|
Description:
|
Determining the distribution of porosity and permeability is one of the main challenges in carbonate petroleum reservoir characterization and requires a thorough understanding of pore type and origin , as well as their spatial distributions . Conventional studies of carbonate reservoirs require interpretation and analysis of cores to understand porosity . This study investigates the use of NMR logs in the determination of pore type and origin . This study is based on the analysis of both thin section petrographic and NMR data from a single well that cored the Lower Cretaceous (Aptian ) shelf carbonates belonging to the Shuaiba Formation of the Middle East . Photographs of thin sections were used to determine pore type and origin according to Ahr's genetic classification of carbonate porosity . Descriptive statistics and modeling were used to analyze the NMR T2relaxation time distributions . Descriptive statistical analyses included estimating arithmetic average , standard deviation , skewness , median , mode and 90th percentile . T2modeling was performed by fitting multiple log -normal distributions to the measured T2distribution . Data from thin section petrography and from NMR measurements were then compared using conditional probabilities . As expected , thin section analysis revealed the predominance of mud -supported fabrics and micropores between matrix grains Vugs and dissolved rudistid fragments account for most of the macro porosity . Descriptive statistics showed that the mode and th percentile of the T2distribution had the greatest power to discriminate pores by origin . The first principal component (PC1 ) of the mode -90th percentile system was then used to compute the probabilities of having each pore origin , knowing that PC1 belongs to a given interval . Results were good , with each origin being predictable within a certain range of PC1 . Decomposition of the T2distributions was performed using up to 3 log -normal component distributions . Samples of different pore origin behaved distinctively . Depositional porosity showed no increase in fit quality with increasing number of distributions whereas facies selective and diagenetic porosity did , with diagenetic porosity showing the greatest increase . |