In Australian sugar factories one operator typically manages the high-grade fugalling, sugar-drying and low-grade (C) fugalling stations. The C fugals are managed least effectively as there is no process instrumentation to monitor on-line C sugar purity or final molasses purity. Conditions can change rapidly in the C fugals without the operator being aware, and poor performance can persist for several hours. Tight control of the C sugar purity is important to avoid high sucrose losses to final molasses or an excessive recycle of impurities in the C sugar (magma or remelt) to the pan stage. For the 2017 season, Isis Mill purchased a Neltec ColourQ 1700CC transducer that had been recently released on the market to measure the colour (inferred purity) of the total C sugar magma production of the station. The transducer proved effective for the operators to pragmatically achieve tighter control of the purity of the C sugar magma. For the 2018 season, Isis Mill purchased a second ColourQ 1700CC transducer to monitor the colour of the C sugar on the screen within their large capacity fugal. Described are the results of extensive testing of the transducer mounted on the fugal and the use of the transducer to assist operators achieve tighter control of the magma purity. The experiences with the use of the transducer on the magma screw for monitoring the purity of the total C magma production from the station are also described.
The centrifugalling of C massecuite is an important step in raw sugar manufacture, as it directly influences the sucrose loss in final molasses and affects the quantity of impurities returned to the pan stage with the C sugar magma. Currently, there are no instruments to measure in real time the purities of either the final molasses or the C sugar magma. Pope et al. (2009) reported on trials undertaken at Mulgrave Mill to measure the magma colour and molasses purity (on-line and at-line, respectively), but the techniques were never commercially adopted. In those trials, the magma colour measurement involved the flow of magma across a prism located in the monitor casing.
The operators of the C centrifugal station generally have a range of other responsibilities, including the high-grade fugals and sugar drying. Thus, the C fugal station is an area where off-specification performance can persist for several hours if the operator is distracted to other duties.
For the 2017 season, Isis Mill purchased a Neltec ColourQ 1700CC transducer that had only recently been released on the market to measure the colour (inferred purity) of the total C sugar magma production of the station. The transducer was mounted above the magma screw. The transducer proved effective for the operators to pragmatically achieve tighter control of the purity of the C sugar magma. For the 2018 season Isis Mill purchased a second transducer that was installed to measure the colour of the sugar on the basket in their high capacity machine. Here, we describe our experiences with these two instruments and present data for the in-fugal application.
C massecuite processing at Isis Mill
The low-grade station at Isis Mill (Figure 1) operates at a C massecuite rate of 5–6% on cane, depending on seasonal conditions. That rate requires the use of the Silver 52/30 (No 4) fugal, which typically processes massecuite at 14–18 t/h and of the BMA K1100 (No 3) fugal processing at 8–12 t/h. The two BMA K1000 machines (4 t/h each) are only used when preparing for maintenance activities, during machine failure or shutdowns of the larger machines, or when extra capacity is needed to clear stocks.
Figure 1. Layout of the low-grade crystallizer and fugal station at Isis Mill.
The crystallizer station operates in batch mode with six 70-m3 crystallizers providing a residence time of 18–20 hours. Under normal processing conditions, each crystallizer holds the contents of a single C massecuite strike. Molasses lubrication is utilised and the rotation of the cooling coils is normally stopped prior to fugalling in order to avoid aeration of product when massecuite is at low levels. No pre-heating of massecuite is carried out within the crystallizers prior to fugalling. All reheating is conducted in the finned-tube reheater.
A Neltec ColourQ 1700CC transducer was installed on the magma screw to measure the colour of the magma for the whole station for the 2017 season. A second unit was installed for the 2018 season; this time on No 4 fugal. Figure 2 shows the unit above the magma screw.
Figure 2. Neltec unit above the magma screw.
