Dynamics of a Beer Pasteurization Tunnel
With more than 500 projects completed with less than one year to payback, Cargill Optimizing Services understands how to decrease costs and increase yields in the food and beverage industry.
Pasteurization is a critical step in many food processes. While it is relatively easy to heat-integrate continuous pasteurizers to reduce the overall energy consumption, disturbances such as startup of the equipment spike utility demand significantly. It can also be challenging to set up temperature profiles for transitions that successfully kill unwanted microbes without damaging quality attributes such as color and flavor.
In a previous article, we looked at a steady-state model of a brewery. In this article, we will review another brewery component - a pasteurization tunnel for a bottling / canning line. We examine how dynamic simulation predicts the behavior of the system during a startup/ shutdown cycle. In particular, we will look at opportunities to quantify peaks in overall energy demand and maximize product consistency.
Peaks in pasteurization are a pain. Dynamic modelling can be used to predict when the peaks occur so that different operating schemes can be tested to manage peak loads, reduce product loss and save on operating costs.
In many breweries, beer is pasteurized after bottling or canning in a tunnel. The beer moves along a conveyor belt through a series of chambers in which recirculating water is sprayed onto the containers. The water circulated so that the hot beer leaving indirectly heats incoming cold beer. The central section is heated to pasteurization temperature, and each stage has trim heaters / coolers that maintain the desired profile. For the purposes of this study, a 7-stage pasteurization tunnel for bottled beer was considered. The flowsheet is available upon request.
The performance of pasteurization is typically quantified by the Pasteurization Units (PU) that the beer encounters – 1PU is accumulated in 1 minute at 60°C. As temperature increases, the PU’s build exponentially faster.
Our model closes the mass and energy balance to quantify temperature and PU profile, as well as heating and cooling loads. In steady state, the model can predict these profiles when factors such as water temperature and recirculation rate change. The two graphs below compare these factors at different flow rates and temperatures. As can be seen, a 1°C setpoint change increases PU by 35% (22.4 to 29.7) and increases heat load by 1.8%. Increasing recirculation reduces heat load by 0.7% with a slight increase in the final PU (22.4 to 24.8).
This model is particularly useful in understanding dynamics disturbances, such as a temporary stoppage of incoming beer. The incoming flow of bottles is stopped at 2 min, and the bottles in the tunnel allowed to flow out. The final bottles leave the tunnel at 78 min, and the tunnel equilibrates. At 80 min, the flow of beer is restarted, and the tunnel refills until beer starts exiting the tunnel again at 156 min. For the base case, the temperature profile during shutdown is maintained.
The graphs below show the cooling (blue) and heating (red) duty and the final beer PU (green) as a function of time.
As incoming cold beer stops, the load on the cooling system increases and peaks when the last of the fresh beer enters the pasteurization section. This coincides with a drop in heater load as less beer is pasteurized. As bottles leave the pasteurization section, the heating load builds as a futile cycle of heating and cooling maintains the profile in the empty tunnel. Following the restart, incoming beer reduces the cooling demand. Once zones 1-3 are full, the cooling load is eliminated, but additional heating is required to warm the beer entering the pasteurization zone. The point at which that zone is filled (135 min) determines the peak load – ~2.5x the normal operational load.
During shutdown, the heated zones have a lower than average holdup, and a slight increase in the PU’s is observed. Far more significant is restart – a peak of 80PU is seen as the first beer through the tunnel experiences higher than normal temperatures.
Maintaining the temperature profile allows the tunnel to heat up while empty, and so causes overheating of the beer during restart.
Testing alternative strategies
The beauty of a dynamic simulation is that alternatives to mitigate such effects can be investigated without the need for physical testing. Both automated optimization and manual iteration can be used to find improved approaches. In this case, an alternative strategy was developed in which the temperature setpoints are adjusted, in order to avoid overcooking the beer without increasing the frequency or magnitude of utility peaks (see figure, right). The final solution successfully reduces both energy waste and product loss.
While the examples provided here are simple, they illustrate a model’s ability to predict and optimize performance of batch operations in brewing and the wider food and beverage industry. Armed with such tools, one can undertake tasks such as scheduling, control system enhancement, transition and load peak improvement and, ultimately, cost minimization.
When to consider
The tool used to develop this case study is directly applicable to most tunnel pasteurizers, and with limited rework, could be applied to products such as food cans or soft drink pouches. As well as being useful for understanding the impact of changes in packaging and/or product mix on temperature profiles, this sort of model can be helpful in developing operating schedules that avoid undue peak loads on utilities equipment, such as the boiler. The underlying approach has been applied to other parts of a brewery, such as the brew house or cellar, as well as to a range of other batch processes in other industries.
About the Author
Charles Sanderson is the Technical Director of CPO, based in Minnesota, USA. He has led the development He has led the development of Cargill's simulation capabilities for over ten years. He and the team have deployed the tools in projects across Europe, Asia, Africa and the Americas. Charles graduated in Chemical Engineering from Imperial College, London and received his PhD from Sydney University in Australia.
With more than 500 projects completed that have less than one year to payback, Cargill Optimizing Services understands how to decrease costs and increase yields in the food and beverage industry.