Back in the old days, when industrial brewers had less technology and worse malt to work with, they used a method called Parti-Gyle to extract as much of the sugar from their grains as possible. One technique that was used, is the Parti-Gyle. Essentially, brewers would mash their grains, and run off a certain amount of high-gravity wort, for use in something like a barley wine or a strong ale.
Now, if you take a full batch-sparge session (where you rinse the grains a few times successively), the first third of of the wort you run off contains about half of the sugars. Which means, if you run off the first third, you’re left with a second third that is weaker, but still usable for making a second beer.
Or, you can “cap” the mash–add more grains, and re-mash ’em–and change the character of the second beer to be a little farther away from the original. For example, you could make a very pale, strong beer from the first runnings, and then cap the mash with some darker grains and make something darker for the second beer. And indeed, depending on how often you want to cap the mash, you could go on for a while this way. I think, though, once you get past three times, you end up dealing with excessive amounts of grain. But, that said, when I was thinking about trying this method out, I thought that three could be a very fun number to do.
Don’t get me wrong, I knew that three would be pretty intense, and that I’d better really plan it out. (For example, maybe doing a Kentucky Common styled sour mash on the third runnings, if you have an appropriate Mash-Lauter Tun (or MLT), because after boiling up two beers, you might not have the energy, or time, to boil up the third. Unfortunately, I don’t, because I used my brewpot as an MLT, and it’s big enough to handle a fair bit of grain, but too big to do a smaller sour mash in–there’s too much head space.)
So I realized that if I were to try this, I’d have to get up early in the morning, and work through the day until all three brews were safely ensconced in their individual fermentation vessels. At the very most, I could leave one batch of wort to sit overnight, either inside if I wanted to sour it, or outside in the cold if I didn’t, and then boil it and hope it didn’t sour too much.
The thing is, it’s a bit complicated calculating how much of which malt is contributing what to the second- and third-runnings batches of wort. A little judicious planning can get you somewhere, of course, but I wasn’t sure how far. So I opted for a very simple grain bill of the first runnings, to simplify recipe formulation for the second runnings; I decided to sort of just hope and trust in terms of the third runnings, though I did end up calculating what I think might reflect a single-batch version of the grain bill later on.
Using the Batch Sparge and Party Gyle Simulator, it’s possible to figure out how much grain, and how much water, to add at each step of the process. And if you make the first runnings mash very, very simple, it’s trivial to formulate a recipe for the second runnings– you simply formulate a recipe with the same base grain, add the other grains in the proper proportions, and make up the difference with more of the base grain. The real question is how to formulate the recipe for the third runnings. I think the bottom line there is you just have to calculate, or wing it. And for me, I’m willing to wing it and see what I get, see it as an opportunity to experiment a bit…
… but whaddaya know, a part of me wanted to try it, just the same.
So here’s the rough plan I wrote up for the brewing I did yesterday. First, the grain bill(s):
First Runnings: Belgian Golden Strong Ale
7.5 kg of Weyermann Pilsner Malt, mashed at 63 degrees Celcius, after a twenty minute protein rest at 50-ish (I can’t quite recall exactly the temp). (I’d have rather used Belgian Pale Malt, but you use what you can.) Ran off ~12.5L of wort, which came out to 1.056 pre-boil. Since I don’t have a gravitometer that can handle gravities higher than 1.060, I’m guessing (by my evaporation rate) that post-boil it was likely to be 1.066; I added a kilo of table sugar (not inverted, since my burners were all in use) to bring it somewhere in the mid-1.070s.
I ran the first runnings off into a bucket, and then closed it and let it sit. After about twenty minutes, I transferred the wort via the spigot into a second bucket, leaving behind about a half-liter or so, including most of the proteinaceous stuff that had precipitated out of the solution; I dumped the whole lot of it this leftover wort/goo back on top of the mash. Worty goodness stays, proteinaceous gunk gets filtered out. That’s the theory, anyway.
Second Runnings: On a Jag Abbey Dubbel
Here, I decided to cap the mash to match the revised version of my On a Jag Abbey Dubbel recipe. To get the On a Jag right, I needed a final gravity of approximately 1.054, and the recipe calls for 0.2 kg of Quaker Quick Oats, 0.5kg of Medium Crystal Malt, and 1.5kg of Munich Malt. Now, to get to an FG of ~1.054, assuming 15% volume loss to the boil, I needed to add 4.5kg of grains in total. Minus the Oats, Crystal, and Munich, that came to 2.3 kg of Pilsner Malt, so I added all of that and and mashed again, this time at about 65 degrees Celcius. (Note: I may have omitted the oats. I don’t know for sure, but I may have done so.)
