Identifying productive zones of the Sarvak formation by integrating outputs of different classification methods

Abstract

The Sarvak formation is the second major carbonate reservoir in Iran. There are several geological, petrophysical and geophysical investigations which have been carried out on this important reservoir. In this paper, the Sarvak is studied in order to determine productive zones. First, four different methods are used to identify producing intervals from well log data and well test results. Then, final zoning is generated by integrating the outputs of these four methods. One of them is the conventional cut-off-based method; the other three are based on flow equation, Bayesian and fuzzy theories. Thereafter, by considering the classification correctness rate of each classifier in each well and technique of majority voting, a unique zoning for the Sarvak formation is presented. Based on the final zoning, the whole Sarvak interval is divided into seven zones. Three of them are classified as oil-producing zones, two of them cannot be classified as conventionally producing zones, and the remaining two are water producing. Zone 2 not only has the highest production rate, but is also the most homogeneous zone among the productive zones. The novelty of this research is using well test results in defining productive classes, which improves the certainty of classification in comparison with previous works that were based on core analysis and log data.