Pore Water at High Pressure

Scholz et al. [1993] reviewed the literature on new faults, and concluded that they are self-similar, implying that the strain-weakening that created the fault is followed by further slip-weakening. However, since major slip-weakening is not seen in the laboratory, local anomalous pore pressure is a more popular explanation for fault weakness. Such pressures would have to approach lithostatic pressure, which is the weight/area of the rock overburden.

Saline hot springs in the California Coast Ranges expell ancient fluids from Cretaceous shales (or deeper sources) and imply some degree of anomalous pore pressure [ Unruh et al., 1992]. Fluid inclusions in the exhumed footwall of the Dixie Valley fault in Nevada record essentially lithostatic pore pressure at (305C, 1.5710 Pa) [ Parry et al., 1991]. Internal structures of the San Gabriel and Punchbowl faults, exhumed from 2-5 km, show up to 50% deformed hydrothermal vein material in the central ultracataclasite zones, which are only 1-10 m thick [ Chester et al., 1993]. The presence of veins becomes significant if one accepts that these faults probably slipped at low shear stresses; with small shear stresses, pore pressure that is equal to the least-compressive principal stress (to open a crack) cannot be very much less than lithostatic.

The source of high pore pressures is less clear. Byerlee [1993] proposed that interseismic compaction of fault gouges creates these high pressures, and that some earthquake precursors are due to fluid flow when barriers are breached between separate reservoirs. Rice [1992] presented a different model, in which water from the mantle or lower crust rises preferentially along faults, creating haloes of high pore pressure, due to pressure-dependent permeability. (The self-sealing of unidirectional hydrothermal systems due to silica precipitation may also be important.)

If high pore pressure is localized only along faults, it should cause local rotation of the principal stress axes, and of secondary shear surfaces; in fact, this is observed along the San Andreas fault [ Byerlee, 1992].