In a production environment with ever-changing brewing schedules, downtime and delays, the continuous expansion of yeast is a challenge. The timer-based expansion protocol is not flexible. It requires additional external laboratory samples and labor to correct or adjust the expansion protocol. One solution is to use a sensor that provides accurate and reliable data for propagation according to specification control and operation.
The Boston Brewery Company (BBC) uses a "pitch & pull" strategy to propagate various yeast strains, i.e., the propagation does not CIP after multiple deliveries until it is switched to another strain. In this case, after trying several sensors, Boston Brewery successfully tested Anton Paar's sensors, which met all expectations, and decided to install them at all Boston Brewery locations.
1 Introduction. The Boston Brewery Company (BBC) operates a state-of-the-art expansion system at all breweries, which was installed between 2013 and 2016. The expansion system was provided by Esau&Hueber (Schrobenhausen, GER) to address the complexities of yeast management at Boston Brewery, using more than 6 standard and special yeast strains. The expansion space occupies 10-15% of the total volume of the fermenter, which means that one expander provides enough space to feed 2-3 batches of wort per fermenter. There is no need to delay the drauflassen procedure and the wort can be added continuously without delay.
2 "pitch & pull" method.
The Boston Brewery Company uses a unique method of expansion, the so-called "pitch&pull" method. The first wort entering the fermenter is delivered in the corresponding volume (10-15% of the full volume of the fermenter) and 10% of the yeast expansion volume is also left in the expander. In the second brewing, the fermenters are filled, and the fresh wort is replenished again to continue the expansion. Depending on the yeast strain and brewing rhythm needs, this "pitch&pull" scheme can be repeated 14-16 times. Due to the hygienic design of the system and the highly active yeast after refilling, no contamination of beer biodeterioration has been observed so far. This "pitch & pull" method has several advantages: - Up to 6 full fermenter deliveries per expander per week (4 expanders in total = up to 24 complete expander drop-offs per week).
The yeast is permanently in the growing phase (no lag), similar to the "drauflassen" - the highly active yeast inhibits any potential beer spoilers at the initial stage - Less labor (automation and yeast cultivation from the laboratory to small-scale) - Significant reduction in the use of lye, acids and disinfectants due to continuous expansion without CIPS The system operates according to the protocol preserved by strain and the desired duration of expansion. As shown in Figure 1, the three-stage protocol contains a variety of parameters such as temperature, time, pump speed, and other parameters, and gives the winemaker the flexibility to adjust the parameters based on cell counts and extracts. There are other parameters that remain the same, such as air flow rate and back pressure.
Add a **comment of 140 words or less (optional).
Propagation protocols for various yeast strains were developed using input of ** and multiple profiles including extract, cell count, viability, pH, and other factors. These protocols provide a solid basis for consistent scale-up in a controlled environment. However, it is important to recognise that production environments vary widely, characterized by frequent schedule changes, brewery wort delays, unforeseen downtimes, changes in wort composition and temperature, different filling volumes, holiday seasons, etc. To accommodate these dynamic conditions, brewers often find themselves increasing the number of laboratory samples or inspections. This allows brewers to adjust protocols and address specific changes encountered. However, this approach transforms expansion control into a labor-intensive process that involves winemakers and laboratory personnel.
3 Convert a static protocol to a dynamic protocol.
In search of a solution that would allow the static protocol to adapt to a variety of external factors, the Boston Brewery Company explored options that did not require additional labor. There is consensus that the implementation of tanks or sensors can provide a potential means to achieve dynamic protocols that can effectively respond to changing conditions.
Of the various methods considered by Boston Brewery, the most promising is the use of **extract measurements. This method is preferable to cell counts by turbidity assay, which may exhibit significant variations in wort quality (e.g. cold truffle retention) and lack of complete uniformity in the tank due to floatation.
In addition, the starting cell counts for the "pitch&pull" method may be different, which further increases variability. On the other hand, extract measurements have shown a strong correlation with cell counts, as shown in Figure 2. This correlation is particularly relevant when a minimum utilization of extracts is required to take into account a certain biological quality. Therefore, utilizing extract measurements appears to be the most effective way to achieve accurate and consistent results.
Add a **comment of 140 words or less (optional).
The expansion system is equipped with a recirculation system, and three sanitary sensor housings have been installed in each expander. Next, multiple extract sensors were tested in a single expander to evaluate their accuracy, robustness, handling, calibration, cost, and integration with existing platforms. A total of five different sensors were tested.
The study found that two of them required more frequent CIPS to maintain accuracy, one required multiple calibrations, and the other could not be integrated into the platform due to the limitations of cloud data management. In the end, the fifth and final extract sensor from Anton Paar in Austria successfully met all of our criteria. Anton Paar's L-RIX 5100 (Fig. 3) is a refractometer that can be immersed directly into the production solution.
It provides continuous measurement of extract concentration and temperature, enabling round-the-clock monitoring and control of raw materials, intermediates and final liquids, as well as high-solids particle processes. With EHEDG certification and the ability to clean and sterilize at temperatures up to 145°C, the unit is the ideal device for controlling yeast propagation.
Add a **comment of 140 words or less (optional).
Figure 4 shows a comparison between the ** extract measurements and the apparent extracts obtained in the laboratory. Initial measurements already show good correlation (indicated by a blue dot), but after a calibration, the correlation improves further (indicated by an orange dot).
