Oomoldic Reservoirs of Central Kansas, Controls on Porosity, Permeability, Capillary Pressure and Architecture

Oomoldic Reservoirs of Central Kansas, Controls on Porosity, Permeability, Capillary Pressure and Architecture

Kansas Geological Survey
Open-file Report 2002-48

Conclusions

Geology and Architecture

Lansing-Kansas City oolitic reservoirs exhibit geometries and architectures similar to modern oolites in the Bahamas although Pennsylvanian oolite reservoirs usually contain multiple stacked, or en echelon shoals which models indicate coalesced in response to sea level fluctuations.

Oolites form across the entire Kansas Pennsylvanian ramp. However, thicker, porous and permeable oolite deposits are commonly associated with the flanks or crests of small and large paleostructural highs. These structural highs may have influenced the intensity of early diagenesis and may have been responsible for development of good reservoir properties. Grain size variation, location on oolite buildups and interbedded carbonate mud (aquitards) influenced the nature and extent of diagenetic overprinting.

Oolite beds with porosity in excess of 8% porosity can reach several tens of feet in thickness, representing cross-bedded stacked shoals.

Subaerial exposure and meteoric water percolation led to cementation around the aragonite ooids and often dissolution of the ooids and variable development of matrix and vuggy porosity. Resulting oomoldic grainstones, the principal reservoir lithofacies, underwent variable degrees of early or later fracturing and crushing, providing connection between otherwise isolated oomolds.

Wireline logs signatures commonly exhibit low gamma ray, porosities ranging to greater than 30% and water saturations in the low teens with bulk volume water (BVW) as low as 0.03 based on an Archie cementation exponent of 2.

Petrophysics

Pay zones typically have BVW <0.05.

permeability (0.01-400 md) is principally controlled by :

Porosity
Oomold connectivity

Other variables that exert influence but are colinear with the above variables or are random include:

Oomold diameter
Oomold packing
Matrix properties
Matrix fracturing/crushing

Although permeability correlates with several of these variables, multivariate linear regression methods only improve prediction from a factor of 6.9X to 5.4X by inclusion of information concerning connectivity index, as measured on rock pieces. Few variable are utilized because of vaiable autocorrelation. The critical role that a single variable may play in controlling permeability hinders linear associations.

Non-parametric Regression analysis, utilizing the variables above, provides a mathematical tool capable of predicting permeability in all oolites studied within a factor of 5.3X, only slightly better than MLRA prediction.

Individual wells exhibit k-f trends with less variance than the overall trend, however, to date, the variables that cause the samples from these wells to exhibit different trends has not been identified. Development of models for specific fields, representing unique associations of conditions, allows the most accurate prediction

Irreducible water saturation (Siw) and residual oil saturation after waterflooding (Sor,w) are also strongly controlled by connectivity and correlate highly with permeability

The oomoldic reservoirs rocks studied provide insight into the interactions of rock fabric-architecture-diagenesis and better understanding of the universal influence of certain variables on oomoldic reservoir properties. Work is still needed to improve predictive tools.

References

Bohling, G.C., And Doveton, J.H., 1999, Manual for Kipling.xla: An Excel Add-in for Nonparametric Regression and Classification”, Kansas Geological Survey

French, J.A., and Watney, W.L., 1993, Stratigraphy and depositional setting of the lower Missourian (Pennsylvanian) Bethany Falls and Mound Valley limestones, analogues for age-equivalent ooid-grainstone reservoirs, Kansas: Kansas Geological Survey Bulletin, p. 27-39.

Watney, W.L., French, J.A., and Guy, W.J., 1996, Modeling of Petroleum Reservoirs in Pennsylvanian Strata of the Midcontinent, USA, in, Forester, A., and Merriam, D.F., eds., Spatial Modeling of Geologic Systems, Plenum Press, p. 43-77.

French, J.A., and Watney, W.L., 1990, Computer modeling of Mid-Continent cyclothems and its application in the prediction of hydrocarbon reservoirs: American Association of Petroleum Geologists, Bulletin, v. 74, p. 657. (Annual Meeting of AAPG-SEPM, San Francisco)

French, J.A. , and Watney, W.L., 1996, KANMOD A 2D carbonate modeling program: Numerical Experiments in Stratigraphy - An International Workshop, May 15-17, Lawrence, Kansas, p. 91-92.

Harris, P.M., Satellite images and description of study areas, in Harris, P.M., and Kowalik, W.S., eds., Satellite Images of Carbonate Depositional Settings, AAPG Methods in Exploration Series, No. 11, American Association of Petroleum Geologists, p. 29-139.

Feldman, H., Franseen, E.K., Miller, R., and Anderson, N., 1995, Drum Limestone project, p. 74-90+figures, in Watney, W.L., compiler, Depositional Sequence Analysis and Sedimentologic Modeling for Improved Prediction of Pennsylvanian Reservoirs, DOE Fossil Energy Final Report, DOE/BC/14434-13.

e-mail : webadmin@kgs.ku.edu
Last updated March 2004
http://www.kgs.ku.edu/PRS/Poster/2002/2002-48/P3-07.html