Abstract In recent years three developments which have evolved more or less independently, when related, may be of value to the petroleum industry. First is the recognition, through normal oil field development, that fractures are significant to both reservoir capacity and performance. Second is the fact that controlled laboratory experiments have produced, in increasing quality and quantity, empirical data on rupture in sedimentary rocks. These data have been segregated to demonstrate the individual control on rupture of several important parameters: rock type, depth of burial, pore pressure, and temperature. The third development consists of the discovery of new methods to recognize, evaluate, use, and, in some cases, see fractures in the subsurface. This discussion of these three developments may help geologists and engineers to find new approaches to exploration and exploitation of fractured reservoirs. Reservoir and production engineers presently make the greatest use of fracture data, but geologists should find this information useful in exploration for oil and gas trapped in subsurface fractures. Except in the search for extensions to proved fracture reservoirs, there is in the literature a paucity of clear-cut examples of the use of fracture porosity data in advance of drilling. For this reason, several speculative exploration methods discussed herein implement mapping of fracture facies as well as stratigraphic facies.

In order to make a quantitative study of a purely cataclastic deformation, jacketed cylindrical specimens of dry St. Peter quartz sand of different grain sizes have been deformed under uniform confining pressure and differential load. Results presented illustrate the influence of the deformation on porosity, pore size, grain size, grain fracturing and orientation, and strength. The present experiments, conducted under deliberately restricted conditions, are intended as a first step toward an adequate understanding of the complicated problem of sandstone deformation. Undeformed St. Peter sand grains are unfractured to slightly fractured; contain planes of liquid and gas inclusions which exhibit a decided preference for the prism zone; exhibit a predominance of grains with no undulatory extinction as well as grains which are slightly, moderately, and highly undulatory; contain very few grains showing “deformation” lamellae; have an intercept ratio of 1:1.3:1.7, and a high intercept sphericity (0.77 to 0.79); and appear to have grain shapes controlled by r {101̄1}, z {011̄1}, and the prisms. The undeformed sand aggregates exhibit an apparent elongation so that the long axes of the grains tend to lie parallel to the circular section. In addition, the aggregates show a tendency for the optic axes of the grains to form a girdle parallel to the circular section. Fracturing is the most conspicuous feature of the deformed sands. The fracture pattern of specimens deformed under uniform pressure is random. Fracture-orientation patterns reflect the symmetry of the deformation under differential loading conditions. In compressed specimens the fractures tend to lie at small angles to the direction of loading and are probably shear fractures. In extended specimens the late-formed fractures tend to lie normal to the axis of extension and are most probably tensile fractures. Apparent elongation and optic-axis orientations for compressed samples are much like those of undeformed sands. In extension, however, the apparent elongations of the grains undergo a progressive rotation through 90° from their initial positions. This is accompanied by a similar shift in optic-axis orientations and therefore indicates that bodily rotation of grains does occur. There is no evidence that the experimental deformation has produced any “deformation” lamellae or that the over-all undulatory extinction index has been changed. The percentage of moderately and highly undulatory grains decreases, however, indicating that these grains are particularly susceptible to fracturing. For sands of uniform grain size the coarsest sand (250–300 microns) exhibits the greatest compressibility (uniform confining pressure), which correlates with the facts that this sand also exhibits the greatest reduction in porosity, median pore size, median grain size, and the lowest percentage of unbroken grains. The coarsest sand is the strongest in triaxial tests (differential loading). On the other hand, the finest sand (105–125 microns) is the least compressible; it exhibits the least reduction in porosity, median pore size, and median grain size, and the largest percentage of unbroken grains, and it is the weakest in triaxial tests. It is significant that uniform loading (approximately simulating an overburden pressure alone) does not reorient the fabric of the sands, whereas differential loading results in preferred orientations of fractures, optic axes, and apparent grain elongations, all of which reflect the orientations of the principal stresses across the boundaries of the specimens. This suggests that petrographic studies of natural sands may be used to distinguish between the effects of overburden and of tectonics.