(237b) Approaches and Software Tools for the Development of Molecular-Level Kinetic Mega Models

Many energy and petrochemical processes involve the reactions of large numbers of species. Modeling these systems at the molecular-level can therefore involve model sizes reaching the mega level in terms of reactions and/or species. Model development at this scale becomes time consuming in each of the building, solving, optimization and editing phases. Advances in computer science have allowed for the development of new modeling frameworks that can be tailored to decrease the time spent and dynamically scale to variations in reaction system sizes that reach up to and past the mega level, all while retaining chemical engineering fundamentals. The Dynamic Model Builder (DMB) is a C++ based modeling framework developed to account for these needs. DMB functions independently of program compilation, allowing for easy use and application portability.

To illustrate the robustness of this modeling framework a lignin structure was developed for modeling pyrolysis between the temperatures of 300-500 °C. To bypass the combinatorial issues created by the large number of reactive sites on the lignin structure a network merging approach was developed and utilized. In the merging approach, reaction chemistries were split into two reaction networks that were later merged into one. For lignin pyrolysis, the α-O-4 and β-O-4 linkage cleavage reaction network (RN1) and the vinyl degradation reaction network (RN2) were merged. The products from RN1 were used as the input to RN2. It was found that RN2 is the root cause of a combinatorial explosion, so its growth was limited by reactant carbon number. Multiple tractable reaction networks were produced based on increasing the carbon range. Greater carbon ranges led to mega level growth of the reaction network. Therefore, DMB was used to model the reaction networks produced from the merging approach. In DMB, linear free energy relationships were used to decrease model kinetic parameters. Optimization of kinetic parameters was achieved from literature experimental data and produce a model output that shows good agreement with experimental values.