H33A-0450 1340h
A Study of Oil-Water and Air-Water Interfacial Areas Measured Using Synchrotron-based CMT
Studies of NAPL dissolution in porous media have demonstrated that the macroscale measurement of saturation alone is incapable of describing the rate of dissolution. Quantification of the NAPL-water interfacial area provides a measure of the expected area available for mass transfer between the phases and will likely be a primary determinant of NAPL removal efficiency. To measure the interfacial area, we have used a synchrotron-based CMT technique to obtain high-resolution (17 micron voxels) 3D images of flow in a Soltrol-water-glass bead system. An array of image analysis techniques is used to compute the interfacial area, excluding the presence of films, which is found to increase as the wetting phase saturation decreases, reach a maximum, and then decrease as the wetting phase saturation goes to zero, supporting the findings of a previous air-water-glass bead study. These results are also supported by pore scale numerical studies presented in the literature. An REV analysis was conducted, demonstrating that the volume used was likely at the lower limits of an acceptable REV size for porosity, saturation, and wetting-non-wetting interfacial area. The wetted fraction of the solid surface was also measured and found to be a non-hysteretic function of the saturation.
H33A-0451 1340h
Estimation of NAPL/Water Interfacial Areas in Well-Characterized Porous Media
The NAPL/water interfacial area is an important parameter which affects the rate of NAPL dissolution in porous media. We generated a set of baseline data for specific interfacial area in a well-characterised laboratory system, and subsequently used these data to evaluate current models that seek to predict this parameter. The interfacial tracer technique was used to measure specific NAPL/water interfacial areas at residual NAPL-saturation in four grades of silica sand wet-packed into a 28cm-long, 3cm-i.d. column. The two-phase system contained water and hexadecane as NAPL. The first model tested distributes entrapped NAPL over the pore classes based on Land's algorithm and assumes spherical geometry for the resulting ganglia. The other model is thermodynamically based, assuming that reversible work done on the system results in an increase in the interfacial area, such that the area between drainage and imbibition curves can be related to the interfacial area. The interfacial tracer tests gave specific interfacial areas between 57 cm$^{-1}$ for the finest sand and 16 cm$^{-1}$ for the coarsest, compared to values between 33 cm$^{-1}$ and 7 cm$^{-1}$ for the first model and between 19 cm$^{-1}$ and 5cm$^{-1}$ for the thermodynamic model. The assumption of spherical geometry made by the first model serves to minimise the specific interfacial areas of the ganglia. Computed tomography (CT) scans of similar samples to those used in the column experiments showed that the geometry of the visible blobs was generally not spherical; hence it is reasonable to suggest that this may explain the underprediction by the first model. We believe the thermodynamic model underestimates the interfacial area because it assumes that entrapment occurs only within the largest pores. We also calculated a modified version of this model assuming entrapment across all pore classes; this yielded values between 64 cm$^{-1}$ and 14 cm$^{-1}$, suggesting that this may be a more appropriate method.
H33A-0452 1340h
Characterizing air-water interfacial area in variably saturated porous media
The air-water interface plays an important role in many mass and energy transfer processes in unsaturated soils. In this research, the gas-phase partitioning tracer method was used to measure air-water interfacial area as a function of water content for several porous media. The porous media comprised a series of sands with narrow particle-size ranges, a sand with a wider particle-size distribution, a sandy soil, and a loamy soil. The measurement range was extended to very low water contents in an attempt to determine maximum air-water interfacial areas. The measured values were compared to the normalized surface areas of the porous media. Experiment results showed that the magnitude of the air-water interfacial areas approached that of the normalized surface areas. Generally, air-water interfacial areas were higher for media with higher specific surface areas. The change in air-water interfacial area with changing water content was less near saturated water contents and greater at smaller values. In addition, the change was greater for poorly-sorted media. Measured air-water interfacial areas were compared to values estimated using various models.
H33A-0453 1340h
Saturation Measurements of Immiscible Fluids in 2-D Static Systems: Validation by Light Transmission Visualization
This study is a part of an ongoing research project that aims at assessing the environmental benefits of DNAPL removal. The laboratory part of the research project is to examine the functional relationship between DNAPL architecture, mass removal and contaminant mass flux in 2-D models under well-defined conditions. For this, the characterization of the DNAPL in the model system is carried out using light transmission visualization (LTV). In this technique, images of the entire 2-D chamber are captured using a charge-couple device (CCD) and analyzed pixel by pixel using image analyses software. Water, air and DNAPL saturation across the 2-D model are measured as variations in the transmitted light intensity of these pixels. The focus of this study is the experimental validation of LTV using 2-D flow chamber constructed of two glass plates separated by 1.4 cm. The flow chamber has an equal width and height of 15.24 cm. Naturally translucent silica sands that have sieve sizes of 20/30 and 40/50 are used as packing media. LTV is used to measure the water, air and DNAPL (modeled by PCE) saturation in a 2-D static system. Variable saturation of these fluids across the height of the model in a water-air system and water-DNAPL system is dictated by controlling the capillary pressure of the fluids. Saturation-Capillary Pressure curve results of LTV technique are compared with experimental results of the saturation-capillary pressure curve obtained by Tempe cell measurements in our laboratory. LTV results are also compared to theoretical predictions of saturations for these three fluids.
