Soma at Moon Under Water Brewpub

It is probably the dream of most homebrewers to get to brew a big batch of one of their beers on a commercial system. In October Moon Under Water, one of the local Victoria breweries, asked me to do just that - brew a 10 HL (8.5 barrel) batch of my favorite recipe: Soma, my saison. I was honored to get this offer and both excited and nervous to try it out. This is a recipe I've worked pretty hard on over the last two years. Working on the yeast blending for saisons (post 1 and post 2), tuning the mild acidity levels from pre-primary souring, and in general trying to work out the hop balance. I feel like I have it somewhat close to where I want it on a homebrew level (although the recipe is always evolving) but trying to scale up the recipe definitely presented new challenges. This write up details the brew process, what I learned from scaling up my recipe and brewing it on commercial equipment, and what I would like to do differently next time.

Just hitting a boil.

I have to say from the start here that Clay and Jeff at Moon Under Water were great to work with. They were open to whatever I wanted and gave me complete control over the recipe and process we'd take. And any decisions heading into the brew day regarding making this beer that deviated from my homebrew recipe were my own and were made considering the different equipment, process, and risks involved in a commercial batch. With some of the batches leading up to this brew I had been pushing the acidity of the beer. With some of these brews possibly slightly too acidic, or at least a bit out of balance, I backed down on the target acidity a bit. I also backed down on the hopping load. This was due to both my expectations of increased utilization on a commercial system and my experiences with some rough hop character recently in some homebrew batches (earlier batch of Belgian single and Sour-Worted Soma).

There are definitely some challenges when scaling a recipe up to a commercial size, as commercial brewers and any homebrewers that have done this or looked into it know. I'm really happy with how our hot side went, and I don't think we ran into much trouble there, but the fermentation definitely presented some challenges. These snags sort of acted together as one, with one feeding into the other.

We went with the yeast blend that I normally use, and I think this is really important in producing the character of the beer. I had never grown up the sort of yeast we were going to need for this batch, so I dropped off a pack of each of yeast for the brewers to grow. Generally they either use dry yeast or talk to local brewers for a pitch (the breweries here seem to cooperate well and share yeast/knowledge/etc. freely with each other) when they want to use a new yeast. So it turned out that we both didn't have a ton of experience growing up yeast from packs to ~1000 L pitches. So we didn't quite grow enough yeast and the first challenge we faced is that we pitched at a lower rate than I generally do. This lower pitch rate, and the subsequent slow start to fermentation, fed into the second and main challenge we faced.

Filling the open fermenter.

We fermented this beer in an open fermenter in Moon's 8-vessel fermentation room. As a side note, adding the yeast was a bit of an adventure. We had the yeast in carboys and we added it by swirling up the carboys and carefully pouring it in over the top of the open fermenter. The top of the open fermenter is about at my chin, which made pouring full carboys a bit tricky.

Many of the other fermenters of the fermentation room were sitting at cold-crashing temps, and because the room is rather small and can be isolated from the rest of the brewery by a set of double doors, the fermentation room  was rather cold. Because we used an open fermenter our fermenting beer was not as insulated from this cold room as a normal cylidroconical would have been. So although we didn't have the glycol jackets on, the cold room definitely had an influence on the beer.

We were working with a notoriously tricky yeast (Wyeast 3724, Belgian saison) which is especially temperature sensitive. Starting with a low pitch rate meant that we got off to a slower start and didn't generate heat from fermentation as quickly. This insufficient generation of heat then fed back into lower beer temperature and lower yeast activity, which cycled back into continued insufficient generation of heat. So because out pitch rate was low, the fermenter was sitting in a cold room, and we had no way to warm our fermenting beer (although I knew this was going to differ from my home brew process going in the big batch, the impact of this limitation was greater than I expected it to be) our fermentation never got warm enough. And this significantly suppressed the 3724 character and left more work for the small helper pitch of 3711 to take care of.

The fermenting wort.

This was evident from early on in the fermentation, and it left it's mark on the final beer. So now the point has been driven home that if I use 3724 on a commercial system, I will need to set things up better to keep a fermentation warm, and possibly come up with a way to add a bit of extra heat. I do have the ability (sort of) to see how this batch would have been had the fermentation gone according to plan. We filled one 6 gallon (23 L) carboy which we temp controlled the way I normally would to compare how the commercial sized fermentation (temperature, hydrostatic pressure, etc.) changed the beer. Although I haven't done a side by side tasting yet (the beer from the carboy is still bottle conditioning), at bottling I did prefer the carboy version to the open fermenter beer.

Racking into the barrels.

After 4 days in the open fermenter we transferred the beer to a 10 HL horizontal tank and 2*225L wine barrels. The jackets were turned off on the horizontal tank and the beer was allowed to rise as far as it would (which wasn't much due to the sluggish fermentation) until it reached terminal gravity. The gravity was still pretty high after about 2 weeks (it was at about 1.015) so we decided to prepare a pitch of S. trios (White Labs WLP644 Brett trois was recently determined to be a Saccharomyces yeast rather than a Brettanomyces strain, and I will refer to it as S. trois from now on). We decided to give the beer a couple more days to see if it would finish up on it's own, but after about 2.5 weeks the beer was only down to 1.010 and I was tasting something that I often, but not always, get out of the French saison yeast (Wyeast 3711) that I would prefer to not have. It is sort of a green and yeasty character, a friend of mine recently described it as being a bit like green bananas. So we added the S. trois and gave it a couple more weeks to finish up and clear.

