A survey is given of some crucial concepts in chemical process modeling. Those are the concepts of physical unit invariance, of reaction invariance and stoichiometry, the chromatographic effect in heterogeneous systems, the conservation and balance principles and the fundamental structures of cause and effect relationships. As an example, it is shown how the concept of reaction invariance may simplify the homogeneous reactor modeling to a large extent by an orthogonal decomposition of the process variables. This allows residence time distribution function parameters to be estimated with the reaction in situ, but without any correlation between the estimated residence time distribution parameters and the estimated reaction kinetic parameters. A general word of warning is given to the choice of wrong mathematical structure of models.