Bringing seismic ideas to acoustic logging

As the oil and gas industry transitions to a dependence on the more costly unconventional reservoirs, operators strive to find new and more efficient ways to produce from those reservoirs. There has been much focus on drilling and hydraulic fracturing technologies, but what about identifying natural fractures? Natural fractures exist in the producing zones, and only about 20% of the fracked area actually produces. If the industry was better able to identify the naturally fractured zones, it would be able to focus on that 20%.

Currently, the industry uses acoustic logging or imaging to infer or find fractured zones. Unfortunately, however, the move to unconventional reservoirs has brought shortcomings in those techniques to the forefront. For images, widespread use of oil-based mud and the common practice of drilling high-angle wells render image data mostly unusable. For shear wave anisotropy, fractures are simply inferred and, for many reasons, that inference cannot be relied on, particularly in unconventional reservoirs. Perhaps imaging and shear wave anisotropy need to make way for different processing techniques.

Furthermore, present-day tools are only able to detect anisotropy when logged sections are more than 5% anisotropic, leaving the subtle fracture systems undetected. For this last issue, the industry needs to look no farther than seismic and microseismic techniques for guidance. Due to issues with signal attenuation and a high signal-to-noise (S/N) ratio, seismic has long used techniques such as stacking to improve results and amplify anomalies.

A new methodology adopts techniques from seismic and microseismic to provide useful and informative results on fractures, creating a new processing technique for acoustic logging.

FIGURE 1. RMS energy distribution (right) across a fractured interval indicates where fractures coincide with a sharp drop in energy. (Source: GeoBiz Technology Inc.)

Current limitations

Current techniques such as shear wave anisotropy and imaging still have applications in certain situations. However, they each have limitations. Fractures are anisotropic, so shear wave splitting can be an indication of fractures in subterranean formations; however, not all anisotropic formations are fractures. Therefore, use of shear wave anisotropy is not so much a definitive indicator of fractures but a definitive indicator of anisotropic formations. Those formations may be fractures but may alternatively be other anisotropic or seemingly anisotropic formations. Furthermore, fractures, though anisotropic, can also exist in mildly anisotropic formations that cannot be detected by the present-day tools using current methods since it is difficult to distinguish the subtle fracture patterns from the surrounding anisotropic media using shear waves. Since fracture detection is the ultimate goal, given that fracture patterns indicate producing areas, a move toward a technique that actually identifies fractures as opposed to simply inferring fractures is logical.

The processing technique created by GeoBiz Technology Inc. allows the industry to move toward a more direct method of characterizing fractures since shear wave anisotropy does not allow direct fracture detection. Although several industry tools, including LWD tools, collect sectored compressional (P) waves, there has been little to no use for them in acoustic logging. However, looking at seismic research, work has been done to find fractures using P waves. Adapting the research that has been done to acoustic logging allows the industry to characterize fractures using P waves. Since tools already exist that collect necessary data, only the processing technique needed to be developed. The theory used to develop the GeoBiz processing technique is based on the theory that, across an open fracture, P waves show a significant drop in energy. By mapping this drop in energy, fracture location and azimuth can be calculated. This technique has been tested in a number of wells and compared against core results. Test results show a match of approximately 90% between core data and log data in both fractured and unfractured intervals (Figures 1 and 2).

Elements of characterization

As noted, some tools already exist that collect sectored P waves. The use of these sectored waves is key in fracture characterization. The use of sectored waves allows mapping the waveforms in a manner similar to what is currently done with shear waves but with modified algorithms. The use of raw, sectored P waves is one of three key elements in the characterization method.

FIGURE 2. Across an unfractured section RMS energies show little variation (right). (Source: GeoBiz Technology Inc.)

Although not mentioned in most publications, root mean square (RMS) normalization is essential not only in processing sectored compressional waveforms but also in shear wave anisotropy. In both anisotropy and the new fracture detection method, four quadrants must be combined to compute RMS energies from 0 to 90 degrees. To combine these four quadrants without normalization, it must be assumed that the raw RMS energy output from all four quadrants is matched, which is unlikely and unrealistic. Therefore, the RMS energy must be normalized prior to combining. RMS normalization, then, is the second of the three key elements to the characterization method.

Although LWD tools provide some mapping capabilities, the S/N ratio is inadequate to properly identify subtle fracture patterns. The S/N ratio also is inadequate in traditional acoustic logging tools. Improving this ratio again requires the adaptation of seismic techniques. In seismic data acquisition, stacking of data gathers is routine. Until now, acoustic logging has found stacking techniques unnecessary. However, to find natural fractures, which are the defining characteristic in a productive unconventional reservoir, the S/N ratio must be improved using a modified stacking technique. GeoBiz has developed software to efficiently stack data gathers, allowing the delivery of results within the time restrictions of operators in the field. For a tool with 13 receivers, a 52-fold stack can be achieved. Stacking has led to a significant improvement in the S/N ratio, enabling more confidence in identifying subtle variations in amplitude. This is the third key element to the characterization method.

Fracture pattern characterization

In addition, the method allows not only the identification of fractures but also the full characterization of the fracture patterns. Tests show the azimuth of fractures seen by both the imaging tool and the sectored P tool match well (Figure 3).

FIGURE 3. The azimuth of fractures seen by the imaging tool match well with the azimuths seen in the sectored compressional tool. (Source: GeoBiz Technology Inc.)

In summary, the use of sectored P waves to map energy in the formations, RMS normalization to achieve amplitude independence and stacking of data gathers to increase the S/N ratio show remarkable accuracy when compared to core results in a shale play for both fracture identification and finding the azimuth of fractures.