Molecule-based modeling of heavy oil

A molecular-level kinetics model has been developed for the pyrolysis of heavy residual oil. Resid structure was modeled in terms of three attribute groups: cores, inter-core linkages, and side chains. The concentrations of attributes were constrained by probability density functions (PDFs) that were optimized by minimizing the difference between the properties of the computational representation—which were obtained by juxtaposing the attributes—to measured properties, which were obtained by analytical chemistry measurements. Computational tools were used to build a reaction network that was constructed based upon model compounds and their associated kinetics. For cases with an intractable number of species, equations were written in terms of the three attribute groups and the molecular composition was retained implicitly through the juxtaposition. These modeling methods were applied to the Shengli and Daqing resids. The composition of the simulated molecular feedstock fit well with analytical chemistry measurements. After simulated pyrolysis, both resids showed representative increases in the weight fractions of lighter hydrocarbons. Relevant end-use properties were predicted for the product mixtures.