基于Markov链模型的相控随机建模

Markov链模型在储层随机建模中的作用越来越受到关注,但其多用于类型属性(岩相、沉积相、沉积亚相等)的模拟,对于连续型属性(孔隙度、渗透率、含油气饱和度等)的模拟还比较困难.本文提出用Markov链模型相控建模方法模拟连续型属性的思路,即首先用Markov链模型模拟出类型属性,其次在类型属性约束下模拟出连续型属性,从而解决连续型属性不能产生突变边界的问题.最后应用此方法进行了模拟实验,模拟结果显示不同岩相中孔隙度差异较大,而同种岩相中孔隙度变化较小,证明了此方法的可靠性和适用性.

Facies controlled stochastic modeling based on Markov chain models

The effect of Markov chain models in reservoir modeling is much concerned recently, but models are usually applied to simulating categorical attributes, such as lithofacies, sedimentary facies and subfacies, and very few to continuous attributes, such as porosity, permeability and oil saturation. This study propose the idea about using facies controlled method in Markov chain models , that is ,using Markov chain models to get lithofacies and then simulate the continuous attributes under the constrict of lithofacies, which can deal with the problem of no definite boundary in simulation results of continuous attributes. The simulation tests show that distribution of porosity has an apparent variance in different lithofacies but slight change in each lithofacies,which verify this method is reliable and effective.