Tag Archives: kellerbier

Brewing, Experiment, Recipes

My Kellerbier Experiment 2024

1 Comment

I don’t homebrew that much anymore these days, at most 4 to 5 times a year, and really only the beers that I absolutely want to brew and drink, which includes fixtures like an 8° Czech-style pale lager for summer and a Czech-style dark lager (which I brew with Ben) for winter. So there is really not that much room for experimentation, simply because I don’t have the time, the drinking capacity (I’m 40, it’s all about quality over quantity now) or the resources like free fridge space for fermentation and lagering.

But there have been a few things that I kept wanting to try out, all in the context of Franconian Kellerbier that I had learned about in the last year or so.

The first thing was when my friend Joe Stange visited Brauerei Knoblach just outside of Bamberg end of last year, and came back with the information that Knoblach generally uses a 1:1 blend of Pilsner and Vienna malt as a grist, that their water is fairly hard parts of the year, and that they rely on that hardness (though he couldn’t provide any concrete numbers, nor was I able to find any analytical data about the water of Schammelsdorf, where the brewery is located). I really like Knoblach’s beers, and I know its peculiar taste, so I was wondering whether these were two factors that played into it (Joe also published an excellent article about Kellerbier in the latest issue of Craft Beer & Brewing magazine which does not seem to be online yet, but as a subscriber myself, I highly recommend getting an online subscription).

The second thing was what I learned at HBCon earlier this year about how Mönchsambacher brewed their Weihnachts-Bock. The two things I wanted to incorporate were their mash profile (which I wrote about in July) and the water profile, which has roughly equivalent hardness of calcium and magnesium, and plenty of it as sulphates. Luckily, my local Berlin tap water has about the right calcium hardness, so all I needed to do was to add the right of magnesium to get my tap water roughly where the Mönchsambacher water is in terms of hardness and mineral composition.

The third point on my agenda of things to try out were Aurum hops, a relatively new German hop variety that was launched as more disease- and climate-resistant with a “highly fine” aroma. As a daughter of Tettnanger, it is meant to replace Tettnanger and similar varieties, and German hop growers as well as hop merchants have been promoting it because they see it as a variety better suited to climate change than others, including German landrace varieties. What I wanted to know was well the hops fared in a traditional style.

I know, integrating all three elements in a single experiment is not exactly scientific, as I don’t have a baseline to compare it to, nor do I isolate any of the multiple variables. I’ll still call it an experiment simply because I want to know what a beer brewed that way would taste like.

And that’s how I formulated the recipe:

The grist was simple: 50% Pilsner malt, and 50% Vienna malt. As hops, I used Aurum hops, with additions at 60 minutes for bittering, 25 minutes for flavour, and to the whirlpool for aroma, which should end up at 41 IBU (calculated):

  • 2.3 kg Pilsner malt
  • 2.3 kg Vienna malt
  • 30 g Aurum hops (5.8% alpha acid) @ 60 min
  • 30 g Aurum hops (5.8% alpha acid) @ 25 min
  • 40 g Aurum hops (5.8% alpha acid) @ whirlpool for 20 min

My tap water needed to be enriched with magnesium, so I simply spiked the mash with food-grade epsom salts (MgSO4). According to my calculations, 18g should get me the right amount of magnesium hardness for 22 liters of beer.

As yeast, I used the Fermentis S-23 dry yeast strain. It’s not my absolute favourite, but it’s all I had at hand, also because I had forgotten to order anything else, which is all my bloody fault.

When it came to mashing, I just stuck to the Mönchsambacher mash profile, a single decoction mash. I boiled the wort for 60 minutes, then cooled it down to 6°C, pitched the yeast, and let it ferment at 10°C until it was finished fermenting. I then ramped down the temperature to 1°C for a week, and bottled it with wort I had held back for bottle conditioning.

The resulting beer has an OG of 11.6°P (slightly lower than a typical Kellerbier, but that’s mainly from me buying the ingredients and only afterwards deciding on a mash profile with a slightly lower efficiency than my regular double decoction mash), and fermented down only to 3.5°P FG, resulting in just 4.3% ABV.

A Willibecher of the Kellerbier. The colour is golden with a tiny bit of haze, with a moderately dense head of foam on top.
A Willibecher of the Kellerbier. The colour is golden with a tiny bit of haze, with a moderately dense head of foam on top.

A first taste test showed that the experiment, in my opinion, was successful: the beer has a minerality and a maltiness very much reminiscent of Knoblach and Mönchsambacher. The same goes for the bitterness: while it’s not quite as pronounced as I hoped it would be (I blame the low attenuation which probably leaves just enough residual sweetness to slightly mute it), as it is very lingering: even minutes after, that hop bitterness just stays on your tongue. Which is exactly what I appreciate so much about these beers.

As for the hop aroma of Aurum itself: there’s not that much there. Even though I kept the hops in the fridge and sealed at all times, it was not the freshest batch (2021 harvest), so that may have had an influence. Still, the bitterness the hops provide was quite on point.

