Monday, October 22, 2018

A Recipe for Kriek from 1907

It’s been quiet on the blog for a while, but now I’m getting some time to write again. I wanted to start off with some more historic beer research – this time with info about kriek from Petit Journal du Brasseur in 1907. This volume includes two short sections on kriek production: one is an article from Robert Jordens and one is an answer to a brewer’s question.

Schaerbeek Cherries near Beersel, Belgium.
To start with, I think I should give a bit of general background. I'll be using kriek here to refer to lambic with sour cherries, rather than the cherries in general or any other base beer with sour cherries; however, the Dutch word "kriek" simply means sour cherry. The articles by Jordens notes that kriekenbier, beer with sour cherries, is made elsewhere but that kriekenlambic (much like lambic) is native to Brussels and the surrounding region. This article also refers to kriek lambic as “Belgian Burgundy”.

Oude Kriek and Kriekenlambic do have some legal protection (see also: here), which sets out ranges of fruit, some process/fermentation requirements, and some parameters of the wort. In addition, the "Oude" designation requires secondary refermentation in the package, conditioning on lees, and some characteristics (e.g. minimum acidity) of the final beer. Both designations appear to allow the use of cherry juices or concentrates and, interestingly, set a maximum cherry percentage which at least some commercial producers using the designations exceed. I found these surprising, and perhaps there is some updated legislation surrounding these.

A barrel at Cantillon with a painted square
denoting it has a square bung.
Schaerbeekse cherries are recommended. The article notes that some use Cerise du Nord (perhaps these are what you can now find called griotte du nord or Chatel Morel), sometimes mixed with Schaerbeek cherries, but that Cerise du Nord were not usually used alone as they are less fine and more acidic. The cherries are used whole (with pits) and fresh. Specific ratios of cherries are discussed below. The texts also recommend adding a sugar syrup (1/4-1/3 L per HL of beer) and a small amount of Ceylon cinnamon (10g / HL) with the cherries. The inclusion of cinnamon here is quite interesting. Cherry pits can give a character often described as almond-y, and which I perceive as also being cinnamon-y. So the inclusion of a bit of cinnamon, to “enhance the cherry flavor" (original French: "…cannelle de Ceylan pour relever l'arome de la cerise”) makes sense. In case you aren't aware, there are multiple types of cinnamon. Most cinnamon available, either ground or as sticks is Cassia cinnamon. Ceylon cinnamon sticks are visibly quite different and have more of a woody and complex characteristic. Ceylon cinnamon is also sometimes called "true cinnamon". Randy Mosher's Radical Brewing talks about the differences in these two a bit as well. Back to the krieks - I’m not aware of any commercial krieks currently being made with this sort of addition, but I think it would be cool to see at a light level. And, with producers expanding their current offerings, especially with fruit lambics, I wouldn't be surprised to see something like this in the future.

The Jordens text notes taking care in choosing which lambic to use for kriek production, with preference being given to more mellow lambic and lambic that is clear. The Q&A article states that 3 year old lambic is too old to use on its own, and gives this ratio for lambics of different ages: 20% 3 year old, 50% 2 year old and 30% one year old. The cherries are then left to macerate for 4-5 months in wood, with one source noting that some brewers agitate the barrels for a month after active fermentation. At this time, seeing square bungs in barrels was common. This made it easier to get fruit in and out. Some barrels with square bungs can still be seen in some lambic cellars, though they are a rare find compared to the common round-bunged barrels now in use.

Square bungs on the front barrels - at Oud Beersel.
After 4-5 months of contact time with the fruit, the kriekenlambic would be bottled and bottle conditioning would last around 6 months, with the bottles stored at 15-17° C (59-62.6° F) or cooler. In this time the bottles would develop carbonation. There is no mention of blending in young lambic or adding additional sugar solution for carbonation. The bottles were ready to drink after about 6 months, such that the final beer would be ready around the same time as the production of the next batch. Jordens notes that bottles would age well for up to 5 years.

Edit 22-Oct-18: It was helpfully pointed out that I didn't include any recipe info for the base lambic. I overlooked this as I've written about it before, but I should have included it here as well. These sources didn't specify, but the lambic would likely have been at least 40% unmalted then, perhaps toward 50%. The remainder would have been malted barley, with the malt pale but likely darker than modern pils and made from winter 6 row barley (escourgeon). Hopping would have been a mix of aged and current harvest hops. I've written a good deal about historic lambic hopping here so check that out if you want more info there, and see this post (with some other ancillary info here) for additional historic lambic production (though from a bit earlier - ~1850).

