Detailed kinetic models in the context of reactor analysis: Linking mechanistic and process chemistry

Detailed kinetic models for the modeling of complex chemistries, including thermal cracking, catalytic reforming, catalytic cracking, and hydroprocess ing, offer the compelling advantage chemical significance at the mechanisti c level. They carry a considerable burden, however, in terms of species, re actions, and associated rate parameters. This, together with the batch and the plug flow reactor balances, requires solution of a large system of eith er stiff ordinary differential equations (ODE) or stiff differential algebr aic equations (DAE), for both homogeneous and heterogeneous processes. It i s often faster numerically to solve a stiff system of ODEs and, thus, it ca n be useful to convert a system of DAEs to ODEs for numerical solution sche mes. For heterogeneous PFR systems, the reactor steady-state balances resul t in a set of DAEs, and it would therefore be desirable to construct the as sociated set of ODEs to minimize CPU demand. To this end, we propose that s uch a transformation can be achieved by making the "flowing surface species " approximation. This involves approximating the overall rate of reaction o f surface species, which is identically equal to zero at reactor steady sta te, by a spatial derivative. We show that this approximation becomes better as the system of equations becomes stiffer, and, hence, is a reverse analo gy of the kinetic steady-state approximation in the case of batch systems. To validate the proposition, we analyze various contrived and real examples of mechanistic kinetics for heterogeneous systems.