Still, I’m very happy with the end result. I think it shows that the local water profile of Bamberg’s surrounding area has a large impact on the flavour of the beer, as long as breweries don’t soften or otherwise treat the water and embrace their very local water profile instead (which is one of the points that Joe makes in his article).

Even the yeast played out alright: I didn’t like S-23 in the past because it can produce rather fruity fermentation byproducts. In this case, the beer came out fairly clean, just with a high final gravity. In retrospect, that actually wasn’t too surprising, as S-23 is the closest known relative of the Wyeast 2001 strain, which is purported to be the Pilsner Urquell “H” strain, and is also known for relatively low attenuation.

My take-aways of this brew are the following:

Hard water, especially similar to the Mönchsambacher water profile, can get you a flavour profile in beer that is similar to the slightly rustic flavour profiles of beers like Mönchsambacher, Knoblach and others in the region.

The combination of Pilsner and Vienna malt probably adds to that rustic character.

A lower attenuation seems to help with the style, but another experiment to try out a more highly attenuating yeast should bring more clarity.

Aurum hops are probably okay for standard German styles, but also require more experimentation to understand their exact aroma potential and how to use them. A more recent harvest would be great to try next time. I generally support the idea of hop varieties that are better suited to climate change (which is an inevitability that will hit the brewing industry hard in the decades to come, so good on German hop breeders to be as forward-thinking as that), as long as we understand well enough how to apply the hops to get the same aromas and flavours as with more traditional varieties.

Homebrew Project 2020

Mahrs aU Clone, Iteration 1

1 Comment

As I’ve announced a month ago, my homebrewing project for 2020 is to develop a clone recipe for Mahrs aU. I brewed the first iteration on December 27, and here’s a quick report on how that went:

My grist was fairly simple, 2.4 kg Pilsner malt and 2.4 kg Munich I malt, both from Weyermann. I doughed in with 20 litres of water which I had treated with 6 ml of lactic acid (something something counteract local water’s hardness, I calculated this ages ago and stuck to it since, I’m terrible with water chemistry). I then conducted my preferred double decoction mash where I start with a very large thick decoction, rest it at around 70°C when heating up, then boil it and mix it back, and then pull a thin decoction, boil it, and mix it back again. The temperature steps the main mash was undergoing in this case were 39°C (dough-in temperature), 62°C and 72°C. I then moved the mash to the lauter tun, recirculated to get clear wort and then started lautering, measuring 17.6 Brix on the first runnings. I continuously sparge, and after 30 minutes, I had enough wort collected to bring it to a boil. Pre-boil gravity as determined by refractometer was 11 Brix.

After 90 minute boil, with hop additions (2019 harvest Perle, 7.2% alpha acid) at 90 minutes (32g) and 10 minutes (10g) before the end of the boil, I moved the wort the a fermenter and chilled it down to 10°C before adding two rehydrated sachets of W-34/70 yeast. Fermentation took off about 36 hours after pitching the yeast.

Things didn’t go quite as planned, as the wort’s OG turned out to be 13.3°P. My grist calculation was based on previous batches in which I employed the same mashing regime, but this time, for whatever reason, efficiency was quite a bit better than what I wanted to achieve. I’m not going to complain about it, as it’s only 0.6°P (roughly 0.002 in gravity points) higher than where I wanted to get to.

Fermentation was quite a bit slower than usual. I’m not sure why, but I may have rehydrated the dry yeast improperly, therefore slightly underpitching. Even after four weeks, fermentation did not seem quite finished. I would have waited, but two days ago, another problem came up: the seal on the fermenter’s tap was leaking. Not a whole lot, but in the tray underneath, a few hundred ml of beer had built up. Since this was just the first iteration, and things weren’t 100% right from the beginning, I decided for an emergency plan: instead of lagering properly, I just cold-crashed the beer, hoping not more would leak between Friday and today, and then bottled the beer today. With 4.4°P, I’m not 100% sure the beer had completely finished fermenting (though there wasn’t really much of a noticeable residual sweetness), but I bottled it anyway and just let it naturally carbonate over the next few weeks. To be frank, I’m not particularly happy with that, but given that things weren’t right from the beginning, it’s still an acceptable shortcut for a first iteration.

A quick taste test directly from the fermenter revealed that the beer indeed tasted remarkably close to the original Mahrs aU. While the beer was pretty clean-tasting, the hop bitterness still seemed a bit rough. My concern is that whenever this has aged a bit and smoothed out, the bitterness might not be quite high enough. Bear in mind that I based the hop additions and overall bitterness just on some very flimsy circumstantial evidence. Since then, I’ve received multiple pointers to other sources talking about the Mahrs aU parameters, such as this very helpful comment by DB who mentioned that Michael Jackson had written about aU in 1999, when it was still bittered with Northern Brewer and Hallertauer Tradition to 36 IBU, and this reply on Twitter pointing out that Ron Pattinson’s book Decoction! (seriously, you should buy it) contains OG, ABV, real extract, apparent attenuation, colour and bitterness from the year 2000 for “Mahr’s Bräu ungespundet – hefetrüb” (aka Mahrs aU), setting the bitterness at a whopping 42.5 EBU.

