Examining the use of USEPA's Generic Attenuation Factor in determining groundwater screening levels for vapor intrusion.

A value of 0.001 is recommended by the United States Environmental Protection Agency (USEPA) for its groundwater-to-indoor air Generic Attenuation Factor (GAFG), used in assessing potential vapor intrusion (VI) impacts to indoor air, given measured groundwater concentrations of volatile chemicals of concern (e.g., chlorinated solvents). The GAFG can, in turn, be used for developing groundwater screening levels for VI given target indoor air quality screening levels. In this study, we examine the validity and applicability of the GAFG both for predicting indoor air impacts and for determining groundwater screening levels. This is done using both analysis of published data and screening model calculations. Among the 774 total paired groundwater-indoor air measurements in the USEPA's VI database (which were used by that agency to generate the GAFG) we found that there are 427 pairs for which a single groundwater measurement or interpolated value was applied to multiple buildings. In one case, up to 73 buildings were associated with a single interpolated groundwater value and in another case up to 15 buildings were associated with a single groundwater measurement (i.e, that the indoor air contaminant concentrations in all of the associated buildings were influenced by the concentration determined at a single point). In more than 70% of the cases (390 of 536 paired measurements in which horizontal building-monitoring well distance was recorded) the monitoring wells were located more than 30 meters (and some up to over 200 meters) from the associated buildings. In a few cases, the measurements in the database even improbably implied that soil gas contaminant concentrations increased, rather than decreased, in an upward direction from a contaminant source to a foundation slab. Such observations indicate problematic source characterization within the dataset used to generate the GAFG, and some indicate the possibility of a significant influence of a preferential contaminant pathway. While the inherent value of the USEPA database itself is not being questioned here, the above facts raise the very real possibility that the recommended groundwater attenuation factors are being influenced by variables or conditions that have not thus far been fully accounted for. In addition, the predicted groundwater attenuation factors often fall far beyond the upper limits of predictions from mathematical models of VI, ranging from screening models to detailed computational fluid dynamic models. All these models are based on the same fundamental conceptual site model, involving a vadose zone vapor transport pathway starting at an underlying uniform groundwater source and leading to the foundation of a building of concern. According to the analysis presented here, we believe that for scenarios for which such a "traditional" VI pathway is appropriate, 10