After a long saccharification rest, I sparged out about 21L of wort that was supposed to be about ~1.0443 in gravity, pre-boil. I think I overdid it, adding as much pilsner as I did, though, but I’ll get to that later. For now, I’ll just say that it may be apparent I added essentially the full grain bill of the dubbel to the mash; which is to say, I could have just brewed the Dubbel on its own. However, that would be to ignore the fact that there’s still plenty of good sugar that was left from the pilsner, and that I could get it out of the grains if I was careful; but also, since I used up some Munich as well as some speciality grains (the crystal, especially!) I wanted to see whether everything I added for the Dubbel could contribute (mildly) their characteristics to a third beer. Which brings us to:
For that third beer, I decided to do something rather different from the other brews: I have a bunch of juniper berries I brought back from Indonesia, and juniper beers often have rye in them. So I figured I’d make a kind of pseudo-sahti kind of beer. (I was thinking of using the rye to make a pseudo-kvass, to accompany the true kvass I’ll be making in a few weeks, but timelines and all make this much easier to pull off.) This brew will be lacking the flavor from the branches, I think–there are some juniper trees outside, but I didn’t think to grab some twigs and clean them for the mash. And it won’t have the traditional slight smokiness of a sahti, either. But I was thinking of trying for a lighter, more session-friendly beer, and thought just mild hopping and tasty juniper flavours would be the way to go.
Note that I was thinking of a lighter beer. Things didn’t work out this way, for reasons that (partially) baffle me. I’m not quite sure why, since I ended up with a wort at 1.058 for the third runnings. I think one reason may be that even though I didn’t add much grain to the third runnings (and had to remove some of it) I managed to drain half of the newly added grains very, very well. Next, I used that well-drained wort — boiled for a few minutes — to sparge the rest of the grains, which were, by this point, about nine kilograms’ worth. I think the heat of the boiled wort probably effectively mashed out all that I’d added, and loosened up some of the remaining sugars deeper in the grain bed. I also think, however, that part of why it’s so high is that I ended up adding a gallon or two of water, and didn’t manage to sparge out the last, really thin, wort that might have diluted the third runnings as I’d expected.
So while I was hoping for something about 4-4.5% alcohol, I’m actually likelier to end up with something in the range of 5.2%, depending of course on attenuation. (Since I mashed some of the grains low, and some on the middle of the temperature range, and did a cereal mash with the rye and some of the pilsner from the third runnings cap, I frankly have absolutely no idea what I’m likely to end up with.
For the record, I probably should have just done a cereal mash, and a minimash, and then run the results through the rest of the previous, easily sparged grain bed. But, oh, no, I had to be stubborn and see whether the MLT could handle 12 kilos of grain.
Newsflash: it couldn’t. The MLT is too vertical, and the grain bed was too damned deep. I did manage to get close to 20L out, but it took some doing, and I lost probably a gallon of viable wort that simple could not be sparged out. (Or, rather, I gave up on it. I suppose I could have removed some of the top grains and sparged them in another container, and then run that through the grain bed, but it was already quite late so I decided to just finish up with the 20L, so I could go boil the wort I’d already collected.
For the record, I capped the third runnings grist with a kilo of rye, plus 500 grams of chocolate malt and a kilo of pilsner. The pilsner was probably unnecessary, but I’d mashed out and wanted to have something with enzymes for the saccharification rest the rye was going to go through.
By the way, the gravity on the final runnings was a lot higher than I expected. (I think I must have gotten something wrong in the parti-gyle emulator, or else my efficiency is a lot better than I thought.) Instead of 1.037, like I’d expected, my third runnings were 1.058. And since I broke my better gravity meter, and had to use the poorer, cheaper one I had remaining, I have no idea what the gravity the first and second runnings ended up being.
This left me with a grand total of about 12 kilos of spent grain to pass on to my friend Mark. Which, as I say, makes it seem like the parti-gyle is a lot of work for questionable benefit, but in my book, the fact I can get the most out of the grains (using the caramel malt and Munich to lend character to two brews, not just one), and the fact that I can get three brews going in a single day, is nice. I don’t think I’ll do it again until I have a much more forgiving setup: the kitchen was hell all day, to be honest, and having only two burners on my stovetop makes it harder, even if I did have an immersion heater to get it going. Prior to yesterday, the most I’d ever brewed in a single day was 5 gallons — less than half of what I produced yesterday.