The two wine barrels came from Averill Creek and formerly held red wine followed by blackberry port. This was their first use for beer, but they had been filled with water at the Moon. The barrels were inoculated with a base of Yeast Bay Amalgamation blend (thanks to Jeffery for giving me a good pitch of this) and Wyeast 5223 Lactobacillus brevis. In addition one of the barrels has some ECY02 Flemish ale blend from East Coast Yeast, an excellent source for unique yeasts and funky blends, and the other barrel had an addition of S. trois. They are aging nicely and I am really excited for how they turn out. At this point the ECY barrel stands out a bit with some mild but developing funk but both barrels have a nice complex fruityness from the saison yeasts and Amalgamation blend. The acidity is starting to show up and they can age a while longer without being in danger of pulling too much oak/barrel flavor. We'll see how they change over the next month and maybe there will be some additional cultures added if needed, but I am definitely happy with how they are now. I expect that they will be blended together at bottling.

I marked the barrels with some inspiration from Brussels

Recipe
Brew date: 20-October-2014
Batch size: ~1100 L (9.4 bbl)
OG: 1.049
FG: 1.003 (1.000 on the 23 L carboy)
ABV:
Target IBU: ~25 (some guesswork here as I don't know the utilization of the big system and the process changes that go along with it).

Grist:
87 % Best pilsner malt
7.5 % Best Vienna Malt
5.5 % Weyermann Acidulated Malt

Hops:
Saaz: 800g with 30 minutes left in the boil and 1430 g grams added at whirlpool
Hallertau: 1100g with 30 minutes left in the boil and 1430 g added at whirlpool
Yeast:
Wyeast 3724 Belgian Saison, ~80% of the total yeast pitch
Wyeast 3711 French Saison, ~20% of the total yeast pitch
WLP 644 S trois added to the clean portion after ~2.5 weeks.

Both barrels were inoculated with Wyeast 5223 Lactobacillus brevis and Yeast Bay Amalgamation blend. One barrel also got a bit of ECY02 Flemish red blend and the other got S. trois.

Other:
Breakbright was added following what they normally do (I don't know the exact amounts).
CaSO4 and CaCl2 were added to boost Ca 92 ppm, SO4 167 ppm and Cl- 39 ppm in 10 HL. Victoria's water is pretty low in everything so these are reasonably close to the final concentrations.

In addition to the saison in barrels I have started helping out the brewers at the Moon with another barrel project. And in addition to these two barrel aged projects there will likely be some more clean and funky beers to come out of collaborations with the brewers at the Moon. I'll keep the blog and facebook page updated with these projects.

60 gallon barrel and saison brew day

The grooves on the insides of the staves

Getting caught up on posts of past brews seems to be a never-ending battle. This is a bit of a long post, which is fitting for a long brew day and the long aging to come on this beer. Hopefully you find the run down of our process (both brew day process and thought process leading up to it) helpful if you are planning a big brew to fill a barrel. I'm definitely a proponent of planning brews well, especially when doing something well beyond what I normally do (or have done before) like this. And if you aren't thinking of that sort of thing, hopefully the challenges of brewing such a quantity of beer on normal homebrew systems makes for an entertaining read.

Preamble

Through the work of my friend Jeffery I've recently started aging beers in a second used wine barrel. The first barrel is a 30 gallon Hungarian oak barrel that previously held red wine at a U-brew shop for a year or two before we acquired it (and in comparison to this new barrel, it looks amazingly small). This new barrel is grooved American oak and held wine at Glenterra Vineyards for about 10 years, so the oak character ought to be pretty neutral. Glenterra uses wild fermentation for their wines so we might find some interesting microbes already in the barrel. And either way it ideologically fits well with the sorts of mixed microbe fermentation we are looking for from this barrel. Generally John at Glenterra keeps his barrels around when they are neutral (he isn't looking for lots of newer oak barrel flavor, which I appreciate) so we were a bit lucky to get this barrel. The one catch, and the reason we got it, was that two of the staves were cracked on the outside at the top of the barrel so John didn't want to continue using the barrel. But since the barrel was free we were happy to give it a shot.

The empty barrel with a wax-sealed crack

If you read anything or talk to anyone about barrel selection, cracked barrels are an immediate pass. But that is assuming that you are selecting from a number of barrels and you can find one for the same cost that isn't cracked. When, on the other hand, your options are one free barrel with a couple cracked staves or no barrel, the conditions are a bit different. We figured that if it passed water tests then it should be alright, and if the first batch of beer through tells us otherwise then it wasn't a huge loss. And we could feel the inside of the barrel where the crack was to confirm that the inside wasn't cracked, so it seemed to be an issue constrained to the surface (had the inside also been cracked we would obviously not have used the barrel). The two staves adjacent to the stave with the bung hole are cracked, but the bung hole stave is fine (that likely would have been another deal breaker). To be extra careful we sealed the top of the cracked section with wax, following some of the barrel waxing ideas presented by Funk Factory and Embrace the Funk.

Look how much brew gear we fit in here!

We had three people on board at the start of this barrel: Kyle (who is housing this and the first barrel), Jeffery (who found this barrel through his common interest in what John is doing at Glenterra) and myself (I guess in this context I look like a bit of a free loader). With only three of us to start we decided to take a solera approach. If you are unfamiliar with this, the idea is that on every pull from the barrel we only partially empty it rather than fully emptying it. And we will then replace this volume with new beer. The idea of each of us coming up with 20 gallons of wort for fills every time the beer in the barrel was ready was pretty daunting (and 20 gallons of the same barrel beer at every empty was a bit more than we needed as well). Something 5-10 gallons each at each pull from the barrel is a bit more reasonable.

