The objective of the study was to investigate the effect of strength heterogeneity on fracture patterns observed under laboratory conditions in hollow cylinder fracturing tests. The finite-element method with cohesive elements representing potential fracture sites was used. The number of radial fractures splitting the hollow cylinder decreased as the heterogeneity was increased. The ultimate strength of the hollow cylinder exhibited a similar trend. Some of the fracture patterns obtained in the simulations, e.g. 2 or 3 major radial fractures, looked qualitatively similar to those observed in laboratory tests on rocks and in an earlier modeling study performed with a different numerical method.
Fluid injection in porous reservoir rocks is a central activity for CO2 storage and hydrocarbon production operations. Such operations require a solid understanding of fundamental mechanisms of rock fracturing under high injection pressure.
In an earlier study (Alassi et al. 2012) experiments were performed on two types of rocks, a limestone and a sandstone. During the experiment, a hollow cylinder with the outer-to-inner diameter ratio equal to 5 was subjected to increasing stress on the hole surface while the stress on the outer surface was maintained constant. No fluid injection was allowed from the hole into the rock. The outer surface of the specimen was hydraulically isolated, too. The fracturing of the specimen was thus caused by applied mechanical stresses only.
Two main fracture patterns were observed in the experiments. Limestone specimens failed by developing two fractures opposite to each other. A photograph of the fracture pattern observed in one of the limestone specimens is shown in Figure 1 where one of the two diametrically opposite fractures produced a branch after having propagated some distance from the hole wall.
Sandstone specimens, on the other hand, developed 3 fractures oriented at approximately 120° to each other (Figure 2).