|
The Economy Products project was completed in 2001 and 2002. The project included delineation of the extent of pesticide contamination at the site, and confirmation that the excavation, treatment, and disposal of contaminated soil met removal action objectives.
All three major elements of the Triad (systematic planning, dynamic work strategies, and real-time measurements) were used to bring about the success of the project. As an EPA-lead project that involved only EPA and its contractor, systematic planning was simpler than many projects but provided a strong framework for the dynamic strategy and real-time technologies used. Dynamic work plans with field-based decision-making allowed the fieldwork to proceed efficiently. Real-time technologies were essential to increasing the data density and reducing cost while maintaining decision quality.
Initial contaminants of potential concern were pesticides, herbicides, polychlorinated dibenzodioxins (PCDD) and polychlorinated dibenzofurans (PCDF); however, only pesticides were detected at the site. Soil in the industrial area was excavated to depths of one to twelve feet, and over 7,037 tons of contaminated soil was treated on-site using thermal desorption. Soil in the residential and railroad area of the site was excavated to depths of one to four feet, and 312 tons of contaminated soil was transported to a local landfill for disposal. Excavation and treatment/disposal decisions were supported through a collaborative data set from immunoassay test kits and laboratory methods.
| Site Name | Economy Products Site |
| Location | Shenandoah, IA |
| Site Type |
Pesticide Manufacturing/Use/Storage |
| Site Regulatory ID |
IAD984566208 |
| Project Lead Organization | EPA Region 7 |
| Project Lead Type |
EPA Lead |
| Regulatory Lead Program |
Superfund Removal |
| Triad Project Status | Field Program Completed |
| Reuse Objective Identified |
Yes |
| Proposed Reuse: |
Residential and Commercial/Industrial |
The Economy Products site occupies three acres along Southwest road in southwest Shenandoah, Iowa. Agricultural pesticides were formulated and packaged at the Economy Products facility from 1961 to 1973. Business activities ceased in 1973 when the main packaging area and a storage warehouse were destroyed by fire. The fire dispersed pesticide residue across the site through air dispersion and surface water runoff from fire-fighting efforts. Some of the pesticides handled at the facility are now banned substances or are subject to restricted use regulations.
After the fire, a 14-inch clay cap was placed on the property. The Earl May Seed and Nursery Company (EMSNC), which owned property adjacent to the site on the north, east, and south sides, purchased the property in 1974. EMSNC used the area to store nursery stock until approximately 2000.
Early sampling efforts identified pesticide contamination of soils on the site. Contamination was also found on four residential properties south of the site and a small area of railroad property west of the site. For the purposes of investigation and remediation, the site was divided into two portions:
Previous investigations at the industrial site had identified chlordane at concentrations up to 5,880 mg/kg, and several other pesticides at concentrations up to 3,340 mg/kg. At the residential/railroad site, pesticides were identified at concentrations up to 93.4 mg/kg. In addition, groundwater samples indicated contamination with chlordane (0.36 µg/L) and heptachlor (0.036 µg/L) below maximum contaminant levels (MCL) established by EPA's Office of Water.
Based on correlations between the removal assessment results from immunoassay test kits and fixed laboratory methods, soil from the industrial site was excavated and treated by low temperature thermal desorption. After treatment, confirmation samples were collected and submitted for laboratory analysis. If necessary, soil was treated a second time. After confirmation that the treated soil was clean to below RALs, the soil was used to backfill the excavated areas of the industrial site.
Soil was excavated at the residential/railroad site based on historical data. Final excavation extent was verified through collection of samples for immunoassay test kit and off-site laboratory analyses. A total of 312 tons of soil excavated from the residential/railroad site was disposed of at a local landfill. Soil from one cell was treated thermally in the manner used for the industrial cells, and the treated soil was used for backfill at the industrial site. The excavated areas were replaced with 280 tons of clean topsoil.
After the investigation and removal action, the site was recommended for no further action, and archived from the CERCLIS database in September 2006.
The primary objectives of the project were to remove soil contaminated with pesticides, treat it using low temperature thermal desorption, and reuse or dispose of the treated soil. Thermal treatment was only used for portions of the site where the planned reuse was industrial/commercial, and the treated soil was used for backfill in those areas. Soil excavated from the residential and railroad portion of the site was disposed of in a local landfill after sampling indicated that this was an appropriate disposal method. A decision was made to use site-specific remedial action levels (RALs) to define the extent of removal, and to use federal universal treatment criteria (UTC) to define the endpoint for treatment (for the industrial reuse portion only).
