|
In the fall of 2004, the South Dakota Petroleum Release Compensation Fund (PRCF) conducted a study to evaluate the effectiveness of using the Triad Approach to manage decision uncertainties at petroleum release sites.
The following five sites were chosen for the study: Steve's Amoco, Former T&T Standard, Severson's Service, DM&E Railroad, and Former Husky Oil. Two of the sites remain active fueling locations while the other three have had USTs and associated piping removed. The five sites were considered "legacy sites" because petroleum releases had been discovered years prior to the study, yet none were moving efficiently towards regulatory closure. Several of the sites had remained in the investigation phase for more than 10 years, while others had undergone tank removal with no investigation for over 10 years.
In the fall of 2004, PRCF conducted the study, with assistance from the United States Environmental Protection Agency (EPA) and Columbia Technologies, to rapidly characterize the sites, establish clear and accurate CSMs, and establish cleanup goals to move the five sites to regulatory closure quickly using sound scientific principles.
Results of the study indicated that use of the Triad-based HRSC approach reduced overall data collection and analysis costs by as much as 70% for a given site. The approach also assisted in expediting work and allowing the project team flexibility to collect data necessary for decision making in a single mobilization.
| Site Name | Multiple petroleum release sites in South Dakota (see Background section below) |
| Site Types |
|
| Project Lead Organization | South Dakota Department of the Environment and Natural Resources (DENR), South Dakota Petroleum Release Compensation Fund (PRCF), United States Environmental Protection Agency (EPA) |
| Project Lead Type |
State Lead |
| Regulatory Lead Program |
South Dakota Department of the Environment and Natural Resources (DENR) |
| Triad Project Status | Field Program Completed |
| Reuse Objective Identified |
No |
All of the five sites investigated under the study were known petroleum release sites, with three having undergone previous investigations (T&T Standard, Severson's Service, and DM&E Railroad). Spills of gasoline and diesel range organics were discovered at the sites with discovery dates ranging from 1990 to 2003. A historical summary of past field analysis is provided below for each site.
Steve's Amoco, Watertown, South Dakota, DENR Site# 98.045/PRCF Site #3385: A release was first discovered in March 1998 during upgrading of underground piping associated with the UST system. A site assessment was performed in 1999 and groundwater monitoring was performed from 1999 through August 2003. Additional assessment was required to determine potential risk to receptors in the area.
Former T&T Standard, Woonsocket, South Dakota: The release was discovered in December of 1991, with assessments performed in 1992, 1993, 1994, and 1999. Three USTs (10,000 gallon gasoline, 8,000 gallon gasoline, and 3,000 gallon diesel) and approximately 900 cubic yards of petroleum contaminated soil were removed in 1994. Groundwater monitoring was performed between 1992 and 2003. Additional assessment was required to define the extent of contamination and evaluate potential risk to utilities and city water supply wells.
Severson's Service, Platte, South Dakota: The release was first discovered in 1983 when a leaking underground storage tank (LUST) investigation was performed for the State in response to petroleum vapors in a nearby sanitary sewer. Additional assessment was performed in 1990 for the State in response to petroleum discovered during installation of a water main on a street north of the site. The tank owner did not comply with DENR directives in 1991 and 1993 to complete an assessment. Additional assessment was also performed in 1998 and 1999 in response to vapors detected in a nearby business. Three USTs and 700 cubic yards of contaminated soil were removed and the sump in the nearby business was sealed in the summer of 1998. However, petroleum was encountered again in the fall of 2001 during replacement of a nearby sanitary sewer. Additional assessment was performed in December of 2001 to better define the extent of contamination. Groundwater monitoring was performed periodically between 1998 and 2004, with contaminated groundwater removed from a drain tile on a quarterly basis.
DM&E Railroad, Pierre, South Dakota: A spill report was filed in November 1990 based on staining and fuel odors in the fueling area of a railroad property. Assessment performed in 1991, found contamination in soil and groundwater and free phase product in both on-site and off-site monitoring wells. Four recovery wells were installed in 1991 and approximately 2,100 gallons of product was recovered by early 2003. Additional assessment was necessary to identify source areas and the extent of free product. A corrective action plan (CAP) was also required for remediation of any remaining free phase product.
Former Husky Oil, Pierre, South Dakota: The site operated as a gas station from 1973 to 1986. USTs were reported as removed in 1986; however, there were no records of an assessment or of any spills. Gasoline and diesel constituents were found in borings and wells along the west property boundary during assessment of a release site to the west in 1995. Free product was measured in at least one of the wells and DENR directed the former property owners to perform an assessment in 1995 and 1997. However, the owners disputed being named responsible parties and by 2004 no assessment had been performed.
Steve's Amoco: The membrane interface probe (MIP) investigation was conducted 11/8/2004 through 11/10/2004. Twenty-seven MIP borings were advanced into the subsurface. Seven soil and two groundwater samples were also collected 11/8/2004 through 11/12/2004.
