Identifying Opportunities in Breweries
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.
Cargill has helped improve efficiencies at breweries around the world with top class Brew Masters and a broad range of malt and adjuncts. Cargill Optimizing Services combined this wealth of brewing expertise with unmatched food & beverage applications talent to create state of the art brewery simulation tools.
The result is standardized, computer models that can optimize brewery performance at any scale. These models include energy and material flows for the entire plant, including brew house, cellars, packaging and utilities. This article demonstrates how models can be used to understand the economic and sustainability implications of a wide range of changes on a fictional brewery where process conditions have intentionally been shifted slightly away from normal to avoid any IP concerns.
- Asset Utilization
Breweries face a number of interesting engineering challenges that makes the application of traditional simulation tools tricky:
- Most operate as a series of batch steps, with batch sizes and processing times that vary with each step.
- The quality of the final product is difficult to quantify and predict from processing conditions, and requires frequent checking and review from skilled Brew Masters and Taste Panels.
- There is strong pressure in many geographies to improve sustainability, with a strong focus on reducing water use.
- There are several solids separations, leading to potential yield losses.
- A typical brewery will produce multiple products in the same equipment, and offer that product in a number of packaging formats, leading to product transitions and the concomitant yield and capacity impacts.
To address this range of challenges, COS has built a set of steady state and dynamic computer models for a typical brewery. The steady state model is particularly well suited to approximating the economic and environmental cost of operating the facility as it includes overall heating and cooling requirements, water and raw material use, and losses associated with wet solids and fermentation. Should additional performance data be required, one of CPO’s standard integrated dynamic brewery models would best fit. (Our next COS newsletter will cover this subject.)
The steady state brewery model includes the brew house, cellar and packaging lines using either recycled bottles or cans. For this example, a lager is being produced using a high gravity wort and a combination of malt and maltose syrup. In the base case, very little attention has been paid to heat integration, water use minimization or yield improvement.
An interactive webpage provides a representation of the model, and shows the sort of data that can be generated – for each stream displayed, one can “Ctrl-Click” and see temperature, composition and flowrate. For each unit, one can see information such as heat input or cooling. One challenge of modelling in steady state is that the batch nature of the process is converted into a continuous flow equivalent. Thus while all the mass, volume and heat flows are labeled as “per hour”, they are better thought of as “per batch”. The system is set up such that the hourly average flow corresponds to a single batch in the brew house. While not tied to the batch model, indicative flowsheets for the boiler house and refrigeration plant are also included.
The model anchors the analysis at a unit-by-unit level, but is typically used to investigate the impact of changes in operating and equipment approach on higher level indicators such as operating cost, water and energy use and yield. A summary of some potential changes is provided in the table below. While some of the modifications investigated are common practice, they illustrate how a model could be used.
The following common operating practices and their impact on the Key Performance Indicators were considered:
- Recovering the moist trub from the whirlpool, which significantly improves yields in terms of carbohydrate loss.
- Using a centrifuge to spin down spent yeast after maturation, which fur-ther reduces losses in terms of ethanol and unfermented carbohydrate.
- Reusing and condensing the kettle vapors to provide a heat source for pasteurization, which takes another large chunk out of the heat losses.
The model was also used to compare the operating costs using recycled glass bottles instead of cans for the case where hot wort was cooled with incoming process water, trub recovery was in place, and yeast was spun down. While the bottles are cheaper, they need to be washed before filling, which requires energy and water, and they are far heavier than the cans, which adds thermal load to the pasteurizers. As can be seen in the last 2 columns of data, though, for the simple system modeled, the increased energy costs can be mitigated entirely if the kettle vapors are used to supply the heat required in washing as well as pasteurizing the bottles.
In a real-world analysis, a far greater range of changes would be reviewed, and additional constraints would be placed on the analysis. This simple overview provides insight into the level of detail that can be derived – an understanding of where water, energy, and raw materials are consumed allows the brewery team to prototype opportunities rapidly. The impact on key per-formance indicators can be reviewed before any physical testing in the plant, so only the best ideas get the operating facility. The model predictions can also be compared to actual operating data, and used to identify gaps – this can help pinpoint yield loss points and/or the economic impact of non-standard operations (e.g. product change over).
When to consider
The tools and templates used to develop this case study are broadly applicable. They can be applied in operating plants to review changes in raw material, recipe or operating conditions, and to see the impact of proposed capital or production changes. They can also be used to review the design and configuration of new facilities, or for significant capacity increases.
Of course, in addition to brewing, the underlying tools and general approach described here is broadly applicable to other food and beverage processes.
About the Author
Charles Sanderson is the Technical Director of COS, based in Minnesota, USA. He has led the development of Cargill's simulation capa-bilities for over ten years. He and the team have de-ployed 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.