Upscaled Pseudo-porosity Extracted from Impedance Volume Can Effectively Predict Gas Fingering in a Dynamic Model – a Gas Injection Case Study

Abstract

Abstract Gas injection in an unsaturated carbonate reservoir which is located in south-west of Iran started in 2009, and based on the original dynamic model the gas breakthrough would not have happened earlier than 2017. However, the gas oil ratio (GOR) in some wells increased rapidly before the anticipated time; also the real field data shows different behaviors of gas breakthrough in different wells. The 3D Geocellular grid obtained from 3D seismic cubes was applied to dynamic model, and by making the size of the cells smaller, the heterogeneity was identified as the most important factor in modeling and monitoring the gas movement from the injected wells towards the production wells. Based on the new upscaling, two dynamic models were constructed. The original model had 267,000 grid cells and the new ones have 1,120,000 and 2,400,000 grid cells respectively. Running the new models illustrated the gas breakout from some wells mostly through one layer of the reservoir. Since effective porosity was estimated by collocated cokrigging of well data with seismic porosity, vuggy porosity distribution in this layer is correctly predicted, given that a karstic volume has been detected in some part of the reservoir which its boundary has been defined by seismic data and the cross-plots of neutron porosity with acoustic impedence validate the reliability of this karstic zone. The results show various types of vuggy porosity that mostly governs the productivity of the wells in this reservoir. As a result of this study, both dynamic and static models modified based on pseudo-porosity derived from seismic cubes as the best known distributed heterogeneity in this field to predict the gas movement in the reservoir. Based on this modeling, new maximum efficient rates were defined for wells and field development strategy was reconsidered.