MPC deployments: lessons from the last season

In modern sugar factories, the sugar house is where the margin is decided. Small issues in crystalliser timing, centrifuge loading, or strike receiver levels don’t remain local — they cascade into bottlenecks, energy spikes, and lost tonnes of sugar.

Most plants still run batch crystallisation with manual scheduling or simple rules. Skilled operators must constantly manage:

• When to start the next strike

• How much buffer remains in the strike receivers

• How many centrifuges can safely run

• How to avoid starving the dryer or overfilling the system

On quiet days, this is manageable. During a real campaign with fluctuating syrup purities, cleaning stops, and unplanned trips, it becomes firefighting. More interruptions. More stress. Less sugar.

From Firefighting to Forecasting

At Sucrosphere, we see a clear shift: sugar houses are moving from “best-effort” manual control to predictive, model-based planning.

Our Digital MPC Crystallisation Scheduling Package uses Model Predictive Control (MPC) and real-time plant data to plan hours ahead instead of reacting minute by minute.

The goals are straightforward:

• More sugar through existing crystallisers and centrifuges

• Fewer stops and overloads

• Smoother, more stable operation for the entire team

Instead of operators constantly adjusting schedules, the system proposes — or directly applies — optimal strike start times and centrifuge loadings. Operators move into a supervisory role: monitor, validate, fine-tune.

What the Sucrosphere Package Does

The solution includes three integrated layers.

Crystalliser Scheduler MPC

Predicts crystalliser behaviour and discharge times. Calculates strike start times to keep centrifuge loading in the optimal range and avoid bottlenecks.

Centrifugation MPC

Controls strike receiver levels, anticipates cleaning or maintenance windows, and stabilises the feed to downstream equipment.

Crystallisation SFC Layer

Ensures each batch runs consistently and efficiently. Uses online measurements and anomaly detection to maintain crystal growth rates and product quality.

All features are visualised through a clear HMI: one screen for planning and one for the live situation. No black box. Operators always know what the system is doing and why.

The Impact in Numbers

Based on deployments in beet factories around 10,000 t/day, and using:

• Sugar price: 450 €/t

• Steam: 40 €/MWh

• Power: 100 €/MWh

the Sucrosphere Crystallisation Scheduler & MPC typically achieves:

• +0.2 to +0.5% more sugar throughput through existing assets

• Fewer emergency stops and overload conditions

• 2.5–5% lower specific energy demand due to stabilised operation

This corresponds to 0.2–0.45 €/t beet in total benefit.
For a typical campaign, that equals 200,000–450,000 € directly on the bottom line — without major hardware investments.

Plug In, Don’t Rip Out

The Sucrosphere platform connects to existing DCS/PLC systems via standard interfaces and can be rolled out gradually.

Step 1

Start in advisory mode — the scheduler suggests actions.

Step 2

Transition to closed-loop MPC once the team is confident.

Step 3

Scale from a single crystallisation station to the entire sugar house.

No big-bang change. Just a smarter, more predictable way to run the assets you already have.

Ready for Your Next Campaign?

If last season felt like permanent firefighting, it’s a sign that scheduling complexity has outgrown manual tools.

With Sucrosphere’s Digital MPC Crystallisation Scheduling Package, factories can:

• Turn batch crystallisation into a predictable, high-throughput process

• Increase sugar tonnage without new equipment

• Reduce specific energy use and CO₂ footprint

• Make sugar house operations calmer and more resilient

Curious what this could look like at your factory? Share a slice of last season’s data, and we can prepare a tailored first business case.