consistent with principles of the Triad approach that occurred during Remedial Design/Remedial Action (RD/RA) activities at an industry-lead CERCLA site in Salt Lake City, Utah. Soil and groundwater at the 18-acre site had been contaminated with dioxins and furans, chlorinated solvents, pentachlorophenol, and a variety of pesticides as a result of chemical formulation and packaging activities that had been occurring at the site from 1957 to 1990. Distribution of contaminants in soil was sporadic and largely uncorrelated to specific historical waste management practices or spill areas. The site?s flat low-lying topography, shallow depth to groundwater, and propensity for flooding complicated the identification of a consistent pattern of contamination.
This project was completed well before the Triad approach was officially formulated by the U.S. Environmental Protection Agency (EPA). However, it used principles later promoted under the Triad to streamline remediation of both soil and groundwater at the site, and illustrates some of the benefits of a Triad-like approach. This profile focuses on surface soil and the innovative soil cleanup program that was developed and successfully implemented.
A statistically-sized grid pattern was created for each soil remediation area that had been defined during the prior Remedial Investigation (RI). Grid sizing and configurations were established using statistically-based power curves developed from the RI data. Random sampling was performed within each grid area with two different compositing schemes, and a step-wise approach was further used for systematically isolating hot spots laterally and at depth, inside and outside the grid areas. Highlights of the project from a Triad perspective include the following:
|Site Name||Wasatch Chemical Superfund Site|
|Location||Salt Lake City, UT|
|Site Regulatory ID||UTD000716399|
|Project Lead Organization||Entrada Industries, Inc.|
|Project Lead Type||PRP Lead|
|Regulatory Lead Program||Superfund Remedial|
|Triad Project Status||Field Program Completed|
|Reuse Objective Identified||Yes|
The 18-acre site is located in southwest Salt Lake City in an industrialized corridor adjacent to Interstate 15. The nearest residential area is approximately ¼ mile northwest of the site. The population within a one-mile radius is approximately 5,000. The flat, low-lying site is near the center of the Jordan River Valley and is underlain by a thick aquifer. In the vicinity of the site, the aquifer contains both shallow and deep portions, which are separated by a continuous confining layer. Shallow groundwater is normally approximately 2 feet below ground surface. Surface water runoff in the area is drained by a network of unlined ditches. One of these, adjacent to the west boundary of the site, received surface drainage from the site.
A component of the preferred remedy at the site, as outlined in the Record of Decision (ROD) included consolidating contaminated soil, sediment, process drain lines, and dioxin removal wastes in a former onsite evaporation pond, and treating the contents of the pond via in situ vitrification. This profile focuses on characterization and remediation of surface soil contamination.
Historically, several onsite areas were utilized by different chemical companies at different times to store drums containing a variety of chemicals. In addition, chemical spills during production were not all accounted for and uncontrolled dumping of contaminated wastewater was common. Thus, over a majority of the site, there was no clear source of surface soil contamination from which to work from a characterization standpoint. Consequently, during the RI, surface soil samples were obtained from borings drilled on a grid at spacings between 50 and 100 feet depending on the area. Initial soil remediation areas were defined by comparing the results of chemical analysis of the soil samples with action levels for the constituents of concern identified in the ROD.
The conventional approach to remediating these areas was anticipated to include excavating each area and performing vertical and lateral confirmation sampling and analysis to check that contaminated soil above action levels was removed. For contamination distribution conditions such as these, however, the conventional approach had several disadvantages, primarily cost and schedule uncertainty: For example, if confirmation sampling and analysis after excavation yielded results above action levels, further excavation and confirmation sampling would be required. Thus, the number of excavation/confirmation sampling iterations was unknown, as was the total volume of soil to be excavated and treated. This uncertainty produced a potential for prodigiously high soil excavation costs. Uncertainty would be reduced by substantially increasing the volume of soil excavated between confirmation sampling events; however, such an approach would still have inevitably resulted in greater soil volumes and thus high treatment costs. At the time, there existed no practical field screening methods (like a photoionization detector [PID] to guide excavations of more volatile contaminants) that could have reduced the excavation/confirmation sampling costs, if not their uncertainty.
Contaminated soil and groundwater resulted from over 30 years of chemical manufacturing activities at the site. Constituents of concern included chlorinated solvents (primarily trichloroethene [TCE]) and herbicides in groundwater; and pesticides, pentachlorophenol, and dioxins in soil. As a result of certain risk factors and collocation of contaminants, soil remediation focused on Heptachlor, Hexachlorobenzene, Chlordane, and Dioxins and Furans. Contaminated media at the site included soil, groundwater, sediments (in process drain lines and in an existing former evaporation pond), and dioxin removal wastes (drummed waste remaining onsite, a remnant of chemical manufacturing activities, containing percent levels of pesticides, herbicides, and dioxins and furans). Aspects of the Triad approach were utilized for both soil and groundwater at the site; however, the focus of this profile is soil.
