Basic Research Needs for Geosciences - Energetics Meetings and ...

PANEL REPORT: MODELING AND SIMULATION OF GEOLOGIC SYSTEMShealing, dislocation creep, pressure solution, free-face diffusion) fracture porosity. Crucial in thisunderstanding of chemical effects are the mechanisms in which mechanical deformationcontributes to changes in permeability, and by which, in turn, the mechanical response ismodified. These effects are known to be important at relatively modest stresses andtemperatures, and are typical in systems pushed far from chemical equilibrium, such as ingeothermal or hydrocarbon reservoirs, or around waste repositories. A key challenge is theaccurate representation of multiple processes that vary in strength and exhibit feedback that maybe self-enhancing or self-limiting. Nonlinear processes cannot be modeled with simple scaleaverageddescriptions, but require the resolution of nonlinear interactions spanning a multiplicityof scales. It also requires an understanding of the physics of systems pushed far fromequilibrium. New linear and nonlinear solvers are needed to enable accurate simulation ofevolving geologic systems on petascale computing facilities.Coupling information across scales requires the appropriate governing equations for given spatialand temporal scales, as well as knowing how to appropriately transfer information across scales.For many subsurface applications, the model must adapt to use real-time information to updatepredictions, inform field-scale deployment of monitoring methods, and be able to quantitativelyassess the results. Because of the range of scales, data integration will require understandingmeasurements from the laboratory scale to the field scale. On the laboratory scale, experimentsneed to be designed to access the data required under controlled conditions (e.g., pressures,temperature, material properties, fluids) over known scales (e.g., specific pore structures,molecular scales), and over time to determine how the structure/system/processes evolve. Forexample, the principal challenge of upscaling techniques for multiphase fluid dynamics in porousmedia is to determine which properties on the microscale can be used to predict macroscopicflow and the spatial distribution of fluid phases at core and field scales.First-principles theoretical formulations over the past decade have been derived from volumeaveraging theorems in which microscopic interfacial behavior is explicitly incorporated. Thesetheories have proposed that interfacial area per volume directly affects macroscopic behavior,and that this variable may govern the observed hysteresis in the capillary pressure-saturationrelationship. Direct visualization of interfaces with high resolution is only possible on thelaboratory scale. Transparent two-dimensional porous media (Cheng et al. 2004) with aresolution of 0.6 microns per pixel edge length and total sample length of 600 microns have beenused to visualize and quantify the behavior of interfacial geometry, while concurrentlymeasuring fluid pressures (globally and locally) to test the volume averaging theorem. Anyupscaled models must be calibrated to ensure that fine-scale influences are properly captured.Thus, laboratory techniques can provide a basis for validating theorems and calibratingnumerical methods for complex media over a range of length scales.While the laboratory can describe physical phenomena over two to three orders of magnitude inlength, testing of coupling across multiple scales will require mesoscale data and field-scale dataintegration. Structural information on the field scale is often difficult to obtain, and tends to bespatially sparse and temporally infrequent. A major scientific challenge is the development ofmodels that can examine the state of a system when initial conditions are unknown, but timelapseddata are available (e.g., time-lapsed seismic monitoring, well head pressures, strain fromtilt meters). This requires developing feedback between model formulation and data acquisition.Modeling will inform the field characterization and monitoring process on the scale of data52 Basic Research Needs for Geosciences: Facilitating 21 st Century Energy Systems