Fruiting Ratios
These sources recommend 160-200 g of cherries per L beer (1.34-1.70 lb/gal). To put this in to context, many modern krieks generally have somewhere between 200-300 g/L (e.g. Cantillon and 3 Fonteinen), though in some cases this can be as high as 400 g/L (Boon, Oude Beersel, some 3 Fonteinen). To confuse things, the fruit ratio can be and is calculated differently by different producers. For example, 250 g/L could mean:
  1. For 1 L of beer used in maceration, you add 250 g fruit. This can unambiguously be used, as it is in this article and as noted in this comment (in the context of fruiting ratios in the barrel of roughly 1 kg/L). But this may not be what is reported in the final ratio on packages.
  2. For ~1 L total volume, you have 250 g of fruit (or roughly 750 mL of beer + 250 g fruit, such that the aging vessel is 25% full of fruit) (e.g. here).
  3. For a beer and fruit mixture, in the final product after losses and blending, you get 1 L for every 250 g fruit used (e.g. here). If you aren't blending after aging, this is similar to method #1. The key key difference is losses when removing the lambic, and anyone that has aged a beer on fruit knows that the volume they get off of the fruit is less than what they put on. The first method doesn’t account for this. To me this method seems like the best for reporting final fruit ratios, but it is also the least direct from a brewing/fermenting point of view since you don’t reach the quoted ratio until the end when you know the volume you get off of the fruit and what you blend it down to (if you are blending).
Maceration in oak at 3 Fonteinen:
480 kg cherries and 500 L lambic.
At least the described ratio is clear in this historic article - 160-200 g per liter of lambic. Fruiting ratios are a bit smaller if expressed in methods 2 and 3 than method 1. For example, this historic kriek at 160-200 g/L by method 1 is 138-167 g/L by method 2.

Comparison with modern kriek
Fruit Ratio
Without knowing exactly which method different producers are using, and without knowing the losses assumed in these recipes to convert to method #3 above, it isn't possible to make an exact comparison with modern production. Even without this option, we can see that these fruiting ratios are on the low end or are lower than typically found in modern production (200-300 g/L or more, and with at least some producers using methods 2 or 3 from above to calculate ratios). In addition, with modern production, it is common to age beer at a higher fruit ratio and blend back to the desired level before packaging. There is no mention of blending back in these articles, but going with the assumption that there was no blending back of these fruit ratios from ~1907, these are still lower than modern usage.

I know of a couple modern beers with cherries from lambic producers that are around or below these ratios. Girardin Kriek is reportedly 150 g/L. Cantillon Kriek used 200 g/L cherries, which could be at the high end of this historic range depending on which method to quantify the ratio is used. Additionally, Cantillon Zwanze 2014 (technically not a lambic since it was Iris as a base) used a rate of 120 g/L. While there were other assertive characters to this beer on top of a spontaneous fermentation base (dry hopping and the stronger malt character of Iris), the cherries were definitely present. I think working at these lower fruit ratios are worthwhile and could be a good way to let other interesting characteristics come through while still having some cherry. Especially when the base has more character and/or when other additions are made.

In disclosure, this suggestion serves my own goals/preferences. I prefer a complex base lambic to a fruit-forward lambic. So if I am choosing something for myself to enjoy, it will almost always be unfruited lambic. And, in general, I have rarely had a fruit beer which I thought was better than (or even equal to) the sum of its parts (in fairness though, I don't know exactly what the lambic tasted like before going onto fruit). Not that I am unhappy to drink fruit lambic, but good lambic and good fruit are almost always better to me than good fruit lambic. However, given the general commercial popularity of fruit-forward beers over their base constituents, I may be in the minority here.

Spent cherries from a maceration in steel
at 3 Fonteinen.
Maceration & Lambic Age
These articles recommend an average lambic age of 2 years and 4-5 months maceration time. Modern production methods can vary considerably, and these historic lambic ages and maceration times are within the range of modern producers. Fitting with the lambic-cherry balance suggested by the lower fruiting ratio in the ~1907 recipe, the use of lambic on the older side suggests that the lambic portion of krieks from the beginning of the 1900s may have had more character. Presuming of course that this source is representative. For comparison, here are some modern commercial examples:

3 Fonteinen – Maceration lasts for around 6 months to 1 year (see here and here), depending on the specific product - the 3 Fonteinen Oude Kriek has a maceration around 6 months (see also herewhile maceration for the Schaerbeekse Kriek may last up to or more than 1 year (see also here, with bottle dates suggesting a typical maceration of 6-10 months assuming fresh fruit is used and harvest dates in June/July). The lambic used is roughly 1 year old.

Cantillon – Maceration lasts around 1-2 months and 2 year old lambic is used (see here and here). This is blended back with a small amount of young lambic for bottle conditioning in Cantillon Kriek, while Lou Pepe Kriek is primed with sugar for bottle conditioning.

Others - Boon reports using lambic that is 1 year old for their Oude Kriek, and Lindemans reports using lambic that is at least 6 months old and a maceration time of 6 months for their Oude Kriek Cuvée René. According to Lambicland by Webb, Pollard & McGinn, Hanssens uses 1 year old lambic and a maceration time of roughly 1 year to produce their Oude Kriek.