Description of the Neltec ColourQ 1700CC transducer
Neltec Denmark A/S developed the ColourQ 1700CC sensor to measure the sugar colour inside continuous centrifugals. The instrument essentially consists of an illuminator and a detector. The lamp in the illuminator sends light to the surface of the sugar. Reflected light is collected by the detector and split into different wavelengths – including 420 nm. Based on a calibration, the instrument can calculate colour from the strengths of the signal at various wavelengths (Neltec 2019). Once calibrated the instrument has very little ‘drift’ and a routine calibration check is all that is required to confirm correct operation. The position of the sensor on top of the centrifugal can be shifted inwards or outwards so that any position from the bottom of the basket to the upper rim can be selected for measurement. The preferred location for C centrifugals is close to the top rim in order to provide measurement of the sugar leaving the basket.
The colour measurement system consists of three parts: the measurement head; a local control station near the centrifugal; and the control cabinet with a PLC and computer. The generated 4–20 mA signal is provided to the factory’s DCS. Isis Mill has additionally connected a monitor, a keyboard and mouse to the control cabinet, enabling the operators to see the results graphically (Figure 3). Results of colour measurements are presented as a light brown live plot. The target colour range is defined by the green colour band. The dark blue live plot is a measurement of the high-speed variation in the colour and can be indicative of unstable feed or ‘streaking’ of the massecuite/crystal up the screen.
Figure 3. Graphs displaying the output signal from the Neltec transducer on the magma screw (left), No 4 fugal (right) as 3-minute real-time displays. The bottom display is a combined 3-hour real-time display.
The operator cleans the windows of the measurement head every shift prior to undertaking a simple calibration check using a calibration tile. The entire exercise of cleaning and calibration check takes less than 5 minutes and is the only routine maintenance required.
The signal can be calibrated to align with the ICUMSA colour range; however, Isis staff opted to maintain the data set based on the initial factory calibration. The signal range for the instrument for No.4 fugal was 0 to 800, while the transducer above the magma screw had a range of 0 to 4000. The difference in range for the two devices is due to the nature of the material passing the read head. One is almost dry crystal, while the other is puddled magma.
Experience with the Neltec ColourQ 1700CC in the 2017 season
In 2017 the Neltec ColourQ head was placed on the outlet of the magma screw (Figure 2). The operators used the colour signal (inferred magma purity) to manually adjust fugal settings to hold the signal in the target range. While this proved beneficial, there were a few obstacles to overcome, viz.:
- Vapour from the fugal outlets drifted across the light pathway and occasionally disturbed the signal. An exhaust fan was required above the screw to ensure the light path was uninterrupted by vapour;
- The action of the screw caused small waves in the magma transition trough and that action was evident in the response from the ColourQ unit. A baffle plate was required to effectively smooth the flow;
- Changes in magma brix affected the signal and so melter water application needed to be monitored. At Isis Mill magma brix is ‘controlled’ by utilising the output signal of the magma pump VF drive. Adjusting the melter water to maintain a constant pump speed was effective in controlling the magma brix.
With the introduction of the Neltec device, the sampling of magma for purity analysis was changed from the traditional composite sample taken to the laboratory at the end of the shift to a procedure at the start of the shift, involving:
- firstly, clean and check the calibration of the Neltec instrument;
- secondly, during a period of steady fugal operation record the Neltec signal and collect a snap sample of magma from the mixer for the laboratory; and
- thirdly, based on the purity of the magma sample (usually supplied within the hour) the operator selects the required Neltec operating band for the remainder of the shift.
The operator then makes changes to the fugal settings based on the current Neltec signal and the selected operational band for the shift. Previously, the laboratory analyses of the composite sample only provided an assessment of the production on the previous shift, which was not conducive to good control outcomes.
Installation of the Neltec ColourQ 1700CC for the 2018 season
For the 2018 season the second read head was purchased and installed directly on the top of No 4 fugal. Figure 4 shows the transducer on No 4 fugal and the light beam directed in the top section of the basket.
Figure 4. The Neltec transducer on No 4 fugal and the light beam on the basket.
