Formatted Title
Sources and Reference Concentrations of PAHs in Surface Soil at Eielson Air Force Base
Background/Objectives
Polycyclic aromatic hydrocarbons (PAHs) are commonly detected in soils from Eielson Air Force Base (AFB) in Alaska, particularly near-surface soils, at concentrations exceeding state cleanup levels, including at sites where there is not a known source of PAHs. These PAHs may be derived from anthropogenic sources such as petroleum products (including asphalt); coal; and combustion of fuels including diesel, gasoline, jet fuel, coal, and wood; and from natural sources such as forest fires and decaying biological materials. Particularly in cases where these PAHs are commingled with other contaminants, stakeholders often question whether the PAHs are site related. The objective of this study was to characterize PAHs in surface soils on and in proximity to the base, using forensic techniques and statistical methods, to provide lines of evidence that can be used in evaluating whether PAHs detected in soil are derived from site-specific activities.
Approach/Activities
Two hundred surface soil reference samples were collected from areas without known site-specific PAH sources to evaluate the range of PAH concentrations present on Eielson AFB and in nearby areas. Additionally, 30 surface soil samples were collected from known contaminated sites, including fuel release sites and the perimeter of coal piles associated with the base’s power plant. The reference samples were grouped by characteristics expected to impact PAH signatures and concentrations, including distance from pavement (greater than or less than 30 ft) and land use (developed or undeveloped areas).
A forensic source evaluation was conducted to characterize PAH composition in surface soils using multivariate analysis, specifically polytopic vector analysis (PVA) and t-Distributed Stochastic Neighbor Embedding (t-SNE), as well as diagnostic PAH source ratio analysis, as two independent approaches to support multiple lines of evidence. Statistical analysis was also performed on data from each group, including maximum detections; means; percentiles; and the 95% upper tolerance limits (95UTLs).
Results/Lessons Learned
The multivariate analysis using PVA and t-SNE defined six distinct, statistically-derived end-members that can be used to describe the PAH composition of the samples collected as part of the study. These include the two pyrogenic end-members related to coal combustion and petroleum combustion, one coal-related end-member, and three petrogenic end-members related to used/weathered petroleum products. The diagnostic ratio approach evaluated the ratios of individual PAHs known to be associated with petrogenic and pyrogenic source types, including the more recalcitrant higher molecular weight (3-6 ring) compounds.
Results from multivariate analysis and diagnostic ratios were consistent and indicate that PAHs in samples from the main base area are associated with combustion-related and coal-related sources, which also correlate with the higher total PAH samples. The highest PAH concentrations in surface soil were reported within 30 feet of roadways and tarmac on the main base area. Samples from areas off the main base had lower total PAH concentrations and were more commonly associated with petrogenic sources. PAHs in soil more than 30 feet from pavement on the main base were pyrogenic in nature but had only slightly elevated concentrations relative to undeveloped areas.
The forensic and statistical analyses together provide strong evidence that pyrogenic PAHs, unassociated with specific release events, are present in surface soil across the base, particularly near pavement. Extensive coordination with state and federal regulators occurred prior to approval of the study report and resulted in a proposed path forward for assessing whether PAHs detected in soil at sites at Eielson AFB are related to site-specific sources or consistent with the reference concentrations of PAHs present in surface soil at the base.