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

PANEL REPORT: MODELING AND SIMULATION OF GEOLOGIC SYSTEMSmodels as the foundations of performance assessment. In addition to the challenges pertaining tomultiscale modeling of coupled processes, difficult issues arise when attempting to build detailedmodels for specific sites that are based on a diverse set of hard and soft data. These issuesinclude:1. Determining the optimal level of detail in geological-scale modeling2. Incorporating large and diverse data sets into models3. Quantifying model predictive uncertainty with consideration of data errors, modelapproximations, scale effects, nonlinearities, and coupled processesModeling multiple processes at interacting scalesThe present level of understanding and modeling of complex properties and processes in thesubsurface is often insufficient for predicting potential responses to anthropogenic perturbations.A fundamental gap in the knowledge base is how to accurately couple information across scales,i.e., accurate accounting of small-scale effects on larger scales, and of large-scale forcings onsmall-scale processes. For example, in a fracture, chemical interactions between pore fluids andthe fracture walls occur locally on the submicron scale but affect fluid flow and wavepropagation along the length of the fracture, on a typical length scale of meters or more. For aporous medium that is partially saturated with gas and water, the distribution of these two fluidphases is affected not only by the intrinsic length scales of the pores but by length scalesassociated with fluid-fluid and fluid-solid interfaces, as well as by time-dependent processes thatalter fluid pressures (e.g., pumping, fault displacement, reactive events). In both of theseexamples, the problem is complicated by the fact that there are no obvious scale separations.This leads to the demanding challenge of determining not only the appropriate mathematical andnumerical representations at each scale but more importantly the interactions across scales.In addition to complex processes, the properties of natural geologic formations rarely displayuniformity or smoothness. Instead, they usually show significant variability and complexpatterns of correlation across a wide spectrum of interacting scales. For example, mechanicaldiscontinuities range in scale from a few microns (microcracks) to centimeters-meters (fractures,joints) to kilometers (faults). They can occur singly or in sets producing heterogeneity from thepore scale to the reservoir scale that varies temporally because of participation in hydrogeologic,geochemical, biogeochemical and tectonic cycles, in addition to any human activities orinterference. Geologic characterization models are usually constructed using sparse data, whichcome from different sources with varying quantity and information content. We are faced withthe scientific challenge of developing stochastic mathematical and numerical frameworks thatcapture the nonlinear dynamic (physical and chemical) processes in geologic formations. Anyprobabilistic description must account for uncertainty in models, inaccessible length scales,scale-dependent statistical description of physical phenomena (e.g., diffusion, mechanicaldispersion), and measurement errors.Subsurface systems are not static, but evolve over time because of natural processes or humaninfluences that lead to nonlinear processes with complicated feedback mechanisms. For example,the evolution of transport characteristics of fractures is controlled by competition betweenchemical and mechanical effects. These either generate (e.g., dilatant shear, microcracking,thermal cracking, focused dissolution) or destroy (shear and hydrostatic compaction, fractureBasic Research Needs for Geosciences: Facilitating 21 st Century Energy Systems 51