The project highlights the ability of non-specific methods to achieve usable sample results at significant cost savings to a project. Immunoassay test kits sensitive to chlorinated pesticides similar to chlordane were used at the site. These field-based method results were compared with a yes/no criterion for whether samples exceeded RALs in samples analyzed by fixed laboratory SW-846 methods. Using the comparison between the fixed lab and field test kits, the project team developed field-based action limits (FBAL) that were correlated to the RALs. The FBALs were used in dynamic work strategies to guide excavation of contaminated soil in real time based on the immunoassay test kit results. This approach provided significant cost and time savings.
The project team estimated cost savings of nearly $25,000 in analytical costs alone. Substantially more cost savings were likely achieved by reducing the amount of time required to make decisions about when sufficient soil had been excavated. The project team estimated that at least five full days of excavation and backfill operations were trimmed from the project. However, no further estimate was made for the amount of cost savings based on the time saved.
Continuously monitoring the performance of the field-based data against the laboratory methods provided part of the $25,000 in analytical cost savings. When it was noted that the industrial FBAL was overly conservative, the FBAL was revised to reduce cost while still meeting the target RALs. For more detail, see the sections on Dynamic Work Plans and Decision Logic in this Triad Profile.The project team established the following specific project objectives:
Because the PRP had cashed out under an Administrative Order and was no longer involved with the site, the regulatory agencies were in complete control of the project, simplifying the systematic planning process. The project team agreed to RALs, FBALs, and to the decision logic that would guide the investigative and removal activities.
The initial conceptual site model (CSM) was based on a review of available and relevant historical information regarding past activities at the site. The initial CSM identified the source of pesticide contamination as a fire at the site that led to the facility closing in 1973. The fire dispersed pesticide residue across the site through air dispersion and surface water runoff from fire-fighting efforts. The CSM postulated that the majority of contamination would be encountered near the surface over most of the site.
A removal assessment was performed at the beginning of the project to further delineate contamination. The removal assessment included comparative analyses of samples by both the immunoassay test kits and laboratory methods to develop correlations and initial FBALs for the later excavation program.
For the removal assessment (and later the removal), the industrial property was divided into 103 cells, each generally 21 feet by 21 feet square. Similarly, the residential and railroad properties were divided into 32 cells, each generally 14 feet by 14 feet square. The smaller grid size was used for the off-site properties to reduce the chance of missing isolated hot spots. Soil samples were not collected from a cell if previous investigations indicated that the cell exceeded the established RALs. Soil samples were collected from a total of 77 soil borings. At some borings, samples were collected at multiple depths. Immunoassay test kit field analytical results were used to guide the depths sampled. A total of 119 samples were collected, of which 63 were submitted to a fixed laboratory for analysis.
Site layouts for the industrial site industrial site and the residential/railroad site show the division of the sites into cells. These layouts also show the cells that required excavation and/or treatment to achieve RALs.
At the industrial property, each cell was excavated to a depth where the removal assessment indicated that the remaining soil was below the RAL for pesticides, up to a maximum depth of 12 feet. Of the 103 cells, 73 required excavation to a depth of between 1 and 12 feet (the other 30 cells had been designated clean based on previous investigation results). Two cells were still above the RAL at 12 feet and were backfilled without further treatment in accordance with the work strategy.
At the residential/railroad property, each cell was excavated to a depth where the removal assessment indicated that the remaining soil was below the RAL for pesticides, up to a maximum depth of 4 feet. Of the 32 cells, 19 were excavated to a depth between 1 and 4 feet (the other 13 cells had been designated clean based on previous investigation results). One cell at the railroad property was still above the RAL at 4 feet and was backfilled without further excavation in accordance with the work strategy.
The work strategy also required the collection of soil samples from the bottom of each excavated cell. For residential cells, 9-aliquot multi-increment samples were collected to reduce the chance of missing isolated hot spots. These samples were analyzed by using the immunoassay test kits to confirm that the remaining soil was below the RALs for pesticides. If the results indicated that the soil was not below the RAL, further excavation was conducted.
