C seed crystallization - high quality crystals without water
In cane sugar mills, C seed is a product of key importance. This is the starting point of crystallization having a significant effect on the overall mill performance. Optimizing the C-side performance also helps improving molasses exhaustion thus reducing sugar losses. No wonder that our new client was interested in upgrading this pan first. Below is the story as reported by Zutora’s Matej Rukavina.
Supersaturation-based control of the sucrose crystallization process was implemented in a sugarcane mill in upper northeastern Thailand. The SeedMaster-4 crystallization control system (hereafter referred to as SM4) was implemented on the C2 batch vacuum crystallizer of 60 t capacity, to produce C seed magma. The C2 crystallizer is seeded with 0.01 mm wet milled slurry, the produced seed magma of 0.22 mm target size is pumped to the subsequent C magma horizontal continuous vacuum crystallizer.
Instrumentation for supersaturation-based control
The batch vacuum crystallizer C2 was already equipped with a microwave probe, vacuum pressure and level sensors; only a Vaisala POLARIS PR53 refractometer, for measuring liquid phase concentration and temperature, was additionally installed in the W-shaped bottom. A small diameter, flexible pipeline with hot condensate for the automatic refractometer prism wash was constructed.
Optimal installation location for the process refractometer in W-shaped pan bottom
Process refractometer installed.
The SM4 was connected to the factory DCS through Modbus/TCP industrial protocol. The SeedMaster transmits commands for adjusting the valves of steam, vacuum, feed and slurry seeding to the DCS. By this connection, the control system could decide whether to forward these commands to the control valves – thus giving the control to the SM4, or to take the control in case of unexpected scenarios.
Comparison of the traditional and the SeedMaster control
After the initial configuration of the system, three C2 batches were monitored, and supersaturation was recorded using the SM4. The feed syrup purity, Q = 68, was manually input into the SM4 on the HMI.
It is evident from the plots below that the level changed in a stepwise pattern under traditional control due to the periodic feeding of syrup and water. In contrast, the SM4 fed the 68° Brix syrup continuously throughout the entire boiling phase.
Graphs of pan level when controlled by SeedMaster (green) vs traditional control (trendlines with steps)
Is shown below, under traditional control, supersaturation reached high values in the initial phase of the batch—entering the labile zone and leading to spontaneous nucleation. It is also important to note that, with the old control system, water for washing the fine crystals was added intermittently. This practise has a very negative effect on steam consumption and thermal efficiency.
Graphs of supersaturation, when controlled by SeedMaster (green) vs traditional control
Steam consumption
We all knew that a significant amount of steam can be saved by eliminating the usage of water during the crystallization process. Our client wanted to know the exact amount.
To monitor the amount of vapour that is condensed in the calandria, a condensate flow measurement device was installed by the customer, so the condensate flow was monitored via their DCS. Under traditional control, condensate outflow was in the range of 41–53 m³. With the SM4, condensate outflow was reduced to a range of 27–31 m³. This indicates an average reduction of 18 m³ of condensate per batch, corresponding to a 38.3% decrease.
Steam consumption when crystallizing with traditional control vs SeedMaster
Crystal size and uniformity
During and after each batch, massecuite samples were taken from the C2 crystallizer and observed under a microscope by the factory staff. Below are photos of the C seed at the end of the crystallization process. The first photo shows crystals from a batch controlled by the SM4; the crystals appear more uniform in size, and the largest crystals seen in the traditional control are absent.
Crystals produced with SeedMaster control. No water was used. Formation of fines was avoided by careful control.
Crystals produced with traditional control. A large amount of water was used to wash the fines away.
Conclusion
The implementation of supersaturation-based control using the SM-4 on the batch vacuum crystallizer demonstrated significant improvements in crystallization performance and process efficiency. Continuous syrup feeding enabled stable supersaturation profiles, minimizing the risk of spontaneous nucleation and improving crystal size uniformity. The reduction in condensate outflow by 38.3% highlights enhanced thermal efficiency and reduced steam consumption. Microscopic analysis confirmed a more uniform crystal size distribution, with the absence of oversized crystals observed under traditional control. These results validate the effectiveness of the SM-4 in optimizing C seed magma production and provide a strong basis for further deployment in other crystallization stages.