Using New Reservoir Characterization Technologies on a Giant Old Gas Field (Hugoton Field, Kansas)

Terrilyn M. Olson, Keith Thompson, Kvk Prasad, and Steve Boughton

New technologies can profitably be applied to mature fields. Neural networks, geostatistics, new wireline logging tools, and 3-D visualization/computation have all added value to the characterization of Kansas Hugoton Field. This field was discovered in the 1920's; cumulative production has reached approximately 26 TCFG, and years of productive life remain. Despite this maturity, controls on gas distribution and water production have not been well understood.

Once the geologic framework of the reservoir is understood, various technologies can be employed to solve problems of reservoir characterization, especially in a system as complex as the multilayer, mixed-lithology reservoir of the Permian Chase Group in Hugoton. Neural networks provided a better estimation of both porosity and permeability from logs when compared with core data than more conventional methods. Use of geostatistics resulted in more realistic porosity distributions than those from interpolation, by preserving heterogeneity and allowing constraint of imperfect log determinations by core data. Geostatistics also allows for quantification of uncertainty, which is shown by a range of possible pore volumes.

3-D visualization makes quick quality control of data possible, and promotes efforts to test sensitivities and cutoffs and to communicate results. Computation of reservoir parameters (e.g. water saturation) in 3-D alleviates the averaging problems attendant on such computations in 2-D. These techniques, plus new log measurements such as nuclear magnetic resonance and magnetic pulsed induction, have aided our characterization of this reservoir and increased our understanding of, and ability to manage, gas and water production.

AAPG Search and Discover Article #91019©1996 AAPG Convention and Exhibition 19-22 May 1996, San Diego, California