Integration of Core and Log Information To Improve the Representation of Small/Medium-Scale Heterogeneity

Reservoir heterogeneities exist at several length scales. Reservoir properties will depend on the scale that we are analysing on the reservoir (from core to grid block). Facies are normally identified in cores and classified based on microscopic and small scale features such as lithology, grain size, texture, sedimentary structure, colour and fossil content. However, facies information is only available in a few cored wells or in wells with high vertical resolution logging. Conventional logging is usually available in all wells, but it only provides information at a medium scale. This information can be calibrated with geological facies to generate electrofacies (also called petrofacies), from which reservoir property distributions can be predicted using 3-D modelling techniques. In this paper, core and log information is integrated to create a facies and electrofacies database. Small-scale geological facies are calibrated with logs and used to generate electrofacies by using statistical analysis. More detailed geological models are then generated using electrofacies data. The impact of using an electrofacies scehme over the traditional log-derived facies scheme during stochastic modelling of reservoir facies has been demonstrated. Modelling performance prediction using both methods is analysed for the different stochastic realisations. The use of * Currently with PDVSA S.A. Venezuela electrofacies schemes has been proved to give more accurate prediction of reservoir heterogeneity. Introduction Proper modelling of reservoir heterogeneity is one of the main goals during the reservoir characterisation process. Recovery efficiencies are in many cases lower than initial predictions due to the effect of reservoir heterogeneities that can cause complicated fluid flow patterns and therefore regions of bypassed oil. To properly represent reservoir heterogeneity, it is necessary to collect and integrate all available reservoir data. However, full data integration is difficult to achieve due to time and budget limitations or because of intrinsic data characteristics. Field data are usually coming from different sources (cores, logs, seismic, etc.) and at different length scales (microscopic, small, medium and gigascopic scale) making this integration process particularly difficult (Fig. 1).