In 2003 the U.S. Army Corps of Engineers (USACE) used the Triad Approach to expedite a high-resolution site characterization (HRSC) of contaminated soil at the Former Evergreen Infiltration Training Range in Fort Lewis, Washington. The characterization was designed to determine if surface soils contain significant concentrations of metals, with the focus on collecting a high-density data set for determining appropriate future actions (i.e., risk analysis or soil remediation). A dynamic sampling and analysis strategy based on rapid field based analytical methods was created to streamline site activities and save resources while increasing confidence in remediation decisions. During the DMA at the beginning of the field investigation, concurrent analysis of soil samples using both FP-XRF and laboratory methodologies was implemented to establish a correlation between FP-XRF and laboratory data. Immediately following the DMA, contaminated soil from the impact beam was delineated by collecting both FP-XRF data and fixed-base laboratory confirmation samples. The combined high-density data set provided analytical results that refined the CSM for the range and directed additional sample collection activities to more clearly define the extent and distribution of soil contamination.
|Site Name||Evergreen Former Infiltration Range|
|Location||Fort Lewis, WA|
|Site Type||Small Arms Firing Range|
|Project Lead Organization||Fort Lewis Public Works|
|Project Lead Type||U.S. Army Lead|
|Regulatory Lead Program||State Remedial|
|Triad Project Status||Field Program Completed|
|Reuse Objective Identified||Yes|
|Proposed Reuse:||Military Barracks|
Fort Lewis is a major military facility located approximately 6 miles south of Tacoma, Washington. The facility consists of approximately 34,875 hectares of cantonment areas, natural prairies, lakes, wetlands, and forests. Weapons qualifications and field training has occurred at Fort Lewis ranges since around the time the Fort was established in 1917. The former Evergreen Infiltration Range was initially identified from a 1951 aerial photograph. There are no records to confirm how long the range was active, however, based upon growth of vegetation observed during site visits and analysis of historical aerial photography, it appears that activity at this range was decreasing between 1955 and 1957. Subsequent photographs from 1965 to the present indicate that the range had not been used since that year. Infiltration ranges provided training opportunities for soldiers to move under live fire and under combat type situations. Fixed-position machine guns placed on concrete footings provided the live fire training. The ammunition associated with infiltration range training during this era was the 30-caliber cartridge. The primary constituents in the bullet slugs consist of 97 percent lead and 2 percent antimony with trace amounts of copper. Potential contaminants of concern are lead, antimony, arsenic, copper, tin, and zinc. As an infiltration range, the impact berm was set back approximately 300 feet from the firing discharge area. The impact berm is a constructed earthen bank approximately 40 feet high. Bullet slugs and fragments are evident at the impact berm. Trees, grasses, and shrubs currently cover a large portion of the area since active use for training has not occurred on the site for several decades.
Soil concentrations greater than 250 milligrams per kilogram (mg/kg) (the State of Washington clean-up level for unrestricted use) were found across the front face of the berm with the highest concentrations located at the impact zone. Lead concentrations greater than 250 mg/kg were present down slope along the toe of the berm in the 0 to 12 inch depth interval. Concentrations remain significantly higher in the middle of the impact zone in the 12 to 24 inch depth interval, with decreasing lead concentrations encountered moving away from the impact zone. Bullet fragments were present to a depth of at least 2 feet within the impact zone. Soil lead concentrations greater than 250 mg/kg were present in the 0 to 12 inch depth interval across the back face of the impact berm. Since no bullets were found, the origin of the contamination was unclear. Lead contamination is highly heterogeneous due to the nature of the contamination source. However, highest concentrations are primarily in the 1-foot depth interval with significant decrease of lead concentration in the 2-foot depth interval. Some limited lead contamination was encountered in samples collected within a trench approximately 75 feet southeast from the backside of the berm; the source of the contamination was not definitively identified.
Results from the DMA study indicated that FP-XRF field technology was appropriate for this Site investigation. The linear regression correlation coefficient factor (r2) for the DMA data set was 0.96 between the FP-XRF and comparative laboratory data for lead. The resulting collaborative data set from the Triad field investigation confirmed that the FP-XRF reliably quantitated lead contamination to a concentration as low as 45 mg/kg in soils. The FP-XRF was not only effective in identifying contaminated areas, but when used in an HRSC framework was also able to clearly locate "clean" areas. Laboratory analysis of collaborative soil samples confirmed that lead was the primary driver to define the remedial action, since other metals were not above Washington State clean-up levels when lead was not above the lead criterion. Based on the refined CSM, lead concentrations in soils posed a risk to potential human health and ecological receptors by direct contact, ingestion, root contact, or inhalation of dust.
The objective was to determine the environmental condition of the property to support the development of a cost-efficient strategy for assessing the extent of lead contamination at the former range and remedial alternatives.
The high-resolution, Triad-based approach offered a cost effective approach to gather the information necessary to ensure that confident decisions could be made in an efficient and timely manner. Investigation and mobilization costs were minimized.