Had to keep 70 gallons worth of yeast safe

Though this approach will help us down the line, that still left the first fill as a bit of a challenge. We figured that with our combined equipment we could, in one giant brew day, come up with enough wort to fill the barrel. And so the plan for a ~70 gallon brew day was born...




The Big Brew

In case you are thinking of doing something like this in the future, I'll lay out our approach. We were going to run 3 consecutive brews with mashing of brews 2 and 3 taking place during the boil of batches 1 and 2 respectively. Going into it we assessed the kettle situation (1x20 gallon, 1x15 gallon, 1x13.5 gallon, 1x10 gallon, 2x9 gallon and 1x7.5 gallon). We had one kettle with a false bottom large enough to comfortably mash enough grain for each batch (the 20 gallon). We then assigned pots to either boiling wort (28 gallons total pre-boil volume, plus boil over space) or heating strike water/sparge water (13 gallons for stike water and 18.8 for sparge). It worked out nicely that we could use a 13.5 gallon kettle, a 10 gallon kettle, and a 9 gallon kettle for the boiling. With each of these pretty close to full we could manage our boil. We didn't worry too much about getting equal gravity in each pot and simply blended runnings by feel while we were collecting to try to equalize things. As the brews went on we got better at this. We did measure gravity and volume on each of the boils, and by this we were able to estimate our OG produced from each batch.


That left us with a 15 gallon, a 9 gallon, and a 7.5 gallon pot. The 15 gallon could do the strike water well and with the help of the 7.5 gallon we were set for the sparge water. That left us with a 9 gallon to spare, which was very helpful when it came to juggling volumes into each of these pots and limiting how often we had to carry large full pots.

Our mash tun and it's blanket insulation.

Likewise we were going to need to be able to heat each of those pots. We had 3 burners and an electric system (for the 13.5 gallon). Two of those burners were needed to drive our other two boil pots, which left one for heating strike water. In sort of a circular way, this also informed pot choices for the boil and strike water. When it came time for the sparge, where we had 2 pots to heat, we would need another burner. But this wasn't really a problem because by the time we were ready for our sparge we needed to be done with the previous boil so that our boil kettles were ready to be filled once again. So we had an extra burner (2 actually) for heating the second volume of sparge water. And since this was our smaller volume it could heat up much quicker.

That covered most of our bases A couple other things we were going to need to worry about were carboy space, yeast, and fermentation temp control. The first part was simple enough to plan. Basically it required 70 gallons of total wort divided by ~5 gallons in each carboy = ~14*6 gallon carboys, plus or minus a carboy or two depending on how full we make each one. It turns out that this required just about all of the empty carboys we collectively had, but we made it work. I suppose I should note that we wanted almost 70 gallons pre-fermentation to allow for us to completely fill a 60 gallon barrel while leaving the trub behind, and maybe giving ourselves a bit extra to top up as needed or compare the barrel aged version to the non-barrel version. We learned the hard way about planning for extra wort when filling a barrel during the first couple fills of our 30 gallon barrel, in which all of the beer prepared either just barely filled or didn't quite fill the barrel. Extra beer is definitely better when going into a barrel filling!

Jeffery, master of gravity systems, prepares our chilling setup.

Preparing yeast for 70 gallons isn't fundamentally different from 5 gallons, just a longer lead time and some large flasks on stir plates. I generally use the Mr Malty yeast calculator, but the brewer's friend calculator has a built in function for stepping up starters, which requires fewer transfers and may be easier depending on your setup. For temp control, we had a fermentation chamber that could fit 3-4 carboys, 2 tubs for hot water baths that could fit two carboys each, 3 aquarium heaters to heat the water baths and a heating pad or two. This obviously didn't quite add up to what we needed. Luckily we were able to utilize an unused bathtub in an extra bathroom. This gave us a hot water bath that could hold 3-4 carboys, getting us much closer to our goal.

Our four tier gravity system for cooling.

So yes, I have now 'brewed beer in a bathtub', as the old stereotype of awful homebrew goes (but it was in carboys too). In the end we only had one carboy left out of a water bath or fermentation chamber, and we were able to control that one with a heating pad and by basically making the entire bathroom a warm fermentation chamber. We knew it was just going to be for a week until we could get them into the barrel so this temporary solution worked fine. If this sort of big brew is something you are planning on doing, where you suddenly need to temp control ~60-70 gallons of beer, perhaps you can get away with a  solution like this and temp control a whole small room in your house for a short time.

I should note that I did miss discussing the step of chilling this much, but based on how we brew the chilling demands weren't substantially different from an average individual brew. I was recently reminded of my first batch of homebrew years ago. I didn't have any friends that brewed and I learned everything I knew at that point (which was almost nothing) from a book or two. The brew day went well enough until the end of the boil, when it suddenly occurred to me and my brew partner that we were going to need to cool down this ~20 L of boiling liquid to room temp. And without contaminating it. And from what I had read at that point I believed it had to be done as quickly as possible. Needless to say it was a bit of a disaster. Well you only make that sort of mistake once. For this big brew this was fortunately not an issue. We all individually had our own chilling capability so we had that well covered. Out of simplicity we ended up running the wort from one of our boil kettles into another to avoid needing 3 chillers, to avoid needing to transfer a plate chiller, and to help prevent the transfer of hop material into our carboys.

