Lithofacies estimation by multi-resolution graph-based clustering of petrophysical well logs: a case study from one of the Persian Gulf’s gas fields, Iran

Located in the Persian Gulf, the gas field studied in this research is one of the largest gas fields in the world. Its gas-in-place is estimated to be about 14.2 trillion cubic meters while amount of its condensate-in-place might be around 18 billion barrels. This gas field has also an oil layer containing about 6 billion barrels of oil-in-place. In this study, Kangan and Dalan formations of this field were considered. Kangan formation has three main facies: clean carbonate facies, basic clay and shale facies, and evaporate carbonate facies. Dalan formation contains four facies: shore restricted carbonates, shore organic carbonates, carbonates of the open sea, in-shore carbonate-clastic. These two formations have gas & condensate fluids.  Since this field is a heterogeneous carbonate system, lithofacies characterization is the best solution for overcoming the problem of heterogeneity in determining the petrophysical properties of the reservoir rock, reservoir modeling and identifying producing zones. However, coring as the most robust method of lithofacies identification is very expensive, time consuming and limited to a few number of wells. Therefore, this study is focused on determining the lithofacies of the study formations from available well logs.
For this purpose, multi-resolution graph-based clustering (MRGC) technique which is a dot-pattern recognition method based on non-parametric K-nearest neighbor and graph data representation was applied to sonic, density, neutron porosity and gamma ray logs to define electrofacies similar to core-derived facies determined as eight distinct facies.The cluster of the MRGC method is defined from a model with a specific character associated with the group of lithofacies. Then, Kernel representative index was used to calculate the optimal number of clusters. Small facies groups were formed based on utilizing the neighboring index to determine a K-nearest neighbor attraction for each point. At last, final clusters were constructed by combining the small clusters which lead to identifying eight facies of this gas field from well logs of high accuracy.
When electrofacies of one of the wells is built basd on its lithofacies, its cross-plots will be plotted and the certainty of electrofacies with respect to lithofacies will be checked. If the model is acceptable, it is applied to the data from two other wells and their electrofacies will be obtained. For testing facies, the cross-plots of these two wells were also drawn and painted based on facies. If there are similar petrophysical properties for each facies, the model created in the wells without cores is confirmed. MRGC is a fast method that allows the geologist or petrophysist to analyze and test different combinations of data in a short amount of time. It is also not limited by the dimensions of the data and number of the clusters. The method used in this study has obviated the need for extensive coring in this field which caused saving large amounts of money and time; and it can help to optimize the determination of new well locations and optimum pay zones.