2008 Annual Meeting

(645a) Phase Diagrams of Methanol and Alkane Binary Systems in the Field of Natural Gas Fractionators: Experimental Work and Modeling

Authors

Courtial, X. - Presenter, Ecole des mines de Paris, Paris Tech
Soo, J. C. B., Mines ParisTech
Coquelet, C., Ecole Nationale Supérieure des Mines de Paris
Paricaud, P. L., Ecole Nationale Supérieure des Techniques Avancées
Ramjugernath, D., University of KwaZulu-Natal, Howard College
Richon, D., MINES ParisTech
Natural gas, within reservoirs, is normally saturated with water. The presence of water can lead to formation of gas hydrates, with a high probability of serious operational, economic and safety problems in production, transportation and processing. In order to avoid gas hydrate formation, methanol is injected as a traditional inhibitor in natural gas lines. Therefore, down-stream of pipelines, natural gases can contain non-negligible quantities of water and methanol. Water is classically removed from natural gases using dehydration units. However, methanol remains in natural gas streams, which may cause problems in optimizing the operating conditions of liquid hydrocarbon fractionators.

Accurate data and modelling of vapor-liquid equilibria of methanol-hydrocarbon mixtures are therefore necessary to provide information appropriate for the design and optimization of natural gas processing units. In this communication, we first report the details of an experimental apparatus based on the ″static-analytic″ technique. Thereafter, we present new isothermal VLE data concerning the n-Butane + Methanol binary system at six temperatures above and below the n-Butane critical temperature. The new experimental results and some selected literature data of binary hydrocarbon + methanol VLE systems are modeled by a phi-phi approach, using the Peng-Robinson equation of state combined with Wong-Sandler mixing rule. These data are also represented using the SAFT-VR and PC-SAFT models, which take into account self-association between methanol molecules. A detailed comparison between the three modeling approaches is presented herein.