The 13 objectives were completed, and a CAP was established. The CAP includes replacing a nearby polyvinyl chloride (PVC) water main with a petroleum resistant material and conducting compliance groundwater monitoring. The site was placed into no further action status on July 14, 2009 after the CAP was completed.
Former T&T Standard: The MIP investigation was conducted 11/11/2004 through 11/16/2004. Twenty-three MIP borings were advanced in the subsurface. Two soil and four groundwater samples were also collected 11/12/2004 through 11/16/2004.
The 13 objectives were completed, and a CAP was established. It was determined that groundwater beneath this site is not hydraulically connected with the drinking source water and that maximum contaminant levels (MCLs) were not applicable. The triad assessment identified additional sources in the study area which were removed utilizing state programs. After the CAP was completed, the site was placed into no further action status on January 19, 2009.
Severson's Service: The MIP investigation was conducted 11/16/2004 through 11/19/2004. Thirty MIP borings were advanced in the subsurface. Seven soil and four groundwater samples were also collected 11/16/2004 through 11/19/2004.
The 13 objectives were completed, and a CAP was established that includes mitigating vapor intrusion potential to basements and utilities. The CAP has been completed and the site was placed into no further action status as of December 29, 2009.
DM&E Railroad: The MIP investigation was conducted 11/19/2004 through 11/22/2004. Thirty-one MIP borings were advanced in the subsurface. Two soil, one groundwater, and one product sample were also collected 11/19/2004 through 11/22/2004.
The areal extent of the contaminant plume was delineated, and because of a lack of immediate receptors in the area, the CAP was limited to monitoring the progress of natural attenuation. At this site, because of the properties of the heavier fuel, the MIP had a difficult time "clearing" the trunkline from one boring to the next and the responses were more subtle than for sites with concerns about gasoline. The CAP has since been completed and the site was placed into no further action status as of March 4, 2013.
Former Husky Oil: The MIP investigation was conducted 12/2/2004 through 12/3/2004. Twenty-two MIP borings were advanced in the subsurface. Seven soil and two groundwater samples were also collected 12/2/2004 through 12/3/2004.
The 13 objectives were completed, and a CAP was established that includes removal of a UST and remedial excavation of a source area. After the CAP completion, the site was placed into no further action status on May 8, 2013.
Investigation Conclusions: Additional corrective action was required at two of the sites to eliminate potential risks to receptors. Corrective action consisting of monitored natural attenuation and plume stability was agreed upon for two of the remaining sites. Following some additional vapor monitoring at the fifth site, it was expected that this site would be eligible for a "no further action" (NFA) status.
Three previously unidentified USTs were also discovered during the course of the study due to the density of data provided by the high quantity of direct push technology (DPT) borings. The ability to obtain real time continuous contaminant and lithology information also made identification of the tanks possible. These USTs are considered very likely to have contributed to contamination at their respective sites. Two tanks have been removed and the final tank was scheduled for removal during late summer 2005.
South Dakota PRCF conducted this study to evaluate and report on the effectiveness of the Triad Approach to manage decision uncertainties as they pertain to petroleum release sites in South Dakota. The study was also intended to (1) aid in the way site investigations are conducted both from private and public sector perspectives, (2) determine if this type of approach was technically feasible for other sites in South Dakota, and (3) help reduce or eliminate backlogs that often occur in cleanup programs. Project goals and objectives were specified for each site.
The systematic planning process allowed all stakeholders to be involved in the decision making process to achieve corrective actions at each site. Using real-time, high-resolution data from the MIP allowed development of accurate CSMs for each site in a single mobilization. This allowed all parties to base decisions on the same data set and refined CSMs, helping stakeholders reach a consensus more quickly.
During the study, 133 borings were advanced at the five sites using an HRSC approach, resulting in a total of 349,500 actual data points, or 20 data points per instrument per foot. Each data point included results for conductivity, a photoionization detector (PID), a flame ionization detector (FID), an electron capture detector (ECD), and temperature.
Comparisons with conventional drilling techniques, where a PID reading was collected every 2.5 feet and one soil sample per borehole analyzed at a fixed-base laboratory, resulted in a much lower data density, with a total of 1,403 data points. Use of the Triad-based HRSC approach provided data density 250 times greater than would have been achieved with a conventional approach.
Fixed-base laboratory analytical costs were also reduced by minimizing the number of samples submitted and using MIP results to optimize fixed laboratory sampling locations.
Cost comparisons were made at the three sites where previous assessments had been performed and multiple data gaps still existed. The three conventional assessments cost $35K, $63K, and $103K. The respective Triad HRSC assessments cost $30K (14% savings), $31K (51%savings), and $31K (70% savings). This corresponds to a total savings of $109K for the three sites compared with funds previously spent to assess these sites. This is a less than perfect comparison since the Triad assessments built on the knowledge gained from the conventional assessments. Thus, it may not be appropriate to equate the conventional to the Triad assessment costs and obtain a cost savings. However, the incongruity in the comparison is somewhat offset by the fact that the conventional assessments left the investigators with uncertainty regarding the CSM, while the Triad-based HRSC tended to remove those uncertainties. For the two sites that did not have previous investigations, the Triad HRSC costs were $25K and $32K. In addition, sufficient data were collected at each site to develop CAPs and move the sites out of the investigation phase, where they had been for more than 10 years.
