Tuesday, September 20, 2016

Thoughts on recipe formulation

Grab yourself a beer. Philosophical ramblings on beer are probably best served with a beer as accompaniment. And preferably a beer you like and want to re-brew or brew something similar to. That will help for the following discussion.

I'd like to present a rundown of how I go about formulating recipes, focusing on the thought process I go through for ingredient and process selection. On the logistics/calculations side I use BeerSmith, so I won't do any discussion of calculations/etc. I also won't spend time on ingredient selection in the sense of not using old/bad malt or hops. You should taste and smell your ingredients so you get familiar with what good hops of a given variety smell like compared to bad hops, and what fresh malt tastes/feels like compared to stale malt. In full disclosure, I'm a person that spends an unreasonable amount of time thinking about beer and I try to take pretty detailed notes. In-depth thinking about beer is part of how I approach it and I enjoy it but it might not work for everyone. Not everyone is in a beer or brewing-related hobby/business to spend lots of time thinking about beer (though those people probably don't spend much time reading blogs either). So no worries if this isn't for you. But this is how I do it.

I don't mean to suggest that this is all easy and this is probably geared to the intermediate to more advanced brewer. But I think the considerations I discuss below can be helpful to some degree no matter what level of brewing you are at. If you are just beginning and are making your first recipes, taking initial steps toward thinking in this way will help you ask questions to be more likely to get the answers you are looking for more clearly. When a brewer says something like 'I want to brew a -blank-' and asks for advice, I regularly follow that up with questions about the different directions within that idea/style/etc. which one could go to try to best be able to offer guidance toward their end goal. And if you are very experienced with making recipes, thinking in this sort of way will make active ingredient and process decisions easy (rather than just doing something a certain way because you've always done it that way or because someone else did it that way).

Reading and critically analyzing your beer and commercial beer give you a great tool set for creating recipes. I take lots
of notes to help this: brew logs at bottom left, commercial beer notes at bot. right and travel/brewery visit notes at top left.

Motivation and intention: To me recipe formulation starts with and spends the bulk of the time focusing on motivation and intention. I think one should always go into a brew thinking about their motivation behind the brew and their intention with the finished beer. This may seem fairly rudimentary, but I think that it doesn't happen a surprising amount of the time. For most of my brewing years I didn't really stop to think in detail about what I really wanted in the final beer beyond the base style and following a more or less random recipe. So I mean going deeper than something like I want a saison, or I want a hoppy sour beer, etc.

Perhaps you are brewing to try out a given ingredient or process. What setting do you want to test that in? Something that makes the difference easiest to see, or a setting where you are most likely to actually want to use that ingredient or process? Both have their merits and limitations.

Or perhaps your goals are more finished beer-driven. If you want a saison, what strength? What sort of hop presence? What color? Do you want it to be balanced more toward fruit/esters or spice/phenols? Or an even mix of both? What sort of body are you looking for? Maybe picture a commercial example you know and like. In an ideal world, how might you want the final characteristics changed to make it your ideal beer? Spending more time nailing down your motivation for the beer in question and your desired result in the finished beer are the first big step - you know exactly what you want.

La Petite Princesse from Thiriez. Drinking & discussing
this beer has inspired me considerably in making recipes.
So I start with basics and step forward from that to target details and what ingredients and/or process will get me there (or ask questions/research/set trials to learn what will get me there). This can make recipe formulation an easier step-by-step process and/or help you choose between different already published recipes. Note that building the skills to do this may require similar critical thinking about beers you drink to come up with a good sort of context for the direction you want your soon-to-be-brewed beer to head. Talking to brewers is a good way to augment this critical tasting of commercial or homebrewed beer.

Setting fundamental goals: Okay, so you've thought about the motivation and intention. I break that into a few fundamental goals to work from. To illustrate my process, I'll work through the following recipe formulation with a low-OG saison. I think this sort of thinking is much more valuable than simply giving a recipe, as hopefully it helps others to think in more detail about their brews and improve their brewing. I am planning a recipe/brewday post on the same beer in the near future, so don't worry - an actual recipe is coming.

The basic essential elements that I am looking for in my homebrews of low-strength saisons or my favorite commercial examples are: A) a refreshing/thirst quenching quality, B) sufficiently low strength that I can drink multiple, possibly during the daytime, without feeling the effect of the alcohol, C) something that doesn't taste/feel especially watery or thin and D) good yeast expression, balanced toward floral/fruity but without anything toward bubblegum. These basic motivations will drive all my recipe/process decisions behind making this beer. Keep in mind that all of the following are based on my personal tastes.