Finishing up the brew by firelight

In the end we got 13x 6 gallon carboys, all full a bit past the 5 gallon level. I think that about covers the brew day. On to the recipe.

Recipe
Brew Day: 16 November 2014
Batch size: 23 gallons in carboys
Target OG: 1.050
Target IBU: 20

Grist:
68.1 % Weyermann Pils Malt
20.1 % Flaked Wheat
7.9 % Vienna Malt (Best and Weyermann)
3.9 % Acidulated Malt (Weyermann)

Hops:
295g of a mix of noble type hops at roughly 4% aa targeting ~20 IBU.

The barrel and our 13 carboys of wort

Yeast: Wyeast 3724 Belgian Saison and Wyeast 3726 Farmhouse Ale (a 62% 3724 - 38% 3726 blend)
The barrel was also inoculated with ECY34 Dirty Dozen Brett Blend and Wyeast 5223 Lactobacillis brevis. We opted out of adding funkier stuff (anything with Pediococcus) because we didn't want to let things like ropyness constrain when we could make our first pull of the beer, though we plan to add other microbes down the line.

Other:
Gypsum and CaCl2 were added to reach final concentrations of roughly 90 ppm Ca, 150 ppm SO4 and 50 ppm Cl-.
Each kettle got 1/2 tab of whirlfloc per batch.
Each kettle got about 1.5g Wyeast yeast nutrient per 19 L/5 gallons.

The yeast was pitched at about 68 F (20 C) and the carboys rose or were raised to ~80 F (26.7 C) by the third day. They were held at this temp until the 5th day, when the heat was turned off and they fell down to ambient (~59 F / 15 C). On the 6th day the carboys were racked into the barrel. They were still active and pretty cloudy, but whatever trub had settled out was left behind. 12 carboys were racked in (filling the barrel a bit more than our plan and therefore not leaving much headspace for the continued fermentation), leaving one full 6 gallon carboy out of the barrel. One of the empty carboys was used to collect any barrel blowoff.

Racking week old beer into the barrel

Our (very) full 60 gallon barrel and its carboy for blowoff.

I'll fill more in as we go with notes of the barrel aging to come.

Petite saison rebrew tasting notes

Setting up for the tasting.

With this batch of petite saison I wanted to work out yeast blend ratios. There were 4 treatments total:
A) 70% Wyeast 3724, 30% Wyeast 3711
B) 90% Wyeast 3724, 10% Wyeast 3711
C) % Wyeast 3724, % Yeast Bay Wallonian Farmhouse, % Wyeast 3711
D) 90% Wyeast 3724, 10% Wyeast 3711 (though at a lower pitching rate than the above treatment) with additions of Yeast Bay Lochristi Brett blend and Wyeast 5223 Lactobacillus brevis

To mildly confuse things, I had labeled the bottles of treatment A 60/40 blend and the bottles of treatment B 80/20 blend based on what I though I did during blending, but as I found out from calculating my blends it turned out to be about 70/30 and 90/10 respectively. Through tasting these beers on their own I had formed some general opinions of how I felt about them and how they compared but in order to really figure out what the blend ratio did I needed them side by side. So I set up a somewhat blind tasting. I knew the beers - the three 'clean' (no brett/lacto) treatments form above plus my previous batch of petite saison, leaving out the brett and lacto petite rebrew treatment (D from above). And I poured them myself so I knew a bit about appearance. But then they were re-arranged and arbitrarily numbered for me given to me a minute or two later.

From tasting these beers individually while knowing what they were (and maybe pouring them, though with the exception of the previous batch with a different grist and therefore color, the differentiating characters were mostly taste and aroma based) I was able to correctly identify the different blends but it ended up being hard to determine which was the 90/10 blend and which had Wallonian Farmhouse. They were identifiably different but not necessarily in a way that I could pin to the yeast blends. I am not very familiar with Wallonian Farmhouse at this point and that probably contributed.

The poured glasses. As you can see there are some visual differences

Glass #5 (later revealed to be treatment A from above, the 70/30 blend)

Aroma: Peachy and floral with mild lime stand out. This one has more green fruit (in a slightly under-ripe way). There is a nice minerally character that is pleasant. This has a mild peppery spice and mild apple and it is more herbal than the other glasses

Appearance: Large white head, great retention, nice lacing, clear brilliant copper

Taste: Tangy lime, thinner flavor than the others with sour and green fruit. It has a dry finish and comes across medium-bitter. Overall the flavor is green and under-ripe fruit forward with low pepper.

Mouthfeel: fluffy light body with high fine carbonation, the carbonation feels a bit higher than the other blends and the body may be a bit thinner, but that might be influenced by carbonation.

Overall Impression: 3711 beers have an identifiable character to me that I would describe as green slightly under-ripe fruit with yeast. I know that's sort of a poor description but I don't really know how to convey the flavor. Either way that is what I identify with 3711, and this beer has more of that than the others. From that I was able to identify it as the 70/30 blend. It was the most 'green' and had less peach and orange citrus. There was a nice tang and spice but I want more 3724 flavor in the balance overall.

Glass #12 (later revealed to be treatment C from above, the blend with Wallonian Farmhouse)

Aroma: There is more 3724 character overall and the beer is more doughy/bready with mildly tangy citrus and tropical fruit (like mango). The aroma has a bit of a character I associated with stressed 3724. It has a sweeter aroma and is more tart than the others.

Appearance: Medium head, lower than #5, the head is a bit chunkier, brilliantly clear pale copper

Taste: Tangy citrus fruit, unique, peach, doughy/bready, unique tropical fruit in this treatment. 3724 character is in there but it is less identifiable than in other treatments, it comes across crisper than #9 (which might be influenced by the tangyness). Nice but it could use a bit more hop character.

