Model-based design of transient flow experiments for the identification of kinetic parameters

With recent advances in automated flow reactors and online analysis techniques, transient flow experiments are attracting significant interest as methods for rapidly gathering kinetic data. However, the design of these experiments is challenging and non-intuitive. This work addresses this challenge by using model-based design of experiments (MBDoE) to design optimum transient experiments for the purpose of identifying kinetic parameters with maximum precision. Using the case study of benzoic acid and ethanol esterification with sulfuric acid as the catalyst, the flowrate and temperature of a plug flow reactor were linearly ramped in time to create transient flow experiments. Two types of experiments were conducted, one where only flowrate was ramped while all other variables were held constant, and one where flowrate and temperature were ramped simultaneously. In both cases, model-based design of experiments (MBDoE) methods were used to design the transient experiments in order to choose the initial value and ramp rate of all ramped process variables, as well as choosing the fixed value of process variables that were not being ramped (feed concentration). The model-based designed experiments were compared against equivalent experiments designed by researcher intuition and standard design of experiments approaches, such as trying to cover a wide area of the design space. It is shown that MBDoE led to significantly more precise parameter estimates, and that the identified model was then able to predict with high accuracy the outlet concentration of other experiments.