Determinação de porosidade e tamanho dos poros a partir de imagens petrográfica usando dados de reservatório de gás, Offshore da bacia de Moçambique

Abstract

Digital image analysis has been an alternative method for characterizing reservoir rocks. In this study, the image segmentation technique was applied to determine porosity and pore size distribution in MATLAB® and ImageJ®. The segmentation operation used in this study is the thresholding or binari-zation that was able to find the threshold or boundaries of each region that best separates the peaks. For this, two methods were used namely: Manual (based on histogram inspection) and Automatic (Otsu Method). The combination of these methods allowed the porosity and pore size distribution to be de-termined for four images of two offshore J and K reservoirs in the Mozambique basin. The estimated porosity from image analysis for reservoir J was 26% compared to 28.7% experimental with a -3% error, for the reservoir K was 23% compared to 26.7% with an error of -4%. The analogy for this pore distribution is to say that they are storage pores with predominance of medium-sized pores of approxi-mately 53 μm and 48 μm for the reservoir J and approximately 57 μm and 51 μm in diameter for the reservoir K.