H33A-0454 1340h
X-ray CMT Study Shows that Drainage Boundary Condition is Responsible for Dynamic Effect in Sandy Material
To address some unanswered questions regarding the cause of observed `dynamic' flow phenomena in porous media, minute pore-scale multi-phase flow experiments were developed to non-destructively visualize the flow process in a sample of porous material. The flow experiments were performed concurrently with collection of high resolution tomographic images. The samples were exposed to similar pressure boundary conditions as in a previous experiment; one-step and multi-step drainage experiments where the water was drained from the sample using either one large or several small pressure steps, respectively. Qualitative examination of the obtained microtomographic images shows that the difference in drainage pattern, and resulting residual water phase saturation, for the two drainage conditions is quite dramatic. For the high initial flow rates, drainage tends to take place in the most well-connected and largest pores first, meaning that subsequent drainage of the finer pores is unlikely because they are no longer hydraulically connected to the main flowing continuum. To support this observation we also performed quantitative analyses on the microtomographic data. A nearest neighbor analysis was performed on the air phase distribution found in the images and showed characteristic differences between fast and slow drainage. Under fast drainage the air bubbles tend to be few and far apart, whereas a larger number of closely spaced and smaller bubbles are obtained under slow flow conditions.
H33A-0455 1340h
Quantification of three-phase fluid distribution in a porous medium by X-ray transmission measurements with energy dispersion X-ray spectrometry
A new X-ray transmission measurement with dispersion X-ray spectrometry was developed to quantify three-phase fluid distribution in a porous medium. Here, distilled water doped with 5.0 % barium chloride and oil doped with 7.0 % iodoheptane were used for the experiments. We measured the X-ray transmission intensities through a porous medium with energy dispersion X-ray spectrometry and quantified the spatial distribution of each phase (water, oil, and air) fluid in the medium by using intensities at two appropriate frequency bands. The used bands for the analysis were several kilo-electron volts higher from the absorption edges of barium and iodine, respectively. We first measured the amount of each fluid with known fluid contents and confirmed the applicability of this technique. Then, the technique was applied to quantify the spatially distributed three-phase fluid contents. As a result, we were able to quantify three-phase fluid distribution in the porous medium with sufficient accuracy. However, the measurement time was rather long for applying this method to capture the fluid movement in a porous medium. In this presentation, we suggest the possibility to quantify temporal change of three-dimensional distribution of three-phase fluids in a porous medium by applying energy dispersion X-ray spectrometry to the traditional CT technique.
H33A-0456 1340h
Use of 3D X-ray Microtomography to Observe the Structure of Colloidal Zirconia Deposits in Sand Columns
We are utilizing the tomography capability of the DuPont-Northwestern-Dow Collaborative Access Team (DND-CAT) at the Advanced Photon Source (APS), Argonne National Laboratory for studies of in situ sediment structure. Images of a sediment sample are taken at a number of different angles as the incident x-ray beam passes through it, and a three-dimensional view of the interior of the sample is then reconstructed from these maps using computed tomography. These images allow examination of individual sediment grains and the pore structure with a spatial resolution of as little as a few microns. In addition, the distribution of a particular element can be determined by difference tomography, i.e., by obtaining a series of images both above and below the x-ray absorption edge of the element of interest and then subtracting the resulting tomographic reconstructions. We used this approach to resolve the distribution of deposited zirconia particles in a matrix of silica sand. Zirconia particles were deposited in small sand columns under steady upflow conditions, and the columns were rinsed by several pore volumes of particle-free water. The saturated columns were then sealed, removed from the upflow system, and transported to the APS for analysis. Column experiments were conducted with a range of influent zirconia concentrations and background solution conditions. We will present three-dimensional tomographic reconstructions showing the structure of both the porous medium and the zirconia particle deposits, along with results on the bulk deposition of zirconia.
H33A-0457 1340h
Pore-level Modeling of Drainage: Crossover from Capillary Fingering to Compact Invasion
We have developed a two-dimensional, pore-level model of drainage that incorporates viscous, capillary, and gravitational effects and validated it quantitatively, in the very different limits of zero viscosity ratio and zero capillary number. For this work, two-phase displacement is simulated for a range of stable viscosity ratios (injected fluid more viscous than displaced fluid). The capillary number was varied from very low values (near the invasion percolation limit) to much larger values and the resulting infiltration was studied. The way in which the flows deviate from capillary fingering at small times and eventually become compact was modeled. Results exhibiting this crossover from capillary fingering to compact invasion are presented for the average position of the injected fluid, the fluid-fluid interface, the saturation and fractional flow profiles, and the relative permeabilities. A comparison of these results and earlier theoretical predictions will demonstrate the validity of these general theoretical arguments and are shown to be independent of the details of the porous media in both two and three dimensions.
H33A-0458 1340h
Modeling of Single-Phase Flow in One-Dimensional Poroelastic Media
A nonlinear model for low Reynolds number, single-phase fluid flow in slightly compressible porous media is presented and solved approximately. The model assumes state equations for density, porosity, viscosity and permeability that are exponential functions of the fluid (either gas or liquid) pressure. We show that the governing equation can then be transformed into a nonlinear diffusion equation with a type of power-law diffusivity. It is solved for a semi-infinite domain for either constant pressure or constant flux boundary conditions at the surface. For the particular case where the permeability varies linearly with porosity, the transformation yields a linear governing equation and exact solutions are easily obtained. For other cases, analytical approximations are derived based on existing unsaturated flow analyses. Because the diffusivity function is much more linear than for the unsaturated flow case, a modified theory that is exact in the linear limit is developed. The solutions obtained, although approximate, are extremely accurate as demonstrated by comparisons with numerical results. Existing experimental data are analyzed using the new theory. Model predictions for the surface pressure resulting from constant-flux injection into a porous medium are shown to compare well with the data.