Martinus Oostrom

STOMP Subsurface Transport Over Multiple Phases, Version 4.0, User’s Guide

This guide describes the general use, input file formatting, compilation and execution of the STOMP (Subsurface Transport Over Multiple Phases) simulator, a scientific tool for analyzing single and multiple phase subsurface flow and transport. A description of the simulator’s governing equations, constitutive functions and numerical solution algorithms are provided in a companion theory guide. In writing these guides for the STOMP simulator, the authors have assumed that the reader comprehends concepts and theories associated with multiple-phase hydrology, heat transfer, thermodynamics, radioactive chain decay, and relative permeability-saturation-capillary pressure constitutive relations. The authors further assume that the reader is familiar with the computing environment on which they plan to compile and execute the STOMP simulator. Source codes for the sequential versions of the simulator are available in pure FORTRAN 77 or mixed FORTRAN 77/90 forms. The pure FORTRAN 77 source code form requires a parameters file to define the memory requirements for the array elements. The mixed FORTRAN 77/90 form of the source code uses dynamic memory allocation to define memory requirements, based on a FORTRAN 90 preprocessor STEP, that reads the input files. The simulator utilizes a variable source code configuration, which allows the execution memory and speed to be tailored to the problem specifics, and essentially requires that the source code be assembled and compiled through a software maintenance utility. The memory requirements for executing the simulator are dependent on the complexity of physical system to be modeled and the size and dimensionality of the computational domain. Likewise execution speed depends on the problem complexity, size and dimensionality of the computational domain, and computer performance. Selected operational modes of the STOMP simulator are available for scalable execution on multiple processor (i.e., parallel) computers. These versions of the simulator are written in pure FORTRAN 90 with imbedded directives that are interpreted by a FORTRAN preprocessor. Without the preprocessor, the scalable version of the simulator can be executed sequentially on a single processor computer. The scalable versions of the STOMP modes carry the “-Sc” designator on the operational mode name. For example, STOMP-WCS-Sc is the scalable version of the STOMP-WCS (Water-CO2-Salt) mode. A separate mode containing an evaporation model as a boundary condition on the upper surface of the computation domain has also been included. This mode, STOMP-WAE-B (Water-Air-Energy-Barriers) can be viewed as an extension of the STOMP-WAE (Water-Air-Energy) mode. Details of this particular mode are outlined by Ward et al. (2005)(a). STOMP V4.0 includes the reactive transport module ECKEChem (Equilibrium-Conservation-Kinetic Equation Chemistry) for the STOMP-W (Water) and STOMP-WCS (Water-CO2-Salt) modes. For this particular module, the “-R” designator is included in the operational mode name (e.g., STOMP-W-R, STOMP-WCS-R-Sc). This mode is described in detail by White and McGrail (2005)(b). For all operational modes and processor implementations, the memory requirements for executing the simulator are dependent on the complexity of physical system to be modeled and the size and dimensionality of the computational domain. Likewise execution speed depends on the problem complexity, size and dimensionality of the computational domain, and computer performance. Additional information about the simulator can be found on the STOMP webpage: http://stomp.pnl.gov. The website includes an introductory short course with problems ranging from simple one-dimensional saturated flow to complex multiphase system computations.

In-situ oxidation of trichloroethene by permanganate: effects on porous medium hydraulic properties.

