History matching for characterization of reservoir facies by stochastic inversion methodologies
- Start date: 2011
- End date: 2013
- Funded by: Eni S.p.A.
- Topic: Data assimilation and stochastic inverse modeling for the characterization of reservoirs in the presence of uncertain distribution of hydrofacies and hydraulic properties.
In reservoir engineering there is often the need to represent geological objects which are associated with complex features (i.e., channels). In this context, traditional methods based on two-point variogram statistics do not provide the complete details needed to reproduce fields whose internal architecture is based on multiple juxtaposed geological facies. This motivates us to employ a method based on multiple-point geostatistics, where a large number of points can be used jointly to estimate the conditional probability of observing a facies at a given point of the domain. Our work is largely based on the methodology proposed by Stien and Kolbjornsen (2011), where a Markov Mesh Model based on a generalized linear model is employed to model the facies distribution.
