Abstract: Simulating Geological Uncertainty with Imprecise Data for Groundwater Flow and Advective Transport Modeling

Sean A. McKenna, Eileen P. Poeter

A known cross-section composed of a heterogeneous mixture of three hydrofacies is sampled along vertical lines representing ten wells. These wells provide conditioning data for stochastic geostatistical simulations. Steady state flow and advective transport are modeled across each resulting realization to evaluate effective hydraulic conductivity, first arrival time and duration of particle arrivals. Further ensembles of realizations are produced with both the original ten wells of conditioning data and additional soft data in the form of imprecise estimates of the hydrofacies at every location within the simulated domain. Use of soft data decreases the ensemble variances with respect to the first arrival times, bulk hydraulic conductivity of the domain, and duration of the particle a rivals, relative to the ensemble variance resulting from only ten wells of hard conditioning data. The accuracy of the resulting distributions is influenced by biases in the conditioning data.

AAPG Search and Discovery Article #90986©1994 AAPG Annual Convention, Denver, Colorado, June 12-15, 1994