In-situ oxidation of dense nonaqueous-phase liquids (DNAPLs) by strong oxidants such as potassium permanganate (KMnO4) has been proposed as a possible DNAPL remediation strategy. In this study, we investigated the effects of in-situ trichloroethene (TCE) oxidation by KMnO4 on porous medium hydraulic properties. In particular, we wanted to determine the overall effects of concurrent solid phase (MnO2) precipitation, gas (CO2) evolution and TCE dissolution resulting from the oxidation reaction on the porous mediums aqueous-phase relative permeability, krw. Three TCE removal experiments were conducted in a 95-cm long, 5.1-cm i.d. glass column, which was homogeneously packed with well-characterized 30/40-mesh silica sand. TCE was emplaced in the sand-pack in residual, entrapped form through a sequence of water/TCE imbibition and drainage steps. The column was then flushed under constant aqueous flux conditions for up to 104 h with either deionized water (reference experiment), deionized water containing 5 mM KMnO4 or deionized water containing 5 mM KMnO4 and 300 mM Na2HPO4. Aqueous-phase relative permeabilities were computed from measured flow rates and measurements of aqueous-phase pressure head, h obtained using pressure transducers connected to tensiometers distributed along the column length. A dual-energy gamma radiation system was used to monitor changes in fluid saturation that occurred during each experiment. In addition, column effluent samples were collected for chemical analyses. Dissolution of TCE during deionized water flushing led to an increase in krw by approximately 22% and a local reduction in h. On the other hand, vigorous CO2 gas production and precipitation of MnO2 was visually observed during flushing with deionized water that contained 5 mM KMnO4. As a consequence, krw declined by approximately 96% and h increased locally by more than 1000 cm H2O during the first 24 h of the experiment, causing sand-pack ruptures and pump failure. Conversely, less CO2 gas production and MnO2 precipitation was visually observed during flushing with deionized water that contained 5 mM KMnO4 and 300 mM Na2HPO4. Consequently, only small increases in h (< 15 cm H2O) were observed in this experiment due to a reduction in krw of approximately 53%. While we must attribute changes in h due to variations in krw to our specific experimental design (constant aqueous flux, one-dimensional flow experiments), these experiments nevertheless confirm that successful application of in situ chemical oxidation of TCE requires consideration of detrimental processes such as MnO2 precipitation and CO2 gas formation. In addition, our results indicate that utilization of a buffered oxidant solution may improve the effectiveness of in-situ oxidation of TCE by KMnO4 in otherwise weakly buffered porous media.

Liquid CO2 displacement of water in a dual-permeability pore network micromodel.

Permeability contrasts exist in multilayer geological formations under consideration for carbon sequestration. To improve our understanding of heterogeneous pore-scale displacements, liquid CO(2) (LCO(2))-water displacement was evaluated in a pore network micromodel with two distinct permeability zones. Due to the low viscosity ratio (logM = -1.1), unstable displacement occurred at all injection rates over 2 orders of magnitude. LCO(2) displaced water only in the high permeability zone at low injection rates with the mechanism shifting from capillary fingering to viscous fingering with increasing flow rate. At high injection rates, LCO(2) displaced water in the low permeability zone with capillary fingering as the dominant mechanism. LCO(2) saturation (S(LCO2)) as a function of injection rate was quantified using fluorescent microscopy. In all experiments, more than 50% of LCO(2) resided in the active flowpaths, and this fraction increased as displacement transitioned from capillary to viscous fingering. A continuum-scale two-phase flow model with independently determined fluid and hydraulic parameters was used to predict S(LCO2) in the dual-permeability field. Agreement with the micromodel experiments was obtained for low injection rates. However, the numerical model does not account for the unstable viscous fingering processes observed experimentally at higher rates and hence overestimated S(LCO2).

Oxygenation of anoxic water in a fluctuating water table system: An experimental and numerical study.

A side effect of in situ groundwater remediation techniques that operate by establishing reducing conditions within an aquifer is that anoxic water exits these zones, posing a potential risk to aquatic organisms inhabiting areas of groundwater discharge downgradient from the site. A number of processes have been identified that can attenuate an anoxic plume in an unconfined aquifer with a fluctuating water table. The hypothesis that water table fluctuations increase oxygen transfer from air to water, through enhanced exchange from entrapped air, is tested in an intermediate-scale, fluctuating water table experiment. A dual-energy gamma radiation system was used to measure water saturations while dissolved oxygen (DO) concentrations were measured with flow-through oxygen microelectrodes. A hysteretic multifluid simulator was used to test whether the experimentally obtained water and entrapped air saturations, as well as DO concentrations, could be predicted using the assumptions of two-phase flow and equilibrium partitioning between the gas and the aqueous phases. The experimental results show that zones with entrapped air, formed during the imbibition portions of the experiment, were instrumental in re-oxygenation of the water. The fluctuating water table system also caused significant amounts of dissolved oxygen to be transported deeper into the flow cell. The simulator was able to predict water and entrapped air saturations, as well as dissolved oxygen concentrations reasonably well.

