| Even though cubic equations of state do a good job of characterizing fluid phase behavior, currently there is no simple method available that either looks at the effects of individual components on the phase behavior of a complex petroleum mixture or that can help us understand how different components would behave relative to each other as the composition of the mixture changes.; Here a novel approach to look at these problems is presented. This approach is based on a two-step minimization of Gibbs energy which reveals some new characteristics of two-parameter equations of state. The two-step minimization approach provides simple and linear expressions for equilibrium constants and the variables in these expressions have some special properties. These features have been exploited in this work, to determine, without detailed phase behavior computations, how K-values of different components are ordered among themselves and how components of different chemical types influence the phase behavior of petroleum mixtures.; It is shown that there is an additional loss of physics, besides the expected loss of information, when large systems such as crude oils are characterized in terms of just a few pseudocomponents. It is also shown that the presence of different chemical types in petroleum mixtures influences the phase envelop in such a manner that the phase behavior may not be adequately represented even by a single carbon number representation.; With these issues in mind, studies were undertaken on both synthetic and real petroleum mixtures. While a reasonable phase behavior match (under static conditions) for the synthetic mixtures studied could be achieved by just a few pseudocomponents, the displacement studies on these mixtures revealed that there is considerable spreading of saturation fronts for mixtures made up of a large number of components. Such behavior is not captured accurately by pseudocomponent representations and, as a result, there may be significant errors in recoveries even when a reasonable match with the PVT behavior of the full representation has been achieved. The sensitivity of the displacement behavior to the characterization also brings out the importance of displacement experiments to characterize petroleum mixtures adequately. |
