Depositional Investigation and Analysis of Porosity Development in a Northwestern Oklahoma Cherokee Sandstone, Using Petrographic Image Analysis

Gregory L. Scheffe, W. E. Full

Recent technological advances allow image analysis of porosity networks in rock, using a computer-based Petrographic Image Analysis System (PIA) developed at the University of South Carolina. PIA processes digitized scenes of the porosity network, quantifying the characteristics of each pore. Pattern recognition and/or classification algorithms (CABFAC and QMODEL) are used to determine the mixing components that define the pore complex. Comparisons of the quantified pore attributes to petrophysical values (e.g., permeability) enable us to predict these values with high precision. This procedure identifies the variables that control the value of the petrophysical parameter.

A study of the Bartlesville sandstone member, Cherokee Group, Pennsylvanian System in northwestern Oklahoma, used the procedures as outlined in defining characteristics of this petroleum reservoir. The investigation identified the reservoir as deltaic. PIA and the unmixing algorithms determined the existence of five pore types that contribute to the reservoirs development. Statistically derived relationships predicted the permeability with an R² value of 0.93. The highest prediction resulted from the pore type and quantified pore attributes that characterize small pores occurring in high density and exhibiting small to moderate perimeter roughness. This pore type occurs in quartz sandstone laminations with alternate layers containing a small percentage of calcite, fossil fr gments, and clay. This type of lamination occurs in very fine cross-bedded and contorted sedimentary deposits. This type of sedimentary structure also restricts permeability. Reservoir production seems related to a system of fractures and micro-fractures.

AAPG Search and Discovery Article #91043©1986 AAPG Annual Convention, Atlanta, Georgia, June 15-18, 1986.