Experimental study of crossover from capillary to viscous fingering for supercritical CO2-water displacement in a homogeneous pore network.

Carbon sequestration in saline aquifers involves displacing brine from the pore space by supercritical CO(2) (scCO(2)). The displacement process is considered unstable due to the unfavorable viscosity ratio between the invading scCO(2) and the resident brine. The mechanisms that affect scCO(2)-water displacement under reservoir conditions (41 °C, 9 MPa) were investigated in a homogeneous micromodel. A large range of injection rates, expressed as the dimensionless capillary number (Ca), was studied in two sets of experiments: discontinuous-rate injection, where the micromodel was saturated with water before each injection rate was imposed, and continuous-rate injection, where the rate was increased after quasi-steady conditions were reached for a certain rate. For the discontinuous-rate experiments, capillary fingering and viscous fingering are the dominant mechanisms for low (logCa ≤ -6.61) and high injection rates (logCa ≥ -5.21), respectively. Crossover from capillary to viscous fingering was observed for logCa = -5.91 to -5.21, resulting in a large decrease in scCO(2) saturation. The discontinuous-rate experimental results confirmed the decrease in nonwetting fluid saturation during crossover from capillary to viscous fingering predicted by numerical simulations by Lenormand et al. (J. Fluid Mech.1988, 189, 165-187). Capillary fingering was the dominant mechanism for all injection rates in the continuous-rate experiment, resulting in monotonic increase in scCO(2) saturation.

Rheological behavior of xanthan gum solution related to shear thinning fluid delivery for subsurface remediation

Xanthan gum solutions are shear thinning fluids which can be used as delivery media to improve the distribution of remedial amendments injected into heterogeneous subsurface environments. The rheological behavior of the shear thinning solution needs to be known to develop an appropriate design for field injection. In this study, the rheological properties of xanthan gum solutions were obtained under various chemical and environmental conditions relevant to delivery of remedial amendments to groundwater. Higher xanthan concentration raised the absolute solution viscosity and increased the degree of shear thinning. Addition of remedial amendments (e.g., phosphate, sodium lactate, ethyl lactate) caused the dynamic viscosity of xanthan solutions to decrease, but they maintained shear-thinning properties. Use of mono- and divalent salts (e.g., Na(+), Ca(2+)) to increase the solution ionic strength also decreased the dynamic viscosity of xanthan and the degree of shear thinning, although the effect reversed at high xanthan concentrations. A power law analysis showed that the consistency index is a linear function of the xanthan concentration. The degree of shear thinning, however, is best described using a logarithmic function. Mechanisms to describe the observed empiricism have been discussed. In the absence of sediments, xanthan solutions maintained their viscosity for months. However, the solutions lost their viscosity over a period of days to weeks when in contact with site sediment. Loss of viscosity is attributed to physical and biodegradation processes.

Correlation of Oil–Water and Air–Water Contact Angles of Diverse Silanized Surfaces and Relationship to Fluid Interfacial Tensions

The use of air-water, θ(wa), or air-liquid contact angles is customary in surface science, while oil-water contact angles, θ(ow), are of paramount importance in subsurface multiphase flow phenomena including petroleum recovery, nonaqueous phase liquid fate and transport, and geological carbon sequestration. In this paper we determine both the air-water and oil-water contact angles of silica surfaces modified with a diverse selection of silanes, using hexadecane as the oil. The silanes included alkylsilanes, alkylarylsilanes, and silanes with alkyl or aryl groups that are functionalized with heteroatoms such as N, O, and S. These silanes yielded surfaces with wettabilities from water wet to oil wet, including specific silanized surfaces functionalized with heteroatoms that yield intermediate wet surfaces. The oil-water contact angles for clean and silanized surfaces, excluding one partially fluorinated surface, correlate linearly with air-water contact angles with a slope of 1.41 (R = 0.981, n = 13). These data were used to examine a previously untested theoretical treatment relating air-water and oil-water contact angles in terms of fluid interfacial energies. Plotting the cosines of these contact angles against one another, we obtain the relationship cos θ(wa) = 0.667 cos θ(ow) + 0.384 (R = 0.981, n = 13), intercepting cos θ(ow) = -1 at -0.284, which is in excellent agreement with the linear assumption of the theory. The theoretical slope, based on the fluid interfacial tensions σ(wa), σ(ow), and σ(oa), is 0.67. We also demonstrate how silanes can be used to alter the wettability of the interior of a pore network micromodel device constructed in silicon/silica with a glass cover plate. Such micromodels are used to study multiphase flow phenomena. The contact angle of the resulting interior was determined in situ. An intermediate wet micromodel gave a contact angle in excellent agreement with that obtained on an open planar silica surface using the same silane.