The purity of the C sugar is adjusted by a combination of water added with the feed massecuite (termed Silvortex water) and water sprayed onto the lower section of the fugal basket (termed spray water).
The Silvortex unit on the feed entry into the machine consists of a double cone arrangement at the entry to the bell assembly rather than the traditional probe water application. A steam/water eductor is used to place water at 90°C between the two cones at the point of entry into the bell assembly. The high temperature of the top cone assists with the conditioning of the feed to ensure that massecuite flows freely into the machine. We found that operating with excessively high water rates on the Silvortex (>150 L/h) caused streaking on the working screen of the basket. For normal operation the Silvortex water rate is restricted to below 120 L/h.
Once a colour setpoint is determined using the procedure described previously, the operator places a target setpoint into the DCS colour algorithm. A visual display (Figure 5) allows the operator to make a quick check on the performance of the No 4 fugal and the whole fugal station magma output with respect to magma colour. The algorithm sets a range of a ±100 colour units on the display alert for the No.4 fugal Neltec device and ±200 colour units for the screw Neltec device. The yellow box indicates that the colour is in the desired range. Should the instrument fall outside the set range, it will alarm the operator that the magma is either too light or too dark.
Figure 5. Visual indication of algorithm for optimising magma purity control using colour.
Because of the rapid response of the Neltec device to variations in-fugalling conditions, e.g. water pressure, motor load, the control strategy for No 4 fugal was changed to adjust the spray water rate to respond to the Neltec signal. The Silvortex rate was fixed at 100 to 120 L/h and the spray water flow ranged to vary between 0 to 700 L/h. The machine load was routinely set at 150 A. This control system worked exceptionally well. In previous seasons, Isis staff relied on a manually set spray water flow for a given motor load with visual inspection of magma quality.
For the magma in the screw, the water rate for magma preparation ranged between 12 and 14 L/min and adjusted according to the magma pump speed.
Test program and results in the 2018 season
Test program on No 4 fugal
The test program had the following objectives:
- Determine the accuracy and repeatability of the Neltec device to measure the colour of C sugar produced by No 4 fugal.
- Demonstrate how the Neltec device can be utilised to optimise the sugar and molasses purities of the continuous fugal in real time by:
- adjusting the distribution of water between the centrifugal feed system and the spray water onto the basket (for use of the same total water rate);
- altering the spray water rate and using a fixed Silvortex water rate.
The fugalling conditions that were monitored and controlled during the trials were:
- Fugal load (current);
- Massecuite temperature;
- Feed/wash water temperature;
- Feed water (Silvortex) rate;
- Spray water rate;
- Neltec ColourQ signal.
Magma was produced in No 4 fugal by the addition of hot water within the monitor casing. During the trials, once steady conditions were established, composite samples of C magma and C molasses were obtained directly from the discharge of the fugal.
The process products were analysed as detailed in Table 1 and purities were calculated. There was a minor deviation from the official ICUMSA colour method for some of the laboratory colour analyses. A GF/F 0.7 μm filter was used early in the season for filtering the colour solution rather than the 0.45 µm filter nominated in the method. This resulted in the ICUMSA colour being overstated by around 4% but we considered it to be within the experimental error of the system and that it would not significantly affect the Neltec on-line signal correlation to the full results of laboratory colour. The correct filter was used for later tests.
Routine sampling and analyses across the season
At regular times through the season, Isis staff collected a sample of magma from No 4 fugal for conditions of high, normal and low spray water addition rates and noted the Neltec signal. Figure 6 compares the Neltec colour with the ICUMSA colour. There was a large deviation from the linear relationship at very low colour values (very high C sugar purities). Overall the data across the season were fairly widely scattered.
Figure 6. Comparison of Neltec colour and ICUMSA colour of magma from No 4 fugal across the season.
For normal production conditions the target magma purity was around 89 to 91, depending on whether changes to the C sugar magma purity may assist compliance with the sugar specification for the factory.
