The evolution of crack transmissivity under normal and shear stress before and after slip

Understanding the flow of fluids in the subsurface is of vital importance to geo-energy exploitation and disposal of waste fluids. In most situations, individual cracks can be more effective for fluid transport than fluids flowing through the porous matrix of the rock. The enhancement of a single crack in a low permeable rock can be over 1000 times. However, for both porous matrix flow and crack flow, the bulk permeability and crack transmissivity are all affected by the stress state in the lithosphere.

This study aims to investigate the influence of the Terzaghi effective normal and shear stress on the transmissivity of cleavage cracks under upper crustal conditions. Penrhyn slate was selected as samples for the experimental study because of its low porosity (<1%), permeability and slaty cleavage. The matrix permeability of Penrhyn slate is very low, ranging from 10^-20 to 10^-22 m² when the effective pressure is in the range of 10 to 53 MPa measured by the oscillating pore pressure method (OPPM). The crack transmissivity ranged from 10^-18 to 10^-23 m³ when the effective pressure changed from 10 to 280 MPa. The experimental results show that the evolution of crack transmissivity of a single fracture under several cycles of pressurization and depressurization is similar to the trend found in permeability. The first application of the normal stress on fracture surfaces always produces a nonrecoverable loss in crack transmissivity. In the subsequent pressurization and depressurization, partially recoverable variations in transmissivity were observed, suggesting a linear elastic behaviour of crack closure. The highest peak effective pressure attained in the stress history affects the extent of subsequent recoverable crack transmissivity. When the fracture surface is subjected to a new higher peak stress, the crack transmissivity will no longer recover to its former low level but to a lower level, indicating a permanent transmissivity loss. Thus, the transmissivity has a memory of the previous maximum stress the fractured rock was subjected to.

The influence of shear stress on crack transmissivity was studied in Solnhofen limestone and Carrara marble samples with saw-cut ground smooth fractures and compared with the rough cleaved fractures of Penrhyn slate. The influence of shear stress was studied in two situations: (a) the stable (no-slip) condition at shear stress less than needed to promote slip on fracture (b) at shear stresses high enough to yield slip on the fractures. In situation (a), the cyclic increase and decrease of shear stress led to a continuous decrease in crack transmissivity. The magnitudes of this decrease in crack transmissivity decrease with more cycling. The transmissivity tended to decrease to a lowest value eventually but this lowest value can be regenerated by slip on the fracture. In situation (b), a single slip can decrease crack transmissivity. The decrease in crack transmissivity can be attributed to the formation and smearing of frictional wear products or gouges. Under the progressive compaction, there exists a lowest level of crack transmissivity which is independent of the normal stress.