Mouthfeel: Fluffy body, nice high fine carbonation

Overall Impression: Great citrus and tangy fruit. The character is a bit like what I got out of a saison brewed with Logsdon Seizoen dregs a while back (sidenote, this Logsdon dregs beer was what I sent to the first round of the 2013 NHC which came out a fair bit different from what I sent to the final round). This version is a bit more reigned in than that Logsdon version. The fruit is unique and different from what I usually get. I think this is the Wallonian Farmhouse version and #9 is the 90/10 but it is hard to tell.

Glass #3 (later revealed to be the previous batch of petite saison)

Aroma: Floral, fruity candy, pear stands out. The classic Dupont aroma is there. There is nice malt and a stronger noble hop character but also some slight oxidation.

Appearance: Slightly paler and more of a golden color than the other three glasses. Large white head with great retention and nice lacing.

Taste: This has more of the classic Dupont floral character with a touch of pleasant sulfur. It has a dry finish with nice hop charatcer and a light tangyness. The hop flavor in this one is good and is more in line with what I am looking for out of this recipe.

Mouthfeel: Fluffy/creamy body with high fine carbonation. Less tangy than #12. The hops contribute nicely to the body.

Overall Impression: The other glasses are missing this hop flavor and I like the Dupont-esque mild sulfur. The fruit is slightly more restrained in this and I really like that. It is still pronouncedly fruity but is a bit more refined. Maybe over-expressed fruit is what I mean when I talk about stressed 3724 character. At it's prime this version was the best. I may still prefer this malt bill to the malt bill I used in the re-brew.

Glass #9 (later revealed to be treatment B from above, the 90/10 blend)

Aroma: This version is the funkiest and is seems complex than the others. Strong peach and 3724 character. There is a nice doughy character in the aroma. There is a bit of stressed 3724 with strong floral. A bit of sharpness almost like what cheddar has and if I didn't know what went in here I could believe that there was some mild Brettanomyces.

Appearance: Very similar to #12 with a low white head and good retention.

Taste: Strong fruityness dominated by the classic peach character I associate with 3724 and a mild green fruit that I associate with 3711. There is a touch of brett-like funk and doughyness and a touch of sulfur in the finish (though it is less smooth and pleasant than what I find in Dupont saisons). Nice citrus fruit (tangerine) and floral. Nice malt character but I want more hop flavor and bitterness.

Mouthfeel: Great creamy/fluffy body. High fine carbonation. I want more hop character.

Overall Impression: The darker color in glasses #5, #12 and #9 compared to #3 is nice but I'm not sure the darker malt bill benefits the flavor. Maybe I'll go back to the original malt bill. And if I decide I want the darker color maybe I'll add a bit of Carafa-type malt for color adjustment. The perception of funk is nice here and I'm not sure where it is coming from. It could be that something else got in, and if so it's pleasant, but I'm not convinced about that. It could also be a combination of the tangyness, malt and slight sulfur. In the 4 tasting this is my preference, but only slightly.

The revealed identities of the saisons.

Closing thoughts: Overall these were very similar. that may not be reflected in my notes because I was looking for differences. Especially the 90/10 and Wallonian Farmhouse bottles. With what remained in those bottles I compared them head to head and may slightly prefer the Wallonian Farmhouse in a 2-way tasting. I should redo a 2-way tasting to test that. The WF batch is slightly more assertive than the 90/10 due to the tangy citrus and tropical fruit. One theory I have for this at the moment is that in a 4-way tasting I might be tasting more for what I think is the best saison for my palate but in a 2-way tasting I might choose the beer that stands out pleasantly. Looks like I have to do a new 2-way for these two.

Dark English Mild

Heating up the strike water

There's been a bit of delay in my getting a post up on this blog, but the Hors Catégorie facebook page has been getting lots of regular brewing updates including a nearly 70 gallon brewday of saison which is now waiting to go into a 60 gallon wine barrel, a 10 HL batch of saison which was a collaboration with a local brewery, blending of sour barrel aged beers, and a string of the posts about the general saison-related stuff I find myself doing on a weekly (if not daily) basis. But I finally got a bit of time to put up a new post. I've been brewing a lot of saisons lately and although I love saison, I was ready to try out something different for a batch or two. Having to prepare 10 gallons of old ale for a barrel aging project and a mistake for one of my yeasts in a shipment of an East Coast Yeast order (which fortunately was corrected as soon as possible) meant that I had a couple different English strains around. I like having a nice malty but easy drinking beer here and there, and as with my saison brewing, lately I've been focusing on trying to make lower alcohol beers with great flavor. So a dark English mild was a natural choice.

My ground malt and the dark grains cold steeping

I found a dry toasty/mildly astringent character that I don't really care for in some of my previous batches of mild and I also find it with some commercial milds I've tried. It is kind of like the flavor of sunflower seed husks. While I don't find it awful, this flavor (which I think may be enhanced by the thinness of the beer) is something I am looking to avoid in milds. I wanted to try ways to get a darker malt character without that so with this batch I tried cold-steeping my grains - my first use of this technique. In brief, the theory behind/advantage of this is that you don't pull out as many rougher flavors like tannins with the cooler temperatures which still getting the color and much of the dark malt flavor out. The liquid you get from this can then be added back at late in the boil, to the hot wort, or straight to chilled wort or the carboy. There are plenty of places to learn more about cold steeping of grains, such as the article by Chris Bible in the October 2014 issue of BYO magazine.