Turning those goals into process: Those were pretty basic goals. I don't mean to suggest that the following is easy and obvious. Knowing the results of recipe and process changes can come from trials, research, and discussions with other brewers but they aren't inherent knowledge. The following breakdown of my goals comes from years of experience in my own brewing, discussions with fellow homebrewers as well as Belgian and French commercial brewers, and reading about saison and Belgian brewing. That experience was built slowly, piece by piece, and thinking in this way helped me to build it and connect those pieces.

Taste testing two different base malts.
Goal A means that I want a fairly firm bitterness (keeping in mind the context of the OG). By my experience (commercial beers I like and homebrews) this means a BU:GU ratio of about 1.25 [Edit 21-Sept-16: 1.0 is more representative of a larger number of recent beers. I have done low-OG saisons with 1.3-1.4 that came out with the bitterness I was looking for, but I think the bitterness calculation was a fair amount higher than what I actually got for those, probably due to the hops I was using. By re-brewing the same recipe with the same hops but increasing the hopping, I was able to get the bitterness I wanted by going up to 1.3+]. It also means that, to my taste, I want the bitter/flavor qualities of noble-type hops rather than something like typical Pacific NW IPA hops. I find IPA hops to bring a perception of sweetness with their fruity flavors as well as a resinous quality, both of which detract from a refreshing quality to me. I find a more resinous or pithy bitterness can clash with the dryness of saisons and the yeast to me, such that the finish can be a bit harsh in a way that doesn't leave me wanting another sip. So I want a more delicate fruity/floral and a firmer spice/herbal character and I don't want something resinous.

It also means that I want a pretty forward hop flavor, therefore I'll be loading up on finishing hops and dry hopping. But through my experience I have found that if I go too far in the finishing hop direction, while I may get great hop flavor, something is missing in the perception of a drying bitter finish. The exact breakdown here will depend on aa of my hops and such, but I've been happy recently with a roughly even split in theoretical IBUs from "bittering" and finishing when using low alpha hops to finish. I quote bittering because to me this could mean 30 minutes left in the boil.

Finally it means that I will want to keep the grain bill rather pale to let that crispness come through. That means a base of pils malt is good. Some Vienna, or a move in that direction, would also be fine as the OG is so low. That basically settles the hops in terms of variety, usage rates, and schedules of additions. And it gives me a pretty good constraint of my malt bill. I am looking at grains no darker than Vienna.

Goal B means that I am shooting for something at or below ~3.5% ABV. The corresponding original gravity for this will depend on mashing and yeast choice. I have been working on raising my FGs a bit so that my OG can be up to 1.030. Getting the ABV down to about 3.5% with a 1.030 OG came from changing my yeasts and changing my mash profile.

Goal C means that I want to maximize character from my grain and mashing process, as grain level will obviously be the one factor most restricted by my ABV target. Mashing in a bit more of an intensive way to maximize extraction and development of flavors will also help and, depending on yeast choice, mashing for less fermentability could help. I generally try not to do this with 3711 as the yeast seems to attenuate quite a bit of stuff that others won't so it can just end up extending the fermentation timescale. For the same reasoning I would not mash for low attenuation when making a beer with brett that I also wanted to be a quick turnaround beer. Building body with certain grains like oats or rye and selecting yeast for more glycerin production can also help to build more body if you are looking for it. Finally, dry hopping can add a nice structure in addition to the flavor/aroma that it brings.

Pulling starchy wort during a turbid mash.
So to sum this up, for my tastes I will probably choose a more intensive mashing profile (more on this to come in later posts) and flaked grains and then dry hop with noble or noble-type hops for both the flavor and structure components I am looking for.

Goal D is fairly straightforward - it means that I'll choose a yeast blend or single strain and fermentation temperature profile that I am pretty happy with. Or if I want to try something out or don't know what strain/blend/temperature to use then I'll test some out, likely by splitting a batch to help constrain variables in the test.

A mixed yeast and bacteria blend can be a great way to add fermentation complexity. Think about something like Le Petite Prince with a bit more age or Jolly Pumpkin's Bam Bière (which, at 4.5% might be a bit stronger than the rest of this example focuses on, but that can be scaled back if needed for your goals). Mixed culture expression will of course come as a trade off with hop impression, which again comes back to careful consideration of what you are looking for.

