The success of a hydraulic fracture treatment is greatly enhanced by control of the created fracture geometry. Of particular importance is the migration of the fracture into the boundary lithologies, which is largely determined by in- situ stress. Most of the presently used hydraulic fracturing models assume a constant fracture height confined to the pay zone, which predicts unrealistically large fracture lengths. Currently, a joint research program (involving The Gas Research Institute of Chicago, Chandler and Associates of Denver, and Terra Tek, Inc. of Salt Lake City) is underway that is evaluating, in the laboratory and in the field, several techniques to place (in conjunction with hydraulic fracturing) deeply penetrating fractures into pay formations not boundedby high stress barriers. Field work is taking place in the Douglas Creek Arch area of Colorado.
This paper presents a part of that program, the laboratory model study of hydraulic fracture propagation in one meter cubic blocks. The test block is loaded in a large frame which allows the independent application of minimum and maximum horizontal stressesto the simulated pay zone and two boundary lithologies, anda vertical stress to the entire block. The test block is internally strain gaged to allow the inference of the actual stress distribution both prior to, during, and after the fracture.Following hydraulic fracturing with a dyed fluid, the block is broken andthe created fracture surface area is examined. Through the use of this test set-up, the effect of in-situ stress on fracture geometry has been demonstrated, and a method of fracture geometry control basedon perforation placement has been established.