Formatted Title

Can Density-Driven Transport Affect Chlorinated Vapor Intrusion?

Background/Objectives

Chlorinated volatile organic compounds (CVOCs) present in the subsoil may cause vapor intrusion (VI) into overlying buildings with consequent potential long-term risks to the health of the residents. In many regulatory programs, to assess potential indoor air risks, VI screening levels (VISLs) are calculated using attenuation factors (AFs), defined as the ratio of indoor air to sub-slab or soil vapor concentrations. When AFs are used for VISLs calculations, they are often based on a fixed attenuation factor extrapolated from empirical findings. For instance, U.S. EPA VI guidance (2015) recommends the use of default values of AFs for VI assessments, which are considered independent from the contaminant source concentration. Recently, Lahvis and Ettinger (2021) and Abbasi et al. (2023) conducted an extensive study, which involved paired measurements of indoor and subsurface soil gas concentrations of CVOCs from hundreds of buildings in California. Both studies showed a consistent trend of decreasing AFs as a function of subsurface soil gas concentrations of chlorinated contaminants, indicating a clear correlation between the AF and subsurface vapor concentration. In this study, we examine if this trend can be attributed to the establishment of a downward density-driven transport. Indeed, there is evidence in the literature that for dense chlorinated compounds and high soil permeability, density-driven transport (i.e., advection transport by density gradients) could be relevant. Typically, density-driven transport is not considered for evaluating VI even for heavy chlorinated compounds. However, many buildings in the U.S. are constructed by posing the foundations on a high permeability layer for capillary break, in which density-driven transport could be relevant. In this study, we investigate the role of density-driven transport on VI by CVOCs in such scenarios. To this aim, a two-layer 1-D analytical model at steady state was developed to simulate the VI pathway considering diffusion and density-driven advection for several CVOCs. To validate the model, sand-filled column tests were performed in the laboratory using TCE as the target compound. Next, the model was applied to assess the expected AFs in the presence of density-driven transport, comparing the results with the field data available in the literature.

Approach/Activities

The 1-D analytical model was developed at steady-state conditions considering two layers of soil with different permeabilities. In this way, it was possible to simulate a scenario in which building foundations were constructed on a drain layer. The volatilization of CVOCs from the groundwater source through the two layers was considered affected by upward diffusion and downward density-driven advection. The source to building AF was evaluated by using the developed model from the source of contamination to the sub-slab and by setting 0.03 from the sub-slab to building. Diffusion tests of TCE vapors were performed in the laboratory, filling glass columns with sand to investigate the vertical soil gas concentration profiles.

Results/Lessons Learned

The TCE concentration profiles within the columns were found lower than the ones expected assuming a diffusion-dominated transport. Conversely, the observed profiles well fitted the concentrations expected by the developed model considering density-driven transport. Similar outcomes were also found by comparing the AFs expected by the developed model with the ones observed by Lahvis and Ettinger (2021) and Abbasi et al. (2023) from field paired measures for CVOCs. In particular, the field AFs data were well fitted by the trends expected using the developed model considering a density-driven transport in the layer under the building foundations with soil permeabilities values in the order of 10-6-10-10 m2 (i.e., gravel range), thus simulating the scenario in which a drain layer for capillary break is present under the building foundations. Hence, the obtained results suggest that density-driven transport could be relevant for chlorinated VI, especially in the presence of high permeability layers.