For my recent beers, since it was already summer by the time I was brewing, I focused on quick turnaround. From travel experience visiting the brewery and talking with Daniel, I knew that the Thiriez yeast was a great choice (see this somewhat dated post from my visits there as well as this post from Farmhouse Beer Blog). Note that the fermentation appearance of this yeast, the flavor profile, and the behavior (especially attenuation) are different from Wyeast 3711, of which Brasserie Thiriez is reportedly the origin.

So from these 4 goals, I have set my OG and IBUs, the hops I am going to use and rough hopping schedule, most of the grist, and my yeast choice. I've also got a good direction to go for mashing. And since I've chosen my yeast I have a good idea what sort of fermentation profile to use (or what ranges to test in a split batch). In the coming week(s) I'll wrap this exercise up with the actual recipes this thinking led to in a post to come on the brewdays for some recent low-OG saisons.

Closing thoughts: Following this through exercise is best as a cyclic process. Design the beer, brew it, and then taste. While tasting it, think about your recipe and process, which will help you learn how specific ingredients and process led to your beer. Also think about what direction you would like it to go with the beer for next time. Having a good starting point for directions to grow helps narrow down the goals quite a bit, which gives you a great head start for repeating this exercise.

Ok, that's if for now. As said above, none of this is an easy process. Any level of brewer will learn as they go through this to better understand the results of ingredients and process changes/choices. So don't feel overwhelmed by this and if you don't know the answers you are looking for, read/ask around for advice and/or design trials to figure them out for yourself.

Saturday, August 20, 2016

Tasting Aged Mixed Culture Beers - Summer 2016

It's been a while since I talked about my own beers on here, so I thought I'd give a bit of an update on the beers I have aging. For context, and as an excuse on why my brewing hasn't made more of an appearance, I spent Sept 2015-May 2016 living in Europe (my second ~8 month stretch working in Europe in the last 3 years). I wanted a bunch of beers for long term aging and blending, and these trips gave me a good excuse to stash some beers to come back to at a later date. For both of these trips, I stored these beers (along with a bunch of bottled beer) in the basement of my very kind friend Mark.

As I've noted on the facebook page, I recently grabbed these beers back from storage and have been able to take inventory of the bottles (including blends from my blending session following my first move to Europe) and taste the beers in carboys. The tastings at this point were just to get a preliminary idea of where things are before doing more detailed/thorough tastings of the beers that might be ready to blend. Expanding on the FB post linked above, I'll summarize the beers, these quick tastings, my plans with the current beers, plans to brew for use with the current beers and the future in general, and some preliminary blending plans. This is a long post, but there are a lot of beers...

A mostly up to date map of most of the beers,
when they got oak, and quick tasting notes.
The Beers: To start this off I should note the beers that I have. Most of the beers are lambic-inspired (turbid mashed, brewed with ~30-40% unmalted wheat and ~60-70% malted barley, using aged hops, open cooled). Some of these are fermented only with local ambient microbes (from either starters of local ambient microbes or only from what the wort got during open cooling) but most have dregs and/or commercial cultures on top of the inoculation from open cooling and ambient starters.

There are a couple other beers that aren't lambic-inspired so I'll start with those:

-Barrel Aged Old Ale: This beer (brewed 5-Oct-2014) is discussed in detail in this blog post. I still have ~5.5 gallons remaining in carboys from my 10 gallon share of a 30 gallon oak barrel. It seems this beer was put on a very small amount of raspberries, though unfortunately I don't have any details on how much/when/anything else (at the time this would have happened, I was scrambling to get ready to move and my notes were's up to my standard). The raspberry level is very low and not necessarily identifiable if you weren't looking for them.

Brief tasting notes (Tasting): This is tasting pretty good (nice dark malt, good mellow barrel, light alcohol, mild acidity, good Brett C, some fruityness) but possibly not as good as the portions that I bottled in summer 2015 shortly after it came out of a barrel instead of bulk aging further in a carboy.

Barrel old ale (L) and Raspberry Dark Session (R).
Blending Plans: I expect most of this beer will be blended with the low strength dark beer below. I'm not sure what I'll do with the rest. Probably either bottle plain or brew a lower strength younger beer to blend toward an oud bruin-oriented end (though of course this aged component is fairly strong). I will also have a Flanders red-inspired beer coming out of a barrel soon so some of this could find its way into that beer.