I cold steeped my grains in a 1 qt mason jar, which was an excellent vessel for this sort of thing. The only downside I found in my approach was that my water to grain ratio was way too thick and as a result of this I did not get very complete extraction. I would have needed multiple mason jars for the amount of dark grain I used. In subsequent brews with less dark grain a 1 qt mason jar has worked great, and when I use this much grain again I'll split it into 3 or so mason jars. So anyway, the day before brewday I started cold steeping. With this batch I decided that I would add the liquid from my steeping at the end of the boil, so with 10 min left I poured the contents of the jar into a fine mesh hop sack which I held above my kettle. This is the point where I realized too much of my water was still hung up in the grain and I didn't have a bit of rinsing water around. So my utilization of the dark grains was pretty low on this batch and I'll be ready for next time with a thinner dark steeping.

My (not very dark) mash runoff

The wort was split into 3 carboys with 2 of them getting WLP007 (one of these two also got a small amount of oak) and one getting ECY18. I figured this was a great way to take advantage of the accidentally getting sent a vial of this yeast and to see how the ECY18 English Mild yeast compared to other English strains. By now these three carboys are all bottled up and bottle conditioned (1-1.5 weeks in primary was fine with an appropriate pitch of rather flocculant yeast in a low OG beer like this) and while I've had each of the three treatments at this point, I'll wait a bit more until reviewing all three.


The Recipe

Brew Day: 19 October 2014
Batch Size: 13.5 gallons in carboys
OG: 1.034
IBU: ~14, Tinseth formula
FG: 1.011 for the two WLP007 batches and 1.012 for the ECY18 batch
ABV: 3.0%
Color: about 20 SRM (I haven't really looked too closely to get an exact number yet).

Grist:
10.6   lb (4.81 kg) Doehnel #24 (UK style light malt, ~1.9 L) -                     63.9%
1.75   lb (790 g)    Doehnel #26 (North American style light malt, ~1.9 L) - 10.5%
1.5     lb (680 g)    Brown Malt -                                                                      9.0%
1.25   lb (567 g)    Crystal 60 L -                                                                     7.5%
1.0     lb (454 g)    British Crystal 135-160 L -                                                6.0%
0.375 lb (170 g)    Hugh Baird Chocolate -                                                     2.3%
0.125 lb (57 g)      Weyermann Carafa II Special -                                          0.8%

*Note: I talk a bit about Doehnel malts in earlier posts like this one. These malts are grown and produced locally by a small maltster. The color and general style allow closest substitute malts to be chosen.

Hops:
33 g Slovenian Aurora 8.0% aa boiled for 60 minutes

Yeast:
WLP007 Dry English and ECY018 English Mild
1 tsp Wyeast yeast nutrients

Other:
Victoria's water is very low in minerals so salt additions are basically the final concentrations.
9 g CaCl2, 5 g CaSO4 in 30qt sparge
13 g CaCl2, 7 g CaSO4 in 43 qt sparge
1 tab whirlfloc

28 g oak - a single piece which was cut from a section of an oak chain from the now defunct Okanagan Barrel Works. This oak went into the primary of one of the carboys and stayed in until the beer was bottled (11 days). I prepped it by rinsing it a couple of times with boiling water and then soaking it overnight in a mason jar full of boiling water. This definitely strips out some of the oak but I was more worried about over-oaking than under-oaking. Also I didn't want to pick anything up in a beer like this, which due to it's low hopping, low alc, and low percent of attenuation is especially susceptible to contaminant microbes.

Mash and Fermentation:
I mashed at 156 F (68.9 C) for 60 minutes. For the fermentation I pitched at 64-65 F (17.8-18.3 C) and kept it there for the first 48 hours. I raised the temp to 68 F (20 C) over the next day (the ECY18 carboy got warmer because I neglected to check the setting of my aquarium heater) and I held this temp for 2 days before letting the beers return to room temp and bottling (at day 7 for the WLP007 carboy, day 10 for the ECY18 and day 11 for WLP007 + oak). The difference in bottling times was driven by how much time I had to bottle rather than each carboy needing different amounts of time.

2016 Update: I'm long overdue in linking the tasting notes into this recipe post. Here are the tasting notes for this brew.

The WLP007 plus oak took first place in the first round (Seattle) of NHC 2015 and scored in the upper 30s in the final round but didn't place. I stored bottles in the fridge but didn't rebrew it, or rather, changed some parts of the recipe in the rebrew and the rebrew wasn't better than the original. Although there was time I didn't want to try again going back to the initial recipe (I had my fill of dark milds for a bit) so the bottles I sent were well past their prime by the time the final round came.
 

My finished wort was much darker than the mash runnings due to the cold-steeped malt

Sour-worted Soma tasting

I've been experimenting a bit with ways to add mild acidity to certain beer styles before beginning primary fermentation. For this recipe I fermented a portion of a previous beer with a similar grist with Lactobacillus until I was happy with its acidity and then blended that portion back into the boil for this brew day.


Aroma: Peachy and citrus, the tartness is there, tangerine, strong Wyeast 3724 floral aroma. There is more Wyeast 3724 character in the balance than I feel I usually get from the yeast blending. There is a pleasant mild minerally character and a mild funkyness/rustic aroma (enough that one might wonder if there is brett in here).

Appearance: Thick fine very large white head (I gave a bit of an aggressive pour) with excellent head retention. The beer is a finely hazy straw color. It is pretty and hits what I am looking for appearance-wise in a saison.

