Some discussions on RGibbs model in Aspen Plus Dynamics...
1) PFR configuration for RGibbs:
PFR-RGibbs additionally calculates response-delay and equipment-volume for given geometry data. We could tell it is a minimum effort to take account the holdup contribution, but it is not done in rigorous way. Personally I don't suggest PFR configuration in RGibbs since it doesn't look theoretically sound. When a holdup is considered for RGibbs, it perhaps better to combine an Instantaneous-RGibbs and a Dynamic-Mixer (before or after the RGibbs), instead of PFR-RGibbs. Basically PFR-RGibbs is not a distributed model therefore there won't be real Plug-Flow effect at all.
2) Slow performance due to PFR configuration:
Such drawback in integration performance may be because of the complexity in delay function and volume calculation. It is not directly related to RGibbs calculation. In Aspen Plus Dynamics, RGibbs is (only one!!) the procedural model, which reuses A+ RGibbs calculation. It means its computational performance in dynamics (for RGibbs block) is identical to what A+ exhibits. Basically it carries out series of SS calculation continuously along integration time.
3) Why RGibbs?
Probably the most merit of RGibbs is that it does not require reaction stoichiometry. Without such detail reaction data, RGibbs can determine phase equilibrium by minimizing Gibbs free energy, subject to atom balance constraints at given process condition (T,P or P,H). For instance, when you model a burner, which is fundamentally a reactor but you perhaps don’t know details of what’s happening inside, RGibbs may be the one to use. There is no suggested criteria to choose RGibbs, but data-availability may be one of the major reason when choosing RGibbs.