-Dark Session on Raspberries: This was extra wort from a dark mild (brewed 8-Feb-2015) and was a subsequent batch of a recipe similar to this beer. I had about 3 gallons that I added a mixed culture to and then, after a few months, racked onto the leftover raspberries and blend from Blend 4.
Quick trial blends of Raspberry Session and Old
(L to R: 33/67, 47/53, 67/33 Session/Old).

Tasting: The acidity is forward (makes sense with the lower hopping) and there is a mellow touch of acetic but not to an obtrusive degree. It is a bit too thin overall - not surprising given the OG of 1.038 - but I think it could make a good blending component. The raspberry level is good and there is a cool chocolate raspberry thing going on.

Blending plans: I figure almost all of this beer will be blended with the barrel aged old ale to come out with an end beer of a more reasonable gravity between the two components while preserving the dark malt and fruit combination. Some extra, if any is left over, may make it's way into other beers to increase the fruityness. But only if those beers also could benefit from an increased acidity. I did some quick blending trials with this beer and the old ale. The winning ratio was about 45/55 to 40/60 session/old.

-Dark Strong Mixed Culture: This beer (brewed 5-Aug-2013) was inspired by Oerbier Reserva. The goal was something red wine-like with a mild-medium acidity and that I could have around for a while. The OG was around 1.080, with a sugar feeding during fermentation of dark candi syrup.

Tasting: The beer is pleasant and the acidity is there (maybe even slightly on the high side for my target of a more mild acidity, but ok) but it isn't quite there. I think much of the problem is that, unlike Oerbier Reserva, this was not aged in a red wine barrel. So I am not getting the wine-like character that I was looking for and I am not getting the oak. I added a bit of oak after tasting (not first use oak, and after soaking in boiling water).

Blending Plans: I think it could be likely that I'll end up blending at least some of this with a red wine to approach the character I am looking for. If so I'll probably choose something a bit lower in tannins and ABV and higher n fruityness. This beer is ready for those test blends with wine whenever I get around to it.

Having Rodenbach's Vin de Céréale in 2011.
-Strong Flemish-Inspired mixed culture: This beer (brewed 26-Oct-14) was inspired by Rodenbach's Vin de Céréale. I've had the pleasure of having this beer a couple times and it is amazing to me. Flanders red-like in character but strong and pale for the strength, age, and the general Rodenbach lineup.

Tasting: I didn't expect to actually hit my target on the first try, but I am a bit bummed about how far off I was. The beer is not bad, but it was not the goal. I think what the beer is lacking is a combination of fruity flavor and oak, and possibly a bit of oxidation. It is more malty and dry/toasty instead of fruity. Part of this may have been my choice of malt (100% Vienna, rather than pils and some darker malts for the same color).

Blending plans: This is another beers that may see some blending with wine. It may also see some blending with the 1/3 share of a Flanders red-inspired beer I have had in a 30 gallon oak barrel for about 13 months now. I could also try to encourage more fruityness by the actual addition of fruit, but if so that will only be a small portion as I want this beer to be around for a while and I don't want to heavily fruit something only to let the fruit fall away before I have even half of the beer.

Turbid mashed beers:

The Dec 2014 Spontaneous Starters carboy.
-December 2014 brew: This beer was brewed with 30% unmalted wheat and 570 g/HL (0.76 oz/gal) aged hops. The batch was split into 3* 6 gal / 23 L carboys. One was pitched with ambient microbe starters, one with various bottle dregs of multiple spontaneous beers, and another with ECY-20.

Tasting: The spontaneous starters carboy is quite bitter and phenolic. This needs some time to mellow out. I think it has a good chance of doing that as younger lambic/spontaneous beer can have similar characteristics. The ECY-20 carboy is sweeter with more fruit but also quite bitter. It is possible that some of this is coming from the hopping rate, which might be a bit high. But I think a fair amount of it is also brett-derived. I added some more cultures (from the yeast cakes from previous good turbid mashed homebrews) to this batch. The dregs carboy more acidic, has more citric fruit, and less bitterness/phenol. This is closer to being ready for blending but I still find it a bit more bitter than ideal.

-May 2015 brew: This was brewed with 41% unmalted wheat and hopped at about 600 g/HL (0.80 oz/gal). The batch (60 L) was split into 4 carboys which were pitched with either lambic dregs, spontaneous starters, ECY-20 or left to ferment with only what it got from open cooling.

