Thursday, 22 October 2015

Autumn 2015 Blending: Pale Sours

[Edit: The pale sour described in this post won gold in the European Sour category at NHC 2016.]

Over the last year or two I've changed the way I think about brewing sour beers. In the past I'd brew a batch, wait till I thought it was ready, perhaps add fruit, and then package.  But as I read more about the practices of professional sour brewers in books like American Sour Beers, and interacted with other homebrewers in forums like Milk the Funk, I started to think that this wasn't the best way to approach sour brewing.  Instead of thinking about each batch as an individual beer, I switched to thinking of them as elements in potential blends.

In some cases this involved simply saving batches for this purpose, but I also deliberately started a number of five and six gallon soleras, using different combinations of brettanomyces and LAB in each with the intention of always having a range of different sours available for blending.  I've already started using some of these beers as a small element in my saisons (i.e. as part of a bière de coupage), but this month I finally got round to creating my first proper blends.  In this post I'll talk a bit about my experience and describe the pale sours I made in my first session.  In future posts I'll talk about dark sours, and a few bières de coupage as well.


I was always planning to do this blending at the start of Autumn, so in the weeks and months prior to the final session I made various preparations.  First, I had to brew top-ups for the soleras.  I stuck to the same base recipe, but used aged hops, added some oak cubes to the primary, and used US-05 as my yeast instead of my preferred choice of Wyeast 1318.

About two weeks before blending, I took a small sample from each beer, and spent an hour writing some basic tasting notes.  My main goal was to check that each beer was ready for blending, but this also gave me some idea of what I was working with, and gave me a basis for beginning to think about potential blends.

I knew that blending for the first time was going to be pretty difficult, so I decided to set some basic parameters.  First, I decided on the beers I would be blending in advance: five gallons of pale sour, to be bottled immediately; four gallons of pale sour, to be transferred onto cherries in a five gallon carboy; and three gallons of pale sour, blended or separate, for use in cutting beers over the next few months.

With this baseline in place, I also decided to think of the blends in terms of one gallon units, with the possibility of going down to half gallons if I thought it was necessary.  For instance, the five gallons of pale sour would consist of five parts.  All I had to do on blending day was decide what those five parts would be.

A few days before blending I took samples that were large enough to provide gravity readings from each beer.  These provided me with plenty of beer to check my tasting notes and start experimenting with possible blends.  I measured blends using a syringe, usually taking 4ml of beer for each part of the blend.  With the pale sour, for instance, I would take five 4ml samples from the various elements, and blend them in a single glass.  So one blend might have been 8ml Roeselare, 8ml Mélange, 4ml ECY20, and so on.

To be honest, I found it quite difficult to settle on a final blend.  It was usually easy to tell if something didn't work, in part because I already felt like some of the base elements were better than others, and tended to prefer blends where these made up most of the whole.  But beyond this it was difficult to find criteria for choosing between acceptable blends, especially since, given the relatively high finishing gravity on at least one element, I expected further fermentation to take place in the bottle.

An added difficulty came from the speed with which I found myself getting palate-fatigue, even with regular breaks.  Sometimes the same blend would taste completely different ten minutes apart---I wonder if this was partly due to the base beers opening up a bit as they sat out on my desk.  I mitigated this to some extent by testing the blends a few days before bottling.  This meant that I could taste my final blend with a fresh palate when I packaged the beers, and check that I hadn't gone horribly astray.

I eventually settled on blends that were identical to the ones I'd envisaged on my initial tasting.  They tasted fine, and I figured that observing how the base elements contributed to the character of the final beer after some conditioning time would help give me a better basis for future attempts.

The Base Beers

Roeselare Solera
Gravity: 1.007
Start Date: 13/10/13
Notes: Stone fruit, honey, pencil eraser.  Light to medium sour.  Strong component in blends.

ECY20 (2014) Solera
Gravity: 1.001
Start Date: 28/11/14
Notes: Grainy, lemons, slight plastic.  Medium+ sour (mouthwatering).  Works well as sour note in blends.

Mélange Solera
Gravity: 1.001
Start Date: 2/4/15
Notes: Strong aspirin/medicinal note, soft fruitiness behind it.  Medium sour.  OK component in blend but medicinal note is a little strong.

ECY01 Solera
Gravity: 1.000
Start Date: 28/8/14
Notes: Woody, minty, distinctive.  Bitter.  Light sour.  Works well as a small component in blends but decided to leave this out to age longer.

ECY20 (2013) Adjunct Sour
Gravity: 1.004
Brew Date: 17/12/13
Notes: Soft barnyard, apples, slight plastic.  Light sour.  Nice component in blends, though perhaps a little oxidized?

Pale Sour Blend

This ended up being 2 parts Roeselare Solera, 1 part ECY20 Solera, 1 part Mélange Solera, and 1 part Adjunct sour.  The Roeselare provided a nice base flavour; the Adjunct sour rounded this out with some soft barnyard funk; the Mélange added another layer of complexity; and the ECY20 enhanced the sourness without making it overwhelming.

Since the Roeselare, which made up two out of five parts of this blend, had a relatively high final gravity of 1.007, I decided I should allow for further re-fermentation in the bottle.  To accommodate this in my priming sugar calculations, I made use of Jeffrey Crane's very useful spreadsheet.  I had to guess the highest temperature the beer had been stored at (my apartment does not have air-conditioning, so it might have been quite high).  I also decided to aim a little high in my desired carbonation, to around 3.5 volumes.  As far as I can tell, the calculation is based on the assumption that the blend will attenuate to the F.G. of the driest component.  I have no reason to believe that's incorrect, but I didn't want to rely on it entirely for a decent level of carbonation, so I aiming a little high made sense.  The various components were combined in a CO2-purged bucket, and bottled right away.


This consisted of 2 parts Adjunct Sour, 1 part Roeselare, and 1 part Melange.  I was aiming for something less sour (anticipating the contribution of the cherries), so I decided to leave out the ECY20 entirely, and use the lightly tart Adjunct Sour as the main component of the blend.  I was also hoping that the soft barnyard funk of this beer might emerge behind the cherries and provide a nice backdrop.

This blend was combined in a CO2-purged bucket, and transferred onto around 7lbs of cherries: a combination of Montmorency and Bing cherries that I picked up at the farmer's market earlier this year and stored in my freezer.  Ideally I would have like to use 8lbs, with more sweet cherries in the mix.  My reason for using a combination like this is that I tried some sour cherries grown on my Uncle's farm in England this summer, and I felt that they had a slight sweetness and depth that was missing in the American cherries I'd bought.  I hoped that adding in some sweet cherries might help approximate this flavour, but I realized at the last moment that I didn't have quite enough in the freezer.

Future Coupage

This left three gallons of beer, besides the 3 gallons that was carried forward in each solera: one gallon of Mélange, and two gallons of ECY20.  I transferred these to CO2-purged 1 and 1/2 gallon containers.  I'll be using them to cut saisons that I want to add a little tartness to.  I think the Mélange might bring out the fruitiness I've seen in some of my buckwheat saisons, and the ECY20 is tart enough to add an interesting dimension to a dry and hoppy beers.


  1. I definitely think that the beers open up over time. I have had the same experience as you multiple times. A few weeks ago I was also putting a blend together, and got pulled aside to handle some other family business. When I returned, the notes I had on the blend components were off the mark. Dissipation of stronger funky notes is one that I seem to notice pretty frequently.


  2. Great post, Amos. I still find blending difficult, particularly because the beer isn't carbonated.

    1. Carbonator caps and empty plastic soda bottles. Try it out.

  3. Nice write up. Im now planning my own solera project. Cheers.