ABSTRACT: Evaluation of Two Techniques for the Initial Quantification of Reservoir Heterogeneity-Examples from Appalachian Basin

MCDOWELL, RONALD R., K. LEE AVARY, DAVE L. MATCHEN, and ANA G. VARGO, West Virginia Geological and Economic Survey, Morgantown, WV

Reservoir heterogeneity of the Big Injun sandstone (late Early Mississippian) is being modeled geostatistically in two West Virginia oil fields (Granny Creek and Rock Creek) as part of a U.S. DOE-funded study. Preliminary quantification of heterogeneity is necessary because this dictates boundary conditions, complexity of the model, and computer power required to solve it.

Two indirect methods (production history and completion date/well location analysis) of assessing heterogeneity were tested. Yearly Big Injun oil production for each field was graphed against time. Resulting plots showed local maxima believed to result from a combination of effects including economics, new production/recovery techniques, offset drilling, and discovery of new production. Discovery of new production should reflect reservoir heterogeneity. Filtering out all other effects requires a detailed history of the field. Production records prior to 1940 are generally unavailable-a serious handicap for fields which may have produced half of their oil prior to 1940.

Next, numbers of completed wells in each field were graphed against year of completion. These plots showed "clusters" of completions through time. To determine if time clusters represent periods of offset drilling and/or new discoveries, well location maps were produced for each cluster. The Granny Creek plot showed six clusters of completions from 1920-1990. Each of the corresponding location maps revealed two distinct sets of wells: (1) those scattered throughout the field adjacent to existing wells (offset), and (2) those tightly grouped in parts of the field not previously drilled (discovery). The Rock Creek plot indicated five clusters of completions from 1900-1990, but only three of five location maps showed discovery well groupings, implying greater reservoir heterogeneity with n Granny Creek than in Rock Creek. Analysis of cores and geophysical logs supports this conclusion. Three stratigraphic units (six lithofacies) have been identified within Granny Creek. Only one unit (two or three lithofacies) has been identified within Rock Creek. Thus, when production data are incomplete, completion/location analysis may provide an alternative predictor of heterogeneity.

AAPG Search and Discovery Article #91013©1992 AAPG Eastern Section Meeting, Champaign, Illinois, September 20-22, 1992 (2009)