As for the Hop Schedule for each brew, I decided to use up some of the hops I had sitting around, especially since a lot of the flavor/aroma character would be lost in terms of the hops added at first wort. This was definitely me rooting through the freezer for hops to get rid of. I should note, though, that some of them probably lost some of their alpha acids, from being frozen for a while. I’m not sure how much, though. Probably not a really appreciable amount, but it’s worth noting.
For the Golden Strong Ale (which I am dubbing Abi Gyuhwan Strong Ale — click the link for the recipe): I hopped as follows on the 60 minute boil, shooting for around 40-42 IBUs:
First Wort Hopping: 10 grams Amarillo, 25 grams Green Bullet
30 mins: 13 grams Saaz, 24 grams Cluster
10 mins: 25 grams Belgian Saaz/Motueka (hybrid, 7%)
I also added the adjunct of 1 kilogram of white table sugar.
For the Dubbel (which is a reworked version of the On a Jag Abbey Dubbel, click the link for the approximate recipe): I had a a target of ~30-35 IBUs on a 90 minute boil:
First Wort Hopping: 15 grams Cluster, 4 grams Hallertau, 22 grams Northern Brewer.
I also added the adjuncts of 700 grams of dark date honey (double the original, for double volume), and 1 kilogram of Indian jaggery.
For the Rye Brew (the work-in-progress name of this brew is Supposedly Sour Pseudo-Sahti): I haven’t boiled this one up. Instead, I was so tired I put the wort into a sealed fermentation bucket with an airlock on the top, and left it on the hot floor (ondol-heated) to see what would happen. Rye is one of the grains that goes into spontaneous fermentation the most easily, and there was a kilo of it in the mash, plus lactobacillus lives on the husks of most barley, so it stood to reason that there was a good load of viable microbiota in the mash.
Sure enough, in the morning it had started to undergo lactic fermentation: there was some sourness and funk already, and a thin head of bubbles on the surface. So I moved the wort into another clean bucket with less headspace, and pitched some raw pilsner grain (more lactobacillus, to boost the microbe culture) and then covered the wort in some sanitized foil. The foil is to keep out the oxygen: basically, if you want a nice clean sourness, you limit headspace and cover the wort if possible with something like this. The foil very nicely floats on the surface, instead of sinking, and holds out oxygen that would be used by other bacteria we don’t want propagating in the wort.
We’ll see how it goes. I am thinking of boiling this up sometime tomorrow, but since I’m doing this a bit like a Berliner Weisse–the traditional, no-boil method of souring–I’m very curious to see just how far it can go. However, I don’t think I want to go all the way. I was looking for some sourness, not brown lemonade, and the wort was so beautifully delicious as it was, right out of the mash tun, that I had second thoughts about souring it at all. So I figure probably tomorrow morning, or evening at the latest, I’ll be boiling this up with its juniper goodness. I may scorch or dry roast a dozen or so of the crushed juniper berries, for just a touch of that acrid smokiness, but again, I’m going for a faint effect, not something overpowering.
The hopping is going to be very, very simple… and minimal: 18 grams of Amarillo at first wort (shooting for about 10-12 IBU), and an amount of crushed juniper berries yet to be determined, but likely around 40-50 grams. I may add other herbs, we’ll see. Randy Mosher recommends bread yeast for a sahti (as well as smoking some of the rye malt), but I did without the smoke, and also without the juniper branches in the mash tun, so I think I’ll do without the bread yeast too.
(If I’d gotten 30L, I might have done a half-and-half to see whether bread yeast made a drinkable product, or even soured half and fermented with Brett C, and done the other half straight with Cry Havoc… but as it is, I think I’ll stick with souring it mildly, and then boiling it and pitching Cry Havoc. I may also reserve a gallon of it, after putting it in secondary, to inoculate with Brettanomyces Claussenii; I’d like to see how it handles secondary fermentation.)
And that, my friends, is an insane Monday Brewday. But hey, I have Mondays off. Mondays are my Sundays.
Still, I don’t think I will be doing parti-gyle for a while… at least, not alone.
UPDATE (18 December 2010): I bottled the Pseudo-Sahti last night. It is a bit odd–too little hopping, and the juniper flavor is pronounced but not overwhelming, with a mild sourness in the background, behind a strong maltiness. I’m very curious how this one will turn out, but I’m afraid it might end up being on of those beers I wish I’d made a half-batch of, instead of a whole. Which makes me think that in future, I’ll likely end up investing in a few 3-gallon carboys and making such test batches for all the new recipes I try.
