Predicting Reservoir Facies Distribution in the Subsurface: Hodgeman County, Kansas

Abstract

Abstract

Previous studies have shown that Bindley Field, Hodgeman County, Kansas, produces from Lower Meramecian carbonate rocks. Stratigraphic subdivisions for carbonate rocks within the field are divided into discrete intervals which can be correlated over much of western Kansas. These intervals are broken down even further into vertically heterogeneous individual parasequence-scale chronostratigraphic units (Johnson 1994). The producing interval for Bindley Field and surrounding areas is difficult to identify in areas where facies distribution has not been extended. This facies is believed to exist to the northwest of Bindley Field, but the actual distribution is unknown. The research for this study examines wire-line data along with petrologic evaluation techniques to determine a more accurate prediction of where a certain producing facies may occur within Bindley Field, and to the northwest. A suite of geophysical logs (Neutron/Density, Gamma Ray, Sonic, Induction, CMB) along with a neural network is used to create an accurate prediction of producing intervals within the subsurface. Values are derived from wire line log data and used to develop a neural network of Bindley Field. Producing facies were interpreted, and identification of reservoir quality zones were made for data poor areas. The neural network model was then examined for accuracy and precision, which may allow the model to be applied to fields where the specific reservoir facies may exist. The use of a facies prediction software allows for immediate evaluation of the accuracy of the model and allows for an overall more fluid process of modeling the subsurface in this area. The model predicts that a reservoir facies may exist to the northwest of Bindley Field.

AAPG Datapages/Search and Discovery Article #90260 © 2016 AAPG/SEG International Conference & Exhibition, Cancun, Mexico, September 6-9, 2016