OIL-BEARING RESERVOIR OPTIMAL CHARACTERISTICS DETERMINATION FOR CARBON DIOXIDE INJECTION BASED ON A COMPLEX CLUSTER ANALYSIS OF DEVELOPMENT FACILITIES

Carbon dioxide injection to enhance oil recovery by improving the physical and chemical properties of the reservoir environment is one of the effective methods of increasing oil production, and every year it is more and more actively used in the oil industry around the world. Carbon dioxide entering the formation increases the pressure and improves the rheological properties of formation fluids, which contributes to the further oil displacement from the pore space. An important advantage of carbon dioxide injection for enhanced oil recovery is the possibility of its geological storage (utilization) in reservoir rocks, which meets the goals of decarbonization – achieving carbon neutrality of the global and Russian economy. The scientific article presents the methodology of objects-polygon selection for carbon dioxide injection to enhance oil recovery. The main objective of the research was to summarize and systematize the geological and physical characteristics of reservoir systems represented by carbonate reservoirs of the fields of the Volga-Ural Oil and Gas Province (deposits of Bashkortostan, Tatarstan and Perm Krai) and their subsequent clustering according to a set of attributive geological and geophysical characteristics using complex geological and statistical analysis. Within the framework of the study the object clustering procedure was performed using the principal component method (PCM), on the basis of which five groups of productive objects with characteristic features were identified. Various machine learning algorithms and types of artificial neural networks (ANN) were applied to process a large volume of specialized data, which significantly improved the accuracy of analysis and the quality of data classification. Four key geological and physical parameters were identified as a result of comprehensive neural network analysis. Clustering of development objects of carbonate reservoirs of oil fields of Perm Krai, Republic of Bashkortostan and Republic of Tatarstan, as well as their subsequent identification using PCM and ANN allowed to establish a set of criteria ensuring the success of carbon dioxide injection technology. In each group a typical object-analog, which satisfies the applicability criteria of the carbon dioxide injection method, was selected. As a result of the research, a database was formed for the operational selection of objects-analogs for carbon dioxide injection to enhance oil recovery at depleted oil fields.