Determination of Hydraulic Fracture Azimuth by Geophysical, Geological, and Oriented-Core Methods at the Multiwell Experiment Site, Rifle, CO

Abstract

ABSTRACTA comprehensive program for prediction of hydraulic fracture azimuth is being conducted in the tight, lenticular, gas sandstone reservoirs of the Mesaverde Group in the Department of Energy's Multi-Well Experiment near Rifle, Colorado. Methods used in the program can be divided into two groups: 1) indirect, predictive techniques which determine either the principal paleo-stresses or in situ stresses prior to fracturing and 2) direct observation from either geophysical detection and mapping of hydraulic fracture propagation or wellbore impression-packers of open-hole hydraulic fractures. Predictive methods which determine the paleo-stress directions include geologic observations of the orientation of normal faults and fractures at the surface and in oriented core and twinned calcite strain analysis of oriented core. These methods show a consistent direction of N 75°W ± 10° for the maximum horizontal paleo-stress. Predictive methods which determine the in situ stress directions include 1) oriented core analysis using anelastic strain recovery measurements and differential strain curve analysis, 2) oriented caliper and televiewer logs to infer stress directions from wellbore breakouts, and 3) computer modeling of the horizontal stress directions due to gravitational loading of topographic relief. The results of these methods are fairly consistent, and are in general agreement with the paleo-stress results, but also suggest a possible 20° clockwise rotation with depth of the maximum horizontal in situ stress from N 89°W ± 10° in the upper fluvial zone starting at 1330 m to N68°W ± 9° in the marine zone ending at 2450 m. The clockwise rotation of the horizontal stresses with depth may be due to large, local topographic relief superimposed on the regional stress field of the basin. Confirmation of the predicted fracture azimuth in the paludal zone at 2100 m was made using a borehole seismic detection method during hydraulic fracturing, which incorporated an observation well that was within 80 m of the stimulation well. The fracture height and azimuth was mapped for 3 different fracture tests and showed a mean azimuth of N 67°W ± 8°. In addition, an open-hole hydraulic fracture test in which impression packers were used, showed a fracture azimuth at the wellbore ranging from N50°W to N 70°W in the marine zone at 2306 m.