membranes that were industrially used for a longer period of time (“aged membranes”)
to study fouling and cleaning phenomena. During this study, reduced cleaning (a onestep
CIP instead of a three-step CIP) was investigated, leading to the suggestion that
the aim of a cleaning procedure should not necessarily be perfect hydraulic cleanliness
in order to restore processing performance. Consequences of reduced cleaning could
however be observed in subsequent CIPs; hydraulic cleanliness reached a lower level.
Further research is required to assess the practical significance of these consequences.
An overview of possible methods to investigate fouling and cleaning of membranes is
given. Various spectroscopic methods have been screened to assess whether they can
be used for online measurement of membrane fouling and cleaning. Also, methods
with a possible capability of offline characterization of fouling residues were
investigated. Both online and offline methods were found unsuitable to follow the
whole process of fouling and cleaning as the concentration of protein residues on the
membrane surface sooner or later reaches a level below the limit of detection of the
respective methods.
The main differences between full scale filtration and small scale simulations have
been evaluated and are reflected in the conclusions. These insights will prove useful for
future development of optimized small scale filtration units that could then be used for
improved simulation of full scale fouling and cleaning of membranes in bench scale.
The results obtained in this thesis and the given literature overview is expected to help
optimizing future membrane cleaning in the dairy industry.