Project objectives included use of expedited site characterization (Triad-based) principles to (1) minimize sampling and analysis costs, (2) minimize the quantity of contaminated soil to be excavated and treated, and (3) quantify decision uncertainty in a single mobilization.
Utilizing systematic planning and dynamic work strategies, the volume of soil requiring excavation was determined with regulatory approval prior to excavation thus eliminating the need for verification sampling and analysis. Systematic planning assured the project decision objectives formed the basis of the field program design and data collection activities. The dynamic field program allowed near real-time determination and regulatory approval of the extent of soil contamination while the excavation crews were mobilized. By eliminating confirmation sampling and analysis:
Overall, a Triad-like approach produced sound decision logic that assured the attainment of remediation objectives at a substantial time and cost savings, and further demonstrated this decision-making strategy to stakeholders prior to the conduct of the work.
The project team assessed a verifiable cost savings of $300,000 based on estimated versus final soil excavation volumes. The lead potentially responsbile party (PRP) further estimated that the project saved them approximately $30 million in anticipated liability based on the buyouts received from other potential PRPs for the site.
Through systematic planning and the development of dynamic work strategies, decision uncertainties were managed that resulted in a specific soil excavation quantity and configuration with a quantifiable level of uncertainty acceptable to project stakeholders. The first step in the systematic planning process included development of a CSM. On the basis of data available from the RI, the CSM for surface soil contamination assumed (1) shallow but random lateral distribution, and (2) collocation of certain constituents of concern (that allowed streamlining of the analytical program). The next step included formation of an integrated project team (discussed below). Finally, agreement among stakeholders regarding acceptable levels of uncertainty was established. As embodied in the work flow diagram Figure 1, initial results from the field program guided the remainder of the program. The CSM was updated on a real time basis as data were collected and evaluated. This process was continued until the investigation objectives were achieved.
The project team was comprised primarily of technical and management staff from HLA (that was later incorporated into Mactec, Inc.), and a contractor hired by the PRP, Entrada Industries. The project was managed by a remediation engineer, who was supported by key professionals such as other design engineers, a field geologist, a geochemist (with supporting analytical and quality assurance (QA) oversight staff), and a data manager/statistician. Subcontractors included construction contractors and local analytical laboratories.
EPA Region 8 and the Utah Department of Environmental Quality (UDEQ) provided technical and regulatory oversight of the project. The EPA Remedial Project Manager (RPM) was a key to the success of the project. He approved the concept, was kept thoroughly informed during the conduct of the work, and intermittently obtained and analyzed split soil samples to independently verify results. EPA was supported by technical staff from EPA?s contractor, PRC Environmental Management Inc. (which was later incorporated into Tetra Tech Inc.).
The dynamic work strategy allowed the project to be completed "faster" and "cheaper" than would have been possible under the traditional static work strategy contemplated in the ROD. The top 4 inches of soil were excavated without sampling. Deeper soil was excavated based on the following dynamic work strategy:
Statistically identify sample size ? Statistical evaluations were performed to assess the required number of samples for each of the soil remediation areas necessary to achieve, within a specified level of confidence, the appropriate risk factor. The statistical procedures utilized are described in Methods for Evaluating the Attainment of Cleanup Standards, Volume I: Soils and Solid Media, EPA 230/02-89-042, February 1989.
Design the sample grid ? The results of the statistical evaluation to assess sample size were utilized to design a sampling grid for each soil remediation area. A random number generator was used to establish the location of one sample for each grid sector. The sampling grids for each soil remediation area and the location of the sample within each grid was laid out in the field by a licensed land survey contractor. A typical grid for a given soil remediation area is shown in Figure 2. Once the sample grids were field located, the dynamic work strategy decision logic shown in Figure 1 was implemented.
Collect soil samples ? Continuous soil samples, 3 feet deep, were collected within each grid sector at the designated sample location. Samples were transported to a nearby commercial analytical laboratory.