I’m still waiting for the bottled beer to condition and slightly mature before I’ll start a proper taste test the results of which will then be the basis for my next step. Based on the brew day itself, I will probably slightly reduce the amount of malt the next time so that I can get closer to 12.7°P in OG than the 13.3°P of the first iteration. Also depending on how the bitterness shines through, I may slightly increase the bittering hops.

I also chucked out that particular fermenter and ordered myself a 30 litre Speidel fermenter instead.

Homebrew Project 2020

My Home-Brewing Project 2020: Cloning Mahrs aU

12 Comments

Possibly one of my and my wife’s favourite beers from Bamberg in Mahrs aU. It’s an ungespundetes (unbunged) pale-amber Kellerbier, meaning that it was carbonated without spunding, taking up only the amount of CO2 the liquid would do under the current temperature and atmospheric pressure, and is normally served unfiltered.

Brewing ungespundetes Bier is traditional for Bamberg, and has already been described as being particular to the city’s brewing practice in the 19th century (e.g. Das Bierbrauen in allen seinen Zweigen by Johann Muntz from 1840, p.121). It also shows in the name: “aU”, which some people call the shortest beer order in the world, is local colloquialism for “ein Ungespundetes”. The Mahr brewery in Bamberg, located in the city’s district “Wunderburg” and just opposite of Keesmann brewery, has been around since 1670, and has been owned by the Michel family since 1895. They are one of the traditional Bamberg breweries, and not just because of aU, they have a faithful followership locally and around the world.

A glass of aU, straight from the source.

If I want to brew a clone of Mahrs aU, I first need to find out about the beer’s specifications. Fortunately, I found a video from 2010 about the production of aU that contains some information: https://www.youtube.com/watch?v=Ubmm2j-2Uq4 (embedding is unfortunatey disabled, so you’ll have to click the link to view it)

From there, we can extract quite a bit of information about the beer. It’s a bottom-fermented beer, with an original gravity of 12.7°P. The grist consists of Pilsner malt and Munich malt of unknown ratios. The sole hop variety used is Hallertauer Perle with an unknown hopping schedule or bitterness. The final beer contains about 5.2% alcohol by volume and is carbonated with 4 grams per litre of CO2.

This is already quite a bit of information that we can work with and construct a recipe out of. Of course, I don’t expect the result to be perfect, so this is just a rough first outline. Further changes to this are up to the result of the first brew.

For the grist, I will start with simply 50% Pilsner malt and 50% Munich I malt. I couldn’t find anything useful about the mashing regime employed at Mahrs, so I’m simply sticking to my preferred double decoction scheme, hopefully ending up with a wort of the right OG.

For the hops, I’m sticking to just Perle. I managed to buy fresh 2019 harvest cone hops, which in my case have 7.2% alpha acid. I estimate the bitterness of aU to be roughly around 30 IBU. This is not just based on taste testing, but also on a rough bitterness scale that Mahrs provides on their website: the Helles has 3 hops, the aU has 4 hops, and the Pils has 6 hops. The following is a lot of conjecture, but bear with me: if the Helles has roughly 20 IBU and 3 hops, then 1 hop in their proprietary scale is equivalent to 6.66 IBU. If we assume the Pils to have around 40 IBU (it is rather bitter) and 6 hops, we get the same result: 1 hop is equivlaent to 6.66 IBU. In that respect, the scale seems fairly consistent. Based on this, a beer with 4 hops bitterness should be equivalent to about 27 IBU. Now I dare anyone to correctly blind-taste the difference between 27 and 30 IBU.

So for a first test brew, I decided to simply add to hop additions, one at the beginning of the boil (90 minutes) at 1.6 g/l, and a late addition just 10 minutes before the end of the boil at 0.6 g/l. This should add the desired bitterness of about 30 IBU, and leave behind some hop flavour and aroma, without being too hop-forward.

I will ferment this beer like any of my other bottom-fermented beers at 10°C, and then briefly lager it. In the beginning, I will simply use W-34/70 yeast. I do think though that Mahrs house yeast strain has quite the impact on the overall flavour of the beer. This is noticeable in direct comparisons of their aU and their Helles which I’ve conducted, and the Helles has some underlying subtle spiciness that is very present in the aU and can’t simply be explained through the malt (the Helles most likely contains no darker malts, unlike aU) or the hops (the Helles is hopped with Hallertauer Tradition and Mittelfrüh, whereas aU is all Perle). A future point in experimentation is to see whether there is any live yeast in bottles of aU which could be grown and used to ferment later batches, or whether other commercial strains show better results, in case straight-up W-34/70 doesn’t cut it.

This is my hypothesis how an aU clone could be constructed. Whether any of speculations hold true will show when the first test brew is ready to drink. My plan is to brew it with the next few days, which means it will be ready in about 6 to 8 weeks time. I’ll keep you updated here.