Vitenskapelig foredrag   1993

Egeland, Thore; Georgsen, Frode; Knarud, Ragnar; Omre, Henning




A stochastic model for the facies architecture of fluvial reservoirs is presented as well as a computer program called SISA designated to simulate from the model. While previous papers (Georgsen and Omre [7] and Georgsen et al. [6]) have offered complete mathematical presentations, the one at hand emphasizes the geological setting and intuitive descriptions of the model. The model includes totally 4 different types of facies; background, channel sand, sheetsplay sand and barriers. The translation of geological information into input to SISA is explained. Two kinds of examples are shown: One synthetic with a few number of facies-bodies and one real from a fluvial north sea field. The aim of these examples is to emphasize the following features; location and behaviour of facies-bodies; relation between different types of facies and conditioning on well-observations. SISA produces realizations of the models honoring well observations.


This paper presents a stochastic model for the facies architecture of fluvial reservoirs as well as describing an algorithm and a computer program called SISA to simulate from the model. The program computes permeability values and prepares input for a reservoir simulator by upscaling methods. However the paper emphasizes the modelling of the facies architecture. Similar ideas have been presented previously in Georgsen and Omre [7] and Georgsen el al. [6] with an entirely different focus. The mentioned papers offer complete mathematical descriptions of the procedure, the latter emphasizing the simulation algorithms. Our goal for the paper is to down weight technicalities trying to convey the main ideas to a broad audience. The geological setting is emphasized.

Heterogeneities in the characteristics of petroleum reservoirs have considerable impact on fluid flow, and hence on the recovery potential.