Characterization of clastic sedimentary enviroments by clustering algorithm and several statistical approaches — case study, Sava Depression in Northern Croatia

Abstract

This study demonstrates a method to identify and characterize some facies of turbiditic depositional environments. The study area is a hydrocarbon field in the Sava Depression (Northern Croatia). Its Upper Miocene reservoirs have been proved to represent a lacustrine turbidite system. In the workflow, first an unsupervised neural network was applied as clustering method for two sandstone reservoirs. The elements of the input vectors were the basic petrophysical parameters. In the second step autocorrelation surfaces were used to reveal the hidden anisotropy of the grid. This anisotropy is supposed to identify the main continuity directions in the geometrical analyses of sandstone bodies. Finally, in the description of clusters several parametric and nonparametric statistics were used to characterize the identified facies. Obtained results correspond to the previously published interpretation of those reservoir facies.

Cluster defined sedimentary elements of deep-water clastic depositional systems and their 3D spatial visualization using parametrization: a case study from the Pannonian-basin

“…This clustering method was ap plied because NNC was used in similar problems to characteri zation of clastic sedimentary environments (e.g. HORVáTH, 2015;HORVáTH & MALvIć, 2013).…”