Sample extraction and analysis ? Under the direction of the project geochemist, the analytical laboratory composited samples, performed extractions, and analyzed the samples for soil indicator chemicals, including chlordane, heptachlor, hexachlorobenzene, and dioxins. The contract laboratories prepared and analyzed the initial composite samples (i.e., a composite of all samples obtained in the upper 6 inches of each grid area) and reported the results within 48 hours of sample receipt. As shown in Figure 1, when initial composite sample results were above the action level divided by the compositing factor, the laboratory prepared and analyzed five detailed composite samples. Turnaround time for the detailed composite sample analyses was 72 hours after completion of the initial composite sample analysis. Analytical results for the remaining discrete samples were received within seven days after completion of the composite sample analysis. Supporting project QA chemists validated the chemical data and the validated data was entered into the project database.
Manage and track sample data ? A data management system was designed and implemented to allow storage of survey or geographical information system (GIS) information for sampling grids and tracking of the soil samples collected from each of the soil remediation areas. The data management approach is discussed in greater detail in the following section entitled "Data Management Approach and Tools."
Perform hypothesis testing ? Hypothesis tests consisted of comparing the mean concentration of the sample population to the indicator chemical action levels from which the spatial delineation of soil to be excavated in each soil remediation area is defined.
Delineate areas of excavation ? The results of chemical analyses, statistical evaluations, and hypothesis testing were used to delineate the extent of soil to be excavated for each remediation area. GIS was used to define these areas vertically and laterally and thus facilitate soil excavation. Figure 3 and Figure 4 show examples of how the lateral and vertical extent of excavation were defined, respectively.
This process was performed in a single mobilization with ongoing decisions regarding expanding the grids and extracting and analyzing soil samples until the extent of the soil to be excavated in each remediation area was determined.
As described under dynamic work strategies above, fast turnaround analyses at local analytical laboratories provided near real-time data to support decision-making for additional sampling and excavation in the field. The objective was to have a sufficiently short turnaround time to ensure that characterization of each area of contamination was complete at the end of a single field mobilization. The primary analytical methods used for decision-making included EPA Method 8080 (organochlorine pesticides by gas chromatography with electron capture detection [GC/ECD]) and EPA Method 8280 (dioxins and furans by GC with mass spectrometry detection [GC/MS]). Cost and time savings were attained for the laboratory analyses by incorporating two constituents of concern normally analyzed by other methods (hexachlorobenzene and pentachlorophenol) into EPA Method 8080.
Sample homogenization and compositing were performed by the analytical laboratories using detailed site-specific protocols that had been prepared by the project geochemist and documented in the quality assurance project plan (QAPP). The analytical program required careful management of sample and sample extract storage by the project geochemistry/chemistry staff to maximize laboratory efficiency (extraction and analytical batches) while ensuring that holding times weren?t exceeded.
The local analytical laboratories for this project included:
American West Analytical Laboratories
463 W 3600 S
Salt Lake City, UT 84115
960 West Levoy Drive
Salt Lake City, UT 84123
TQRS not prepared -- project relied on fast TAT analytical data from local laboratories using EPA Method 8080.
A critical component of the soil verification program was quantitatively assuring data quality. For this project, the project geochemist performed a full validation of only those analytical results that defined the limits of contamination in a remediation area. Like the analyses themselves, the validation was performed on a near real-time basis (daily or semi-daily) as results were received. The evaluation of results other than those indicating the limits of contamination were validated through two real-time audits of the project laboratories. Data defensibility documentation also included project-specific composite tracking sheets to verify that sample compositing was performed correctly by the laboratory.
The use of EPA "definitive" laboratory methods for the grid sampling, combined with the auditing and data validation program that was performed, eliminated the need for confirmation sampling (bottom and sides) after excavation was complete.
A key to the success of the program was the approach utilized to track and manage data. A data management system was developed that allowed storage of the sampling grids and tracking of the soil samples collected from each of the soil remediation areas. The data was entered and managed using Arc/Info, a GIS that allowed efficient tracking of linked spatial and tabulated databases and directly linked a given soil sample to the corresponding chemical value reported in the laboratories? electronic data deliverables. Data generated from Arc/Info was imported into a sample tracking system and statistical evaluation software for performing hypothesis testing. Data printouts and maps were faxed to EPA, after which the EPA RPM was called by the project team to attain concurrence on the next steps in the characterization and excavation process.
The soil verification program was implemented between August 5 and October 30, 1993. The Sampling and Analysis Plan and associated QAPP were finalized in August 1993 as the field program began. A report documenting the completion of the program, including analytical results, quality control and quality assurance data, and the spatial extent of required soil excavation, was issued on May 29, 1994.
|Further discussions of the site cleanup for the Wasatch Chemical Site|
|Record of Decision and other Superfund documentation for the Wasatch Chemical Site|
To update this profile, contact Cheryl T. Johnson at Johnson.Cheryl@epa.gov or (703) 603-9045.