Consider that the data come from different expansion ranges (5 different yeast strains, with their own multiple protocols). In addition, we observed a higher correlation with the real extract, as shown in Figure 5. Despite the wide variety of "pitch&pulls" and variations in wort quality (including low to high gravity, different truffle loads, and different cell densities and temperatures) in the absence of the CIP process, the sensors consistently provided reliable results. However, it is important to note that the values obtained from the extract sensor cannot be directly compared to laboratory measurements on a one-to-one basis. The difference in the measured apparent extract concentration can be attributed to the basic measurement principles and sample handling employed by the sensor (refractive index and no filtration in our case) and the lab equipment (in our case the density with sample filtration).
Add a **comment of 140 words or less (optional).
4 How to incorporate extract values into the propagation protocol.
The reliance on extract data provides a reliable indication of the current state of the expander, especially when using consistent wort quality, especially relative to the original gravity, but additional consideration is required to determine the best way to incorporate these extraction values into the protocol. For example, for wort at 12°P and 15°P, 100% extract values are different. The wort at 15°P has a significantly higher cell count, which is due to the greater extract difference between the original gravity and the current extract. The best way to capture this difference is to use True Fermentation (RDF), using Real Extract (RE) expressed with this simple formula:
Add a **comment of 140 words or less (optional).
It is important to note that the extract value at full time of the expander is used, rather than the raw gravity of the wort is introduced. This method is used because refilling the expander results in a change in the value of the expander's extract. This difference occurs when the incoming wort is diluted by the remaining 10-15% volume in the expander after the initial delivery. This dilution can vary greatly depending on the length and dilution of the previous expansion (e.g., 10% vs. 15%). Figure 6 shows how the RDF calculated from the extract values matches the RDF of the laboratory sample.
Add a **comment of 140 words or less (optional).
As shown in Figure 6, the implementation of RDF significantly improves the correlation with laboratory measurements. Therefore, in order to enhance the dynamic nature of the propagation protocol, the "hours" parameter has been replaced by an RDF parameter based on real-time data from the extraction sensor. This adjustment allows for more precise and adaptive control of the expansion process. In the 12 months of use, there was only one extract reading that deviated from the expected results. After a 6-week "pitch & pull" protocol, the refractometer prism was covered with residue, resulting in a stable false overweight extract reading.
Anton Paar recommends using a deflector on the other side of the sensor. In this way, the liquid is directed to the sensor, ensuring a continuous mechanical "cleaning" of the prism and preventing the build-up of any coatings. Since then, no deviations have been observed, and Boston Brewery has decided to equip the remaining expanders with Anton Paar's ** extract monitors.
Currently, the Boston Brewery Company is fine-tuning various protocols for different strains to put protocols in place once all expanders are equipped with extraction monitors. 5 Yeast requires aerobic monitoring to enhance propagation The next step in fine-tuning the propagation protocol is the aeration regime with the following objectives:
Ensure that there is enough oxygen** at each stage to support yeast biomass and the formation of ergosterol, which is essential for cell wall development. - Create an aerobic environment by eliminating carbon dioxide from the expander to reduce toxic stress from yeast The Boston Brewery Company asked two questions: - Can we control aeration by measuring oxygen content and need to achieve a certain oxygen level for efficient expansion? Or is the CO2 content a more reliable value by trying to achieve the lowest possible CO2 content?
For this purpose, Anton Paar's optical dissolved CO2 sensors and optical dissolved oxygen sensors have been added to the recirculation line next to the installed extract sensors. The Carbo 6300 (Figure 7) is a maintenance-free and EHEDG-certified dissolved CO2 sensor based on the attenuated total reflection (ATR) principle, while the Oxy 5100 (Figure 8) is an EHEDG-validated oxygen sensor based on the fluorescence phase shift principle. To prevent residual layers, as mentioned earlier, a deflector is installed opposite the Carbo 6300 and Oxy 5100 sensors. Figure 9 shows the respective CO2 levels over a 10-day expansion time frame with 3 "pitch & pull" activities, while the results for oxygen content are still inconclusive. Both Boston Brewery and Anton Paar are actively exploring ways to incorporate oxygen content measurements into their scale-up protocols.
Add a **comment of 140 words or less (optional).
Add a **comment of 140 words or less (optional).
The CO2 values consistently indicate the presence of three distinct"pitch & pull"Cycle, the initial expansion occurs after the incubation period. During each expansion, the CO2 level is initially low due to the filling of the wort. The concentration is then increased until the maximum value is reached. After reaching the maximum, CO2 levels decreased due to lower yeast activity, stronger aeration, and lower temperatures in Phase 3. The insights gained from these profiles help minimize CO2 concentrations in the expander and help optimize the aeration protocol.
6 conclusion
Anton Paar's L-RIX 5100 extract monitor was successfully tested in Boston Brewery's expansion system, showing a strong correlation between extracts measured in the laboratory and externally measured. It has proven reliability and robustness, is able to be used for long periods of time without cleaning the tank, and can be adapted to a wide range of expansions. Boston Brewery is transitioning from a static to a dynamic protocol using RDF calculated based on extract values provided by the monitor. RDF is particularly valuable in the "pitch&pull" mechanism, which minimizes cell counts and adaptively controls propagation to consistently achieve target cell counts. Tested with dissolved oxygen and dissolved CO2 sensors from Anton Paar.
While the results for dissolved oxygen are currently inconclusive, monitoring CO2 shows different patterns, making it suitable for optimizing the aeration protocol. Next, Boston Brewery will use data from extracts and CO2 sensors to fine-tune the yeast protocol. By implementing more consistent and efficient expansion, including increasing deliveries per week, Boston Brewery aims to achieve greater consistency in deliveries and fermentation times.