Sunday, 20 September 2009

Weird brewing practices (part 92)

Every now and then I stumble across something in the brewing records I just don't understand. That's understand the why not the what. The latest is from Truman.

They started it in April 1917. When they scrappped their whole beer range. But they were still at it in the 1930's. For the life of me, I can't see the point. It's the weirdest form of party-gyling I've seen.

Truman's new range of beers was mostly made up of X, XX and XXX. As you would expect, they party-gyled them in various combinations: X and XX. X and XXX, XX and XXX.

Nothing weird about that. You get three or four gyles from the mash and sparge that you blend together to get two beers of different gravities. It's bog-standard stuff. But Truman's took it a step further.

It's probably easiest to explain with an example. This is a brew from the 6th June 1917.

The first four rows show the three gyles that were combined to create XXX.

26 barrels @ 31 lbs per barrel (1085.87)
22 barrels @ 14 lbs per barrel (1038.78)
21 barrels @ 6.1 lbs per barrel (1016.89)
blended to make a single wort of:
69 barrels @ 18 lbs per barrel (1049.86)

The next four rows are the blend for XX:

26 barrels @ 31 lbs per barrel (1085.87)
44 barrels @ 14 lbs per barrel (1038.78)
65 barrels @ 6.3 lbs per barrel (1016.89)
blended to make a single wort of:
135 barrels @ 13.6 lbs per barrel (1037.67)

The weird bit is what happens next. It's on the next two lines. The giveaway is the comment "Rack as"

XXX 135 barrels @ 15.8 per barrel (1043.76)
XX 69 barrels @ 13.6 per barrel (1037.67)

What they've done is take 66 barrels of XX and mix it with the XXX, dropping it's gravity by about 6º. That's the effective OG, because the blending was done after fermentation. That's clear from the rows at the top right of the image. They give tun details. 7, 8, 9, 13, 14, 15 are the tun numbers. The middle number is, I think, some sort of dip measurement. The last number is the SG. For some reason given not given in pounds per barrel like all the others. You can see that, in the fermenters XXX was 1049 and XX 1037.

My question is this. Why didn't they just blend the worts to the strength they wanted before fermentation?


Graham Wheeler said...

There is nothing in that snippet that convinces me that it is a blending exercise. I would call it an un-blending exercise if anything.

It would have helped if an example had been chosen in which the brewer concerned could actually multiply and transcribe accurately.

What it tells me is that xx and xxx were made from the same physical mash, split into two worts or gyles, one for the XX one for the XXX, and the rest is an accounting exercise to calculate the yield from the malt, and apportion that yield to the two beers for excise purposes. It seems that the brewer screwed up.

The top three-row group and the next three-row group are confusing because it is not clear what exactly is going on. They could not have been triple mashing at that late date. They seem to be successively weaker worts taken off the mash during sparging. May be, during sparging, when they had collected a certain volume (more likely a certain gravity) in the underback, sparging was stopped and the collected sweet-wort was split in a fixed proportion into two copper-loads. Sparging was then restarted and at another point it was stopped again and the sweet-wort apportioned accordingly, and so on. There is a slight difference in the gravity of the last wort for each beer, but nature does not always oblige with the expected numbers so a bit of adjustment would have necessary.

Alternatively it could be a single collected sweet-wort of 1085, and liquored down, but why in three stages.

It is all a bit unusual for either, and I have no idea what is going on in the two right-most columns.

Anyway, you end up with two copper-loads, one for each beer.

What happens next is not a blending exercise, but a multiple cock-up.

I guess the brewer was trying to back-calculate what was supposed to be in the fermenters from the information already obtained, as a theoretical exercise perhaps, but he transcribed the numbers from the totals to the bottom columns incorrectly, so it went wrong.

