Richard H. Groshong1, Jack C. Pashin2, Robert D. Schneeflock3
(1) University of Alabama, Tuscaloosa, AL
(2) Geological Survey of Alabama, Tuscaloosa, AL
(3) Paramount Petroleum Co, Ridgeland, MS
A reservoir that appears to lack faults at one scale of resolution or at one sampling density may nevertheless contain faults that are below the resolution of the observations. Is it possible to predict the abundance of sub-resolution faults from the available data? Predictions using fractal fault distributions tend to imply large numbers of faults and correspondingly large hidden fault strains. The question is applied to a cross section of the Gilbertown graben in south Alabama, a growth structure within the regional peripheral fault trend formed along the updip limit of salt at the northern margin of the Gulf of Mexico. Are there more faults than those seen in the wells used to construct the section? Area balance is chosen as the prediction technique because it is akin to mechanical modeling but more practical, capturing the mechanical response of the structure without requiring knowledge of the mechanical stratigraphy or the stresses. Bed-length change is calculated from the area balance and correlated to the heave on unresolved faults. For growth units, bed-length changes are determined using the lost area in the graben and the depth to the lower detachment. The method is illustrated and verified with an analytical model of a growth full graben. Applied to the Gilbertown cross section, the area-balance method indicates that few sub-resolution faults are required in the Selma and Eutaw reservoir units. These predictions are confirmed using the more complete spatial coverage afforded by a 3-D seismic survey.
AAPG Search and Discovery Article #90906©2001 AAPG Annual Convention, Denver, Colorado