18 - Process modeling for hydrocarbon fuel conversion

Abstract

A description is given of the basic equations of computational fluid dynamics and mesh types. Reynolds and Favre averaging and models for dealing with turbulence flow models are then described, including the standard k-ε model. The modeling of combustion chemistry is discussed, describing probability density function methods, the eddy dissipation method, radiation modeling and empirical techniques using regression analysis and neural networks. Process flow models using spreadsheets and commercially available modeling packages are then described. The chapter then deals with methods for modeling chemical kinetics. Hydrocarbon reactions can become very complex, making the kinetic relations difficult to describe, not conforming to any power law. Kinetic models can be lumped, molecular and mechanistic in nature. Molecular and mechanistic models are described. The complexity of hydrocarbon reaction systems can be reduced with a combination of molecule-based kinetic modeling with existing software engineering. Molecular representation of the hydrocarbon feedstock is generated from analytical chemistry that is used to build a reaction network using graphic techniques to estimate rate parameters. These correlations are used to create a model that is then solved with process or experimental data for different reactor systems. Specific applications are listed with examples.