No formal cost comparison was performed. Because the high-resolution, Triad-based approach minimized the number of mobilizations necessary to characterize the site, it is estimated that months of time and thousands of dollars were saved.
Systematic project planning (SPP) began with team formation. Coordination was accomplished with meetings, telephone conferences and e-mail. After initial project team discussions, the USACE prepared a SPP memorandum and data quality objectives (DQOs) to guide Triad implementation. These materials included a preliminary CSM that was based on information from reports developed for similar sites in the area. Uncertainty regarding the presence of hot spots and volumes of soil exceeding regulatory criteria was managed with a strategy of progressive high-resolution sample evaluation based on the preliminary CSM. Field techniques were used to reduce sampling uncertainty, while laboratory analysis was used to manage field analytical uncertainty for lead by providing a collaborative data set for comparison to the FP-XRF data. Additionally, the laboratory methods were used to determine the presence of analytes other than lead, and for field data interpretation.
The project's decision addressed the project objective of delineating the nature and extent of metals contamination in soil at the site. A systematic grid was used to delineate the vertical and horizontal extent of contamination. Starting with the areas most likely to be contaminated (e.g., impact berm), sample locations were stepped out laterally until lead FP-XRF values were detected below the clean-up level. Samples were collected from both the 0 to 12 inch and 12 to 24 inch depth intervals. Sample location density was initially determined using process knowledge of site usage; it was modified as real-time data was collected. The HRSC Sampling and Analysis Plan (SAP) allowed for the collection of additional samples to define the extent of contamination. Once real-time data were obtained from the FP-XRF, the sampling density was evaluated and increased in certain areas based on the decision logic as the investigation progressed.
The potential need for additional sample collection was to be determined after evaluating the FP-XRF data obtained from the initial sampling grid. For example, if results from several grids indicated lead contamination greater than the clean-up level, no additional information was required; however, higher-resolution refinement of the CSM was required where uncertainty regarding the boundaries of the lead-contaminated soil existed. The project decision logic was incorporated into the SAP. The logic began with the CSM's premise that lead contamination if present would be at the highest concentrations in the impact berm. The initial sampling grid was designed to ensure detection of potential contamination. The logic included provisions to delineate the vertical and horizontal extent of any contamination discovered. FP-XRF was the primary analytical method used to address sampling uncertainty and to determine the pattern of contaminant distribution. Laboratory analysis was used to obtain results for metals other than lead, to aid interpretation of the data, to confirm the FP-XRF data, and to gather data below the FP-XRF detection limits.
Following approval of the SAP, soil samples were collected from a pre-established grid. The initial sample collection was performed at both the 0 to 12 inch and 12 to 24 inch depth intervals. Additional depths were considered, however, the instability of the berm soil did not allow the collection of samples below 2 feet using hand tools. Starting with the areas most likely to be contaminated, sample locations were stepped out laterally until lead FP-XRF values were detected below the clean-up level. Additional soil samples would have been collected if it were found necessary to delineate the extent of contamination. The soil samples were analyzed using FP-XRF with collaborative samples sent for laboratory analysis. The FP-XRF analysis was performed both in the field and in the controlled conditions of the Seattle District soils laboratory. The laboratory method was EPA 6010/6020 for metals of interest with Method 3050B for predigestion.
During the DMA, 40 samples were collected from various areas of the berm in order to obtain a potential range of concentrations. All were measured by both FP-XRF and fixed-base laboratory analyses to determine the correlation between measurement methods.
All samples were sieved through a #10 sieve and homogenized in a stainless steel bowl before being placed in a gallon-sized plastic baggie.
Quality control (QC) measures included well defined and implemented sampling procedures as well as collection of collaborative samples, co-located field duplicates, and precision samples. Collection of collaborative samples was greatest in the region of decision uncertainty around the clean-up levels; the collaborative samples were analyzed by fixed-base laboratory methods to provide definitive measurements of contamination levels in the sample. The linear regression correlation coefficient factor (r2) for the FP-XRF and the collaborative sample data was 0.96. The collaborative samples showed good agreement at the action levels. The co-located field duplicates were collected to assess site heterogeneity and the precision samples were measured to determine within-sample heterogeneity. Both the co-located and the precision sample results indicated heterogeneity, which provided information on sample variability and guidance for decisions based upon sample results.
SADA, Spreadsheets. Data visualization using SADA was used to maintain close communication with team members as work progressed and to evaluate statistical uncertainty.
September 2 - October 3, 2003
December 2 - 4, 2003
|DMA Memorandum (1.2 MB)|
|DQO Memorandum (68 KB)|
|Draft Site Investigation Report (45 MB)|
|Fort Lewis Evergreen Plans-Final (1.3 MB)|
|Fort Lewis Final Workplan (4.7 MB)|
|Fort Lewis Interim Cleanup Action Plan (4.6 MB)|
|Path Forward Memorandum (3.7 MB)|
|Project Planning Memorandum (67 KB)|
To update this profile, contact Cheryl T. Johnson at Johnson.Cheryl@epa.gov or (703) 603-9045.