255971 Automating Molecular Modeling of Gasification with the Kinetic Modelers Toolbox

Tuesday, October 30, 2012: 10:35 AM
306 (Convention Center )
Scott R. Horton1, Yu Zhang2, Craig A. Bennett1, Frank Petrocelli3 and Michael T. Klein4, (1)Chemical and Biomolecular Engineering and Energy Institute University of Delaware, University of Delaware, Newark, DE, (2)Tonnage Gases, Equipment and Energy - Technology, Air Products and Chemicals, Inc., Allentown, PA, (3)Tonnage Gases, Equipment and Energy - Technology, Air Products and Chemicals, Inc, Allentown, PA, (4)Chemical and Biomolecular Engineering, University of Delaware Energy Institute, Newark, DE

Automating Molecular Modeling of Gasification with the Kinetic Modelers Toolbox

 

Scott R. Horton*, Craig A. Bennett,  Michael T. Klein

Department of Chemical & Biomolecular Engineering, University of Delaware

 

Francis P. Petrocelli

Air Products and Chemicals, Inc.

 

 A molecule-based kinetics model was developed for gasification, a prevalent chemical process used in the production of syngas from a variety of feedstocks including coal, biomass, and municipal solid waste.  It is envisioned that this modeling approach will have applicability across a wide range of gasification reactor types and operating conditions, and represents a more-rigorous treatment for modeling gasification than most previous efforts described in the literature, in that it is not reliant on the definition of arbitrary, “lumped” pseudo-components nor on the assumption that the reaction trajectory achieves chemical equilibrium. 

The generally macromolecular feeds to a gasification unit were described, structurally, as independent molecules or a collection of various interconnected cores (following either a polymeric distribution or confined to a lattice). Gasification reaction processes fall into three major categories: pyrolysis, gasification, and combustion.  The feed structures were subjected to a series of reaction and logistical rules via the Interactive Network Generator (INGen) software in order to obtain a reaction network comprising all three reaction classes. 

Polymeric and lattice-type species were conceptually broken into “cores” and “substituents”, reacted independently, and reformed into molecules by following the Attribute Reaction Model (ARM) credo.  This methodology removed the permutative explosion of species from the balance equations, and placed it on the post reaction analysis.  Isomeric detail was collapsed into representative species. Each species was broken into a series of composition components (such as ring, group, and element counts). 

The balance equations for the reaction network were automatically generated and solved using the Kinetic Model Editor (KME) software. The specifics of the reactor design led to the modeling of interconnected ideal reactor zones, wherein the conditions of reaction could be independently varied. It was then possible to use process, experimental, and published data to optimize the kinetic parameters.


Extended Abstract: File Not Uploaded