(38g) Surface Tension Predictions for High-Pressure Column Operations – III: Recommended Methods

So said Gupta et al. (2023, doi:10.1021/acs.iecr.2c03153) in a recent review article on current practices and limitations of thermophysical property development; it underscores the need for an accurate model for surface tension when designing columns for applications such as distillation, absorption, and more. Software most often used in column design typically offer limited choices of models that can be used to estimate the surface tension of mixtures. It is frequently the case that none of the available models perform satisfactorily, especially when applied to high-pressure operations. At high pressure it is possible – likely even – that one or more than one compounds is beyond its own critical point. Operations where this may occur include demethanizers and nitrogen rejection units (NRUs). In these operations the surface tension may be very low indeed (we have seen cases where the estimated values are lower than 1 dyne/cm). Consequently, estimated surface tension may significantly exceed the actual surface tension; this has important consequences for column design.

At the Kister Distillation Symposium in 2022, we presented a two part paper that looked at the consequences for column design of an erroneous estimate of the surface tension. We also discussed possible ways forward that included the use of new methodologies, in principle at least, of estimating the surface tension of mixtures that included supercritical components. In this work, we present our progress in identifying the best techniques (both predictive models as well as models that must be fit to data) for modelling mixture surface tension. We discuss the integration of these models into column simulation programs, with particular emphasis on rate-based simulations. Included in our consideration is the choice of thermodynamic model, surface tension calculation method, and computational efficiency and reliability. We also examine the effect of the choice of surface tension model on the column designs and performance of several commercially important high-pressure processes. Finally, we describe how our findings may be implemented by other engineers.