The concept of residence time distribution (RSD) is used in process engineering to characterize the hydrodynamics of a chemical reactor or any other installation through which a circulating fluid passes. It can be used to :
The main properties of residence time distributions are reviewed, followed by experimental methods for acquiring DTS curves, and then methods for developing DTS models for ideal and real reactors.
The parameters of a DTS are identified by comparing experiment and model. Signal processing methods are numerous, ranging from the simplest, i.e. the method of moments, through Laplace transforms, then Fourrier transforms, to the most complex, i.e. the method of non-linear regression directly on the curves. Emphasis is placed not only on the methods of implementation, but also on the precision that can be expected.
Two examples are presented:
the first on a compartmental model, the cascade mixer model;
the other on a distributed-parameter model, the axial-dispersion plug-flow model.
Two programs written in Matlab
®
enable the reader to easily implement the most precise processing mode.