Profitable processing through model based optimisation: Production capacity increased by 40%

Food process modelling is an essential ingredient for process optimisation. The use of modelling allows the exploration of numerous processing conditions, while evaluating the effect on process efficiency and product quality. Instead of separate costly trials for each processing condition, only relatively few experiments are needed to validate a specific model and apply it. For the development of useful food process models, besides the basic knowledge on heat and mass transfer, a deeper understanding of the food-process interactions is needed. It is useless for process optimisation or development to know the heating time of a product without knowing the effects on product quality. Over the years, NIZO has developed several models covering all the standard equipment which is used in general food processes. By combining fundamental understanding of heat and mass transfer with expert knowledge in protein functionality, microbiology and dairy science, these models can be used to model all aspects of process optimisation, from process settings to product quality.

NIZO was approached by a customer, who had a problem with one of their new products: Due to fouling in their heating equipment, the runtimes was shorter than 2 hours, and cleaning was required after each runtime. Based on NIZO’s modelling capabilities and expert knowledge on proteins and fouling behaviour, the customer decided to work with NIZO to find a solution. NIZO setup an optimization trajectory consisting of a process scan, model based optimisation and final implementation in practice, with as goal to increase the runtime while maintaining/increasing product quality. The result was a 40% increase in capacity, due to reduced fouling. Meanwhile, the microbial and product quality where either maintained or improved.

1. Process scan

With the process scan, NIZO aims to collect all plant data required for the specific project scope. Information is gathered by interviewing operators and managers, getting equipment measurements manually or from P&IDs, collecting samples for analysis or using data from the company sampling schemes. In this case, data on the equipment size and operation parameters like product flowrates and temperatures were gathered. This gave an accurate description of the process and the heatload on the product. Also, product samples from several points in the production were collected for analysis. This sample data was used to validate the models for the process-product interactions.

2. Model optimisation

NIZO has models in place to calculate heat and mass transfer for all major heat transfer units in the food industry. Additionally, NIZO has developed several process-product interaction models for microbial growth/inactivation, protein denaturation and fouling among others. These heat and mass transfer models and process-product interaction models are incorporated in the NIZO Premia software. This software was also used for calculations in this case. Since every process is unique, NIZO first validated all process and product models in Premia for the specific process at the customer. This resulted in an accurate description of the process and consequent effect on the product in the form of microbial inactivation, fouling and Maillard products.

3. Model simulations and implementation

With the validated model in place, NIZO performed several process simulations with different heating regimes. The best result showed a significantly reduced fouling rate while also giving a better inactivation of microbes and less Maillard products. Implementation of the simulation settings in the actual process at the customer resulted in a 40% increase in capacity due to increased running times and reduced cleaning frequency. After the success of the optimization performed by NIZO, the customer purchased the modelling tool developed by NIZO and realised an even greater production capacity through continuous model based optimisation.

NIZO has recently setup a collaboration with PSE, a software developer offering model solutions for processes. The collaboration is intended to combine the sophisticated solvers and flowsheeting capability of PSE with the food oriented models and expertise of NIZO, offering even more flexibility and solutions for future customers.