Hydraulic fracturing is the most dominant method of in-situ stress measurement at great depths. It is based on the fundamental concept that the extension fluid pressure is relatively constant inside the hydraulic fracture and slightly greater than the least in-situ principal stress. The various methods of in-situ stress measurement are all based on this same principle. These include measurement of fracture re-opening pressure and various techniques of shut-in pressure analysis.
Experiment data from actual field measurements indicate that, contrary to general belief, the fluid pressure drops rapidly and significantly along the fracture length. The cause of the pressure drop is the rapid tapering of the fracture width along its length that consequently creates a large frictional pressure drop during fluid flow, as well as roughnesses on fracture faces due to formation inhomogeneieties. This large pressure drop, coupled with formation physical and mechanical anisotropies, promotes creation of numerous secondary branches along the tip of the fracture, as well as shear failure of the formation along existing planes of weakness. The net effect is a very complex network of tensile and shear fractures that extend randomly around and inside the fractured domain, and further complicate the fluid flow inside the fracture. After the fracturing treatment, fracture closure occurs slowly, and incompletely (hysteresis), often without a distinguishing signature on pressure fall-off data.
The paper concludes that the least in-situ principal stress as measured by hydraulic fracturing is always higher than its actual value, with the difference being highly dependant on the formation mechanical and physical anisotropy, as well as the manner by which the experiments are performed. This has a domino effect on the other principal stress measurements based on fracture breakdown pressure. It proposes replacing the constant pressure assumption with a parabolic distribution. This would mean that the least in-situ principal stress is equal to 2/3 of the fracture re-opening pressure. The paper also discusses the conditions for the applicability of shut-in data for stress measurement.
The appeal of hydraulic fracturing for in-situ stress measurements comes from its operational simplicity and ease of interpretation. As first introduced and developed by Kehle1, Fairhurst2, Haimson and Fairhurs3, von Shoenfeldt4 and Roegiers5, the technique consists of creating a hydraulic fracture inside a borehole by injecting a fluid inside it. The plane of the fracture is shown to be perpendicular to the direction of the least compressive in-situ principal stress, s1 (s3> s2 > s1). The shut-in pressure provides the magnitude of s1, while the breakdown pressure is related to the formation mechanical properties, its tensile stress and the second principal stress which lies in the plane perpendicular to the wellbore. The literature contains a vast collection of papers by many authors who have since studied, analyzed and contributed to the details of testing as well as the interpretation of the fracturing data.
As the use of the technique became more widespread within the industry, more refined processes were developed to address various abnormalities observed during its use. These abnormalities included the following.
