Taking mash filtration to the next level

Besides the lauter tun, the mash filter(opens external link in a new tab) constitutes the second important alternative for separating solids and liquids in breweries. It is used for an estimated 30 to 40 per cent of global beer production because its technology offers a number of persuasive advantages, particularly for countries where beers are not brewed in line with the German Purity Law.

One important benefit: A mash filter allows brewers to use raw materials other than the conventional malt strains, including grist loads with a greater content of local cereals like sorghum, which is popular in African countries.

What’s more, a greater yield coupled with high original extracts can be achieved, an attractive advantage especially for high-gravity brewing. All of these pluses result in reduced consumption of both primary energy and water, thus fitting in ideally with current efforts to design lines and systems for maximum energy- and resource-economy, in other words: for enhanced sustainability.

New plate design developed

Since Steinecker is also fully aware of mash filtration’s huge potential, the brewing technology experts conducted an extensive series of trials in their pilot plant. Based on the large number of results obtained and a comprehensive statistical data analysis, they then developed a new design for the proven mash filter plates, not only achieving the advantages mentioned above but also increasing the filter’s throughput and reducing its footprint.

Following the successful trials in Steinecker’s in-house pilot plant(opens external link in a new tab), planning is currently underway for the installation of an industrial-scale pilot filter intended to validate the insights gained.

Trials at Steinecker’s pilot plant – the details

The preconditions

Various process-engineering prerequisites must be met to ensure a successful filtration and wash-out operation. To this end, homogeneity must be achieved for a number of parameters which can be controlled by fluid mechanics. Firstly, the same mass of solid matter must be evenly deposited inside all chambers. Secondly, a filter cake of even thickness must be created which at all points exhibits the same filter cake resistance. That is achieved not only by the cake’s thickness but also by its particle composition. So it is likewise necessary to ensure a uniform particle-size distribution over the entire filtering area.

Based on these considerations, Steinecker(opens external link in a new tab) drew up an analysis plan and conducted trials in its pilot plant. Initially, 100 per cent of pilsner malt was used for these trials. After that, trial brews were run with wheat, rice, maize and sorghum with different liquor-to-grist ratios and different amounts of enzymes blended in.

The results

Building on the results of these pilot-plant trials, Steinecker developed the final design for the upgrade of its mash filter(opens external link in a new tab), first and foremost of its membrane plates. The most important outcomes are described below:

• Higher throughput: A mass of 36 kilograms per square metre of malt equivalent turned out to be the ideal grist load (as compared to 30 kilograms per square metre loaded previously). This is made possible by a larger chamber depth of 50 millimetres. These revamped chambers can also be retrofitted to existing Steinecker mash filters, thus increasing throughput by 20 per cent for the same filter size.
• Smaller footprint: If the vital focus is not on upsizing performance but on the space required to install the system, the shorter plate package is likewise a worthwhile option because if the throughput remains the same, the mash filter can be up to 2.6 metres shorter.
• Flexibility in regard to type of filtration: The tests conducted have shown that for most raw materials single-side filtration is the most suitable method. However, two-side filtration can make sense for certain raw materials. The overall process would then run significantly faster.
• Short process times: Process times permitting up to 14 brews a day were achieved in the trials. Occupancy time (without spent-grains removal) for a brew under acceptance-test conditions (100 per cent pilsner malt, 22 degrees Plato first wort) was 84 minutes, a timespan also expected for large-scale production. The process time is longer for higher original extracts.
• Outstanding technological parameters: In the trials, a gravity of less than 3.5 degrees Plato (with reference to spent-grains dry matter) for the washable extract was reliably achieved in the process time mentioned, which according to the MEBAK (Mitteleuropäische Brautechnische Analysenkommission; Central European Commission for Brewing Analysis) corresponds to the requirements for this target variable.

First-wort concentrations of up to 24 degrees Plato were processed in the trials. The turbidity values were determined using the Imhoff-cone method and were lower than 5 millilitres per 1,000 millilitres.

It was possible to dry the spent grains cakes down to a moisture content of 64 per cent. However, the length of the final process step (emptying) is key to achieving this. With a normal process time, the average moisture content was 70 to 74 per cent.

• Cakes released in full: In the over 100 trials conducted, it was possible to release all spent grains cakes without any residues to speak of on the filter cloths. Even when hot break was re-used, no residues were found on the filter cloth.

The mash is passed into the filter chambers through an exterior mash channel (1). The requirement for low oxygen pick-up during filling is thus reliably met. The exterior-mounted infeed channel provides two advantages: an increase in filtering area and elimination of any components that may affect flow conditions in the filtering area. In combination, they make for homogeneous cake wash-out.

The mash transfer channel to the rear (2) provides the field-proven option for filling the mash filter from the front and from the rear. That is important for achieving uniform mash deposits in all chambers. If one of the main lower infeed channels is blocked, there is also an equilibrium channel (3) to make sure that all chambers are nevertheless filled with solids. In the standard process, wort is discharged through the top (4) and bottom (5) wort channels.

There are likewise two channels (6, 7) for feeding water to the wash-out process. The washing water flows through the boreholes in the water feed pipe (8) onto the membrane’s nubbed surface, from where the water passes through the spent-grains cake during sparging. Similarly, the wort is discharged through the boreholes on the filter plate (9). The washing-water channels also have a second, optional, function: As mentioned above, the new plate design enables two-side filtration for some raw materials. In this case, the washing-water channels serve as additional wort-discharge channels.

Moreover, the type of filter cloth was likewise modified. In future, overhang cloths with a sewed-on sealing edge will be used, which is made of foodgrade PVC fabric. With these overhang cloths, any soiling sprayed through the filter fabric can be swiftly removed via the movable opening at the plate’s bottom, without any clogging.