Vadose Zone Transport of 1,1,1‐Trichloroethane: Conceptual Model Validation through Numerical Simulation

Abstract

Wright, 1992). As pore spaces become drier, air pathways become better connected and vapor can more eas-A conceptual model is developed to better understand vadose zone ily migrate under molecular diffusion or pressure driven vapor-phase diffusion within the mesas of the Pajarito Plateau at Los Alamos National Laboratory. We focus on 1,1,1-trichloroethane (TCA) flow. Rapid vapor transport can also be caused by large vapor transport from a liquid-waste disposal site. The conceptual model variations in barometric pressure (Auer et al., 1996) and incorporates several physical processes, including partitioning of TCA pressure gradients resulting from topographic features into the liquid phase, saturation-dependent diffusion, diffusion through (Weeks, 2001). asphalt, and interaction with the atmosphere. Three-dimensional nu-Several factors can act to reduce the ability of vapormerical simulations that use the conceptual model of TCA transport phase VOCs to migrate in the vadose zone. As pore spaces are then calibrated to pore-gas monitoring data. Adjustable paramebecome saturated, vapor-phase tortuosity increases and ters in the numerical simulations are limited to (i) the vapor-phase both the effective vapor-phase diffusion coefficient and diffusion coefficients for the different geologic units, asphalt cover, and the gas-phase permeability are reduced (Jury et al., 1991). land-atmosphere boundary layer and (ii) the fixed concentrations inVolatile organic compounds tend to have low solubility, the two shaft clusters. By including all of the components of the conceptual model in our numerical simulations we were able to achieve but some partitioning into vadose zone water occurs and a reasonable match between the simulated plume and site data for two can slow the migration of VOC vapor (Jury et al., 1990). alternate conceptual models of asphalt, one with asphalt as a diffusivePartitioning is often assumed to be an equilibrium phebarrier and one without asphalt as a diffusive barrier. A goodness-of-fit nomenon; however, Thomson et al. (1997) reported that analysis shows that the best-fit simulations are highly correlated to this may be an oversimplification of field conditions. data points from 21 boreholes. The simulations demonstrate thatAnother process that slows VOC migration is sorption diffusive behavior describes the general characteristics of the current onto mineral surfaces, which may or may not be reverssubsurface vapor plume. Effective vapor-phase diffusion coefficients ible. Slow desorption of VOC in the vadose zone can used in the simulations that best fit the data suggest that barometric lead to long-term sources that are extremely difficult to pumping is not contributing to diffusion in the deep vadose zone; howremediate. Finally, degradation of VOCs by both abiotic ever, it is likely that barometric pumping is occurring in fractures near the mesa edge. We conclude that asphalt is most likely acting as

Evidence for High Rates of Gas Transport in the Deep Subsurface

“…MDA L operated from the 1960s to 1985 as a liquid chemical waste disposal site (Stauffer et al, 2005). Waste drums emplaced in shafts subsequently leaked, forming a subsurface VOC vapor plume that extends beyond the boundaries of the 2.5 acre MDA L site and to a depth of approximately 90 m (Behar et al, 2018).…”

“…The mesa is bounded by canyons: Cañada del Buey to the north and Pajarito Canyon to the south (Stauffer et al, 2005). The mesa is bounded by canyons: Cañada del Buey to the north and Pajarito Canyon to the south (Stauffer et al, 2005).…”

“…The model is built within the FEHM (Finite Element Heat and Mass) porous-flow simulator, developed at LANL, and used successfully to simulate barometrically pumped contaminant transport in fractured rock (Harp et al, 2018;Jordan et al, 2014Jordan et al, , 2015Neeper & Stauffer, 2012a, 2012b. Previous publications describing simulations of the MDA L plume can be found in Stauffer et al (2005) and Behar et al (2018). The tracer test data can help to constrain aspects of the physical system such as permeability, porosity, and dispersivity.…”

“…The tracer test was conducted at Material Disposal Area L (MDA L), the site of a subsurface VOC vapor plume ( Figure S1) at Los Alamos National Laboratory (LANL) in New Mexico, USA. While VOC movement within the tuff units that contain the plume can be understood using a diffusive transport model, there remains considerable uncertainty regarding the transport of vapors through the deep, underlying fractured and rubblized basalt (Behar et al, 2018;Stauffer et al, 2005). To address this uncertainty, we employ data from the tracer test to develop a numerical model that simulates gas flow and tracer transport in the deep basalt under barometric pumping conditions.…”

“…The MDA L site is located on Mesita del Buey, a narrow finger mesa of the Pajarito Plateau in New Mexico, USA. The mesa is bounded by canyons: Cañada del Buey to the north and Pajarito Canyon to the south (Stauffer et al, 2005). The plateau was formed between 1.6 and 2.8 Ma (million years ago) by eruptions of the Cerros del Rio (2.3 and 2.8 Ma) and Jemez (1.6 Ma) volcanic fields (Broxton & Vaniman, 2005).…”