PNNL

Abstract

Carbon tetrachloride (CT) was discharged to waste sites that are included in the 200-PW-1 Operable Unit in Hanford 200 West Area. Fluor Hanford, Inc. is conducting a Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) remedial investigation/feasibility study (RI/FS) for the 200-PW-1 Operable Unit. The RI/FS process and remedial investigations for the 200-PW-1, 200 PW-3, and 200-PW-6 Operable Units are described in the Plutonium/Organic-Rich Process Condensate/Process Waste Groups Operable Unit RI/FS Work Plan. As part of this overall effort, Pacific Northwest National Laboratory (PNNL) was contracted to improve the STOMP simulator (White and Oostrom, 2006) by incorporating kinetic volatilization of nonaqueous phase liquids (NAPL) and multicomponent flow and transport. This work supports the U.S. Department of Energy’s (DOE’s) efforts to characterize the nature and distribution of CT in the 200 West Area and subsequently select an appropriate final remedy. Previous numerical simulation results with the STOMP simulator have overestimated the effect of soil vapor extraction (SVE) on subsurface CT, showing rapid removal of considerably more CT than has actually been recovered so far. These previous multiphase simulations modeled CT mass transfer between phases based on equilibrium partitioning. Equilibrium volatilization can overestimate volatilization because mass transfer limitations present in the field are not considered. Previous simulations were also conducted by modeling the NAPL as a single component, CT. In reality, however, the NAPL mixture disposed of at the Hanford site contained several non-volatile and nearly insoluble organic components, resulting in time-variant fluid properties as the CT component volatilized or dissolved over time. Simulation of CT removal from a DNAPL mixture using single-component DNAPL properties typically leads to an overestimation of CT removal. Other possible reasons for the discrepancy between observed and simulated CT mass removal during SVE are differences between the actual and simulated 1) SVE flow rates, 2) fluid-media properties, and 3) disposal history (volumes, rates, and timing). In this report, numerical implementation of kinetic volatilization and multicomponent DNAPL flow and transport into the STOMP simulator (White and Oostrom, 2006) is described. The results of several test cases are presented and explained. The addition of these two major code enhancements increases the ability of the STOMP simulator to model complex subsurface flow and transport processes involving CT at the Hanford site.