Development of equations of state for gas condensates for compositional petroleum reservoir simulation

In this paper, we describe our experience in the development of an equation -of-state (EOS) model for compositional reservoir simulation studies of gas condensates, especially those near the critical point. We focused on the f ollowing tasks: (1) heavy-end pseudoization; (2) selection of the measured data to match and the EOS parameters to adjust; and (3) evaluation of model predictive capability. We compared several methods for each of the above tasks for a number of gas condensates, one of which is a near-critical gas condensate showing experi mentally significant compositional change with depth. The combination of th e different methods offers up to 44 different heavy-end descriptions. The P eng-Robinson (PR) EOS was then used with these fluid descriptions to predic t the phase behavior of gas condensates. We observed that none of these 44 heavy-end descriptions is significantly better than the others, although th ey give very different properties of pseudocomponents. Tuning EOS-parameter s to match experimental data is necessary in order to well describe the pha se behavior of gas condensates. The matched experimental data should includ e both phase amount and volume information. The initial guesses significant ly affect the quality of the matches; those initial guesses using the Gauss ian quadrature method for lumping lead in general to satisfactory matches. Tuning molecular weights of pseudocomponents (critical properties and acent ric factor vary with them through correlations) in combination with indepen dently tuning pseudocomponent critical properties is found to be a better p rocedure than other available tuning procedures. The model developed using this procedure can well capture the compositional variation and other assoc iated property changes with depth for near-critical gas condensates tested in this work. The volume-shift parameters need to be used after the model m atches phase-amount data.