Automatic Reaction Network Generation using RMG for Steam Cracking of n-Hexane Kevin M. Van Geem, Marie-Francoise Reyniers, and Guy B. Marin Laboratorium voor Petrochemische Techniek, UGent, Krijgslaan 281 (S5) B-9000 Gent, Belgium Jing Song and William H. Green Dept. of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139 David M. Matheu National Institute of Standards and Technology, Gaithersburg, MD 20899 DOI 10.1002/aic.10655 Published online October 12, 2005 in Wiley InterScience (www.interscience.wiley.com). A new reaction mechanism generator RMG is used to automatically construct a pressure dependent kinetic model for the steam cracking of n-hexane. Comparison between simulated and pilot plant data shows that RMG is able to generate detailed reaction networks that accurately predict the conversion and the yields of the major products although none of the kinetic parameters are fit to the experiments. RMG generates reaction networks based on minimal assumptions, making it possible to test commonly used assumptions such as the -hypothesis and the quasi steady-state approx- imation (QSSA) for -radicals, traditionally used in steam cracking, 1,2 as well as in pyrolysis. 3 The RMG-reaction network for n-hexane confirms that no bimolecular reac- tions of heavy radical species are important at the examined conditions (COT: 953 K - 1090 K; COP: 0.20 MPa -0.24 MPa; 80% conversion), and that the QSSA for the group of -radicals leads to negligible errors. RMG also offers the possibility to estimate the error introduced by neglecting the pressure dependence of most of the reactions. In the case studied, this frequently made (but seldom tested) approximation appears to be justified. © 2005 American Institute of Chemical Engineers AIChE J, 52: 718 –730, 2006 Keywords: steam cracking, reaction network generation, pressure dependence, sensitivity analysis Introduction Large-scale detailed kinetic models find increasing use in the modeling of combustion processes, atmospheric chemistry, soot formation, and other areas of industrial or environmental interest. Because such reaction networks may contain up to thousands of reactions and species, constructing them by hand can be tedious and error-prone. Therefore, many research groups have developed computer tools to automatically gener- ate these mechanisms. 2,4,5,6,7,8,9,10,11,12,13,14,15,16 A key difficulty with mechanism generation programs is that they can be com- binatorial, producing large numbers of kinetically unimportant reactions and species. Moreover, sometimes nonphysicochemi- cal criteria and/or expert user involvement are employed to limit mechanism growth, thereby risking the possibility that important reactions are not included. In contrast, rate-based termination of computer-generated reaction mechanisms pro- vides a physicochemical criterion for including reactions and Current address of D. M. Matheu: Cabot Corporation, 157 Concord Rd., Billerica, MA 01821. Correspondence concerning this article should be addressed to K. M. Van Geem at kevin.vangeem@ugent.be © 2005 American Institute of Chemical Engineers 718 AIChE Journal February 2006 Vol. 52, No. 2