Abstract:

Development of oil and gas field with carbonate reservoirs is quite challenging due to its high heterogeneity (diagenesis, fractures, wetting etc.). There is no universal rock-typing methodology in order to define homogeneous groups of rocks in petrophysics, geology and diagenesis. This work proposed integrated approach of rock-typing by using clustering algorithms. Based on the data of one of the carbonate reservoir 3 rock types were defined. For further development it is better to try different clustering algorithms and use data from several wells simultaneously.

Keywords: machine learning;, rock-typing, Carbonate reservoir