A Compositional Rescaled Exponential Model for Multiphase-Production-Performance Analysis of Boundary-Dominated Gas/Condensate Reservoirs

Will, Ryan (Pennsylvania State University) | Sun, Qian (Pennsylvania State University) | Ayala, Luis F. (Pennsylvania State University)



Hydrocarbon-reservoir-performance forecasting is an integral component of the resource-development chain and is typically accomplished using reservoir modeling, by means of either numerical or analytical methods. Although complex numerical models provide rigorous means of capturing and predicting reservoir behavior, reservoir engineers also rely on simpler analytical models to analyze well performance and estimate reserves when uncertainties exist. Arps (1945) empirically demonstrated that certain reservoirs might decline according to simple, exponential, hyperbolic, or harmonic relationships; such behavior, however, does not extend to more-complex scenarios, such as multiphase-reservoir depletion. Because of this limitation, an important research area for many years has been to transform the equations governing flow through porous media in such a way as to express complex reservoir performance in terms of closed analytical forms. In this work, we demonstrate that rigorous compositional analysis can be coupled with analytical well-performance estimations for reservoirs with complex fluid systems, and that the molar decline of individual hydrocarbon-fluid fractions can be expressed in terms of rescaled exponential equations for well-performance analysis. This work demonstrates that, by the introduction of a new partial-pseudopressure variable, it is possible to predict the decline behavior of individual fluid constituents of a variety of gas/condensate-reservoir systems characterized by widely varying richness and complex multiphase-flow scenarios. A new four-region-flow model is proposed and validated to implement gas/condensate-deliverability calculations at late times during variable-bottomhole-pressure (BHP) production. Five case studies are presented to support each of the model capabilities stated previously and to validate the use of liquid-analog rescaled exponentials for the prediction of production-decline behavior for each of the hydrocarbon species.