The first obvious whoopsie is that the quantity of XXX suddenly jumps from 69 barrels to 135 barrels, and the XX vice-versa; that cannot happen. The second obvious whoopsie is that 135 x 15.8 does not equal 2140 (or 938 for that matter). The total volume and the total pounds extracted are identical to the totals given above, so everything should tally, it doesn't. The brewer noticed that something was wrong and switched two numbers, but that didn't fix the problem either, because he switched the wrong numbers. Had he transcribed 135 @ 13.6 into the XX row, instead of 69, things would have tallied.

The whole exercise was about the very last number on the sheet. You will be pleased to know that the brewer thinks that he got an extract of 93.3 lbs/qtr , which means that he mashed 33 quarters perzactly. Poor sod didn't even get that right, because 65 x 6.3 does not equal 416. I wonder how long he held down his job.

It does not represent a real-world brewing; there is no accounting for evaporation loss during the boil, nor any liquoring down, etc. They are calculations either before or after the fact. It may be pre-brewing calculations, but my guess is that it is just an efficiency calculation, and, from that calculation, the extract for each beer can be calculated as a double-check. "Racked As" may mean "Declared As" for the Excise "Red Book" that the brewer had to fill in.

Ron Pattinson said...

Graham, this isn't a brew that went wrong. I've dozens of examples of them brewing this way spread over many years.

The first stages are bog standard. Mash and sparge giving three worts which were then boiled separately and blended post-boil to get an XX and an XXX wort.

It's a particularly confusing example because of the volumes. I should perhaps have picked one where they didn't exactly flip.

Do the calculation. (Admittedly, it is a nominal calculation as the worts had already been fermented.) But 66 barrels of XX at 1037.67 and 69 barrels of XXX at 1049.86 gives XXX 135 barrels at 1043.76.

There are a couple of slight errors in the maths, but nothing significant. And they did mash exactly 33 quarters.

Gary Gillman said...

Ron, there are a couple of reasons at least why they might have wanted to blend post-fermentation.

First, you might get a more accurate final gravity. This may have been important for legal or taxation reasons.

Second, you would have more control over the final palate. Even in those days, I believe some beers were designed to have a "profile", to use modern packaging goods terminology.


Ron Pattinson said...

Graham, "Racked as" is pretty damn specific. I've seen "Declared as" in records before.

No need to take evaporation into account, because the blending is post-boil.

Gary, in this period final gravity and ABV were irrelevant for tax purposes. Only the OG counted. For tax purposes the number of standard barrels were what counted. A standard barrel is 36 gallons of beer with an OG of 1055. Tax was calculated on the number of standard barrels brewed in a month. Some records give the number of standard barrels in the brew.

The blends are made to hit a specific notional SG. The FG's vary between brews. (I have multiple examples.)

Palate reasons? Could be. Though I've examples of mixing to dilute and to strengthen.

Tomorrow I'll post a clearer example.

Gary Gillman said...

Clearly tax was not the reason then, but they may have wanted to hit with accuracy the final ABV - and this may be connected to palate too.


Graham Wheeler said...

I was not suggesting that it was a brew gone wrong, but calculations gone wrong.

There are so many anomalies that I though it was important to determine the purpose of the document to get an understanding of what is going on. My conclusion was that it was all about that magic number of 93.3, and column four was the only important column, the other columns are a means of calculating the numbers for column four. The sub-total volumes and sub-total lbs/brl are not important and they seem to have been filled in later, in pencil perhaps, at the point when the brewer discovered the final numbers were wrong. The density of the print is different for those sub-totals.

So we are now saying that it is a dilution exercise to bring the gravity of XXX down to within wartime restrictions or wartime conditions. That at least makes some sense, but still strange. Presumably the XXX at the top of the page is its pre-war gravity. We have to ask why did they keep the same name for a weaker XXX. It should have been called "two-and-a-half-X". By diluting this way they have ended up about face, with more XXX than XX. I would have thought that their sales figures would have been the other way round; the cheaper beer being the best seller. Mind you, a 1.037 beer would have tasted like water to a bloke of 1917, war or no war, so perhaps they did sell more XXX. Dread to think what X must have been like.

Ron Pattinson said...

Graham, neither X, XX nor XXX were brewed before April 1917.