Abstract
Recently, fracture mapping has contributed a vast amount of data on hydraulic fracture geometry showing in general a fairly strong containment of fractures, but it is unclear what explains observed height containment. Re-analysis of published fracture mapping data yields a rule-of-thumb for expected fracture geometry and gives insight into the role of reservoir pressure in the observed containment.
Fracture containment is important for designing stimulation treatments that cover the entire pay, without breaching into aquifers or gas caps. Although modern fracture mapping provides the ground truth for post-treatment fracture geometry, it is still important to forecast fracture height growth based on pre-treatment data.
Fracture mapping shows that on average, fracture length is five times height. Some, (often depleted) reservoirs show even more extreme containment effects. In addition to stress differences, new mechanisms have been proposed to explain the strong observed containment, such as layer interface opening. Although such mechanisms are quite plausible in some formations, it is unlikely that they provide a universal explanation. New developments in fracture propagation modeling provide a simple mechanism for stronger containment than predicted by conventional models. Laboratory testing indicates that fracture propagation must be described by a cohesive zone at the tip. In such a model, rock ruptures at the fracture tip due to effective stress exceeding strength, which introduces the difference between stress and pore pressure into fracture propagation (Schmitt et al., 1989; Visser, 1998). In the first place this propagation model readily explains high net pressure, while a relatively small stress difference can yield much slower height growth compared with length growth. Furthermore, pore pressure changes upon failure can yield a strong containment effect.