Improved Batch Pan Monitoring and Control – a soft sensor approach

Conductivity-based vacuum pan controllers dominate the Australian sugar industry. A new scheme, based on Brix control of the massecuite and mother molasses is proposed. Multi-loop control of the process is demonstrated, whereby the massecuite Brix is controlled by manipulating the liquor/A molasses feed rate to the pan and the mother molasses Brix is controlled by manipulating the steam flow rate to the pan. A single loop controller is also proposed, which controls massecuite Brix by manipulating the feed flow to the pan. The steam flow rate is set manually (i.e. mother molasses Brix is uncontrolled). A gPROMS simulation of a pivot pan under typical operating conditions demonstrates the efficacy of this approach. Introduction To date the Australian raw sugar industry has clung to a simple, conductivity-based control scheme for vacuum pan operation. In this scheme, liquor is fed to the pan such that a predefined conductivity trajectory is maintained, so that a suitably ‘heavy’ massecuite results. The pan boiler chooses an appropriate set point for the steam flow rate controller, in order to set the overall process rate. While this system is inexpensive and reliable, it is not always easy to relate the measured conductivity output with process-relevant variables, since variations in conductivity may result from factors that are not linked to the crystallisation process. Examples of this would include changes in cane variety or maturity, changes in geographical region of the cane supply, dirt loading in the cane and temperature excursions during vacuum pan operation. This paper reports on an alternative pan control scheme, which is based on process measurements that are arguably more process relevant, compared with conductivity. The first is a process refractometer, which gives a measure of mother molasses Brix, and the second is a microwave moisture device, which delivers an estimate of the massecuite Brix. Two multi-loop controllers are proposed in this scheme. One controls the process refractometer output, by manipulating the steam flow rate to the calandria. The other controls the microwave output, by manipulating the feed flow rate to the pan. Results from a series of computer simulations are presented, which demonstrate the feasibility of the proposed method. Traditional vacuum pan control Ideally, massecuite crystal content should be controlled as high as possible, but not so high that pan circulation suffers. This ensures that the amount of exposed crystal surface per unit massecuite mass is maintained at the highest possible level, leading to enhanced net deposition rates of sucrose. Additionally, the mother molasses oversaturation should be controlled to a point just below the secondary nucleation limit, in order to sustain the highest possible crystal growth rates. Since it is currently not possible to measure directly either crystal content or oversaturation, they must be controlled inferentially. In Australian sugar factories, pan control (whether batch or continuous) typically employs electrical conductivity probes to determine the condition of the massecuite within the pan. Adjusting the flow rate of fresh liquor/molasses to the pan controls the conductivity to its desired set point. The set point for massecuite conductivity is usually scheduled against the tonnage of massecuite in the pan. This resulting setpoint trajectory attempts to anticipate (i.e. model) changes in the process, including the build-up of impurities in the mother molasses and the boil back of A molasses, which is a higher Brix/lower purity feed. This scheme is pictured below in Figure 1. Steam flow to the pan is normally controlled by feedback, using an operator-selected set point and in that sense is manually controlled. The conductivity controller maintains the massecuite at an appropriate “weight” over the duration of the strike, while the steam flow controller affects the rate at which the process proceeds. Fig. 1: Schematic of typical batch vacuum pan conductivity control scheme PV Process Variable SP Set Point CO Controller Output Feed control valve