--> Abstract: An Empirical Approach to Understanding Rock Lithology and it’s Relationship to Minimum Horizontal Stresses for Fracture Stimulation Design, by T. H. Leshchyshyn, R. Barba, K. Baird, and K. Beadall; #90088 (2009)

Datapages, Inc.Print this page

An Empirical Approach to Understanding Rock Lithology and it’s Relationship to Minimum Horizontal Stresses for Fracture Stimulation Design

T. H. Leshchyshyn1, R. Barba2, K. Baird1, and K. Beadall3
1Knowledge Reservoir/OXY, Bakersfield, CA, [email protected], [email protected]
2CS Resources, Houston, TX, [email protected]
3Beadall Services, Houston, TX [email protected]

It has been shown empirically that rock lithology has a direct relationship with minimum horizontal stress gradients. The higher the grain or bulk density, the higher the minimum horizontal stress gradient. In Alberta, Canada, the order of increasing stress gradient is sandstone (15 kPa/m or 0.663 psi/ft), siltstone (18.5 kPa/m or 0.817 psi/ft), claystone (18.8 kPa/m or 0.828 psi/ft), shale (19.5 kPa/m or 0.861 psi/ft), limestone (20 kPa/m or 0.883 psi/ft), dolomite (21.5 kPa/m or 0.95 psi/ft), etc. The weight of overburden is assumed to be about 21.8 kPa/m or 0.96 psi/ft.

Rock mixtures of the above lithologies have interpolated minimum horizontal stress gradients relative to the amount of each lithology mixture. Accuracy of the final stress values has been estimated at about ±3%. Previously, a Canadian Rock Table was created from the above data for a total of about 10 pure lithologies (as above) and 70 rock lithology mixtures. All were generated from actual fracture stimulations performed between 1970 and 2003 (a total of over 30,000 stimulations were investigated) in Canada, Russia, and a few in China. A separate set of Rock Tables were created for southern California, different from the rest of the world due to the young age of the reservoirs (about 15 million years old) and the occurrence of diatomite and it’s diagenetic burial lithologies, siliceous shale, and opal CT.

Sonic logs with added shear wave data are also used for fracture stimulation design, and if corrected for coherency (bedding effects) and tectonic strain and tied to minifrac closure stress values, give excellent stress distributions with lithology changes, resulting in accurate estimates of created fracture heights and lengths.

Both methods, if used properly, give good stress values for fracture model predictions. Older depleted reservoirs in the US have become candidates for refracturing to increase productivity, requiring better stress values for fracture height control.

AAPG Search and Discovery Article #90088©2009 Pacific Section Meeting, Ventura, California, May 3-5, 2009