Effect of vapour-solid interfacial adsorption on benzene multiphase partition and its implication to vapour exposure assessment of contaminated soil in arid area.

The partitioning of volatile organic compounds (VOCs) in soil multiphase system is a critical process for vapour intrusion, however, the importance of vapour-solid interface adsorption doesn't receive the due attention, which causes the exposure assessment too conservative particularly in arid conditions. This paper proposed a multiphase partitioning equilibrium (MPE) model establishing the quantitative relationship between VOCs and its various partitioning phases in soil, including solid-liquid interface adsorption phase, vapour phase and dissolved phase and vapour-solid interface adsorption phase. Taking benzene as the targeted pollutant, the model was found in good agreement with the experimental data while the errors were within one magnitude basically. The role of vapour-solid interface adsorption under different soil moisture conditions was also investigated by the model. The results reveals that a) soil moisture is the conspicuous controlling factor that affects the benzene partitioning in soil; b) vapour-solid interface adsorption dominates benzene uptake when soil relative saturation (RS) is under 20% among three typical soils; c) as adsorption by soil minerals (vapour-solid interface adsorption) is reduced by increasing amounts of humidity (RS > 20%), uptake by partitioning into the soil organic matter (OM) increasingly becomes a controlling factor; d) the common sense that vapour concentration of benzene is particularly high with low level of RS may not occur since the vapour-solid interface adsorption dominates benzene uptake in arid environment. The MPE model is suitable for prediction of VOCs partitioning and vapour exposure risk assessment of contaminated soil in arid area.

Peculiar attenuation of soil toluene at contaminated coking sites.

In the soil of contaminated coking sites, polycyclic aromatic hydrocarbons (PAHs) and benzene, toluene, ethylbenzene and xylene (BTEX) are typical indicator compounds. Generally, PAHs are enriched in the topsoil layer. BTEX, with higher water solubilities and lower organic carbon-water partitioning coefficients (Koc), are distributed deeper than PAHs. However, current models have employed predictions using single compounds to mimic the migration of BTEX at contaminated coking sites. Such models have not considered the influence of the upper soil layer, where PAHs are enriched. An attempt to fill this gap was made by setting up a control soil column experiment in this study. One column was filled with undisturbed soil (column #1) and the other with PAH-contaminated soil (column #2) to simulate the theoretical and actual surface soil layers, respectively. The results showed that in column #2, the toluene gas concentration of the headspace and time required to reach steady state were notably greater than those in column #1. High-throughput sequencing revealed that there were large microbial community structure differences between the two soil columns throughout the experiment, while some genera that degrade toluene with high efficiency emerged noteworthily in column #2. This implied that the upper soil layer enriched with PAHs was conducive to the degradation of toluene vapor. Applying this finding to human health exposure assessment of toluene suggests that the potential exposure level should be reduced from the current predicted level given the unanticipated attenuation at contaminated coking sites.

A source depletion model for vapor intrusion involving the influence of building characteristics.

If volatile organic compound (VOC)-contaminated soil exists underneath a building, vapors may migrate upwards and intrude into the interior air of the building. Most previous models used to simulate vapor intrusion (VI) were developed by assuming that the source was constant, although a few recent models, such as the Risk-Based Corrective Action (RBCA) Tool Kit (TK) model, have been developed to consider source depletion (SD). However, the RBCA TK model ignores the effects of building characteristics due to its assumption that the ground is not covered by the actual building it models, which leads to incorrect results since the presence of the building affects the SD. In this study, a SD model is developed based on the three processes of VI while considering the impact of key building parameters on SD. The proposed model (i.e., the SD model) still follows the law of mass conservation, and the sensitivity analysis shows that the soil-building pressure differential (dP) is an important building characteristic that affects SD. Taking trichloroethylene (TCE) for simulation in the case of a soil concentration below the saturation concentration, as the soil permeability decreases, the differences in the results between the SD model and RBCA TK model decrease; as the Peclet number decreases, the effect of the dP on the results of the SD model decreases. The new model only accounts for the migration of contaminants at the source of depletion; therefore, the model is more applicable for these contaminants, which are considered to have low-biodegradable characteristics. Furthermore, since the model emphasizes the impact of buildings on the source, it is applicable when there is a considerable building area above the source, such as large commercial buildings or residential communities with underground parking lots, which exist in most cities.

Applicability of Soil Concentration for VOC-Contaminated Site Assessments Explored Using Field Data from the Beijing-Tianjin-Hebei Urban Agglomeration.

A total of 128 available soil-soil gas data pairs of benzene were collected from 5 contaminated sites in the Beijing-Tianjin-Hebei urban agglomeration. Soil gas concentrations predicted by the linear model and the dual equilibrium desorption (DED) model were compared with measured values. Although the immersion of soil samples in methanol during sampling and preservation was specified to minimize volatilization losses and biodegradation, the study still found that many points with high soil gas concentrations correspond to unreasonably low soil concentrations. Further analysis revealed that the soil matrices of these points are basically composed of sandy and silty soils, given that soil gas collected may migrate from more contaminated soils nearby due to the large porosity and soil benzene escapes more easily during sampling in the coarser soil particles. Therefore, for sandy and silty soil, collecting soil gas would be more reasonable for screening the vapor intrusion (VI) pathway. For clay, the combination of bulk soil concentration and the DED model will be more convenient. Defaulting f as 1, as recommended by previous studies in the DED, would not be suitable for all cases, and this value needs to be further explored to revise the DED model for future applications.