An integration of conventional rock typing methods and fuzzy Cmeans clustering algorithm: a case study

Determination of rock types in hydrocarbon reservoirs (whether carbonate or clastic) is veryimportant. Therefore, various methods for identifying rock types have been introduced and developedin recent years. In this study, we have applied some conventional rock typing methods in combinationwith a clustering machine learning algorithm in order to optimize the number of clusters and eventuallygroup input data in accurate rock types. The results show that flow zone indicator (FZI) methodintegrated with the fuzzy C-means (FCM) algorithm was the best approach for detecting rock types withthe range of correlation coefficient (R۲) from ۰.۸۱ to ۰.۸۷.

Keywords:

rock types , clustering machine learning algorithm , flow zone indicator , fuzzy C-means algorithm