About This Item
Share This Item
Petroleum geologists are continually searching for new and better techniques of exploring for hydrocarbons. This study uses one such technique, discriminant function analysis, to predict trends of potential hydrocarbon production of Mississippian rocks in Cowley County, Kansas, Discriminant function analysis is a method of studying differences between two or more groups, using a set of discriminating variables. The two groups used for this study were (1) wells producing from the Mississippian and (2) wells not producing from the Mississippian.
Variables used were thought to be potentially significant to hydrocarbon accumulation in Mississippian rocks within the study area. These variables included structural tops and thicknesses, surface elevation, presence of tripolitic Mississippian chert, trend surface residuals, and four types of satellite lineation data. A total of 564 wells, each with 21 distinct parameters, was used in performing the two-phased analysis.
The two phases, consisting of a principal components analysis and the discriminant function analysis, were accomplished through use of computer programs and resulted in a single discriminant score value for each well. These values were used to generate a discriminant score surface map. Statistical analysis of the map resulted in 80.6% of the producing wells being correctly classified within producing areas. Alternately, 75.9% of the nonproducing wells were correctly classified. These favorable results suggest that this technique could be successful in establishing trends of production and potential production while simultaneously delineating areas of nonproduction.
End_of_Article - Last_Page 314------------