That said, I don’t think it turned out badly, and maybe with some carbonation I’ll actually find it more refreshing.
The Golden Strong Ale and the Dubbel are waiting to be bottled. I’ll probably try to borrow a capper and put them all into smaller bottles, saving my big glass bottles for the Berliner Weisse that I’m about to rack to secondary. The advantage of the smaller bottles is that these beers are higher-alcohol (both in the range of 7-8%, I think) and it’s hard to find an occasion on a random evening where one would want to drink a full liter of beer that strong. Also, when beer is in bigger bottles one tends to be less able to hold onto what one really likes… giving a friend a sample means giving a friend something like 5% of the batch (one liter) at a time. It’d be nice to be able to give a 300 ml bottle, to see if they like it, and to be able to give samples to more people.
Anyway, I won’t even think about bottling those until after exams are done.
UPDATE (20 May! 2011!): Wow, I didn’t realize I had let the Golden Strong Ale sit that long. I kegged it today, and it tastes just gorgeous. Bulk aging — not just for meads, I’m thinking now, though who knows — maybe I’ll change my mind when I taste the chilled, carbed final product. To be honest, next time I’ll bulk age it in a keg, preferably purging out some of the cloudiness at the bottom so it ages without risk of autolysis. Having switched to kegging, I won’t be bottling this after all, though I will attempt to get it to Belgian-styled levels of carbonation.
It has a fruity, rich flavor which I was very surprised to discover, having expected something to have gone wrong in the carboy. (Which was why I left it sitting so long: I’d given up on it.) The attenuation was all the way down to 1.004, which makes this a nice dry beer, with an ABV of approximately 9.8% alcohol. When it is chilled sufficiently, I’ll add gelatin to clear it, run the gunk off, and post a shot of it. I plan on photgraphing more of my beers from now on.
I have to say: when I read the recipe, and saw that it called for months and months of conditioning, I was dubious: I thought waiting that long would be onerous, but it really wasn’t, except for how the carboy took up space. (Were it in a keg and tucked away, it would have been completely lacking in onerousness.)
On the other beers mentioned here: the sahti turned out, well, disappointing. I think sahti needs smoke, and this had none. I might try a small experimental batch again, if I can get fresh juniper berries/branches. But I’m far from eager to rush into it. I have many bottles of the stuff sitting, aging, though I doubt they’ll be improving much.
The dubbel, on the other hand, has been aging nicely. It’s dry, and has a caramely quality to it that I like, without the harshness of the quadrupel I ended up with the first time I tried to make it. But the standout is definitely the beer made with the first runnings, as far as I can tell at the moment.
UPDATE (15 Oct. 2011):
Well, the Sahti is long gone. Not sure I even have a photo of it to post here.
On the other hand, the Abigyuhwan Belgian Golden Strong Ale is now just about ready to be bottled. In small bottles. It came out very strong, at an estimated 8.9% ABV — though it tastes more like 10-11% or higher — but I’ve had a hell of a time clearing it. Several additions of gelatin and several weeks at near 0°C still haven’t quite gotten it clear. I may have to settle for it being slightly cloudy, since I want to free up this keg. (OR maybe it’s just that my tulip glasses make everything look a bit fuzzy… looking down through the beer it’s clear as glass. Abigyuhwan came out okay, but it’s a bit fusely, I think because I didn’t have any system in place to control the fermentation temperature so it got a bit hot when I started it, just a little over a year ago. That said, it’s still quite fruity and it’s nice and (relatively) dry, so I think I’m pretty happy with it overall.
As for the dubbel, I have a few bottles left, which I am guarding jealously. It’s pretty much exquisite now, a fact I attribute to the date honey (or date syrup, I think it really is). It’s drier, and much darker than the recipe suggests (again, because of the date honey/syrup), and the kind of beer one saves for a special occasion. In fact, though I’m obviously biased, I’d day it’s the best Dubbel I’ve ever had… though, well, it’s also a higher gravity than I originally imagined. (I’m guessing 8% or more.)
What can I say? It was one of my first all-grain brew sessions, and also my first attempt at parti-gyle. I didn’t know any of the things I know now… but it was an important step to getting the understanding I have now. (Such as it is.)
UPDATE (13 Nov. 2011):
Got tired of Abigyuhwan taking up a whole keg, and decided to blend the last remaining amount (maybe a gallon or less) into a remaining 6-6.5 liters of my Shiktakju. The result, so far, is quite glorious — a sweet, rich brownish ale with the fruitiness of the Abigyuhwan and none of the fusely pain, and all of the color and body of the Shiktakju.