Modelling identification and control of High temperature short time milk Pasteurisation unit

Abstract

Dairy farming is traditional farming that entails the long-term production of milk, which is then processed for sale as a dairy product. Processing raw milk into finished goods entails more plant operations and techniques. Dairy industries are forced to improve and manage their production schedules due to the rapid degradation of milk and milk products. Pasteurisation is a thermal heat treatment process that can be achieved using VAT or heat-exchangers to destroy most of the pathogenic organisms present in milk. Moreover prolonged heat treatment, on the other hand, may deteriorate/denature the protein composition. Hence, temperature and exposure time must be carefully controlled. A state-of-the-art review on milk pasteurisation shows that there is a need for enhancement in the plant efficiency and also the propositions on the best control to improve the process operability. And also, the best methods and procedures for safer practices to manufacture milk products employing precise and perfect heat treatment procedures are recommended. The lethality effect of pasteurisation must ensure the eradication of undesired microorganisms and all harmful bacteria, without causing damage to the product. This can be accomplished by developing a suitable mathematical model, validating it using real-time data, and implementing an appropriate control system for the proposed model. newlineIn this work, two types of modelling are proposed, namely mathematical modelling based on physical laws and statistical modelling based on plant experimental data. Phenomenological models based on first principles are capable of predicting both transient and dynamic behaviour of the milk plant. The mass and energy balance equations are initially framed in the process of deriving models based on the first principle laws. These model equations can be used to scale up or down the plant, based on the production rate demanded. In order to develop equations in the form of transfer functions, the obtained equations are linearized and then Laplace transformed.