2014 AIChE Annual Meeting
(31a) Physical Properties Needs for the 21st Century – Databases, Models and Uncertainty
Authors
Physical property calculations are at the core of process design, process simulation and process optimization. These activities are normally executed within a process simulation environment and users are presented with a highly polished, “pasteurized” environment where exceptionally complex calculations involving thermodynamic models proceed with little or no actual understanding from the user related to the models or the data mused to develop the models used in the computations.
Notwithstanding advanced materials and advanced models required to simulate such materials, common, apparently well understood substances and mixtures are now being processed under very high pressures and temperatures, stressing existing thermodynamic models and moving us towards areas where little is known a-priori about the quality of the models or the underlining data. In this paper we examine the interplay between databases, models, the uncertainty in the thermodynamic data and the consequent uncertainty on calculated thermodynamic properties and what would be the desired characteristics of models to be developed in the future where quality assurance parameters are built in the model development from the ground up with the aim of developing introspective software systems that can provide the user with estimates about the quality of values it computes.