Figure 7 shows the magma purity versus the Neltec colour value. Again, there is a fairly wide scatter of results which indicates that control of a fugal across the season to a single Neltec signal is likely to result in magma having a wide range of purity values. At a Neltec signal of 600 (average for 90 purity magma), the range of purity values was ± 5 units about the mean.
Figure 7 shows the linear regression to the entire data set and also for data for magma purities between 85 and 95 purity. The data show that a change of 100 units in Neltec colour equates to an approximately 2 unit change in magma purity.
We note that these data include the influence of many factors that may affect the colour of the molasses layer retained on the crystal surface, including changes in the colour/impurity relationship attributable to variations in impurities in the cane supply or variations in colour generated in process (such as during a maintenance stop or wet weather). Other factors that may influence the colour value for a constant magma purity or the way in which the image is reflected include the aeration of massecuite in the crystallizers and massecuite properties such as crystal size distribution.
Figure 7. Relationship between magma purity and Neltec colour for No 4 fugal across the season.
Figure 8 shows the relationship between magma purity and ICUMSA colour across the season. This relationship is much stronger than shown for the relationship with Neltec colour and is linear across the full range of values for the season.
Figure 8. Relationship between magma purity and ICUMSA colour for No 4 fugal across the season.
Test results on a single crystallizer
The data in Figures 6–8 are across the season and include the effects of varying massecuite conditions. Trials were also undertaken on massecuite from a single crystallizer (i.e. massecuite from a single pan strike) in order to limit the variation that may be caused by varying C massecuite properties.
The data in Figure 9 show the typical relationship between Neltec colour and ICUMSA colour when processing massecuite from a single crystallizer. There is a very strong correlation between the Neltec colour and ICUMSA colour when processing massecuite of consistent characteristics.
Figure 9. Neltec and ICUMSA colour correlation when fugalling massecuite of consistent composition.
Effect of the distribution of the same total water
A trial was undertaken to determine the optimum distribution of water between the Silvortex water and the spray water, with a constant total water flow of 600 L/h selected. Figure 10 shows the Neltec colour signal and the purity of the molasses collected for each combination of water use. As the Silvortex water rate increased from zero, the C molasses purity increased linearly. Being a centre–feed centrifugal with a well-designed massecuite conditioning pot and bell, increased dissolution of sucrose occurred with increased Silvortex water rate. The Neltec colour response decreased initially, indicating a higher magma purity, but then increased. Isis staff investigated the cause and discovered that the massecuite feed became unstable at the high Silvortex water rates, with significant streaking of massecuite/crystal up the screen. As a result, a Silvortex water rate of 100–120 L/h was identified as the optimum for massecuite feed stability and low sucrose loss in molasses.
Figure 10. Effect of water distribution on Neltec colour and molasses purity.
Effect of grain size in the C massecuite on the Neltec signal
The grain size of the C sugar did not significantly affect the Neltec signal. Consecutive C massecuites were boiled with one having a grain size of 0.34 mm and the other 0.25 mm. This range of sizes covers the typical range for C massecuite in the industry. The centrifugal parameters were controlled to achieve a similar Neltec signal for each massecuite. The difference in the laboratory magma colour and purity of the large and small grain size was less than 2% (Figure 11).
Figure 11. The response of the Neltec colour signal to grain size.
Effect of molasses lubrication quantity to the crystallizers on the Neltec signal
A trial was undertaken where a pan-drop C massecuite (80 m3) was split between two crystallizers and different quantities of lubrication molasses (viz. at the rate of 1 and 5% on massecuite) were used prior to centrifugalling. The massecuites in the two crystallizers were fugalled successively in automatic control with the Neltec signal manipulating the spray water flow. The Neltec set-point was 650 for fugalling both crystallizers. Table 2 shows the results for the magma purity and ICUMSA colour based on three samples taken during each test.
Changes in molasses lubrication quantity did not confound the Neltec signal, possibly because it was located near the top of the rim of the fugal. A consistent magma purity was produced within the usual scatter of results when processing massecuites of similar properties other than having a change in lubrication quantity.