The May 2015 Fully Spontaneous carboy.
Tasting and Blending Plans: The ECY-20 batch us nicely fruity (citric mostly) and has a more balanced level of phenols. This I think could be ready to blend soon, or I may wait on it if I am not coming up with a blend that I'm happy with.

The dregs carboy is more fruit balanced and with lower phenols than the ECY-20 batch. It has a bit more acidity (possibly slightly on the high side of what I want, or at least I don't want much more) but the acidity is clean/not acetic. This is another carboy that could be incorporated into a blend soon if I come up with something I like.

The spontaneous starters batch has more bitterness than others of this brew, but lower than the Dec 2014 and Sept 2015 brews. Overall this is pretty mellow yet, with a low level of sulfur and some lower fruityness and acidity. Perhaps not surprisingly, my ambient microbes may need a bit more time to develop the same sort of character intensity as lambic dregs or lab blends.

The fully spontaneous batch had more astringency/phenolic bitterness and a pleasant soft tropical fruit. I'm not too worried about the phenols at this point as spontaneous beer producers have anecdotally noted that their younger beers can be pretty rough in this regard and it mellows with time. So I'll wait this out for a bit.

-June 2015 brew: (this is even more true for the following brew) I know brewing in warmer months breaks from tradition for lambic. The reasoning given here being that unwanted microbes are more populous in warmer weather and the beer will head in a non-ideal direction. Another factor, which isn't discussed as much but which I think is quite important, is that warmer temperatures are insufficient to properly cool a commercial sized batch in a coolship to pitching temp (in the ballpark of the high teens C / mid 60s F).

So I wanted to try warmer month open cooling out for a couple reasons. First, the weather wasn't too warm (nights near 10 C / low 50s F). This was sufficient to cool my volume of wort to the same sort of temperatures that a commercial batch of lambic is cooled to in winter. Second, most of these beers were pitched with something, which means I am less reliant and inoculation from cooling. And finally, a batch are two aren't much to dump if it turns out that the warm weather microbes are a problem. So I will see if the summer microbes and microbe balance are problematic, at least in this small sample set (spoiler alert - so far this isn't the case).
The 30 gal barrel with a red beer (L) and
60 gal headed in a lambic-inspired direction (R). 

This batch was brewed with 39% unmalted wheat and hopped at a rate of about 225 g/HL (0.30 oz/gal). It was split into 4 carboys, with 2 of them going into a 60 gallon barrel that I share with some other great homebrewing friends. This barrel started with a saison with brett and lacto and we taking a solera-type approach and encouraging it to move in a lambic-inspired direction. The barrel is tasting quite nice so far and I am excited to incorporate the next pull (planned for early October) into blending as it will provide some of what I am generally short on so far - oak and fruit. The 2 carboys I kept were pitched with some combination of ECY 20, the yeast cake from previous beers, and bottle dregs.

Tasting: At this point there is nothing drastically unpleasant about these but at the same time they are not my favorites from the aged beers. They are pretty mellow and overall understated, which makes them currently better than the highly phenolic beers, but there is a bit of a character I am not too keen on. It is something I have noted in certain commercial lambics before, and that a friend describes as a bit 'pool water-y', which I think is quite apt (though it makes it sound worse than it actually is). This could be related to chloramines/chlorines (I carbon filter my water and did the same process with the same filter cartridge for other aged and non-aged beers that aren't showing that character) or it could be something picked up in cooling, or something else. Anyway, my beers are not anywhere near this intensity, but they have hints of something that remind me of this.

Otherwise they are pretty mellow and muted compared to both older and younger batches. I don't really know why this is. The bitterness and phenolic levels are lower and there is a mellow and pleasant sweet fruityness. Overall I think these beers are on a good track, especially as most of my other beers are higher in the phenolics and lower in the fruit, but they need more time to develop a more full character of their own.

The Sept 2015 Spontaneous Starters carboy.
September 2015 brew: If warm weather was going to be a problem, this would definitely show it. This beer (~50 L) was brewed with 27 % unmalted wheat and hopped at a rate of about 370 g/HL (0.49 oz/gal). The temperature of the wort in the morning was 19 C / 66 F so the cooling was certainly sufficient. I didn't note nighttime lows, but I suspect it was in the 10-15 C range given that final temperature of the wort the morning after cooling. The beer was split into three carboys with one receiving spontaneous starters, microbes from a bit of the trub of my favorite batches from my previous blending, and the third one receiving dregs from various mixed-culture saisons.

