Monday 26 April 2010

Parti-gyles

I'm easily annoyed. Very easily. And by all sorts of stupid things. Like party-gyling. And the misinformation spread about it on the web. Time to try and put that right.

Despite what many homebrewers believe, party-gyling (or is that parti-gyling? not quite sure of the correct modern spelling) is not using each separate running to make a different beer. That method of brewing disappeared about 1762.

And, despite what many beer writers have claimed, party-gyling didn't pretty much die out by 1800. It's a common feature of all the brewing records I've looked at from 1805 to 1965. That's real party-gyling. Where you blend 3 or 4 worts of different strengths in differing proportion to create several beers. The important point is that even the weakest beer will get some of the strongest wort. For reasons which Lloyd Hind will soon explain.

"Parti-gyles

Brews divided in the fermenting vessels to make worts of different gravities are known as parti-gyles. That described in the previous Section could, for example, be divided among other ways to produce:

100 barrels at 1050 = 5,000 degrees
100 barrels at 1040 = 4,000 degrees
200 barrels at 1032 = 6,400 degrees
                     15,400 degrees


The extra length can be obtained by adding more liquor to the coppers or by boiling the necessary length of trated liquor and passing it over the refrigerators to add to the wort in the fermenting vessels. It is possible to produce many parti-gyles from one copper length by adjusting the gravity of the weaker wort with liquor in the fermenting vessels. It is, however, more usual to collect more than one copper at different gtravities, and blend them in the fermenting vessels. The second and third coppers must then be of considerably lower gravity than the first. If the worts were divided in the coppers as they ran from the mash tun, they would necessarily be of abnormal composition, with proteins, salts, etc. very unequally divided. As a result, the break in the last copper might be defective, the fermentation of the weaker worts likely to suffer and the flavour of the beer not as good as it should be. A boiling fermentaion frequently occurs in a weak wort made in a parti-gyle with a strong one. It is consequently advisable to hold back some of the stronger mash tun worts for the later coppers, even though it may be inconvenient to equalise their gravities.

If the mash tun worts are run as in Table 193, the first copper would yield 9,484 degrees of extract and the second 3,814; 8 cwt. of sugar added to the first and 16 cwt. to the second, give 10,284 and 4,214 degrees, respectively. The "copper in" lengths, with 100 barrels of liquior in the first, would be 215 and 180, or about 206 and 173 barrels of cold worts, with gravities of about 1050 and 1024.4. The last gravity is rather low for the second copper. It would be better to hold back the first 20 barrels of mash tun wort for the second copper, when the "copper in" lengths would be 195 and 200 barrels. With 8 cwt. and 16 cwt. of sugar in the first and second coppers, the "copper in" gravities would be, on cold worts,

187 barrels at 1045.7 = 8,546 degrees
192 barrels at 1037.25 = 7,152 degrees
                        15,698 degrees

The hops are divided between the coppers in proportion with the extracts, say 245 and 203 lb. The worts collected in the fermenting vessels might be expected to be as follows, with 25 barrels of hop sparge and 80 barrels of liquor, taken at a specific gravity of 1004,

146 barrels 1st copper wort at 1056 = 8,176 degrees
149 barrels 2nd copper wort at 1046 = 6,854 degrees
105 barrels hop sparge and liquor at 1004 = 420 degrees
                                            15,450 degrees

It is difficult to work out the proportions of the two worts and liquor to make the three lengths and gravities required, since the gravity of the first copper only is known when the proportions in which it must be divided have to be decided. A first approximation can be found by the following simple method.

The differences between the gravities of the first copper wort, second copper wort and sparge, and the required gravity, taken in reverse order, give the proportions in wjhich the two worts and sparge must be mixed to give that gravity. Thus to obtain 1050 from a first wort of 1056, with second wort and sparge estimated at 1046 and 1004, the proportions are: 1st wort 46, 2nd wort 4, sparge 6. The respective lengths to make 100 barrels at 1050 would be 83.1, 7.1 and 10.8 barrels. This device cannot be used for the 1050, 1040 anmd 1032 worts simulataneously, since it takes no account of the available copper lengths.

The following figures are obtained by adopting it for the two stronger worts and assigning the surplus copper and sparge lengths to the 1032 wort.