Effect of magma brix on the Neltec signal
Isis staff observed that the signal from the Neltec device located above the magma screw was affected by the brix of the magma. Tests where the melter water rate was varied demonstrated the change in Neltec signal. This result was expected, as visually magma is of darker colour when the brix is reduced. Isis staff use the speed of the magma pump as a guide to the operators to select the appropriate melter water rate.
Figure 12 shows practically instantaneous response of the Neltec signal to incremental changes to the spray water rate in the region circled. The changes are marked by a positive ‘blip’ on the chart that was introduced by opening the top inspection hatch on the centrifugal monitor case. The subsequent windage created a disturbance of steam and water that was registered by the Neltec transducer. Each major division on the horizontal scale of Figure 12 is 10 minutes and each point a 1.06 minute average. The vertical scale major divisions are 50 units, which is equivalent to around 1 unit of purity and 2000 IU. The changes in the bracketed region were created by decreasing the spray water rate by the equivalent of 0.3% massecuite in each step, i.e. 50 L/h.
Figure 12. Neltec ColourQ display demonstrating the response to changes in spray water rate.
Use of the Neltec ColourQ 1700CC to optimise fugal performance
Because of the rapid response of the Neltec transducer, the tight relationship between magma purity and ICUMSA colour (Figure 8) and the strong relationship between Neltec colour and ICUMSA colour for massecuite of consistent properties (Figure 9), the instrument enabled optimisation of the low-grade fugal performance in terms of C magma purity and molasses purity. The optimisation of the fugal is a trade-off between the recirculation of impurities to the high-grade pans and the sucrose loss in molasses. The recirculation of impurities impacts shipment sugar quality, production loadings on the high-grade pans and the loading on the low-grade stations (pans, crystallizers, fugals). The loss of sucrose in molasses, measured by the molasses purity and molasses loss, impacts the overall economics of the factory in terms of the raw sugar make as measured by overall recovery of sucrose and PSI (Pool Sugar Index). The test data described below demonstrate the interactions and selection of the appropriate Neltec colour set-point.
The test data for a trial on a single crystallizer to optimise fugal performance are shown in Figures 13–14. For this trial, the Silvortex water rate was held constant at 100 L/h and spray water rates varied to effect changes in magma and molasses composition. Both the magma purity and molasses purity increased with decreasing Neltec colour (Figure 13) as a result of increasing total water flow (Figure 14). The total water flow consists of the sum of the Silvortex water and the spray water. Data from this trial indicated that a Neltec signal range of 600 to 650 would be an appropriate set-point for automatic control of the magma purity to within the target range of 89 to 91 purity and achievement of a relatively low molasses loss. Control at these conditions would ensure that high-grade massecuite production suited premium shipment sugar quality and the low-grade pan, crystallizer and fugal station loading was appropriate for good exhaustion performance.
Figure 13. C magma and molasses purities versus Neltec colour for trials on a single crystallizer using different water rates.
Figure 14. C magma and molasses purities versus total water rate for trials on a single crystallizer.
It is interesting that the data in Figure 14 show that at high total water rates (> 400 L/h) the magma purity did not increase. Similarly, increased dissolution of sucrose into final molasses did not occur. Closer inspection of Figure 13 shows that there could be a flattening in the Neltec response at lower than 500 Neltec colour, but there are too few data points to be conclusive. We surmise that at a high spray water rate (>400 L/h) the crystal bed may be disturbed (perhaps with streaking of the massecuite flow over the basket) and the spray water is then less effective in removing the molasses layer from the outside of the crystals.
Application of the Neltec ColourQ 1700CC
Currently, in Australian factories C sugar purity is not measured in real time and so is not controlled as a process variable. Feedback of C magma and molasses purities from the laboratory generally takes several hours and, in the ensuing time, there is the potential for substantial losses of sucrose to final molasses or recycle of an excessive quantity of impurities to the pan stage. The magma purity reported is often not representative, being only a snap or short-term composite and is of historical value, e.g. for the previous shift.
