Theory of Gas Injection: Interaction of Phase Behavior and Flow

Abstract:

The theory of gas injection processes is a central element required to understand how components move and partition in the reservoir as one fluid is displacing another (i.e., gas is displacing oil). There is significant amount of work done in the area of interaction of phase-behavior and flow in multiphase flow conditions. We would like to present how the theory of gas injection is used in the industry to understand/design reservoir processes in various ways.

The tools that are developed for the theory of gas injection originates from the fractional flow theory, as the first solution proposed by Buckley-Leveret in 1940’s, for water displacing oil in porous media. After 1960’s more and more complex/coupled equations were solved using the initial concept(s) developed by Buckley-Leverett, and then Welge et al. and others. However, the systematic use of the fractional flow theory for coupled set of equations that involves phase relationships (EOS) and phase appearance and disappearance was mainly due to the theory developed by Helfferich in early 80’s (in petroleum literature) using method of characteristics primarily for gas injection process and later on by the systematic work done by Orr and his co-researchers during the last two decades. In this talk, we will present various cases that use and extend the theory developed by Helfferich and others (Orr et al., Lake et al. etc.).

The review of various injection systems reveals that displacement in porous media has commonalities that can be represented with a unified theory for a class of problems originating from the theory of gas injection (which is in a way generalized Buckley-Leverett problem).

The outcome of these solutions can be used for (and are not limited to):

1) Benchmark solutions for reservoir simulators (to quantify numerical dispersion, test numerical algorithms)

2) Streamline simulators

3) Design of laboratory experiments and their use (to invert the results)

4) Conceptual learning and to investigate the microscopic displacement efficiency.

5) MMP/MME for various processes (not only for gas injection)