Fracturing in Chalk Completions (includes associated papers 14483 and 14682 )

Abstract

Summary Chalk reservoirs are an exploration objective in parts of the North Sea and elsewhere but, because of their low permeability, usually require stimulation to produce economically. However, conventional stimulation methods frequently fail to give the expected improvement in productivity. Research has shown that this is caused by certain specific properties of chalk: low hardness, homogeneity, lack of fracture barriers, and pore collapse under stress. Acceptable increases in productivity have been obtained when these properties were accounted for properly in the design. Introduction Hydrocarbons have been found in chalks in the North Sea, the Middle East, the U.S. gulf coast and midcontinent regions of the U.S., and the Scotian Shelf of Canada. Chalk is typically very porous but impermeable and soft; it is also rather homogeneous and massive. Because the permeability is so low, effective fracture stimulation is normally a prerequisite for the development of chalk reservoirs; even where secondary permeability is present, fracturing may give an attractive increase in productivity. During the 1970's, production from chalk, particularly in the North Sea and Middle East, became of interest to the Shell Group. Our initial experience with conventional acid and hydraulic fracturing techniques, however, was not very successful; the postfracture productivity frequently was poorer than expected and declined rapidly. Similar experiences have been reported by others. The failure to obtain sustained productivity improvement initially was attributed to the softness and homogeneity of the chalks we were dealing with. It was reasoned that softness would lead to the embedment of proppant and the loss of fracture conductivity in hydraulic fracturing, while the homogeneity of the chalk would give evenly etched walls in an acid fracture treatment, resulting in very low fracture conductivity when the fracture pressure is released and the fracture closes. Moreover, any ridges left after the acid fracturing process may be too soft to withstand the stresses imposed by production. Therefore, research to improve the available stimulation techniques focused on overcoming these specific problems; we found, however, that two other properties also play a role and must be considered in the design. These arethe ductile behavior of chalk, which can cause pore collapse and loss of permeability, andthe lack of significant fracture barriers, which leads to radial fracture development. Etched-Fracture Conductivity We use the apparatus described in the Appendix to measure acid-etched fracture conductivity in the laboratory. A typical result is shown in Fig. 1, where the dashed line shows the fracture conductivity of a chalk sample etched with an acid emulsion, plotted as a function of closing stress. It can be seen that the initial fracture conductivity is low and disappears under increasing stress. Such a result is typical of chalks, no matter what kind of acid is used (plain, gelled, or emulsified). This is caused by the homogeneity of the chalk, which results in evenly etched fracture walls, and by its softness, which causes any ridges left after the acidization to collapse under the stresses imposed by production. To obviate the problems of homogeneous etching, we make use of the phenomenon of viscous fingering. A high-viscosity, nonreactive preflush is used to create the fracture through spaced perforations, and is followed by the injection of low-viscosity acid, which fingers through the viscous preflush as indicated in Fig. 2. This etches the fracture walls irregularly; the technique, WISPER, derives its name from these widely spaced etched ridges. An example of the acid-etched fracture conductivity induced in a sample by means of the WISPER technique also is included in Fig. 1. Comparing the results it is clear that the viscous-fingering technique induces a higher etched fracture conductivity. Unfortunately, in this particular example, the acid-etched fingers collapse and the conductivity is lost under stress. This problem is common in chalks and, as illustrated by Fig. 3, is dependent on the formation hardness [throughout this paper we use the Brinell hardness number (BHN) as a definition of hardness]. To produce these data, we milled a groove in the sample as shown in the top left corner of the figure and measured the conductivity of this groove as a function of time and stress. Fig. 3 shows the fracture conductivity measured after 1,000 hours at a given stress. We chose to mill the groove, rather than acid-etch the sample, so that all the three samples used, taken from the same North Sea well, would have the same initial fracture conductivity. The results demonstrate that increasing stress will result in a strong decline in fracture conductivity, which indicates that the conductivity of acid-etched fractures is sensitive to reservoir depletion. In the field from which these samples were taken, the effective stress acting on the fracture face after closure is estimated to be about 130 bar [13 MPa]. JPT P. 73^