The Neltec colour transducer provides real-time, continuous monitoring of C sugar colour that allows for automation, objective decision making by operators and supervisors, and continuous monitoring of the performance of continuous low-grade fugals.
The Neltec device provides for the first time the necessary real-time process feedback signal on individual continuous centrifugals to manipulate the spray water to automatically control C sugar purity. Many factories do not use spray water on the basket but use a combination of steam rate to the basket, water rate to the feed probe and motor load (least favoured process variable to change) to achieve the desired magma purity. The Neltec signal could similarly be used to regulate the C sugar purity for fugals operating in this manner.
If the C sugar colour drifts outside a pre-set range, the operator can be warned to investigate. Issues that may raise this warning include: mixed or fine grain massecuite; highly viscous massecuite (such as from stale cane or insufficient heating); blocked fugal screens that require cleaning; low crystal content massecuite; unstable fugal feed; loss of water pressure or reduction in water temperature.
The Neltec ColourQ transducer can enable labour saving either on the centrifugal station or at the laboratory by providing continuous indication of C sugar purity. It will free-up operators to allow closer attention to other aspects of their responsibility and so improve overall factory performance.
While the Neltec signal provides an excellent correlation to ICUMSA colour and magma purity when processing C massecuite of consistent characteristics, the experience at Isis Mill is that a cross-reference between magma purity and the Neltec signal should be undertaken at the start of each shift, e.g. just following the calibration of the Neltec transducer with the supplied tile. These results will then define an appropriate Neltec set-point for the shift to achieve the target C sugar purity. For factories with continuous crystallizers the variation in massecuite properties may be less than those experienced with batch crystallizers. A tighter correlation between magma purity and Neltec signal across the season could be expected compared with that given in Figure 7. Nevertheless, the cross-check at the start of each shift is still likely to be required to determine the appropriate set-point for the control loop using the Neltec signal, in order to take account of variations in massecuite properties.
The Neltec transducer proved to be beneficial for measuring the colour of the magma in the screw but variations in magma brix affected the signal. For an effective indication of magma purity, the brix would need to be controlled, e.g. by using conductivity in a feedback loop. However, the experience at Isis Mill is that the transducer is most effective when installed on individual centrifugals and used for automatic C sugar purity control. It is proposed that for the 2019 season, the screw Neltec device will be mounted on No 3 fugal at Isis Mill.
The trials at Isis Mill with the Neltec ColourQ 1700CC have shown strong benefit for the operators of the C massecuite centrifugal station for both the measurement of the colour of the total magma stream of the station and when used to directly control magma colour produced by an individual centrifugal. The signal from the colour transducer was highly responsive to changes in processing conditions, e.g. changes in spray water rates to the basket, and so is suitable for feedback control of magma colour. Alarms should be used to define upper and lower limits for the colour signal to warn the operator if a closer inspection is required. We recommended that the appropriate set-point for the control of the colour of the magma is determined at the start of each shift by cross-checking the signal to a sample of magma analysed in the laboratory.
We thank: Isis Mill for their support and initial capital funding for the project; laboratory staff at Isis Mill, Jenny Magdalinski and Tanyia Rainbow, for their efforts in completing the many analyses during the 2018 season to support the trials; the funding provided by Sugar Research Australia Limited under the Small Milling Research Program; and Neltec Denmark A/S who provided advice during the course of the trials.
*This paper was presented at the 2019 Australian Society of Sugar Cane Technologists annual conference and is published here with the agreement of the Society.
Neltec (2018) Neltec ColourQ 1700 CC. Technical sheet. http://www.neltec.dk/get_file.php?v=1&f=neltec_1700CC_en.pdf (Accessed 5 March 2019).
Pope G, Accatino P, Vidler G, Tack D, Kerchner S (2009) Full automation of a Western States continuous low grade fugal. Proceedings of the Australian Society of Sugar Cane Technologists 31: 460–469.