Physical refinery modeling

The previous sections in this chapter address the creation of pseudocomponents by cutting an assay curve into a set of discrete components based on boiling-point ranges. We also briefly alluded to physical properties and process thermodynamics selection in the earlier workshops of this chapter. In this section, we consider, in detail, the problem of how to represent these components in the process modeling software. There are two major concerns in this area physical properties of pseudocomponents and selection of a thermodynamic system that can deal with these hydrocarbon pseudocomponents in the context of refinery modeling. 

A neural-network-based simulator can overcome the above complications because the network does not rely on exact deterministic models (i.e., based on the physics and chemistry of the system) to describe a process. Rather, artificia] neural networks assimilate operating data from an industrial process and learn about the complex relationships existing within the process, even when the input-output information is noisy and imprecise. This ability makes the neural-network concept well suited for modeling complex refinery operations. For a detailed review and introductory material on artificial neural networks, we refer readers to Himmelblau (2008), Kay and Titterington (2000), Baughman and Liu (1995), and Bulsari (1995). We will consider in this section the modeling of the FCC process to illustrate the modeling of refinery operations via artificial neural networks. 

According to NCUT s definition, the product quality model links the molecular representation of the process effluent to its physical properties (e.g. density and viscosities at two different temperatures) and refinery-type characterization (e.g. smoke point and cetane number). In this section we will illustrate the development of product quality correlations for diesel-range materials using density of liquid and cetane number as examples. 

Modern refineries deal with a multitude of complex systems that may require different thermodynamic models for each refinery plant and its associated process model. For example, we cannot model the sour gas units that deal with acid gases and water with the same thermodynamic model that we use for the crude fractionation system. In fact, reasonable thermodynamic models form the heart of any process model. Chen et al. [7] have documented the variety of thermodynamic models available for frequently encoxmtered chemical and physical systems. Agarwal et. al [18] present a detailed account about the pitfalls of choosing a poor thermodynamic system for process models and the undesired consequences of using these poor models to modify plant operations. Process model developers and users must be aware of the underlying thermodynamics and its limitations. 

While this chapter has focused extensively on the requirements for modeling fractionation systems, we can use the same techniques in the context of modeling refinery reaction process as well. We illustrate this process in Chapters 4 through 6 of this text It is possible to obtain good predictive results for fractionation systems provided that we make reasonable choices for the thermodynamic models and physical properties of the pseudocomponents involved. 

---