(16b) Estimating Reid Vapor Pressure for Mixtures of Oxygenates Blended in a CA-Rbob

The Reid Vapor Pressure (RVP), the vapor pressure measured at 37.8 ^oC, is a critical quality attribute for gasoline set by EPA regulations to reduce the evaporative loss of gasoline to the environment. Various components of gasoline can impact air quality through the formation of smog and increased ozone formation. When blended with gasoline, some organic oxygenates such as ethanol, disproportionately increase the RVP beyond its own pure component vapor pressure because of non-ideal mixing; some oxygenates can decrease the RVP when blended into an E10 gasoline-ethanol mixture, as well. The deviations from ideal mixing characteristics are dependent on the formulation of the gasoline and the molecular interactions between the many compounds present in gasoline.

The Co-Optimization of Fuels and Engines initiative supported by the U.S. Department of Energy (Co-Optima) is investigating the concurrent optimization of fuels and engines to achieve higher efficiency with lower environmental impact. A variety of new oxygenated blend stocks have been proposed as candidates for further study. The RVP of the oxygenates/gasoline mixtures will be a critical characteristic of each candidate that will help determine the optimum blend level, and to some extent its market value. The ability to predict the RVP of new mixtures facilitates the evaluation process by eliminating the need to experimentally measure every possible candidate and allows researchers to identify possible mixtures that have beneficial (decreased) RVP changes when blended in the right proportions. Promising candidates can then be validated experimentally.

A California Air Resources compliant Reformulated Blend stock for Oxygenate Blending with 10 volume percent ethanol added (e.g. - E10 CA-RBOB) was blended with various oxygenates including higher alcohols, ketones, acetates, and acetals; the RVP was measured by a commercial laboratory using the ASTM D6378 test method. An ASPEN Plus model was used to emulate the ASTM method, and the RVP values were estimated for each mixture at 10, 20 and 30 volume percent oxygenate in an E10 CA-RBOB. The physical property models in ASPEN Plus (version 10) that were used for the RVP estimates were the Cubic Plus Association equation of state (CPA), and the Non-Random Two Liquids (NRTL), Wilson, UNIFAC and UNIF-DMD activity coefficient models. Compared to experimental measurements, the RVP estimates made using the CPA, NRTL and Wilson equations had average relative errors over all 57 samples of -1.5%, -6.4%, and -9.6%, respectively, with worse case errors of -11%, -19%, and -22%. The UNIFAC and UNIF-DMD models had relative errors of -2.6% and -0.6% over the 45 samples for which there were default parameters in the ASPEN Plus databank; four of the chemicals used in the study had missing parameters for the UNIFAC equations, so their average errors cannot be compared directly with the models for which all components were included.

The results of this study were based solely on physical property data and binary interaction parameters that come with the ASPEN Plus software package with no attempt to provide new interaction parameters based on new data regressions. Better binary interaction parameters, where applicable, could potentially improve model estimates.