Individual components may not be able to represent the structure of heavy hydrocarbons because these materials are formed by several chemical species that are difficult to characterize with the current analytical techniques. Lumped component techniques can be applied to model these types of hydrocarbons; this procedure is often based on combining many pure compounds into groups with average physical properties. Nevertheless, this technique fails for separations that are chemically driven due to the lack of chemical information in the lumped component groups. A new approach of the lumped characterization technique is shown in this work. This technique consists of using constant slates of selected compounds to cover the carbon number ranges of interest for the modeling of different refinery reactors. The different combinations of these component slates allow matching the experimental distillation curve of a given feed and calculating its chemical characteristics ranging from simple properties such as molecular weight and standard density to PIONA (n-paraffin, iso-paraffin, olefin, naphtene, and aromatic) characterization data. The key advantage of this new method is the capture of the essential chemistry of the feedstock that affects property calculations while keeping a constant and consistent component list.