Search

Geomechanical reservoir modelling is one of the key tools used to predict reservoirdeformation in response to exploitation of the subsurface, whether for hydrocarbon orgeothermal energy production, or for the injection of CO2 and waste water. However, to makesuch predictions, a better understanding of the reservoir state of stress is necessary. Pressureleak-off tests and wellbore break-outs are generally used to assess the in-situ stateof stress of a reservoir. These methods require the creation of fractures near thewellbore, which may cause damage to the wellbore and potentially cause drillingproblems. We are investigation the potential to use core material to assess the in-situ stress state byperforming true triaxial experiments on reservoir rock, coupled to measuring P- and S-wavevelocities and microstructural analysis. We performed experiments on outcrop material fromthe Flechtingen sandstone (Beberthal quarry, Germany; ∼6% porosity) and the Bleurswillersandstone (Bleurville, France; ∼25% porosity). True triaxial experiments were performed atthe Christian-Albrechts University of Kiel, under dry and room temperature conditions.During the experiments, acoustic P- and S-wave travel times will be measured in the threeprinciple stress directions (σ1, σ2 and σ3). Microstructural analysis using optical microscopywill be coupled to the experimental results to quantify the microcrack densities ofthe samples prior to- and after deformation and to look at the orientations of themicrocracks. The experiments consisted of two stages each: 1) differential stress stage to inducedamage in the sandstones, and 2) hydrostatic stress stage to assess if the stresses at which thedamage was induced can be inferred from changes in wave-velocity and anisotropy. Theseexperiments aim to provide a proof of concept for using true triaxial tests to assess in-situstress states of reservoir rocks. [ABSTRACT FROM AUTHOR]

Joints, fractures, and fault zones are ubiquitous on all scales in the brittle upper crust. The presence of these features significantly alters the effective mechanical and hydraulic behaviour relative to intact rocks. In hydraulic fracturing for example, this relation between physical properties and damage structure is exploited and high fluid pressures are used to create new fractures in order to enhance a rock formation's hydraulic properties. Increasing the fluid pressure in the subsurface may also lead to shear failure of pre-existing fractures which might promote further enhancement of hydraulic properties. Our goal is to further investigate the relation between shearing and modification in hydraulic properties. We used a method that allows for the continuous measurement of hydraulic properties of cylindrical sandstone samples cut at an angle of 35° from the cylinder axis to model a fracture. A central borehole drilled through the upper sample half provides a hydraulic connection to the fracture. To determine hydraulic properties, we apply a periodic pore pressure variation. The phase shift and amplitude ratio between the fluid flow and the pressure oscillation are representative of the effective hydraulic properties from a comparison to known analytical solutions. The samples containing the artificial fractures were prepared from two adjacent sandstone formations from the triassic Buntsandstein Unit of Central Germany with high and low matrix permeabilities, respectively. The large permeability contrast between the samples enabled us to study fracture flow with and without matrix leak-off. Our results for hydrostatic conditions match previous findings that normal stress on the fracture surface controls hydraulic properties. By changing the frequency of the pressure oscillation we identified a pronounced frequency dependence of the resulting effective hydraulic properties. We suggest that tests at higher frequencies yield properties of highly conductive features (i.e. the fracture) and the influence of less conductive features (i.e. the surrounding rock) increases with decreasing frequencies. Depending on frequency we therefore probe different subsets of the sample. For both host rocks we found that shear displacement and associated surface processes reduce effective hydraulic properties compared to hydrostatic measurements at similar stress conditions. [ABSTRACT FROM AUTHOR]