Modeling of performance behavior in gas condensate reservoirs using a variable mobility concept

Abstract

The proposed work provides a concept for predicting well performance behavior in a gas condensate reservoir using an empirical model for gas mobility. The proposed model predicts the behavior of the gas permeability (or mobility) function in the reservoir as condensate evolves and the gas permeability is reduced in the near-well region due to the "condensate bank". The proposed model is based on observations of simulated reservoir performance and predicts the behavior of the gas permeability over time and radial distance. This model is given by:

The proposed concept has potential applications in the development of a pressure-time-radius solution for gas condensate reservoirs experiencing this type of mobility behavior. We recognize that the proposed concept (i.e., a radially-varying gas permeability) is oversimplified, in particular, it ignores the diffusive effects of the condensate (i.e., the viscosity-compressibility behavior). However, we have effectively validated the proposed model using literature results derived from numerical simulation. This new solution is presented graphically in the form of "type curves." We propose that the "time" form of this solution be used for applications in well test analysis. Previous developments used for the analysis of well test data from gas condensate reservoirs consider the radial composite reservoir model, which utilizes a "step change" in permeability at some radial distance away from the wellbore. Using our proposed solution we can visualize the effect of the varying gas permeability in time and radius (a suite of (dimensionless) radius and time format plots are provided). In short, we can visualize the evolution of the condensate zone as it evolves in time and radial distance. A limitation is the simplified form of the kg profile as a function of radius and time - as well as the dependence/appropriateness of the α-parameter. While we suspect that the α-parameter represents the influence of both fluid and rock properties, we do not examine how such properties can be used to calculate the α-parameter.