InTech

JAN-FEB 2018

Issue link: http://intechdigitalxp.isa.org/i/935787

Contents of this Issue

Navigation

Page 43 of 53

44 INTECH JANUARY/FEBRUARY 2018 WWW.ISA.ORG cess in real time. If any values diverge from recognized norms, we know some- thing is going wrong with the batch. We can use this information for train- ing as we look at the characteristics of the most effective batches and most effective brewers. Positive deviations from normal operations can be captured and analyzed, so we can duplicate improvements. Making this kind of thing happen is not complex or expensive. It is the re - sult of several technological approaches working together: l continuously logging critical process variables, with perpetual data reten- tion using the cloud l data collection and reporting using small, cheap, replaceable devices with powerful capabilities l strategically placed process instruments l the ability to recognize when useful information can be inferred from all the data The lesson for process engineers is that you should not be afraid of looking for valid inferences. These are not guesses if they are informed by the data. Data, by itself, does not help. Information comes from understanding the data and seeing what it is telling you. Insight comes from understanding the information and us- ing it to improve what you are doing to gain competitive advantage. n ABOUT THE AUTHOR Michael Koppelman (michael@badgerhill- brewing.com), former head brewer, is cur- rently the CTO of Badger Hill Brewing in Shakopee, Minn. He is responsible for the technical aspects of brewing the company's craft beer, and for other aspects of the com- pany's operation. Koppelman holds a BS in astrophysics from the University of Minne- sota Twin Cities, and a BA in music from the Berklee College of Music in Boston. These profiles document each step and put the process in a form suitable for comparing it to similar batches. This provides 90 percent of the infor - mation we were recording manually, and provides it in greater detail. When we lay profiles from multiple brewing days on top of each other (figure 4), we can see a high degree of consistency with these manual processes. This suggests we have a good recipe, and our brewers know what they are do - ing. It also shows us that the process does not need to be adjusted on the fly, which gives us a basis for plans to automate the process. This allows us to build our craft brewers' know-how into our automation. We manage risk by watching the pro- came clear that much more was possible when looking at more complex opera- tions (figure 3). The process of starting a new batch of beer in the hot-liquor tank follows a set series of steps outlined in the recipe. Usually we try to make two batches, one after the other, over 20 to 25 hours to use energy more efficiently. The hot- and cold-liquor tanks interact as water needs to be heated, and the first batch is cooled by transferring its heat to the second batch. The graph shows the levels on both tanks superimposed with the same time scale. It is easy to see the changes as liquid moves between the two tanks. By following the profile, it is possible to see each step in the process and identify changes. So how do we use this information? AUTOMATION BASICS Figure 4. Multiple brewing batches can be compared, illustrating how consistently the recipe can be applied, and how individual brewers approach their craft. RESOURCES Brewers Association www.brewersassociation.org/statistics/ national-beer-sales-production-data "Brewing Quality Beer while Increas- ing Production Efficiency" www.emerson.com/en-us/industries/automa- tion/food-beverage/beer Figure 3. Watching the level indications from the hot- and cold-liquor tanks during a brewing process shows each step, documenting overall operation. Volume (BBL) 20 30 40 50 60 Volume (BBL) 0 10 20 30 40 50 60 05.00 10.00 15.00 20.00 Time 0 10000 20000 30000 40000 50000 60000 Time

Articles in this issue

Links on this page

Archives of this issue

view archives of InTech - JAN-FEB 2018