New Albany Shale Gas Project
Natural fractures in the New Albany Shale: Implications for development of effective drilling and completion technologies
The objectives of the study are to characterize the natural fractures in the New Albany Shale in the Illinois Basin, and to assess their relevance for natural gas production. Our focus is on the thermogenic part of the play, in southern Indiana and western Kentucky and on whether the sealed natural fractures that are present represent weak planes that reactivate during hydraulic fracture treatments. The possibility of permeability enhancement through open natural fractures is investigated.
These objectives were met through the following tasks:
- Synthesize published fracture characterization data and augment with fracture description in cores and outcrop for 12 cores from southern Indiana, 1 core from western Kentucky and 4 outcrop locations.
- Qualitative assessment of the tensile strength of natural fracture planes.
- Measurement of the subcritical crack index (SCI) of core samples with the aim of understanding fracture spatial distribution and how fracture patterns differ from one layer to the next.
- Consideration of the relative orientation of SHmax and natural fractures, and whether this geometry is favorable for reactivation.
- Assessment of whether part of the fracture population is open and would enhance permeability in the reservoir.
Summary of Results (poster presentation at AAPG 2010)
- Natural fractures are common in the New Albany Shale, but vary in character and have different properties with respect to their effect on gas production.
- There are both opening-mode fractures and faults in core and in outcrop.
- In many cores there is more than one opening-mode fracture set. On the basis of aspect ratio, mineral fill, orientation and whether the fracture has been shortened during sediment compaction, the fracture sets have different origins.
- The majority of fractures observed in core in the Clegg Creek member (the usual gas target) are similar to the demonstrably weak planes present in the Barnett Shale in the Fort Worth Basin (Gale et al. 2007; Gale and Holder, 2008) where calcite mineral fill is only weakly attached to the fracture walls. These fractures are undeformed, straight-sided, with large height to width aspect ratios, and are commonly arranged in en echelon arrays.
- Fractures in the lowest part of the section in the Blocher Member are partly open and have low height to width aspect ratios and irregular fracture walls. They are sealed with dolomite and have residual bitumen and large open vugs.
- The most common sealing cement is calcite, but some samples are sealed by both calcite and quartz. We speculate that quartz cements result in stronger fracture planes because of bonding between wall rock and cement as quartz overgrows broken quartz wall rock grains.
- We tested three samples for subcritical index, fracture toughness, and Youngs modulus. The samples were from cores from Sullivan County, southwestern Indiana. Sample preparation was challenging because the cores are both slant cores and are oblique to bedding. Most of the measurements are considered reliable; stress-strain (2 load/unload cycles per specimen) behavior was consistent, fracture toughness results (three specimens for each core loaded to failure) are consistent, and decay curves for almost all subcritical index tests have good fits to theoretical curves. There are differences in the subcritical indices. The highest subcritical indices were measured in the samples from the two horizons where the apparent fracture intensity is lower. Higher subcritical index generally indicates a more clustered fracture pattern. In a highly clustered pattern the chances of sampling fractures decreases, so that it is quite possible that the lower horizon in the Solsman well (2630 ft) and the core from the Osburn Trust well are fractured, but that the clustering meant the fractures were not sampled.
- The structural grain of the two areas – southern Indiana and western Kentucky – is fundamentally different. Southern Indiana is dominated by the Wabash Valley normal fault system, which is currently active and trends approximately N-S. Western Kentucky is dominated by the E-W trending Rough Creek Graben. The major and minor faults and opening-mode fractures in part reflect this difference. However, preliminary core data suggests that some opening-mode fractures in western Kentucky may not be parallel with the larger faults.
- The present-day in situ stress must be determined on a site-specific basis. The world stress map database suggests a swing in the maximum horizontal stress direction from ENE in southern Indiana to E-W in western Kentucky. The large difference in underlying local structure might have a significant perturbation effect on the far-field stress orientation.
For more information see Reports