Search for: All records

Abstract Air–water interfacial adsorption complicates per‐ and polyfluoroalkyl substance (PFAS) transport in vadose zones. Air–water interfaces can arise from pendular rings between soil grains and thin water films on grain surfaces, the latter of which account for over 90% of the total air–water interfaces for most field‐relevant conditions. However, whether all thin‐water‐film air–water interfaces are accessible by PFAS and how mass‐transfer limitations in thin water films control PFAS transport in soils remain unknown. We develop a pore‐scale model that represents both PFAS adsorption at bulk capillary and thin‐water‐film air–water interfaces and the mass‐transfer processes between bulk capillary water and thin water films (including advection, aqueous diffusion, and surface diffusion along air–water interfaces). We apply the pore‐scale model to a series of numerical experiments—constrained by experimentally determined hydraulic parameters and air–water interfacial area data sets—to examine the impact of thin‐water‐film mass‐transfer limitations in a sand medium. Our analyses suggest: (a) The mass‐transfer limitations between bulk capillary water and thin water films inside a pore are negligible due to surface diffusion. (b) However, strong mass‐transfer limitations arise in thin water films of pore clusters where pendular rings disconnect. The mass‐transfer limitations lead to early arrival and long tailing behaviors even if surface diffusion is present. (c) Despite the mass‐transfer limitations, all air–water interfaces in the thin water films were accessed by PFAS under the simulated conditions. These findings highlight the importance of incorporating the thin‐water‐film mass‐transfer limitations and surface diffusion for modeling PFAS transport in vadose zones. 

Abstract Per‐ and polyfluoroalkyl substances (PFAS) are surface‐active contaminants experiencing strong retention in vadose zones due to adsorption at air–water and solid–water interfaces. Leaching of PFAS through vadose zones poses great risks of groundwater contamination. Prior PFAS transport studies have focused on homogenous or layered vadose zones that significantly underrepresented the impact of preferential flow caused by soil heterogeneities—a primary factor known to dominantly control the subsurface transport of many contaminants. We conduct numerical simulations to investigate the impact of preferential flow on PFAS leaching in stochastically generated heterogeneous vadose zones. The simulations show that while shorter‐chain PFAS experience accelerated leaching similar to non‐surfactant solutes, the accelerated leaching of more surface‐active longer‐chain PFAS is uniquely amplified by 1.1–4.5 times due to reduced accessible air–water interfacial areas along preferential flow pathways. Our study highlights the criticality of characterizing soil heterogeneities for accurately predicting the leaching of long‐chain PFAS in vadose zones. 

Abstract Dynamic pore‐network model (PNM) has been widely used to model pore‐scale two‐phase flow. Numerical algorithms commonly used for dynamic PNM including IMPES (implicit pressure explicit saturation) and IMP‐SIMS (implicit pressure semi‐implicit saturation) can be numerically unstable or inaccurate for challenging flow regimes such as low capillary number (Ca) flow and unfavorable displacements. We perform comprehensive analyses of IMPES and IMP‐SIMS for a wide range of flow regimes under drainage conditions and develop a novel fully implicit (FI) algorithm to address their limitations. Our simulations show the following: (1) While IMPES was reported to be numerically unstable for lowCaflow, using a smoothed local pore‐body capillary pressure curve appears to produce stable simulations. (2) Due to an approximation for the capillary driving force, IMP‐SIMS can deviate from quasi‐static solutions at equilibrium states especially in heterogeneous networks. (3) Both IMPES and IMP‐SIMS introduce mass conservation errors. The errors are small for networks with cubic pore bodies (less than 1.4% for IMPES and 1.2% for IMP‐SIMS). They become much greater for networks with square‐tube pore bodies (up to 45% for IMPES and 46% for IMP‐SIMS). Conversely, the new FI algorithm is numerically stable and mass conservative regardless of the flow regimes and pore geometries. It also precisely recovers the quasi‐static solutions at equilibrium states. The FI framework has been extended to include compressible two‐phase flow, multicomponent transport, and phase change dynamics. Example simulations of two‐phase displacements accounting for phase change show that evaporation and condensation can suppress fingering patterns generated during invasion.