No abstract

No abstract

No abstract

We consider liquid infiltrating by gravity flow into a system of parallel, regularly spaced fractures in an unsaturated porous medium. The position of the fracture liquid front as a function of time, under some simplifying assumptions, is shown to obey a nonlinear integrodifferential equation. Approximate analytic solutions are developed, showing that the movement of the liquid front exhibits three major flow periods: (1) at early time, the frontal position is determined by the fracture inlet boundary condition and the gravity‐driven flow behavior of the fracture with negligible influence by the matrix; (2) at intermediate time, matrix imbibition retards the frontal advance against the pull of gravity; (3) at late time, the matrix approaches saturation and the frontal velocity approaches a limiting value. A two‐dimensional numerical model is used to confirm the approximate solutions. Implications of the model for nuclear waste storage are discussed. The analysis is applicable not only to fractured rock but also to lateral infiltration into coarse‐grained sediments lying between layers of fine‐grained soil.

The results of the first two cycles of the seasonal aquifer thermal energy storage field experiment conducted by Auburn University near Mobile, Alabama in 1981–1982 (injection temperatures 59°C and 82°C) were predicted by numerical modeling before their conclusion with good accuracy. Subsequent comparison of experimental and calculated results provided important insight into areas of model improvement and alternative experimental designs. Key factors influencing energy recovery appear to be aquifer heterogeneity (layering) and strong buoyancy flow in the aquifer. An optimization study based on second‐cycle conditions calculated a series of scenarios, each using a different injection and production scheme, to study possible ways to improve energy recovery. The results of this optimization study were used by Auburn University in the design of the third‐cycle experiment.

The purpose of this work is to study the transport of water (both the liquid and vapor phases) in tuff from the Topopah Spring welded unit under conditions expected in the near‐field environment of a high‐level nuclear waste container. A naturally fractured sample of Topopah Spring tuff (approximately 8 cm in diameter and 10 cm long) from Yucca Mountain at the Nevada Test Site, Nevada, was studied using, as a pore fluid, natural groundwater recovered from a well in which the principal producing horizon is the Topopah Spring Member. Confining pressure, sample temperature, and pore pressure were held at values that simulate expected in situ near‐field conditions shortly after emplacement of the container. Results of this work are comparable with results from previous similar experiments on 2.54‐cm‐diameter samples of Topopah Spring tuff. During the more than 6‐month experiment duration, water permeability decreased about 3 orders of magnitude. The most rapid measured permeability change (between about 600 and 100 μdarcy) occurred when the sample temperature was increased from room temperature to 89°C. Subsequently, water permeability decreased in a fairly monotonie manner to a value of about 1 μdarcy. This behavior probably resulted from a decrease in the aperture of the natural fracture, possibly reflecting transport and redeposition of silica (which was already present in the fracture). This process also resulted in a weak bonding of the sample halves. Previous experiments on two smaller naturally fractured Topopah tuff samples resulted in fracture healing presumably by the same mechanism. The degree of healing, indicated by bonded tensile strength and change in fluid permeability, seems to be sample dependent. The distribution of electrical resistivity indicates that the sample dehydrates and rehydrates nonuniformly. This is consistent with our findings on smaller samples. Computed impedance tomographs indicate that during each of the five subsequent dehydrations of the sample, water first left the matrix adjacent to the fracture surface, then escaped the sample through the fracture aperture. Drying then progressed into the matrix around the fracture. Rehydration of the dry sample was not quite the reverse of the dehydration process. Water initially entered the fracture along the fracture edge, and the matrix was then wetted progressively from these points. However, the matrix was also wetted along paths apparently unrelated to fracture flow.

The risk associated with storage of carbon dioxide in the subsurface can be reduced by removal of a comparable volume of existing brines (e.g., Buscheck et al., 2011). In order to avoid high costs for disposal, the brines should be processed into useful forms such as fresh and low-hardness water. We have carried out a cost analysis of treatment of typical subsurface saline waters found in sedimentary basins, compared with conventional seawater desalination. We have also accounted for some cost savings by utilization of potential well-head pressures at brine production wells, which may be present in some fields due to CO 2 injection, to drive desalination using reverse osmosis. Predicted desalination costs for brines having salinities equal to seawater are about half the cost of conventional seawater desalination when we assume the energy can be obtained from excess pressure at the well head. These costs range from 32 to 40¢ per cubic meter permeate produced. Without well-head energy recovery, the costs are from 60 to 80¢ per cubic meter permeate. These costs do not include the cost of any brine production or brine reinjection wells, or pipelines to the well field, or other sitedependent factors.Keywords: carbon capture and storage, desalination, brines, reverse osmosis, osmotic pressure, produced waters 3 IntroductionA major risk associated with geologic sequestration of carbon dioxide is the buildup of pressure in the subsurface due to injection. The maximum sustainable pressure is limited by the ability of the cap-rock to contain the CO 2 as a low permeability barrier, and also the potential for the overpressure to create new fractures, or to reactivate existing fractures, which would then open new flow paths for CO 2 escape (Rutqvist et al., 2007). Moreover, emplacement of CO 2 could drive the existing brine into useful aquifers where potable water would be contaminated, and could also increase the likelihood of induced seismicity (Bachu, 2008).A method to avoid these potential problems is to withdraw the saline fluid existing in the subsurface and thus reduce the amount of overpressure. However, the outstanding problem with this approach is the disposition of the withdrawn fluids. Because of their high salinities (>10,000 mg/L TDS), these fluids are generally not useful for either domestic or agricultural use. The large volume of CO 2 to be emplaced implies a similarly large fluid volume must be produced. Such a large volume of brine cannot be disposed of as-is without significant cost.Desalinating the produced brines using membrane-based technologies is a potential solution that would allow brine withdrawal, with disposal partly accounted for by production of useful low salinity water. Moreover, some of the energy needed to drive desalination could be obtained directly from the overpressure present in the subsurface generated by the emplacement of CO 2 for some sequestration sites. This would allow a re-capture of some fraction of the energy used to pressurize and inject CO 2 into the subsurface to...

[1] We present a multiscale model that simulates coupled thermal and hydrological behavior driven by radioactive decay heat from a potential nuclear waste repository at Yucca Mountain, Nevada. We use this model to evaluate repository performance for different designs with respect to major thermal design goals (e.g., keeping waste packages dry). A locally boiling or globally boiling design uses rock dry out to create dry (low relative humidity (RH)) conditions around waste packages for a long period of time. A subboiling design eliminates boiling in the host rock (possibly reducing uncertainty) but is less effective at maintaining dry conditions. The addition of backfill increases the duration of RH reduction significantly without added boiling in the host rock but at the cost of higher waste package temperatures. The interaction between engineering design variables and natural system factors that affect thermohydrologic behavior is highly nonlinear. As a consequence, designs that differ by seemingly small details can result in markedly different thermohydrologic behavior and consequently repository performance, whereas with regards to other details, large differences may have little or no effect.