Read more
The detailed geological models of hydrocarbon reservoirs require a process of upscaling (averaging) to flow models to reduce computational costs and to allow multiple realizations be used for optimal operation design under uncertainty. A drawback of upscaling, however, is the loss of information and an increase of inaccuracy in predictive capacity of complex subsurface processes. This project examines two approaches to tackle the problem. First method is to use naturally multiscale wavelets for operator upscaling, and the second approach is to devise a post-processing step of downscaling of simulation results for upscaled models by use of special functions and operators. It was shown that the first approach encounter shortcomings and deficiencies due to the definitions of the wavelets. However the second technique can drastically reduce conventional errors of upscaling due to representation of simulation properties at coarse scale resolution. The application was examined for enhanced oil recovery processes such as thermal, polymer and surfactant flooding demonstrating the versatility and applicability of the proposed technique.
About the author
Masoud Babaei is currently a lecturer in Petroleum Engineering at the School of Chemical Engineering and Analytical Science of the University of Manchester, UK. Before starting his PhD at Imperial College, London, he obtained an MSc degree in Reservoir Engineering from Sharif University of Technology, Tehran, Iran in 2008.
