Geostatistical Integration of Outcrop and Geophysical Data to Describe Diagenetic Heterogeneities for Flow Modeling, Frontier Formation, Wyoming

Hong Tang
Louisiana State University, Baton Rouge, LA

Recovery programs for deltaic reservoirs should consider depositional and diagenetic heterogeneities, and their effects on flow. A study of the Wall Creek Member of the Cretaceous Frontier Formation, Powder River Basin, Wyoming examines the spatial distribution and effects on flow of diagenetic calcite concretions in near-surface and exposed fluvial-deltaic deposits.

The dataset includes interpreted outcrop photomosaics, permeametry, gamma logs, ground-penetrating-radar (GPR) surveys, and cores. GPR data are calibrated to core data for concretion modeling. However, imperfect correlation of GPR attributes and core observations causes uncertainty in concretion modeling.

Two new algorithms (sequential indicator Bayesian simulation and truncated Gaussian Bayesian simulation) are used to model the nonstationary spatial distribution of concretions. Indicator semivariograms computed from outcrops can be converted to equivalent Gaussian semivariograms. Observed spatial trends in concretion occurrence are incorporated using proportion curves. Instead of using deterministic methods, Borehole and GPR data are calibrated using logistic regression and cluster analysis. These calibration models update the truncated Gaussian estimates with Bayes' rule.

Two-dimensional flow simulations validate the variogram and trend models. A Student t-test indicates that stochastic models using outcrop statistics are not different from the flow model using direct outcrop observations (with 95 percent confidence). Concretions decrease average permeability (63.44 percent) and recovery (5.41 percent) and decrease breakthrough time (17.65 percent). Three-dimensional geospatial models integrating GPR data and associated flow simulations indicate that even weakly informative GPR data significantly affect concretion models and flow behavior.

KEY WORDS: diagenesis, calcite concretion, geostatistical modeling, flow simulation, nonstationary, ground penetrating radar, Cretaceous, heterogeneity

AAPG Search and Discovery Article #90039©2005 AAPG Calgary, Alberta, June 16-19, 2005