Investigation of density-dependent gas advection of trichloroethylene: Experiment and a model validation exercise

An experiment was conducted to evaluate whether vapor-density effects are significant in transporting volatile organic compounds (VOCs) with high vapor pressure and molecular mass through the subsurface. Trichloroethylene (TCE) was chosen for the investigation because it is a common VOC contaminant with high vapor pressure and molecular mass. For the investigation, a 2-m-long by 1-m-high by 7.5-cm-thick flow cell was constructed with a network of sampling ports. The flow cell was packed with sand, and a water table was established near the lower boundary. Liquid TCE was placed near the upper boundary of the flow cell in a chamber from which vapors could enter and migrate through the sand. TCE concentrations in the gas phase were measured by extracting 25-μl gas samples with an air-tight syringe and analyzing them with a gas chromatograph. The evolution of the TCE gas plume in the sand was investigated by examining plots of TCE concentrations over the domain for specific times and for particular locations as a function of time. To help in this analysis, a numerical model was developed that can predict the simultaneous movements of a gas, a nonaqueous liquid and water in porous media. The model also considers interphase mass transfer by employing the phase equilibrium assumption. The model was tested with one- and two-dimensional analytical solutions of fluid flow before it was used to simulate the experiment. Comparisons between experimental data and simulation results when vapor-density effects are considered were very good. When vapor-density effects were ignored, agreement was poor. These analyses suggest that vapor-density effects should be considered and that density-driven vapor advection may be an important mechanism for moving VOCs with high vapor pressures and molecular mass through the subsurface.

Modeling fluid flow and transport in variably saturated porous media with the STOMP simulator. 1. Nonvolatile three-phase model description

Abstract STOMP is a multipurpose engineering simulator for investigating remediation technologies for the cleanup of organic compounds and associated radionuclides within the soil and groundwater in the arid and semiarid western United States. Fundamentally, the simulator numerically solves flow and transport through variably saturated geologic media using an integrated-volume finite-difference scheme to discretize the governing conservation equations. Modeling capabilities include those for multiple phases, compositional NAPL, and nonisothermal conditions using a variable configuration source code to apportion execution speed and memory requirements to the problem specifics. This paper presents a brief overview of the entire simulator and then focuses on solution techniques for solving problems involving nonvolatile three-phase systems with hysteretic, fluid entrapment, and nonhysteretic forms of the saturation-relative permeability-capillary pressure relations. The solution techniques concentrate on developed algorithms for phase-transitions, which are demonstrated for a hypothetical problem involving NAPL appearances and disappearances. Mass-balance errors for these simulations were less than 0·06%. In a companion paper, saturations computed from simulations using STOMP are compared against experimental data for two one-dimensional, nonvolatile three-phase flow systems.

Single-source gamma radiation procedures for improved calibration and measurements in porous media

When dual-energy gamma radiation systems are employed for measurements in porous media, count rates from both sources are often used to compute parameter values. However, for several applications, the count rates of just one source are insufficient. These applications include the determination of volumetric liquid content values in two-liquid systems and salt concentration values in water-saturated porous media. Single-energy gamma radiation procedures for three applications are described in this paper. Through an error analysis, single-source procedures are shown to reduce the probable error in the determinations considerably. Example calculations and simple column experiments were conducted for each application to compare the performance of the new single-source and standard dual-source methods. In all cases, the single-source methods provided more reliable data than the traditional dual-source methods. In addition, a single-source calibration procedure is proposed to determine incident count rates indirectly. This procedure, which requires packing under saturated conditions, can be used in all single- and dual-source applications and yields accurate porosity and dry bulk density values.