Comparisons of analytical cost savings were not conducted; however, evaluation of fixed-base laboratory data indicates that soil and water samples were collected at approximately one third of the MIP locations, resulting in significant savings and more efficient use of project resources.
The PRCF project manager allowed all team members and stakeholders, including PRCF, DENR, their consultants, and Columbia Technologies, to attend a systematic planning meeting. Most members agreed on the concept of the Triad Approach but there were concerns expressed regarding the collection of the necessary quantitative data and the number of fixed-base laboratory samples required. Ultimately, the field crews were given sufficient decision making power to determine the number of fixed-base laboratory samples necessary to achieve the project objectives.
During the initial systematic planning meeting, team members were also given a summary of the existing state of knowledge concerning their sites to assist in identifying the work that had been conducted, existing data gaps, establishment of cleanup goals, and objectives to reach those goals. Project goals were achieved through systematic planning and stakeholder buy in. Early and continuous communication between field teams and regulatory authorities resulted in team consensus.
CSMs for each site were developed using refined and existing information and high-resolution, real-time results from the MIP. Data was presented in 3-D using Columbia's SmartData Solutions software to visually represent instrument responses from the MIP. Using these representations, field crews and stakeholders could identify potential data gaps in real-time and use dynamic work strategies (DWS) to delineate contamination and optimize sampling programs.
PRCF and EPA provided the funding mechanisms, while Columbia Technologies, site consultants, and Matrix Technologies (the drilling subcontractor) provided the crews necessary to complete field work. Regulatory oversight was conducted by the South Dakota DENR.
At least two MIP borings were advanced in the location of known areas of high soil contamination at each site. These initial borings provided information on the expected vertical extent of the plume as well as providing instrument response factors that could be used to provide comparisons with respect to contamination at other boring locations.
Following placement of these initial borings, historical site information was used to optimize other boring locations in an effort to horizontally delineate the extent of soil and groundwater contamination. In additional to contaminant delineation, the focus of the field crews was on potential pathways and receptors in the immediate area of the plume to evaluate the level of risk associated with those receptors.
Daily planning meetings among team members were held at each site as the field work progressed. Meetings included reviews of the site objectives and the previous day's data to identify any new data gaps. If uncertainties were identified, with respect to the data, information was collected the following day to eliminate or limit those uncertainties.
At one site, a source that the initial CSM identified as impacting a SWPA, was actually determined to be a separate source and not the result of site contamination. Additionally, the conductivity results indicated that the general lithology in the area was not consistent with that of aquifer materials, and that groundwater beneath this site was not hydraulically connected to the aquifer currently used for drinking water. The new information allowed an evolution of the CSM and the corrective action was changed from compliance with drinking water standards to evaluation of the plume stability, elimination of risk to potential receptors, and monitored natural attenuation.
Decision logic diagrams were not completed for this investigation, however daily project meetings allowed all stakeholders to identify any new data gaps as a result of the previous day's activities. The project team could then agree on a course of action (for example, step out delineation of contamination or increasing data density in areas of concern) to address data gaps and collect information crucial to achieving site objectives and making project decisions. The MIP investigation continued until project goals and objectives were met to the satisfaction of stakeholders.
The MIP and Columbia's SmartData Solutions software were used as real-time measurement technologies. Using the software, Columbia was able to present 3-D presentations of subsurface soil and groundwater contamination. Although a demonstration of methods applicability (DMA) was not conducted, initial MIP borings at each site were used in conjunction with expected soil and groundwater concentrations to estimate detector response. Soil and groundwater samples were sent for fixed laboratory analysis using EPA Methods SW-846 8260 and SW-846 8015M.
The MIP uses a heater block to heat soil and water near the membrane to 121°C allowing volatile organic compounds (VOCs) like BTEX and TPH-gasoline to partition across the membrane. Nitrogen is used as a carrier gas to bring the sample to the surface for exposure to a series of detectors. Detector configuration for this study included a PID, ECD, and FID in series.
TQRS not prepared
Data quality reviews included validation of fixed laboratory data using the National Functional Guidelines
Data visualization tools were used during this project. Data management consisted of using Columbia's SmartData Solutions software to produce cross sectional, three dimensional, and planar representations of contaminated soil and groundwater to help stakeholders conceptualize the data. Raw data in the form of spreadsheets and graphs were reviewed by stakeholders as they were generated while "pushing" the MIP. This raw data stream was processed overnight into stakeholder-requested representations of the data in preparation for sampling activities the following day. Real time data generation and review allowed the team to agree on how to close data gaps and sufficiently refine the CSM during a given site-specific mobilization.
605-773-3769To update this profile, contact Cheryl T. Johnson at Johnson.Cheryl@epa.gov or (703) 603-9045.
