The key operating objective in the production of Casein powder is to maximise moisture content while still remaining within final specification. Operating too close to upper quality limits may result in out-of-spec end product, along with the risk of lost production due to blockages in the dryer, cyclones or mill. Dryers are often run with a wide ‘comfort’ margin; final product moisture may be well below customer specification.
We constructed a robust model of dryer performance from process metrics, including raw material quality, dryer air temperatures, discreet data from laboratory analysis, dryer operating parameters. Then we developed an inferential sensor providing accurate, real-time estimation of casein moisture under all conditions. This integral part of the model predictive control system enabled the plant to maintain very tight control of product moisture.
RETURN ON INVESTMENT
The project required 12 weeks to implement. Real-time inference enabled powder moisture to be maintained closer to constraint, improving yield by 0.65%. Higher moisture was exploited as increased dryer capacity, enabling higher throughput. Improved start-up control delivered an average of 7 hours per week of extra on-spec production.
The benefits delivered a full return on investment within 4 months of commissioning, sustained via an ongoing support programme provided by our industry-experienced team.
“Advanced control and the ‘soft’ sensor from Perceptive helped us to greatly improve our process efficiency. Exit moisture capability in the Casein dryer increased from 1.8 to above 3.5 Sigma and the Perceptive system is a key part of our optimisation strategy.”
Niall Herlihy, Process Technologist, Glanbia