Upscaling of water-based EOR methods

This project is part of the research area of reservoir simulation, which in simplified terms is the use of computers to simulate the flow of oil, gas, water and other substances through a reservoir.

Creating a computer representation of an oil reservoir is difficult partly because of the wide range of physical scales. Whereas the flow through the rock takes place in tiny pores the size of a micro-meter, the distance between two wells can be more than a kilometer. To include all this information into a full size reservoir model is not even close to possible, and so the physical properties from the smallest sizes must be upscaled to larger sizes, which is a form of averaging the properties. A large part of this project has been considering upscaling of oil and water properties [2,5].

As many oil reservoirs today are mature, enhanced oil recovery has become increasingly important, where different techniques are used to increase the production. Polymer injection is one such technique, in which polymer is added to water andinjected into the reservoir. Simulations including polymer have been studied in this project, and one of the issues considered is upscaling of polymer properties [1].

Another method of tackling the issue of scales is multiscale methods. In such methods, a solution is first estimated on a less detailed representation of the reservoir, which is fast to calculate. Then, this solution is projected onto a detailed reservoir representation. By repeatedly going back and fourth between these two models during a simulation, one can obtain an accurate and detailed solution much faster than by just considering the detailed model. This project has included polymer into simulations using a recent multiscale method [3].

Also studied in this project is a specific physical property of polymer flow. Because the polymer molecules are relatively large, parts of the small pores in the oil reservoir are not accessible to the polymer, even though water and oil may be able to flow in this region. How to best create a mathematical model of this phenomenon has been studied in detail [4].

 - - -

The technical director for the project was Prof. Knut-Andreas Lie (SINTEF) and the research was primarily conducted at SINTEF's offices in Oslo.

Publications:
1. S. T. Hilden, K.A. Lie, X. Raynaud. Steady state upscaling of polymerflooding. ECMOR XIV, Catania, Sicily, Italy, 8-11 September 2014.

2. In review: S. T. Hilden, C. F. Berg. Rate dependency and upscaling of an unsteady flooding process, Transport in Porous Media.

3. In review: S. T. Hilden, O. Møyner, K.A. Lie, K. Bao - Multiscale Simulationof Polymer Flooding with Shear Effects, Journal of Computational Physics.

4. Unpublished: S. T. Hilden, H. M. Nilsen, X. Raynaud - A Well-Posed Model ofInaccessible Pore Volume for Polymer.

5. Unpublished: S. T. Hilden, C. F. Berg - Steady-State Two-Phase Upscaling of Realistic Field Models.