Laboratory observations of fault transmissibility alteration in carbonate rock during direct shearing

Abstract

We present the results of hydromechanical changes across and along the evolving rupture surface of carbonate rocks during direct shear experiments. Direct shear experiments were performed on a laminated travertine of continental, microbial origin with calcite content of 99 wt%, chosen as a lithological analogue for Aptian presalt oil reservoir rocks found in South Atlantic presalt basins. Medical X‐ray CT images show that the porosity (~9–13%) is mainly composed of subplanar pores and vugs. Permeability is high along the laminations (~50–200 mD), controlled by interconnected pores and fractures, and extremely low across the laminations (≪1 mD). Six intact samples of travertine were sheared across the bedding direction to short displacement (20 mm), and a further three samples were sheared to a much longer displacement (120 mm). A constant effective vertical stresses of either 35, 40 or 45 MPa was applied throughout the tests. Fluid flow response across and along the fault zone was monitored continuously during both shear deformation (dynamic transmissibility) and hold periods (static transmissibility), while keeping a constant pore pressure throughout the measurement. While the samples show some microstructural variability, the set of samples sheared to 20‐mm displacement followed the same early kinematics and show very similar mechanical response to that seen in the same period of the samples where shear was continued on to 120‐mm displacement. After 20‐mm displacement, the microstructures are dominated by fractures in a zone 10–20 mm wide resembling a typical fault damage zone, with only thin and patchy development of gouge on the principal displacement plane. In all of the samples sheared to a high displacement of 120 mm, a continuous layer of gouge (6.2–13 mm thick) was formed, defining a distinct core of the shear zone. The dynamic transmissibility across the fault decreases progressively in all the sheared samples regardless of the applied effective stress. In general, the volume decreases as well, though at a tapering rate, throughout shear. A small vertical dilation occurred near the peak strength for travertine sheared under 35 MPa vertical stress, but this was not accompanied by an increase in transmissibility. From the results, we conclude that once the gouge material is developed in the core of the fault zone, the dynamic transmissibility across the fault is permanently decreased.

Fault reactivation in travertine and its impact on hydraulic transmissibility: laboratory experiments and mesoscale structures

“…The thickness of the cataclastic gouge layers in fault zones has previously been reported to increase with displacement (Engelder 1974;Robertson 1982;Scholz 1987;Hull 1988;Michie et al 2014). Delle Piane et al (2016) and Giwelli et al (2016) illustrated that experimentally produced faults in travertine with small shear displacement (c. 20 mm) do not display a fault core, as seen in Figure 10c, but do contain a dense fracture network. The porosity and permeability reduction that occurs during the first 20 mm of shear is mostly explained by localized compaction and asperity shearing, which can therefore create a narrow, but relatively efficient, barrier to fluid flow.…”

“…We propose that reactivation of faults with reversed shear sense can be an important control on fault-core architecture and hydrogeology due to accumulated deformation events. Giwelli et al (2016). A colour chart is added to the X-ray CT image to highlight density contrasts, where red shows low-density zones, pale blue shows high-density zones and black shows pores.…”

“…Fluid flow during the experiments was measured from the base/bottom to top compartments with a constant pore pressure of 0.1 MPa throughout the test. For convenience in describing the experiments and following the geographical orientation of the laboratory, both the configuration of the shear cell and samples used in the experiment were divided into four compartments (top east (TE) and top west (TW), and bottom east (BE) and bottom west (BW): Giger et al 2011;Giwelli et al 2016). …”

“…Prior to the direct shear tests, the samples were saturated by immersion in tapwater under vacuum for 2 days. More details on the experimental procedures and sample preparation can be found in Giwelli et al (2016). A shear displacement rate of 0.17 mm min -1 was used initially, which was necessary to capture flow behaviour during the peak and immediate post-peak stress in detail.…”

“…Figure 10 gives an example of a longitudinal X-ray CT image (computed from the acquired transversal X-ray CT images) of B2T2N40D12 and B2T3N45D12. A sample that sheared to 20 mm of displacement and under a constant effective stress of 45 MPa ( presented in Giwelli et al 2016) is shown in Figure 10c for comparison. Note that the longitudinal X-ray images of the sheared samples show a longitudinal slice taken close to the drill hole parallel to the sample margins.…”