Optimization of the Markov chain for lithofacies modeling: an Iranian oil field

Reconnaissance and interpretation of underground heterogeneity, particularly lithofacies, always plays an important role in evaluation and management of hydrocarbon resources. Between various methods presented for modeling discrete characteristics of hydrocarbon reservoirs such as lithofacies, one with a more proper conformity with actual condition of reservoir facies is of great advantage. Formed on the basis of probability and presenting transition matrix, the Markov method is widely applied as a powerful tool for modeling the facies. In the present study, first, the method is introduced in details; then, in order to optimize it for conditions with insufficient well data, two suggestions are made based on changing the motion direction of the chain and increasing the conditional boundary in simulation procedure. The case study is a 12-km-long 110-m-thick section of anhydrite and three major members of the Asmari Formation from an oil field in southwest Iran. This section is modeled through Markov classical procedure, changing chain motion direction and finally adding one seismic horizon as another conditional boundary. The models set indicated that on the basis of using the data from three wells and seven seismic horizons, the best result with 86 % accuracy is for the state of using two conditional boundaries.