A demonstration of methods applicability (DMA) was conducted in March 2001 during initial removal assessment sampling at the site to demonstrate the ability of the immunoassay kits to provide useful data. Comparisons of the immunoassay kits to the laboratory methods were made to establish FBALs based on optical density (OD) for the industrial and residential/railroad portions of the site. The kit?s immunoassay result was reported as the direct read-out of OD as measured by the photometer instrument. High OD readings correspond to lower concentrations of pesticides. Conversely, lower OD readings correspond to higher concentrations. Initially, the FBALs were set to an OD of 0.45 for the industrial area, and 0.9 for the residential/railroad area, corresponding to a lower concentration threshold for residential areas than the threshold for industrial land use. Laboratory results were classified based on whether any chlorinated pesticide was present at a concentration greater than the project RAL. There were on-going comparisons between the field and lab results during the course of the investigation.
Based on the DMA, the QAPP specified the following decision logic for samples from the residential/railroad and industrial areas of the site:
Partway into the project, it became clear that the industrial FBAL was initially set too conservatively, therefore it was refined to 0.35 for the rest of the investigation (see Data Quality section below). The QAPP was amended to reflect this change.
A graphical representation of the soil sample data showing the relationship of the OD measurements to the laboratory findings (i.e., whether or not any pesticide was present at a concentration greater than the RAL) is shown here . This chart shows the "gray region" (as termed "a region of decision uncertainty") where the OD measurement does not clearly identify whether a sample exceeds or does not exceed the RAL for any individual pesticide, as well as the FBALs used on the project. The same data set is plotted for the residential/railroad RALs and FBALs, and the industrial RALs and FBALs.
After cells were identified for excavation based on the above decision logic, the excavation program was completed, with the confirmation samples from the final excavation depth falling into the following categories:
Ultimately, 203 samples were screened using the immunoassay test kits to document achievement of RALs for excavation or disposal, or UTCs for treatment. Of those samples, 56 were also analyzed by the laboratory method.
Field based methods for soil included:
Laboratory-based methods for soil included:
The IA test kits measure the OD (that is, the color intensity) of a solution after extraction of soil, filtering, and exposure of the extract to an antibody that binds the contaminants. As noted previously, the OD is inversely proportional to the concentrations (the darker the color, the lower the concentration).
The IA test kit results were compared to the laboratory results. Two charts show the comparisons. The comparisons were revisited several times during the project to ensure that the FBAL for the test kits was still accurate. At one point during the project, the FBAL was revised for the industrial portion of the site. See the Data Quality section for more information.
Prior to a full startup of the removal action, a proof of performance test was conducted on the thermal treatment unit. The results indicated that the thermal treatment unit was able to achieve the targeted destruction removal efficiency for the chemicals of concern. Air samples were also collected as part of this test.
TQRS not prepared
Sample Handling QC: Disposable sleeves were used with the Geoprobe soil sampling apparatus to minimize the risk of cross-contamination of samples; therefore no rinsate samples were collected. No duplicate samples were required because it was decided that evaluation of total method precision was beyond the scope of the removal action. However, one duplicate sample was noted in the removal reports, with a 12% relative percent difference (%RPD).
Analytical Method QC: Negative and positive controls were also used as quality control measures for the immunoassay method. Results for these control samples were not published in the removal reports for the site.
The performance of the immunoassay test kits relative to the laboratory samples was continually monitored during the project. As additional data were obtained and reviewed, it became apparent that the industrial FBAL was overly conservative. No sample with an OD above 0.35 had shown any pesticide concentration in laboratory confirmation analyses that was above the corresponding industrial RAL. The QAPP was amended in August 2001 to change the FBAL to an OD of 0.35, reducing the cost of cleanup while continuing to meet removal goals.
During the project, a concentration range of decision uncertainty (also called a "gray region ") was defined for the performance of the immunoassay method relative to both FBALs.
Data was managed in accordance with EPA Region 7 requirements. The project did not feature the use of statistical or other decision support tools, or the use of web-based data sharing or communication protocols, due to the small size of the stakeholder team.
913-551-7600
garvey.daniel@epa.gov 816-412-1754 ted.faile@ttemi.com 913-851-3919 dkinroth@seagullenvirotech.comTo update this profile, contact Cheryl T. Johnson at Johnson.Cheryl@epa.gov or (703) 603-9045.
