Cereals processing technology

36 Cereals processing technology
supplied by a small number of equipment manufacturers. However, most recent
development work has been centred on the optimisation of machine design and
capacity and then the application of these machines to existing processing
strategies. The result has been the development of more compact flour mills in
recent years.
3.5 On-line process measurement
The importance of accurate and reliable data for control, optimisation and
information systems cannot be over-emphasised. These data are the first step in
the loop that makes process optimisation possible. Mill processes use many tools
to measure performance, but many of the methods are unsuitable for on-line
application. The Farrand 13 method for starch damage measurement is an
example. To be suitable for on-line application, measurement methods need to
be accurate and capable of producing results that can be interpreted by personal
computer-based systems. Verification of sensor data is an essential and realistic
demand of these systems. However, problems with measurements are one of the
main reasons for the failure of process control. The problems include incorrect
readings, unreliable readings and even the complete absence of readings.
Most complex instruments depend on statistical data analysis and a lot of
work has to be performed in the area of assembling enough test data to develop
and maintain calibrations (Fearn and Maris 1990, Graf 1994). This work has to
be performed in every application of instrumentation.
McFarlane (1992) documented the large number of sensors employed in the
food manufacturing industry. However, Near Infra Red (NIR) analysis currently
forms the backbone of on-line process analysis in flour mills and is likely to
remain so for some years to come. Psotka (1999) mentioned the emergence of
X-ray fluorescence as a technology that may supplant NIR in some applications
in the future. This technology appears to be superior to NIR for ash and starch
damage measurement and is currently being evaluated in the USA.
The NIR equipment employed by Fearn and Maris (1990) is typical of the
equipment in use throughout the milling industry today. The reasons for its
popularity are: the instrumentation is rugged and simple, there are few
manipulations (no weighing, titration, calculation) required and the devices
deliver rapid results. A small increase in error over classical methods can be
tolerated for most quality control applications because this error is counterbalanced
by the analysis of larger numbers of samples. The NIR determination
of protein and moisture is now comparable with that of classical methods
(Osbourne 1980), while ash determination is readily available in laboratory and
on-line instruments. In addition plant breeders and other specialised users have
developed their own calibrations for NIR (Davies and Grant 1987).
13 AACC method 76-30A.