The 22-acre area of concern encompasses the 'Old City Landfill,' an auto impoundment where petroleum and metal contaminants may have been released, and a municipal snow piling area potentially contributing polycyclic aromatic hydrocarbons (PAHs), total petroleum hydrocarbons, and metals from melt water. From 1951 until 1965, the unregulated landfill was used for municipal debris potentially containing multiple contaminants of concern (COCs). The site subsequently was covered with 5-20 feet of clean fill and developed for recreational use.
The underlying soil consists of alluvial sand and gravel deposits with interbedded silty overbank deposits up to 500 feet thick. Ground water is shallow (8-15 feet bgs) and generally flows toward the adjacent river. Limited Phase I investigations in 2005 focused on ground-water and soil conditions in and below the landfill. Results indicated PCB and lead concentrations in soil within the landfill exceeding regulatory criteria, as well as slightly elevated manganese and thallium concentrations in ground water.
The Alaska Department of Environmental Conservation (ADEC), U.S. EPA Office of Superfund Remediation and Technology Innovation, and U.S. Army Corps of Engineers-Seattle District assisted FNSB in Triad planning and implementation, including establishing and refining the CSM. Key issues concerning site development on the landfill included gas emissions, potential leachate, and the vertical and horizontal extent of debris. A high-level decision flowchart was developed to correlate release/risk identification to site redevelopment strategies.
Investigation activities included geophysical surveys, in situ soil-gas monitoring, test pit excavation and soil sampling, monitoring well installation, and ground water sampling. A detailed flowchart was established to allow CSM updates in the field and provide criteria for decisions such as sampling locations. For example, preliminary soil-gas sampling plans were designed on standard 300-foot grids in accordance with EPA guidance for investigating landfills (EPA600-R-05-123). Field decisions allowed for adding sampling locations within and outside the original grids and included triggers for additional sampling supporting future modeling of vapor pathways. (Initial ambient air monitoring activity was cancelled due to ground cover by snow and consequent reduction in fugitive gas emissions.)
To maximize application of Triad in a real-time environment, FNSB’s contract mechanism reflected flowchart options on a unit cost basis to allow for maximum flexibility during field work. Field investigations began in October 2007 with mobilization of drill rigs, excavators, and a field laboratory. Work started with a one-week geophysical survey using ground penetrating radar to confirm the landfill’s areal extent (as determined in Phase I investigation) and determine the landfill depth. Survey results confirmed that the landfill was confined to the western part of the site (Figure 3) with soil fill and debris extending 25 feet bgs.
This information was used in real time to adjust sampling locations for soil-gas and ground-water sampling. Soil probes were advanced above the landfill to depths of 7-8 feet, and one soil-gas sample was extracted from each probe. Hand-held flame ionization detectors (FIDs) and photoionization detectors (PIDs) were used to field screen each sample. Those showing hits were analyzed for VOCs using a portable gas chromatograph (GC) housed in a nearby trailer. Due to slow throughput of the field GC, some samples were submitted to a fixed laboratory for VOC analysis; samples were selected from locations of low or non-detect PID readings across the landfill footprint. Over one week, 94 soil gas samples were collected, 41 samples were analyzed with an onsite GC, and 6 samples were submitted to the offsite laboratory.
Results from the PID screening of soil gas predicted higher and more extensive VOC concentrations than the other methods. PID readings indicated concentrations of 0-165 parts per million by volume (ppmv), while field GC and fixed laboratory results indicated only two detections, which were below 1 ppmv. This difference suggested presence of volatile compounds with ionization potentials less than 10.6 EV, which were not included in other analyses.
The six fixed-laboratory results indicated that field screening results may not have quantified soil-gas concentrations. Fixed laboratory results indicated presence of benzene and PCE above ADEC draft screening levels of 3.1 and 8.1 µg/m³, respectively. Other compounds such as trichloroethene and carbon tetrachloride also exceeded screening levels in at least one location. PID measurements were not detected for three of the samples at these locations due to detection limits higher than the fixed laboratory's. Laboratory results also did not agree with field GC data showing non-detects (with the exception of one sample) and generally indicated interference from non-target hydrocarbons. Although the methods did not agree with respect to concentrations, all three methods indicated VOC presence above screening levels throughout the landfill. This information was deemed sufficient for planning purposes.
Traditional, non-Triad methods were used to initially investigate the auto impound and snow piling areas. Six test pits were excavated and soil sample results indicated that concentrations of metals and PAHs in the two areas did not exceed regulatory criteria, and additional planned sampling using Triad methods was unnecessary.
The geophysical survey and soil-gas sampling results were used to select five monitoring well locations on the landfill perimeter for evaluation of flow direction and potential release of leachate into the river. The wells were installed with 5-foot screens set at depths to encompass historical high and low ground-water levels. Only arsenic was detected at concentrations exceeding state regulatory criteria for ground water in one (downgradient) well, where concentrations reached 55.2 µg/L (above the 50 µg/L ADEC action level). Remaining COCs were either below state criteria or not detected in any of the monitoring wells.
Results indicating VOC and methane presence prompted collection of soil and grain-size samples for future air modeling. Soil-gas FID and PID data indicated landfill methane concentrations of 0-8,600 ppmv, while field GC analysis found concentrations reaching 20,700 ppmv. Findings suggest future onsite activities need to account for potential methane concentrations above the lower explosive limit (50,000 ppmv).
Project results indicated that FNSB could:
- Develop the auto impound and snow piling areas without restrictions, and
- Employ soil-gas management strategies such as subslab venting to prevent vapor intrusion into onsite buildings.
The entire sampling program was conducted in a single mobilization over two months at a cost of approximately $200,000, in contrast to a traditional site investigation estimated at $300,000 over five months.