Abstract
An extrapolating model provides values for the thermophysical properties when infeasible specifications (either pressure or temperature) are recognized at a given iteration in a process simulator, applying an EOS technique. The model has a considerable effect on the convergence characteristics, accuracy and robustness of a higher-level simulation algorithm.
The present paper outlines a general approach for evaluating and comparing the abilities of different extrapolation techniques: to promote the convergence of higher-level algorithms; to ensure for them a solution that is accurate and not significantly inferior to the actual one. The latter could have been obtained using for example a different thermodynamic model.
The approach is based on the classical fixed-point theory and its connection with the contraction-mapping principle.
