RAPID AND ROBUST PHASE-EQUILIBRIUM CALCULATION TO MODEL FLUIDS IN RESERVOIR AND SURFACE PROCESSING

Citation
F. Gozalpour et al., RAPID AND ROBUST PHASE-EQUILIBRIUM CALCULATION TO MODEL FLUIDS IN RESERVOIR AND SURFACE PROCESSING, Chemical engineering research & design, 76(A5), 1998, pp. 594-603
Citations number
25
Categorie Soggetti
Engineering, Chemical
ISSN journal
02638762
Volume
76
Issue
A5
Year of publication
1998
Pages
594 - 603
Database
ISI
SICI code
0263-8762(1998)76:A5<594:RARPCT>2.0.ZU;2-7
Abstract
Compositional simulation is widely used in the petroleum industry to p redict the phase behaviour of reservoir fluids within the reservoir, f low lines and process facilities. Computational time is an important f actor in compositional simulation where CPU time exponentially increas es with the number of components. It has been shown that by omitting b inary interaction parameters (BIPs) from the mixing rules of the equat ion of state (EOS), phase split calculations can be carried out much f aster than when using conventional methods, when a large number of com ponents is used to describe the reservoir fluid. However, using the ra pid flash calculation method, the phase behaviour calculation may fail to converge, particularly for gas condensate systems. Omitting BIPs m ay also deteriorate the predictive capability of the EOS for some flui ds. In this paper, the original rapid flash calculation method is modi fied to improve its convergence for gas condensate systems. To compens ate for the omitted binary interaction parameters, the temperature dep endency of the attractive term (or) in EOS has been modified to improv e its phase behaviour prediction over a wide range of temperatures. A number of binary systems were used to determine the or function for su percritical compounds. This improved the EOS predictions for systems w ith high concentrations of supercritical compounds. The objective is t o perform the phase equilibrium calculation as rapidly as possible wit h a large number of components in the reservoir without compromising o n accuracy of predictions. Then, using the same number of components, detailed compositional analysis can be used to design efficient proces s facilities.