Multiphysics processes in partially saturated fractured rock: Experiments and models from Yucca Mountain

Abstract

The site investigations at Yucca Mountain, Nevada, have provided us with an outstanding data set, one that has significantly advanced our knowledge of multiphysics processes in partially saturated fractured geological media. Such advancement was made possible, foremost, by substantial investments in multiyear field experiments that enabled the study of thermally driven multiphysics and testing of numerical models at a large spatial scale. The development of coupled‐process models within the project have resulted in a number of new, advanced multiphysics numerical models that are today applied over a wide range of geoscientific research and geoengineering applications. Using such models, the potential impact of thermal‐hydrological‐mechanical (THM) multiphysics processes over the long‐term (e.g., 10,000 years) could be predicted and bounded with some degree of confidence. The fact that the rock mass at Yucca Mountain is intensively fractured enabled continuum models to be used, although discontinuum models were also applied and are better suited for analyzing some issues, especially those related to predictions of rockfall within open excavations. The work showed that in situ tests (rather than small‐scale laboratory experiments alone) are essential for determining appropriate input parameters for multiphysics models of fractured rocks, especially related to parameters defining how permeability might evolve under changing stress and temperature. A significant laboratory test program at Yucca Mountain also made important contributions to the field of rock mechanics, showing a unique relation between porosity and mechanical properties, a time dependency of strength that is significant for long‐term excavation stability, a decreasing rock strength with sample size using very large core experiments, and a strong temperature dependency of the thermal expansion coefficient for temperatures up to 200°C. The analysis of in situ heater experiments showed that fracture closure/opening caused by changes in normal stress across fractures was the dominant mechanism for thermally induced changes in intrinsic fracture permeability during rock mass heating/cooling and that fracture shear dilation appears to be less significant. Significant effort was devoted to predicting the long‐term stability of underground excavations under (mechanical) strength degradation and seismic loading, perhaps one of the most challenging tasks within the project. We note that such long‐term strength degradation is actually an example of a chemically mediated process governed by underlying (microscopic) stress corrosion and chemical diffusion processes. In the Yucca Mountain Project, such chemically mediated mechanical changes were considered implicitly through model calibrations against laboratory and in situ heater experiments at temperatures anticipated to be experienced by the rock. A possible future research direction would be to simulate such processes mechanistically in a complete coupled THMC framework where C denotes chemical processes.

Geomechanical effects on CO2 leakage through fault zones during large-scale underground injection

“…Faulkner and Rutter, 1998) This work extends previous geomechanical modeling using the TOUGH-FLAC simulator Rutqvist, 2011). The TOUGH-FLAC simulator has been applied and tested over a wide range of geoscientific fields and geoengineering applications, including enhanced geothermal systems (Rutqvist et al, 2013a;, hydrothermal systems and volcanology (Todesco et al, 2004), fault reactivation and hydraulic fracturing during shale gas operations (Rutqvist et al, 2013b), gas production from hydrate-bearing sediments (Rutqvist et al, 2012b), nuclear waste disposal (Rutqvist and Tsang, 2012, and references therein), compressed air enhanced systems (e.g., Rutqvist et al, 2012a), as well as carbon sequestration (e.g., Cappa and Rutqvist, 2012;.…”

Flow, Transport, and Reaction in Porous Media: Percolation Scaling, Critical‐Path Analysis, and Effective Medium Approximation

“…We do, however, point out that much progress has been made (see, for example, Darcel et al, ; de Dreuzy et al, ; Davy et al, , ; Ebigbo et al, ; Maillot et al, ; Pyrak‐Nolte & Nolte, ). Much of the progress has been reviewed by Meakin and Tartakovsky (), Adler and Thovert (), Sahimi (), Rutqvist and Tsang (), and Adler et al ().Meakin and Tartakovsky address complications to modeling flow and transport in porous media due to complex geometries, density contrasts of the fluids, and dynamic changes in contact angles. They do not, however, mention problems due to the importance of the connectivity, the chief motivation for the present review.…”

An overview of TOUGH-based geomechanics models

“…At that time TOUGH2 (Pruess et al, 2012) was the main code used for the analysis of unsaturated zone flow and transport of the Yucca Mountain site, but there was a need to analyze how flow and transport was affected by geomechanical processes (Rutqvist and Tsang, 2003b). The idea was then to link TOUGH2 to a geomechanics code; the FLAC3D code (Itasca, 2012) was selected, because it had the required geomechanics capabilities, was a continuum code compatible with the TOUGH2 continuum approach, and was already qualified and applied in the Yucca Mountain Project (Rutqvist and Tsang, 2012). programing added in a dynamic link library file (Rutqvist, 2011).…”

“…This development started with the development of the TOUGH-FLAC simulator to meet the need for analyzing the effect of geomechanics on multiphase fluid flow behavior and transport properties around nuclear waste emplacement tunnels at the previously proposed U.S. highlevel nuclear repository site at Yucca Mountain, Nevada Rutqvist and Tsang, 2012). The TOUGH-FLAC simulator was developed as a pragmatic approach, linking the two existing codes, TOUGH2 and FLAC3D .…”