1996 Conference Proceedings
Numerical evaluation of groundwater as a supply for radon in dwellings
M. Oostrom, M.D. White
Pacific Northwest National Laboratory, Richland, Washington
R.J. Lenhard
Sultan Qaboos University, Oman
Proceedings of XI International Conference on Computational Methods in Water Resources, Kluwer Academic Publishers, London, pp. 193-200, 1996.
Abstract
The question whether groundwater may be a source of radon arises from observed high radon concentrations in groundwater, and recent findings that advection in the gas phase may be an important transport mechanism for radon into slightly under pressurized dwellings. A mathematical model has been developed to investigate whether radon degassing from groundwater may contribute to indoor radon levels. To investigate whether the radon that partitions from groundwater to soil gas can contribute to indoor radon levels, multi-fluid flow models need to be used that consider interphase mass partitioning. The developed code presented in this paper considers flow and transport in both the water and gas phase. The mass-conservation equations are solved simultaneously. The radon transport equation, which accounts for advection, diffusion/dispersion, retardation, production, and decay, is solved sequentially using the computed water and gas velocities as inputs. The conservation and transport equations are discretized spatially and temporally in algebraic form using an integrated finite-difference method. The nonlinear discretized equations are converted into linear form using a multivariable Newton-Raphson iteration technique. The code includes a transient Dirichlet boundary condition reflecting the radon concentration in a dwelling. The equivalent continuum approach has been implemented in the code to model cracks in concrete foundations. Numerical simulations were conducted to evaluate indoor radon concentrations as a function of depth to the water table, the intrinsic permeability of subsurface strata, and the pressure gradient in the gas phase. The results suggest that radon degassing from groundwater may contribute significantly to indoor radon concentrations. This may have important implications for regions where the aquifer that underlies dwellings passes through porous media high in radium content and the subsurface materials overlying the aquifer are composed of very coarse sands and gravels, such as is common in many alluvial deposits. The simulations help to explain how some dwellings may have high indoor radon concentrations when the radium content of the underlying geologic strata is low.
Simulation of a quantitative multi-fluid flow experiment
M. Oostrom, M.D. White
Pacific Northwest National Laboratory, Richland, Washington
C. Hofstee, J.H. Dane
Auburn University
Proceedings of XI International Conference on Computational Methods in Water Resources, Kluwer Academic Publishers, London, pp. 449-456, 1996.
Abstract
During the last decade, several multiphase flow and transport codes have been published. Few of the codes have been tested against experimental three-fluid (i.e.,water, nonaqueous phase liquid, air) flow data. To establish the capability of a code to accurately predict subsurface flow and transport phenomena, the code should be tested against quantitative data. Additionally, the correctness of the constitutive relations used in a code can only be properly evaluated by comparison results with those of controlled experiments. In this paper, we present results generated with the nonvolatile three-fluid operational mode of the STOMP (Subsurface Transport Over Multiple Phases) simulator with those obtained during a recently conducted quantitative three-fluid flow container experiment. Governing flow and transport equations are solved numerically by following an integrated finite difference approach. The part of the code describing the constitutive functions between fluid pressures, saturations, and relative permeabilities includes hysteretic capabilities and routines to compute NAPL and gas entrapment.
The flow container, with internal dimensions of 167 cm x 100 cm x 5 cm, was filled under saturated conditions with a mixture of three uniform sands. After lowering the water table to a height of 25 cm above the bottom of the container, 825 ml of the LNAPL Soltrol® 220 was injected at a constant flow rate from a small source located on top of the sand. After allowing redistribution of the Soltrol® in the partly saturated sand for four days, the water table was slowly raised to a height of 65 cm. Seven days later, the water table was slowly reduced to a height of 10 cm. Throughout the experiment, water and NAPL saturations were obtained at a number of locations with a dual-energy gamma radiation system.
The experiment was simulated using two nonhysteretic and one hysteretic constitutive theory to describe relations between relative permeability, saturation, and fluid pressure. The input parameters for the simulations were obtained independently. Results show that the initial infiltration and redistribution of the Soltrol can be modeled accurately with both modes. The hysteretic mode, however, provides better results after the water table has been raised because of the effects of nonwetting fluid entrapment and pore geometry hysteresis. Due to space limitations, only results of the infiltration and redistribution part of the experiment are shown in this contribution.