(48g) Development of a Molecular Database for Use with Compound Species in the Process Simulation of Reactor Systems

In the petroleum refining industry, the accurate design and optimization of hydrocarbon conversion processes for gasoil feedstocks demand a growing effort in developing rigorous and reliable kinetic models (Van Geem et al. 2007). Existing techniques rely on the approximation of gasoil stream compositions using a set of pseudocomponents called “oil cuts” characterized by a set of lumped kinetic models.  Although the lumped models are easy to construct and implement, they lack the necessary fundamental molecular information needed to model the kinetics of the process at the elementary step level.

To address this issue, Bryan Research & Engineering (BRE) has developed a Process Reactors Suite (PRS) for the process simulator, ProMax® that when provided a set of user selected species, is capable of automatically generating all reactions sets and their corresponding rate expressions via the application of strict chemical rules, imposed by the underlying chemistry, and proprietary algorithms.  At its basic level, PRS relies on molecular reconstruction techniques that transform characterization data into molecular structures suitable for reaction kinetics modeling.  Unlike most molecular reconstruction approaches that utilize computationally intensive stochastic optimizations at runtime (Van Geem et al. 2007), PRS uses a deterministic molecular reconstruction approach to create a molecular database network that contains all chemical constituents important to gasoil products and streams prior to program execution.  When needed, the database is called and returns the species requested.  The effectiveness of the method is dependent on two competing factors, the completeness of the database and size of the database.  Since most gasoil reconstructions require representations of molecular structures of up to 40 atoms in a hydrogen suppressed, molecular graph, enumerating all chemical constituents would be beyond the capability of state-of-the-art graph generation algorithms (Peironcely et al. 2012, Fink and Reymond 2007, Blum and Reymond 2009).  Furthermore, it would result in a database too large to be efficiently searched.  The PRS addresses these concerns in two ways.  First, a hybrid method is proposed that generates structures using a series of coarse grained sub-graphs which limit the database size without compromising its completeness by employing constraints from molecular group, point group, and graph theories. Second, levels of resolution are introduced such that compound species (e.g C8 aromatics) are manipulated by the end user, while the chemical structures in the database (e.g. o-xylene, m-xylene, p-xylene, and ethylbenzene) are utilized in the reactor modeling.  If desired, the user may toggle between the levels of resolution, but all kinetic modeling proceeds at the highest possible resolution. Case studies will be utilized to highlight the database generation methodology and highlight the use of the database within the Process Reactors Suite. 

Van Geem K.M., Hudebine D., Reyniers M.F., Wahl F., Verstraete J.J., Marin G.B. (2007): “Molecular reconstruction of naphtha steam cracking feedstocks based on commercial indices.”  Computers & Chemical Engineering, 31, pp. 1020-1034.

Peironcely J.E., Rojas-Chertó M., Fichera D., Reijmers T., Coulier L., Faulon J., Hankemeier T. (2012): “OMG:  Open Molecule Generator.”  Journal of Cheminformatics 4(21), pp. 1-13.

Fink T., Reymond J.L. (2007): “Virtual exploration of the chemical universe up to 11 atoms of C, N, O, F: assembly of 26.4 million structures (110.9 million stereoisomers) and analysis for new ring systems, stereochemistry, physicochemical properties, compound classes, and drug discovery.” Journal of Chemical Information and Modeling, 47, pp. 342–53.

Blum L.C, Reymond J.L. (2009): 970 Million Drug like Small Molecules for Virtual Screening in the Chemical Universe Database GDB-13. Journal of the American Chemical Society, 131, pp.  8732–8733.