Taste: There is a nice dry citrus fruityness (tangerine and mild lemon) and floral yeast character. The yeast character is strong and pleasant with some peach in there with the citrus. The hops come through a bit muddled, which is influenced by a rough planty hop character in the finish. I'm not sure if that is due to hop material making it into the fermenter and sitting warm with the beer for an extended period of time or related to the lower boil pH with the hops (or both, and more on this below). But it is definitely something I'll need to work on for future batches. The yeast fruityness becomes more prominent in the balance with warmth and the hop flavor becomes lower in the balance with warmth. The acidity is a bit higher than I'd like it to be for this beer. There is a nice hop bitterness before the roughness comes in.

Mouthfeel: The beer has a pleasantly prickly high fine carbonation. The body is light with a nice smoothness and the lightly tart finish is great. There is a bit of warming in the finish (probably due to both the acidity and the alcohol). That planty character in the finish is noticeable in the mouthfeel as well as the flavor.

Overall Impression: This beer is good but not where I wanted it to be. For a first order try with this method I think I got some good information out of it and I like the way the acidity is working with the yeast character. There is a hint of flavor I've gotten from stressed 3724 before, so the low pH might have stressed out the yeast a bit (or maybe it was something else). Either way it is really low level and I don't mind it. I would probably bring the acidity down a bit on future batches anyway. I don't think the low fermentation pH had many effects on the yeast and I think if I lower the acidity a bit, as I wanted to do independent of it's influence on yeast character, I think the pH wouldn't be a problem for the yeast character. And that seems to be going away with more conditioning time. Overall the beer is also improving with more conditioning time.

The aroma is great and overall the flavor is good. The acidity brightens the beer and lightens the body. The flavor is great when I get past the roughness of the hops in the finish. That is something that I am especially sensitive to in my beers and I think many people would miss it if I didn't mention it. I sent this off to a competition, where it scored better than I expected, and one of the judges said something briefly about some astringency, but didn't talk about rough hop flavor or anything like that.

I think the big thing to work on with this batch for me is the rough hoppyness. I've gotten that recently with another beer without any pre-primary souring and with a comparable hopping regiment, which leads me to believe it is primarily driven by what is happening with the hops and not the pH of the boil. I've had similar hop amounts at similar timings to this batch without any problem, and have also noticed with previous batches that if I get too much hop material carryover into the fermenter than I am prone to getting this flavor, so I think that may e the big issue.

The roughness is in the lingering aftertaste after the finish and if I'm not thinking about it I can not notice it. And then I think the beer is great. After this batch I switched to hop bags to try to limit hop material getting into the fermenter (partially driven by switching boil kettles) and I haven't run across the hop roughness anymore. Some batches have worked great, like the re-brews of my Belgian single and petite saison, but I also have had batches coming out less hoppy than I'd like, leaving something missing. I'm not totally happy with the hop bags and am still looking for a better solution to keep hops out of the fermenter while getting the hop character I want.

So overall the beer is good but not up to my standards. But I don't think much (or possibly any) of that has to do with the method of getting acidity. The 1st attempt at sour-worting was a reasonable success even if the overall beer didn't hit the target.

Yeast Blending continued- A worked example with Petite Saison

To follow up the introduction for how I go about yeast growing and blending, I wanted to give a worked example of how I handled blending in a more advanced scenario. It could certainly get much more complicated than this, but I think this is a good way to show my process and the potential value that yeast blending can add to brewing. As a warning, this is on the more technical/math side of things and this approach may not be what you're interested in. It's pretty dense but if you are really interested in blending yeast, and you want to do it in a controllable way, continue on. I try to break down the steps simply and repetitively so if you get it right away, excellent and sorry for the redundancy. And if you feel like you are missing something let me know. I probably wasn't clear enough and I though about this for a LONG time before actually doing the blending, so it took me a while to understand it. And it took me a while to work out the best way to do it. I think if you want to do yeast blending in a repeatable fashion you are going to want to think about blend ratios and total pitching rates. So here's how I do it in a more tricky situation...

The setup: I have been using a blend of Wyeast 3724 (Belgian Saison) and Wyeast 3711 (French Saison) for almost 2 years now and have been really happy with the results. I get a majority of the flavor profile from the 3724 combined with the reliable and excellent fermentation behavior of the 3711 and a nice complexity from additional flavor. So far this blend had been based on wanting that flavor balance with 3724 dominating, but I hadn't run any trials to find an optimal blend. So batch to batch blend ratios would swing from about 60%/40% Wyeast 3724/3711 to 90%/10% Wyeast 3724/3711 based on what I had available at the time and what splits worked best with the cell count splits I had of the individual yeasts. It was time to run some controlled yeast blend ratio trials to work out what the optimal blend is for what I want my saisons to be.

The yeasts used in this blending

I was also starting to run low on petite saison and I was really happy with how that recipe worked out in my first batch. With summer fading away I thought I would try to get a re-brew in and use that recipe, a beer one can drink larger volumes of and that is on the simple side for recipe, as a base for working on these yeast blends. The goal, as I break down in my Petite Saison Rebrew post, was to target a blend on the upper end and on the lower end of the range that I usually fall into for my yeast blending. In addition I made a blend with a third yeast, Yeast Bay Wallonian Farmhouse, and in a fourth carboy I used the 3724/3711 blend on the upper end of the range I usually fall into with additions of Brettanomyces and Lactobacillus.

