We present theory and experiments on the relation between pressurisation rate and fluid viscosity and the fracture initiation pressure and fracture geometry. We observed a strong increase in initiation pressure with pressurisation rate. The theory explains qualitatively the observed transition from transverse fractures to axial fractures with increasing pressurisation rate and minimum stress along the wellbore.
Understanding hydraulic fracture initiation is important both for optimising the communication of fractures with the borehole in well stimulation and for stress measurements. The geometry of the induced fracture and the initiation pressure must be predicted for practical applications. Earlier experimental work by Haimson and Fairhurst (1967) showed the effect of pressurisation rate for small cylindrical samples. Cleary (1979) showed later that fluid penetration could indeed have such an effect. The objective of this paper is to describe the mechanism of fracture initiation from an open hole and to predict the fracture orientation with respect to the well. We provide a theoretical description for the initiation of a hydraulic fracture and compare this to the results of the experiments. In particular, we concentrate on the influence of pressurisation rate and 22º fluid viscosity on the geometry and initiation pressure. This is a new aspect in the description of fracture initiation.
We choose as extremes for the fracture geometry a transverse and an axial fracture, shown in Figure 1. We use Linear Elastic Fracture Mechanics to predict the destabilisation of an initial micro-fracture. The value of ? 4 depends on the pressure profile within the fracture. It approaches 1 for a uniform pressure in the fracture and zero for zero pressure in the fracture. For an accurate description of the process, the pressure distribution in the fracture has to be coupled with the fluid flow.
The model for fracture initiation that we present here follows the work of Cleary (1979) and Narendran (1986).
To investigate the fracture geometry as a function of stress, pressurisation rate and fluid viscosity, we have performed laboratory model tests. Scaling laws and experimental set-up are discussed in previous papers (de Pater et al., 1992). To obtain pressures in the laboratory in the order of magnitude of the pressures in the field, the viscosity of the fluid has to be scaled up. We used cement blocks of 0.3 m on a side m which a borehole was cast with a diameter of 0.02 m. In the middle there was an open section of 0. l m that was in direct contact with the fluid. The borehole was sealed on both sides with steel packers that were glued to the cement with epoxy. The maximum stress was 23 MPa and the minimum stress was 9.7 MPa in all tests. The intermediate stress was 12.1 for a low stress contrast and 19.4 MPa for a high stress contrast. We used several stress regimes with the minimum stress along the borehole and perpendicular to the borehole.