1050 1040 1032
1st copper 82*56 = 4,592 62*56 = 3,472 2*56 = 112
2nd copper 8*46 = 368 10*46 = 460 131*46 = 6,026
sparge & liquor 10*4 = 40 29*4 = 116 66*4 = 264

100 5,000 101 4,048 200 6,402

100 barrels at 1050 101 barrels at 1040.1 200 barrels at 1032


It is obvious that the composition of the 1032 wort leaves much to be desired, on account of the excess of second ciopper wort. It is, indeed, essential to scrutinise all calculated blends in this way, since it many of the possible variations may be unsatisfactory. A reasonably proportioned bkend in one of the gravities may be accepted or fixed. The other two would be suitably adjusted. In this case, the 1050 wort is taken as the starting point.

The other two are adjusted by reducing the 1st copper length in the 1040 wort and iincreasing that in the 1032, with corresponding alterations in the sparge lengths to produce the required gravities, giving the following figures,



1050 1040 1032
1st copper 82*56 = 4,592 30*56 = 1,680 34*56 = 1,904
2nd copper 8*46 = 368 50*46 = 2,300 91*46 = 4,186
sparge & liquor 10*4 = 40 21*4 = 84 74*4 = 296
100 5,000 101 4,064 199 6,386
100 barrels at 1050 101 barrels at 1040.2 199 barrels at 1032

"Brewing Science and Practice" by H. Lloyd Hind, 1943, pages 738-741.

Is that clear? I thought not. It took a while to percolate through my thick skull..That bit about the proportions being decided by the difference in gravity betwwen the wortss and the target gravity in reverse order. Let me expand the description.

In the example Lloyd Hind gives, there are worts of 1056, 1046 and 1004 and the target gravity is 1050. So starting with the sparge wort, the difference  between thaat and the target gravity is 1050 - 1004 = 46. For the 2nd wort it's 1050 - 1046 = 4. For the first wort 1056 - 1050 = 6. Giving the proportions of 46 1st wort: 4 2nd wort: 6 sparge.

Simple, isn't it? I'll provide some real-world examples tomorrow.

14 comments:

Gary Gillman said...

The writer in the posted link does state you can mix some of the stronger wort with the weaker, or vice versa, to obtain a target gravity, so I think he is aware of parti-gyling in this sense. I think too he is focusing on the older sense of the term because of its relevance in the context of brewing at home or on a small scale.

I don't know the etymology, but seems logical that parti-gyle is a "part" of the gyle which could take in e.g., just the first runnings.

I was struck by his statement that 1/3rd of the wort produces a strong beer and 2/3rds, a beer of average strength, each component having about an equal amount of the available extract. This made me think of another explanation for "three thirds" (supposed antecedent to porter). Mixing the beers from those two fractions gives you all - 3/3rds - of the available extract in the malt.

If you combined the beers from three runs the same reasoning would apply of course, but it works with two beers also and the Bailey definition of three-thirds seemed to allow for the possibility of only two beers, as did the Denneston essay. In this way of looking at it, of course thread would be a derivation from third.

Gary

mentaldental said...

OK, I'll bite.

"That method of brewing disappeared about 1762."

That's quite precise for an about. What happened in 1762?

Ron Pattinson said...

mentaldental, OK, "yonks ago" would have been more accurate.

Ron Pattinson said...

Gary, but in parti-gyling as practised for the last 200 years there's no "can" about mixing the different-strength worts. That's the whole point.

Spencer said...

I'm a bit confused about your historical boundaries. From reading your blog for the last couple years, it appears that usage within the last 50 years is not sufficient to make something "correct", but usage that ended 250 years ago is also not "correct."

I don't think you can have it both ways, Ron.

Gary Gillman said...

I see that Ron. The whole focus of the article is on small-scale brewing but he may not be aware of how the parti-gyle concept evolved in commercial brewing.

Lloyd Hind is masterful as always, I can never read enough of him.

Gary

Ron Pattinson said...

Spencer, it's highly confusing to talk about a method of part-gyling no-one has used for 200 years or more. While at the same time ignoring the method that has been in widespread use in Britain over the same period.

That's why I keep getting into arguments with homebrewers who tell ne the one running per beer is the "accepted definition" of parti-gyling.

Real parti-gyling is a useful and efficient technique which is much more subtle. I just want top inform people about it.

Martyn Cornell said...

