Laboratory Hydraulic Fracturing Test on a Rock With Artificial Discontinuities

Abstract

Abstract The design and subsequent results of a hydraulic fracturing test performed on a large block of high modulus and low permeability rock (Colton sandstone) are presented. The focus of this experimental study was to assess the effects of discontinuities on hydraulic fracture growth. A high viscosity fluid was used in order to provide fracture growth similar to actual field conditions. Fracture growth and its internal fluid pressure were monitored by fixed probes placed normal to the expected plane of propagation. Fracture tip arrivals were captured by the fixed pressure probes and showed a distinct fluid lag region (i.e., a lower fluid pressure region close to the fracture tip). The controlled laboratory experiments showed planar fracture propagation trends as expected from three-dimensional modeling. Introduction Since the inception of the hydraulic fracturing technique as a means to improve productivity of oil and gas wells, the hydraulic fracturing community has determined certain containment mechanisms that influence fracture growth (i.e., in-situ stress, stress gradients, rock mechanical properties, frac-fluid rheology, injection rate, etc.). However, predicting which variable or variables have a decisive impact is still unclear and highly controversial. The work presented in this study addresses part of this issue with a laboratory controlled hydraulic fracturing test performed on a large block of Colton sandstone. Physical model tests are of great value to determine the relevant phenomena of hydraulic fracture propagation because they allow for measurements that are unavailable in the field. However, one cannot generally correlate their results directly to field applications due to scaling issues. Their ultimate purpose is to provide a benchmark for numerical models that may represent the essential physics of the process. Many laboratory-scaled hydraulic fracturing tests on physical models have been conducted1–6 where limited consideration of scaling differences was given in the design and interpretation of results. Often, these tests have shown rather unpredictable, often "wandering" and "branching" fracture propagation, which may have resulted from test conditions unlike (or like) field conditions. This study is one of many research efforts being conducted by an industry-sponsored research group called FAST (Fracturing, Acidizing, Stimulation, Technology), a consortium formed within the Petroleum Engineering Department at the Colorado School of Mines. The group performs practical research in all areas of oil and gas well stimulation. In the pursuit of enhancing our understanding of fracture height growth and containment mechanisms, the experiment attempted to recreate fracture propagation observed at field scale, even though a great difference of scales exists between fractures generated in laboratory tests and in field applications. To account for these scaling issues, model laws that relate experimental parameters of the physical model to field-scale prototype parameters were utilized to perform the block test and interpret some of the experimental observations7,8,9. The experiment was designed to perform two tests in one large block. Two conditions for hydraulic fracture propagation were examined:Fracture #1 extending through a flawless section of the block (tip mechanisms); andFracture #2 (normal to fracture #1) intercepting angled, filled artificial joints placed on each side of the centered wellbore (fracture containment), as seen in Figure 1. Fracture tip arrivals were captured by the fixed pressure probes which indicated that field-like fracture growth rates were generated in the test block. Fracture fluid pressure measurements show a distinct fluid lag region (i.e., a lower fluid pressure region close to the fracture tip). Both induced fractures presented slightly higher net pressures than predicted by the scaling analysis, presumably due to extended lag zones at the tip of the fracture under high confining stress.