PRACTICAL USE OF NIRS FOR FACTORY
PROCESS CONTROL
MADHO S, BARKER B AND RAMARU R
Sugar Milling Research Institute NPC, c/o University of KwaZulu-Natal, Durban, 4041, South Africa
[email protected] [email protected] [email protected]
high molasses factor
low OR low BHR low VR
3
increased centrifugal purity rises
increased m massecuite per tonne MJ brix
high UDL high TPD
low exhaustions
high reducing sugar ratios
small magma crystals
large purity drops across evaporators
Read on
high remelt purity
high molasses purity
Summary
All of the South African sugar mill laboratories have access to Near Infrared Spectroscopy (NIRS) instruments with process stream
calibrations supplied by the Sugar Milling Research Institute NPC (SMRI). The calibrations include fructose, glucose and sucrose (FGS), and
also pol and brix predictions for mixed juice, clear juice, syrup and all the various grades of molasses and massecuites. Further to this, dry
solids can be predicted on C-molasses, and conductivity ash on mixed juice and C-molasses. The rapid and reliable predictions by NIRS
can be used by factories to improve recoveries and rapid reporting of factory performance figures. This poster elucidates some approaches
on how to achieve this.
Determining Undetermined Losses
Boiling House Control
Evaporator inversion losses
Curing of C-massecuites
• Conventional SMRI method of evaporator inversion loss
determinations:
• The SMRI-NIRS predictions can also be used to determine target
purity differences (TPD) for each C-centrifugal.
• Figure 2 below shows that the TPDs can be used to replace
molasses purity rise determinations. Advantages of this include:
Disadvantages: Long lead times (± 2 weeks, excludes results interpretation and
report writing) and once-off tests usually done.
• New SMRI-NIRS method of evaporator inversion loss
determinations:
 Frequent analyses to identify losses timeously (TPD predictions
can be done as often as twice per shift to show centrifugal losses,
whereas purity rise determinations are often only done once daily
and on selected centrifugals). Torn screens and poor operations
(e.g. overwashing) can be identified rapidly.
 Molasses purity rise determinations required a Nutsch sample; this
isn’t needed for NIRS TPD predictions.
82.4% brix
Sucrose lost (%)
NIRS TPD
Result: Many tests under different factory conditions can be performed by mill
staff, with conclusions drawn using SMRI inversion loss toolkit, all within a day.
83.7% brix
87.7% brix
85.3% brix
84.1% brix
True purity rise (%)
Sample time (minutes)
Figure 1. Sucrose lost across 1st effect evaporators in a
South African sugar factory
Figure 2. Results showing that SMRI-NIRS TPD predictions for each
centrifugal can be used to replace C-molasses purity rises. The
results also show the high TPD and purity rise for centrifugal 4 was
as a result of low molasses brix (through overwashing).
Profile of reducing sugars
• A spreadsheet toolkit has also been developed to input the NIRS
analytical results for any evaporator station configuration and
calculate the estimated sucrose losses.
• The devised method and a trial version of the toolkit has already
been used at three mills, using different processing conditions,
and has determined possible losses in excess of R70,000 daily
under certain conditions.
Fructose/Glucose
• The method to determine inversion losses across evaporator
stations using the SMRI-NIRS predictions has been developed and
demonstrated to several mills in 2014 and 2015.
• A standard operating procedure (SOP) and the final version of the
spreadsheet toolkit will be issued to all SMRI member mills in 2015.
© SMRI
2015
Also ask about…
• Sucrose-based factory performance calculations
• Massecuite and seed target purity predictions
Figure 3. Use of SMRI-NIRS predictions to show fructose/glucose
ratios across the boiling house – useful to identify areas of reducing
sugar formation/destruction, Maillard-type reactions, etc.