Application of multivariate statistics in the analysis of hydrocarbon producer well, Rio grande do norte/Brazil

Abstract

The identification of lithologic types along a well is one of the important contributions of geophysics to geological studies in the petroleum industry. However, it is not a trivial task, since the measurements made in the well indicate the changes of the physical properties and not necessarily the lithological variations. Therefore, this work has as objective to make the association between the variations of the physical properties and the lithological types in the well. For this, the gamma rays (GR), microresistivity (MSFL), neutron porosity (NPHI) and sonic (Dt) profiles were used. These are the data obtained from wellbore profiling inthe hydrocarbon-producing field on the Potiguar Basin. In order to analyze the data obtained through well profiling, the multivariate statistical techniques involving cluster analysis and main components (PCA) were applied with the help of the PAST3 software. The results obtained allow to relate the well profile data with lithological types, and associate them at intervals of sealant or reservoir rocks in the wells. Where the upper portion of the well is the sealant rock zone S10 and reservoir R10, in themiddle the sealant rock zone S14 and reservoir R14; and sealant rock zone S30 and reservoir R30 in the lower portion. Applying the classical multivariate analysis to the data matrix by zone, they were separated into two groups, which in the well profile are associated with the sealing and reservoir rocks. In the principal components analysis biplot correlation, the rocks were also separated into two groups. This result obtained through the association of well profiles with rocks is important, since it assists inthe management activities of the reservoirs in the hydrocarbon producing fields, since it serves as an orientation in the decision making to complete the potentially producing intervals in the geographic space of the field.