The mechanical strength of porous rocks is thought to be controlled by two deformational modes, that is frictional pressure-dependent (brittle) deformation, and thermally activated creep. This view of rock deformation leads to the overly famous Brace and Kohlstedt strength concept [1], where the envelope in a differential stress versus depth diagram exemplifies the maximum strength of a rock under a given, though steady, state of loading (constant strain rate approximation). This definition is based on dissipative fluid mechanics of rock deformation, where the transition from a brittle-like behaviour to a ductile-like behaviour coincides with maximum values in energy/entropy dissipation [2]. Following this view, it is the efficiency of ductile creep to accommodate the stored deformation that provides the first-order control on the distribution of maximum dissipation, thus resulting in a sharp brittle-to-ductile transition (BDT hereafter). The assumption at play is that stored elastic energy can be effectively dissipated by viscous creep below the BDT, thus limiting earthquake occurrence to the brittle realm. Experimental based evidence now exists that seems to contradict this common picture of rock strength as derived on the base of the YSE concept, that is, that of strong layering as dictated by a sharp brittle-ductile transition, especially under conditions spanning the transitional range between macroscopic faulting and viscous behavior. This has important socio-economic implications in that, mapping in time and space the strength evolution of rocks would provide us with a conservative estimate on the depth distribution of seismicity within a plate. In addition, we have become aware that dynamic changes in terms of forcing conditions (from natural, tectonic driven to anthropogenic ones) also influence the strength profile of specific rock types thereby the open problem on how to impose self-consistent bounds to the amount of energy which could be released in a seismic or aseismic way. To estimate this amount of energy, attention has been given in physical description of the frictional behaviour of fault zones. Several studies have reported lower values of the friction coefficient in existing fault zones as those predicted by Byerlee’s law. In addition, the evolution of the friction coefficient after onset of faulting or reactivation of existing faults requires to also account for microstructural effects but also for the presence of fluid. In this contribution I aim at discussing the multi-physics coupling characterising the evolution and stability of localised deformation and therefore in a broader sense of faulting mechanics, with particular emphasis on reservoir applications. I will consider two main discussion points: (i) controlling factors on microstructure evolution and their feedbacks onto the macroscopic deformation and frictional behaviour (rock damage), and (ii) the additional role of a non-inert (geomechanical sensu) fluid phase and its impact, mediated by porosity evolution, on the strength of porous rocks and on their deformation modes. The message to convene is that a proper understanding of these two, interacting and non-linear aspects might allow us to gain fundamental insights into the dynamics of localised deformation and therefore onto the strength of rocks and their hydraulic behaviours under a variable stress state, essential co-players in any successful reservoir operation. To assist the discussion I will also rely on dedicated examples ranging from laboratory to reservoir scales.

References
[1] Brace and Kohlstedt, Limits on Lithospheric Stress Imposed by Laboratory Experiments, Journal of Geophysical Research 85 (1980), 6248–6252.
[2] Regenauer-Lieb and Yuen, Multiscale Brittle-Ductile Coupling and Genesis of Slow Earthquakes, Pure appl. geophys. 165 (2008), 523–543.