ABSTRACT
A combined theoretical and laboratory test program is described with the purpose of providing a useful model of the multiple radial fracturing process. A non-quantitative conceptual model is developed as the basis for detailed mathematical modeling and test planning. The model includes borehole pressurization, flaw rupture, and gas- driven fracture propagation phases. The laboratory tests are carried out in blocks of scaled rock simulant loaded to a true three-dimension stress state and with miniature propellant/explosive charges in a wellbore as the fracturing energy source. The conceptual model, fracture initiation modeling, and laboratory test setup have been completed and are described. The remaining work is ongoing.
INTRODUCTION
Multiple radial fracture stimulation (under various names) has long been viewed as a possibly advantageous alternative to conventional hydraulic fracturing in certain petroleum reservoir formations, such as the Eastern Devonian Shales. The formation rocks for which multiple radial fracturing has this potential generally have an anisotropic natural fracture system. It will be difficult to force a hydraulic fracture to intersect this pre-existing natural fracture system. This case exists because both the hydraulic fracture at depth and the natural fracture system are likely to be aligned in the same direction as the greatest horizontal in situ stress. The hydraulic fracture, therefore, has little likelihood of intersecting the conductive natural fracture system. It would be preferable in this case to generate "crosscutting" fractures that propagate across the natural fracture system, even if they are shorter than normal hydraulic fractures. It is virtually impossible to conceive of a method of generating a single isolated crosscutting fracture. However, multiple fractures that emanate from the wellbore in random radial directions, and penetrate sufficiently far into the formation to be in contact with the natural fracture system, might be achievable and should be very effective in stimulating certain reservoirs. Early work in this technology (and some current efforts) tended to emphasize the possible use of high explosive mixtures placed in the wellbore to create a virtually instantaneous source of high pressure gas and a hoped-for resulting random radial fracture pattern. Unfortunately, there is field evidence that in many formation rocks, the use of high explosives applied directly to the wellbore face tends to create an intensely damaged zone near to the wellbore that is only rarely associated with long extending fractures. In fact, the damaged zone by its energy absorption and low permeability due to compaction and fines, may actually defeat the intended stimulation. Therefore, most of the more recent work centers around the use of propellant mixtures placed in the wellbore to create a less rapid source of lower pressure gas to create fracture growth, under the assumption that the damage caused by high explosive gases is related to their high initial pressures and rapid energy release rates. The generally accepted hypothesis is that if a high explosive (high pressure, rapid energy release rate) produces numerically intense, randomly directed, but only local multiple fracturing, and conventional hydraulic fracturing (low pressure, quasi-static energy release rate) produces a long, but adversely oriented single, bi-winged fracture, that a propellant (intermediate pressure, intermediate energy release rate) might be able to initiate an intermediate number of fractures with the desired randomness of direction, and extend several of those fractures without causing near-a wellbore damage.
