A NEW METHOD OF UTILIZING OLD GEOPHYSICAL LOG SUITES TO CREATE REALISTIC GEOCELLULAR MODELS: AN EXAMPLE FROM THE LAWRENCE OIL FIELD IN THE ILLINOIS BASIN

The Illinois Basin (ILB) has been producing oil for over a century, but enhanced oil recovery methods (EOR), such as CO₂ EOR, have not been widely used. Reservoir simulations are a powerful tool capable of reducing uncertainty associated with an EOR project, but they require a detailed, realistic geocellular model of the reservoir architecture. In mature oil fields, such as those within the ILB, the data required for the construction of such models is limited, due to the lack of wells with modern (post-1958) geophysical log suites or core analysis data. In the absence of adequate data coverage from core or modern log suites, it is necessary to use older log suites in the modeling process. A new method was developed to use old log suites to create geocellular models that are representative of the site geology.

The majority of data available for Lawrence Oil Field, Lawrence County, IL is in the form of spontaneous potential (SP) logs, but recent EOR projects have provided some neutron-density porosity logs. Spontaneous potential data were used by converting digital log curves to a sand/shale ratio through normalization. This ratio was then cross-plotted against core data to obtain regression curves relating SP to permeability and porosity. The resulting equation defining the curve was used to convert the normalized SP log data into porosity values. Three datasets were produced to compare the effect of different log types on model development: porosity log data only, converted SP data only, and a combination of the two.

Geostatistical techniques used these datasets and a conceptual geologic model to generate three geocellular models of the Cypress Sandstone within the Lawrence Field. Both the model created with the converted SP data and the model created with the combined suite had average porosities of 14% and a standard deviation of 3%. The model created with the porosity log dataset had an average porosity of 16% and a standard deviation of 6%. These model results compare favorably with core analysis data. This study demonstrates that SP logs can be used effectively to construct geocellular models or be combined with more modern datasets to improve data coverage, which may improve site screening for EOR projects in mature basins.