Gary, I'm not sure struggling with the etymology of party- or parti-gyle gets us anywhere, as looking through the Oxford English Dictionary it could be "party-" in the sense of "divided, separated" (as in "party wall", or it could be "parti-" in the sense of "partly of one sort and partly of another" (as in "parti-coloured": sense one gives us the "each mash brewed separately" meaning and sense two gives us the "mashes combined together" meaning. My own impression, from what I've read of brewers' practices, is to believe that British brewers used the term to mean BOTH those things, but Ron has read very many more old brewing records than I have …

Interesting idea that three-threads could mean strong and weaker mashes combined in the ration 1/3 to 2/3: this would, of course, make "three-threads" a synonym for "entire" as a combination of all mashes. But I'd have to see some early 18th century evidence for that …

Gary Gillman said...

Martyn, thanks, and my thinking is absolutely in the direction of three threads duplicating entire gyle brewing and even more so for the vice versa because entire brewing seems a later historical development.

I believe some evidence is in the Denneston essay. The pub in that case made a three threads from two beers, one strong, one weaker. I inferred - it's not direct evidence I know - that those two beers were made from the same set of goods. I believe the publican in that case was the brewer himself, in which case the inference is even stronger. The extra-strong beer would have been like Mosher's strong first run brew, the weaker one like his second, weaker beer. And, both three threads and entire/porter were sold for 3 p the pot...

Gary

Graham Wheeler said...

As far as I am aware, parti-gyling means, and always has meant, making more than one beer from the same mash or batch of grain. It matters not how this is or was achieved. There certainly does need to be any blending of worts. Indeed, there does not need to be any more than one wort; the other strength(s) can be achieved by liquoring down a single collection.

When blending does occur it is performed for consistency between batches by commercial brewers. It matters little whether it is blended before or after the boil. It is far more flexible if it is blended before the boil such that each beer has its own boil, because then variations in each beer can be made: roasted malts, roast barley and even crystal malt can be added to the copper, along with sugars and different varieties and quantities of hops.

This is quite difficult for small brewers to do, particularly those with only one copper, because they need to hold mash tun run-off somewhere in readiness for the next copper charge. It would probably require at least two underbacks to hold different run-off fractions for blending purposes.

Most brewers, even small ones, have several fermenting vessels, so it is quite common for the blending to be done post boil, in the fermenters. The disadvantage of this is that the resulting beers suffer decreased hop character and decreased colour with decreasing gravity; basically the beers all taste the same - they have no individuality - a criticism often levelled at many regional brewers in the late twentieth century, and some still deserve that criticism today.

Likewise I cannot see why the definition should not apply to beers produced by the old-fashioned triple-"mashing" process; there is little difference in end result between re-mashing and continuous sparging, apart from improved efficiency. Even Hind talks in terms of first, second, and third worts just like the old days.

The fact that Ellis mentions "Entire Guile Small Beer" in his book, which pre-dates porter and certainly pre-dates continuous fly-sparging, means by inference that parti-gyling must also have existed. It was not mentioned specifically because it was the standard way of doing things, and did not need singling out for special mention.

As far as I can see, Randy Mosher's definition is equally as valid as any given here; lacking in detail perhaps, but that's about all.

If you wish to wage all-out war on home brewers, there are far more serious "misunderstandings" to pick on than what parti-gyle actually meant.

Kevin LaVoy said...

Don't blame those homebrewers you're arguing with for the mistake. Practically every how to brew sort of book that we read says that the original is parti-gyle. Few of us have gotten esoteric enough in our brewing history to know about the other.

That said, I'm interested in tomorrow's post.

Ron Pattinson said...

Graham, Randy Mosher's description doesn't explain the method of party-gyling used by British brewers for the last 200 years. Like I've already said, the whole point is mixing the different strength. This isn't something you "can" do, it's the whole bloody point.

So do you think all the Fullers beers taste the same? They're a classic example of party-gyling.

Mike Anson (Killer Ales Webmaster) said...

I am confused about one issue - I am writing a brief article on parti-gyle beers, and was in the middle of a mild rant about the etymology of the term - parti-gyle (a portion or part of the wort) versus "party" (which seems to have originally meant a united group of people meeting for a common purpose). This blog, however, frequently uses "party," not "parti-" ... I am more than ready to bow to an expert opinion (Ron!) ... and I do NOT want to propagate an error. Help? Am I wrong about the etymology?

Ron Pattinson said...

Mike,

not sure which is correct, to be honest. I now usually use parti-gyle.