Estudo das características porosas das rochas reservatório do Pré-Sal através do processamento digital de imagens

Abstract

Characterizing the porosity of hydrocarbon reservoirs is a major challenge in the oil industry, since understanding this important petrophysical property determines the quality of a hydrocarbon reservoir. There are many methods that aim to calculate the porosity in sedimentary rocks, and these can be direct (obtained directly from the rock sample, such as Sedimentary Petrography and Porosimeter) or indirect (such as Geophysical Profiling of Wells and Nuclear Magnetic Resonance (NMR)). A common method of estimating porosity is the porosity modal count, which can be performed according to transverse lines perpendicular to the lamination or grain orientation performed with the aid of an optical microscope, from a petrographic slide. In this way, an estimate of the porosity is performed directly on the petrographic slides, the success and agility of the procedure depend on the experience of the petrographer, the cost of this method is the time spent. In addition, the blades are difficult to access. When requesting a well at the Agência Nacional de Petróleo (ANP), this material is not available. However, photographs of the thin sections are made available along with other well data, such as geophysical profile curves. This is public data provided when requesting well data. Thus, the object of this study are the photographs of petrographic thin sections of the carbonate reservoirs of the Barra Velha and Itapema formations of Campo de Búzios in the Santos Basin, in addition to data from the neutron and magnetic resonance profiles, aiming at the evaluation of the porosity of the rocks of form indirect. In this work, five methods were used to assess porosity, three traditional methods: Porosimeter, Neutron Profile, Nuclear Magnetic Resonance Profile, and two methods of digital image processing using information from the R and B channels of the RGB standard (red, green and blue). The algorithms were developed in python - the only direct method being the porosimeter, which was considered the default porosity value. The porosity value obtained by the method that uses the R channel was the closest to the value considered standard. The average porosity percentage obtained by the R Channel corresponds to 94.53% of that obtained by the porosimeter, for the Barra Velha Formation, and 88.16% for the Itapema Formation.