Accounting for Local Probabilities in Stochastic Modeling of Facies Data

Publisher Website
Google Scholar
Abstract
This paper proposes an annealing-type algorithm for generating multiple realizations of facies distributions that: (i) honor the facies data at well locations, (ii) match global proportions of facies, (iii) reproduce spatial continuity and transition probabilities between facies, (iiii) account for prior information such as trend or gradient in facies proportions or the absence of a facies in any particular area. A case-study shows the realizations to reproduce important features of the reference categorical map. A waterflood simulation indicates that the algorithm allows a reduction of the space of uncertainty of flow response values.

This publication has 0 references indexed in Scilit: