Validity, Verifiability, and Confirmability: A Critique of Multiphase Packed Bed Modeling

Abstract

The pseudocontinuum models of reactions in packed beds are complicated, and an assessment of the reliability of the predictability of their numerical solution is difficult. The predictability reliability depends on validity and verifiability, whereas the numerical solutions of models of reactions in packed beds cannot be validated or verified. Scientific acceptability cannot commence by metaphysics alone, and the truth of the speculative justifications of the results of the numerical models without robust empirical confirmation is a matter of chance occurrence. Adherence to the principles of noncontradiction and mathematical consistency seems to be the minimal criterion if a pseudocontinuum model is to demonstrate a degree of reliability in prediction, simulation, and design. This article is an exposition of the verifiability, validity, and confirmability characteristics of multiphase multidimensional models of reactions in packed beds. It addresses the difficulties of validation and the complexities of construction of models of reactions in packed beds by modeling kinetic data directly to show that often the claims of validity, verifiability, or confirmability of the results of multidimensional or even one-dimensional models of chemical reactions in packed beds, in spite of robust statistical tools, should be viewed with some degree of skepticism.

Keywords

• packed bed modeling
• validation
• multiphase simulation
• computation fluid dynamics
• reaction simulation