CORRELATION AND CLUSTER ANALYSIS OF PRINCIPAL GEOLOGICAL PARAMETERS OF GIANT HYDROCARBON FIELDS

We consider the problem of extraction of relevant geological parameters, related to the formation conditions of major hydrocarbon (HC) fields and estimation of their resource potential. The identification of most informative sets of geological parameters and revealing correlation dependencies between them is of crucial importance for HC exploration. Experience has proven, that 10–15 geological parameters have high information impact and can be used to estimate HC potential in most sedimentary basins. These parameters characterize sedimentation conditions, tectonic environment, organic matter transformation, location of the oil and gas accumulation areas etc. Depending on the geological conditions, relations of various nature can exist between these parameters, which can be characterized by pairwise correlation coefficients. We present the pilot study of the pairwise correlation coefficients between the metric subset of these parameters, calculated from data set of major world basins, including Azerbaijan giant fields. Several alternative graphic models are proposed (including trend curves, correlograms, chord diagrams and clustered dendrograms) to visualize the dependencies and to derive conclusions as to the impact of these parameters on the probability assessment of the HC field formation.