A Comprehensive Fracture Diagnostics Experiment: Part 2— Comparison of Fracture Azimuth Measuring Procedures

Abstract

Summary A great deal of effort has been devoted to finding geophysical techniques for measuring the hydraulic fracture azimuth. This paper discusses a comparison of seven different measurements to determine the azimuth in a sandstone formation at a depth of 1,050 ft [320 m]. The azimuth was determined as N95E, but significant differences existed between some of the results. This is of fundamental importance because in developing new measurements, the limits of these must be found and honored. Of particular interest are the results from microseismic monitoring. The lack of results suggests that remote (e.g., surface) monitoring for seismic events may be impractical for normal, sedimentary, hydrocarbon-bearing formations. Introduction For many years, it has been accepted and verified by research that hydraulic fractures are generally vertical and extend from the wellbore along a single azimuth. It was also noted that planar features, such as fractures, can significantly alter reservoir fluid flow. Little attention was directed to this problem, however, until the development of massive hydraulic fracturing proved a viable technique for producing low-permeability gas reservoirs. For these microdarcy formations, fracture lengths greater than normal drainage radii are justified, making it necessary to determine the fracture azimuth before the infill wells are drilled, which would make determining the azimuth with pressure-transient analysis possible. pressure-transient analysis possible. In response, several geophysical techniques have been developed, including core analyses, surface tilt measurements, electric potential measurements, and seismic monitoring for microseismic events related to the propagating fracture. Several references for each propagating fracture. Several references for each procedure show that interpretable signals, or data, can procedure show that interpretable signals, or data, can be recorded. There are fewer data where several results are compared, which is unfortunate because several techniques are indirect or base their analyses on assumptions about the physical process creating the signal. Thus not every procedure can work in every case; for example, monitoring for microseismic events in a poorly consolidated formation would probably not yield good results. One older study in Colorado and more recent tests in Colorado and Texas have compared several techniques with generally good agreement; all of these tests, however, were conducted in similar formations-relatively deep, hard, low-porosity sandstones. This paper reports the results from seven different azimuth measuring procedures conducted as part of an Amoco/Dowell Schlumberger/Gas Research Inst- fracture diagnostics test. The test was conducted in the Skinner sandstone near Mounds, OK. The formation at this location was found at a depth of 1,050 ft [320 m] and had an 80-ft (24-m] thickness, 25% porosity, 5-md average permeability, 100% water saturation, and normal permeability, 100% water saturation, and normal pressure. Because the primary goal of the experiment was to pressure. Because the primary goal of the experiment was to measure proppant transport, the test was conducted in two stages by first measuring the fracture azimuth to allow optimum placement of surface magnetometers for the second stage, mapping of magnetic proppant injection. The azimuth measurements included on-site and laboratory-oriented core analyses, tiltmeter measurements, downhole television camera logging, remote seismic monitoring, borehole seismic monitoring, and caliper logs for wellbore "breakouts. "Table 1 summarizes the results. While excellent agreement between borehole logs, tiltmeter analysis, and strain recovery measurements on core shows conclusively an essentially east/west azimuth, the disagreement in one sense is good, emphasizing the need for multiple tests in a new area or formation to guard against misleading results. Borehole Logging Two types of borehole logs were used:downhole television to measure wellbore azimuth directly, andoriented caliper logging to measure wellbore ellipticity. Downhole television logging is a simple process 15 because there is no interpretation involved. Such logging can determine wellbore azimuth for vertical fractures very accurately. For this experiment, the wellbore was circular, and prefracture logs showed an unfractured wellbore. SPEPE P. 423

A Comprehensive Fracture Diagnostics Experiment: Part 1—An Overview

“…This fluid had a hydroxypropyl guar polymer loading of 50 Ibm/1,000 gal [5991 g/m 3 ] with a borate crosslinking system. 3 The tiltmeters also showed the breakout of a horizontal fracture centered around the vertical fracture, coinciding with the high BHP response to the arrival of the crosslinked gel. time is given in Fig.…”

“…3 (This was the only real-time diagnostic method used for this test). Twelve biaxial tiltmeters were installed in a circle 760 ft [232 m] in diameter about the 414 treatment well.…”

“…3,12,13 Otherwise, high-frequency passive seismic monitoring appears to be an impractical tool for mapping an evolving hydraulic fracture in a sedimentary environment. Despite the high sensitivity level of the highfrequency seismic monitoring system, no signals comparable in intensity to those observed in the geotHermal tests previously mentioned were detected by the horizontal surface arrays and vertical arrays in the observation boreholes during the injections.…”

“…[3][4][5] This work was performed by the Oas Research Inst. Potential fracture diagnostic methods, including seismic monitoring and tiltmeters, were used in attempts to measure final dimensions of the fracture and the geometry of the fracture as it was propagating.…”

Comparison of stress-measuring techniques at the DNA-UTP site, Rodgers Hollow, Kentucky

“…Determination of the orientation of the stress field is calculated from the principal strain orientations measured during relaxation of the core (Teufel, 1982;Teufel and Warpinski, 1984;Smith et al, 1986;Warpinski and Teufel, 1989b). Determination of the stress magnitudes is more complicated and relies on one of two models.…”