Measuring stress in the subsurface is of prime interest in all areas of geomechanics, with applications ranging from earthquake physics to seal integrity for underground storage or waste disposal to wellbore stability and hydraulic fracturing in the geothermal or oil and gas industry. One principal stress component is usually assumed to be vertical and its magnitude is inferred by integrating density logs. The orientation of the principal horizontal stress directions can often be inferred from drilling induced borehole-wall features such as breakouts. The minimum horizontal stress magnitude can be measured by means of mini- or micro-hydraulic fracturing tests. Measuring the maximum horizontal stress, however, has always proved challenging. Here, we revisit the sleeve fracturing technique to infer both the minimum and maximum horizontal stresses. This technique consists of using the pressure exerted by a packer onto a wellbore to induce, propagate, and reopen multiple pairs of tensile fractures around the wellbore. Measurement of the orientation of the induced fractures and their opening or closing pressures provides the required data to invert for the orientation and magnitude of both horizontal stresses. We show laboratory-scale results from sleeve fracturing experiments subject to low stresses. We use a packer in conjunction with analog materials subject to anisotropic 2D stresses to create two pairs of fractures. We optically monitor fracture initiation, propagation, closure, and reopening together with the evolution of pressure in the packer during multiple inflation deflation cycles. We then use a Markov Chain Monte Carlo inversion approach to invert for the boundary stresses. We observe that the Kirsch solution is appropriate under low anisotropic stresses while using the fracture closing pressure. However, the fracture opening pressures tend to overestimates the horizontal stress.

1 Introduction

At reservoirs depth, the overburden stress is often taken as one of the principal stress components. It is referred to as the vertical stress (σσ) and, in general, inferred by integrating density logs [e.g. Zoback, 2010]. The two other principal stress components are horizontal and referred to as the minimum and maximum horizontal stresses. The orientation of the principal horizontal stress directions can be obtained by analyzing drilling-induced features such as breakouts on borehole-wall images or by using closure pressure in an extended leak-off test, or from analysis of mud losses [e.g. Zoback, 2010]. In general, the magnitude of the minimum horizontal stress, σh, can be measured by micro-hydraulic fracturing, and its azimuthal direction can be obtained by analyzing borehole-wall images [Thiercelin et al., 1996; Desroches et al., 2005]. However, measuring the maximum horizontal stress, σH, is very challenging [Amadei and Stephansson, 1997]. Some methods have focused on inferring σh and σH from the breakdown pressure. This approach has proven difficult, since it requires additional insight into the fracturing process [Ljunggren et al., 2003; Sano et al., 2005; Zoback, 2010]. Stress relief methods, such as under- or over-coring measure the strain when a rock sample is removed from the rock mass [Sjöberg et al., 2003]. These methods require a priori knowledge of the elastic properties of the rock mass and a lengthy process for the strain measurement resulting in stress estimations imprecision of 10% – 20% [Amadei and Stephansson, 1997; Fairhurst, 2003]. Other efforts focused on estimating the maximum horizontal stress are based on the acousto-elastic effects in rocks, which relate changes in the waves velocity to the stressing conditions of the rock mass. This approach relies on the three shear moduli away from the near-wellbore along with a mechanical Earth model and an estimate of σh [Sinha et al., 2008]. Here, we propose to revive the sleeve fracturing technique to directly infer both σh and σH. This approach consists of inflating a packer in a borehole to induce two (or more) pairs of fractures [Stephansson, 1983; Ljunggren and Stephansson, 1986; Sano et al., 2005]. By focusing only on the re-opening and closing pressures of the induced fractures, we avoid the difficulties related to the fracturing process [Serata et al., 1992; Ito et al., 2001]. Once the fractures are created, the packer undergoes cycles of inflation and deflation to re-open and close the fractures. Closely monitoring the location of the fractures (θi) and the pressure (Pi) in the packer allows us to invert for both magnitudes and directions of σh and σH (Fig. 1).

This content is only available via PDF.
You can access this article if you purchase or spend a download.