The Blending: Before I started any of the blending I got out the yeast I had and thought about the cell counts I had in each container. In the petite re-brew post I refer to 80/20and 60/40 blends, and I have also referred to these in the bottles I gave away. In my brewing notebook I refer to 85/15 and 70/30 blends, which were my actual target going into this. The 80/20 and 60/40 must have come out of what I thought I had done after the fact (as it turns out that was not the case).

Anyway, I made a table with my target blend ratios and the total number of cells I would need in each carboy based on the Mr. Malty calculator. I had jars of ~110 billion and ~150 billion cells of Wyeast 3724 and of ~65 billion cells of Wyeast 3711. Conveniently for the main two carboys the total desired pitch worked out to 100 billion cells each based on my OG and volume. I also knew full and empty weights of my jars (I approximated the empty with another jar of the same style/size). I blended the 150 billion cells of 3724 and the 65 billion cells of 3711, making a blend of roughly 70/30 with ~215 billion total cells.

My Target Blends

Knowing the full weight of this jar and an approximated empty weight, I knew the weight of my blend slurry which was ~215 billion total cells (150 of 3724 and 60 of 3711). For a pitch of ~100 billion cells of my 70/30 blend I needed to add a bit under half of this slurry. After adding the appropriate pitch of this slurry I would have about 115 billion cells left in the 70/30 blend (210 to start - 100 needed, approximately 80 billion cells of 3724 and 35 billion of 3711 left).

To this I added the extra 110 billion cells of 3724 from the other jar, making an approximate total cell count of 225 billion cells composed of ~190 billion of 3724 and ~35 billion of 3711, a blend ratio of roughly 85/15. Knowing the new full mass of this blend and the same approximated empty jar mass I could add a bit less than half of what I had to add ~100 billion cells to the target 85/15 blend, leaving me with ~125 billion cells in 85/15 target blend left over.

Okay, so I've added the yeast to my first two carboys and it all worked out pretty smoothly. Now I have 2 carboys remaining: 85/15 blend plus brett and lact and the carboy with Yeast Bay Wallonian Farmhouse (which I'll call WF from here on out). For the WF blend I have a blend ratio of 62% 3724, 27% Wallonian Farmhouse and 11% 3711 in my notebook. I honestly don't remember my reasoning for such a precise and strange ratio. I had no reason to choose a specific amount of Yeast Bay yeast to add, except that I was happy with the character from using it in my Barrel Aged Saison. So the choice probably came from how many cells I needed and how much 3724 and 3711 I had left over. If I actually worked it out that exactly I must have been pretty sharp at that point in the process.

I had ~125 billion cells left and based on volumes and OG I would need ~85 billion cells total in the +brett and lacto (+BL) treatment and ~55 billion cells total in the WF treatment. I couldn't meet my desired pitch counts with the 3724 and 3711 I had left, but that's good because I would be adding other yeast/bacteria. I would slightly underpitch Saccharomyces to the +BL carboy and make up for this with brett and lacto to come to my target, while potentially overpitching total cells a bit. And I would underpitch 3724/3711 and add Wallonian Farmhouse to hit my target of 55 billion cells in the WF treatment. Back calculating based on those targets I wanted to add approximately 75% (~82 billion cells) of 125 billion remaining 85/15 blend to the +BL treatment, leaving ~43 billion cells of 3724 and 3711 for the WF treatment. Since that is a pretty even split, it must have determined the strangely specific 3724/3711/Wallonian Farmhouse ratio from above for the WF treatment.

I find it easier to think about the blending from this perspective after the fact and I would suggest carrying on like this, but during my actual blending I did the last 3 blends in the the opposite order (adding ~125 billion cells of the target 85/15 blend to a combination of WF and +BL treatments and leaving the rest for the normal 85/15 target blend).

I kept the Wallonian Farmhouse yeast separate, since it was only going into one carboy, to make up the difference in case I messed up on one of my pours. The same can be said for the brett and lacto. I didn't add known numbers of cells of those, but I could add more or less based on how much Saccharomyces I added. So with this target plan all thought through, I went ahead and did the blending. My actual blend ratios are shown in the table below. As it turns out, my Saccharomyces blends were 86/14 (Wyeast 3724/Wyeast 3711), 70/30 (Wyeast 3724/Wyeast 3711), and 69/12/19 (Wyeast 3724/ Wyeast 3711/Wallonian Farmhouse). So a pretty good job hitting my targets overall.

My actual blends. Wallonian abbreviated Wall. Cells added from (slurry mass added) / (total slurry mass). Cells of each type added from (total cells) * (% of each type) /100. Total cells remaining determined from (initial counts) - (added).

I think this was as successful as it turned out to be was due to the meticulous planning of target blend ratios and how I would come up with those based on the yeast I had. Keeping target pitches simple fractions of my total blend volumes (in two of the four cases it was about 1/2 of my blend and in the remaining two it was 75% and 25%) made this simpler to execute. And if my jars of yeast did not break down so easily into these blend ratios I would have either adjusted my target (with this blend ratio trial, targets were pretty flexible as long as I had one as a high % of 3724 and one as a lower %) or I would not have added a full jar. Keeping things as fool-proof as possible when it comes to pouring into a carboy is key here.

Hopefully this approach made sense. Definitely ask if you're stuck on something. And hope it proves helpful to you if you decide to try your own yeast blending. It definitely doesn't have to be as complicated as this so don't feel scared away from doing your own blending if this seemed too much. It's something I had to work up to from doing a bit of more simple blending first but I think it is a valuable approach to optimize your blending and to controllably try out different yeasts.