Authors: Oscar J. SANTOLLANI, Rajeev K. AGARWAL, Yau-Kun LI and Marco A. SATYRO
Affiliation: Virtual Materials Group, Inc. Calgary, Alberta, Canada
Reference: Oscar J. Santollani*, Rajeev K. Agarwal, Yau-Kun Li and Marco A. Satyro, Virtual Materials Group, Inc. Calgary, Alberta, Canada; CAPE-OPEN Specialty Property Packages – New Thermodynamic Models and Software Integration Paradigm; Hyprotech 2000 User’s Conference, Amsterdam, 6-10 November 2000
Comments: the authors discuss the need, for a provider of robust software core for the calculation of physical properties and for the calculation of phase equilibria, to allow its manipulation by different users with different objectives. They experience with several interfaces, including CAPE-OPEN ones and provide their assessment. They state that “Our own experience suggests that the CAPE-OPEN interfaces will work properly for standard property packages, since we were able to properly interface with HYSYS, gProms and other process simulators, and had calculations being performed fast and reliably… Nevertheless, we would like to offer our experience in interfacing a non-standard property package to the HYSYS process simulator.” So the authors question the ability for CAPE-OPEN thermodynamic interfaces to deal with non-standard thermodynamic models, like electrolytic ones.
They conclude with an interface specification checklist which remains valid when considering the revision of the Chemical Reactions interface specification by the Thermo Special Interest Group.
General Issues
- Flash interfaces should be completely documented and open, including access to stability test criteria, tolerances, and number of iterations and error codes.
- Flash interface behavior should be completely documented, including but not limited to: handling of special components such as water, how specifications with respect to number of phases are to be interpreted (e.g. what two phases mean? a liquid and a vapor or two liquids ?), how failures are handled and default behavior when failure happens.
Reactive system issues
- Ability to set proper basis for enthalpy calculations.
- Ability to perform calculations in such a way that the composition of the phase changes due to chemical reactions – perhaps by making the proviso that apparent species specified by the user are stored together with actual species calculated by the flash engine, thus honoring the preservation of mass but not necessarily moles.
- Ability to report actual compositions after calculations. This is a serious issue for a bi-directional simulator such as HYSYS.
- Ability to report molar enthalpy or mass enthalpy.
- Flexible scheme to define chemical reactions for each phase.