Tasting and Blending Plans: The spontaneous starters batch is fairly high on the bitterness so this will get at least another winter before it is used. The batch with previous blend microbes has a good fruityness and is on a really nice track but is a bit more bitter than I'd like so this will probably get more time. Though it is possible that I'd pull it as a phenolic component in a blend with a more fruit forward base. Overall though I am happy to let these beers sit a bit more as they are fairly young yet (just under 12 months).

The saison dregs carboy has a high lactic acidity and not much phenolic character. This beer may head toward blending with cleaner saisons to create something in line with blending an acidic old beer with fresh young beer.

Older Turbid Mashed beers: These are left over from previous blending of aged mixed-culture beers (the same blending post was linked earlier in this article too).

Muse #1: I have three gallons of this beer left over from my previous blending (there are some more details on it in that post, linked above). The beer was brewed in January 2013 and at the time of the previous blending it wasn't one of the best. It was a bit more acetic but had a pleasant fruityness. I prefer very low to no acetic acid, so this wasn't my ideal. But it certainly wasn't out of line with what you would find in many well regarded commercial beers so we're definitely not talking extreme levels. Just more than I wanted. This is generally still the case. In the new company of beers that are generally more bitter and less fruity than I want in a final blend (and with a clean acidity), this beer could be very useful for blending.

Most of the carboys safely transported back and in their new home
Wild Rye: I also have 3 gallons of this beer left from the previous blending session. This was brewed in July 2013. During the previous blending it also wasn't one of my favorites as I thought it was too phenolic. Since then my taste preferences have shifted a bit (toward more phenolic) and the beer has developed/mellowed more. At this point I find the phenolic level is low-moderate and pleasant. As a bit of an aside, I need to do a side by side tasting of my previous blending, but my feeling from individual tastings over the past few months is that Blend 2 (the only non-fruited blend with this beer as a component) is my favorite. And I think part of that is due to this beer. With the more balanced and developed character I think this beer will also be a good candidate for blending.

Sour Saison: the term sour saison is not one I am super fond of but for internal consistency I'll continue using this name for this beer while it is around. It is not saison-like and I wouldn't present it as one, but this was brewed mostly as a saison (not turbid mashed, not aged hops, pitched with saison yeast) while knowing it was going to head in another direction (it was open cooled and also pitched with lambic dregs). I liked this beer in the previous blending but it worked out that I had a gallon left over so I saved that.

I was probably going to blend it with something younger and clean, but now that I have the 'saison dregs' carboy this may go into the other mixed culture beers. And that's fine with me as that is more closely what it tastes like. The beer has a higher acidity with a citric and light apricot-like stonefruit character. There is a bit of acetic which has developed since transferring and in storage so it is good that this will only be a small component of a blend, where that acetic will be at an appropriately low level.

Common themes with my beers: One of the common themes in my beers is that I am missing the contribution of oak (at least the beers that didn't come from wine barrels). Some beers have been on a very small amount of oak cubes - usually soaked in boiling water to soften the more aggressive tannin character - but not high enough levels. I should note that while I am only saying oak, I don't necessarily mean an obvious oak flavor. And I certainly don't mean a bright tannic/more 'raw' oakyness (obviously they are toasted so not actual raw oak). But an older/more used oak can add softness and a bit of a smooth/mellowing/almost sweet contribution. And also some structure to aid the body of the beers. So that is more what I am looking for - the influence of a well used oak in mellowing flavor and structure. Low levels of oak (soaked multiple times in boiling water, and sometimes not first use) have now been added to all the carboys.

The other main common theme is a bitterness across most of the beers brewed between 2014 and 2015 (where I had been living in the same place, and which was a different place from the beers brewed in 2013). This shows up regardless of hopping rate, which microbes were (or weren't) pitched, and across different seasons. The one common thread is that they were all opened cooled at the same place.

After tasting the first couple beers I associated it with hopping rate but I am inclined to think that this isn't the case as I finished tasting through all the beers and noted that the bitterness can vary significantly in different carboys from the same batch. So I think this is microbe derived, and may have to do with microbe balance (note that brett pitching rate may not have much influence on brett flavor development) and/or other microbes promoting or reducing the phenolic bitterness produced by one/some. As I mentioned above, commercial brewers of spontaneous beers note that they can have rather phenolc younger beers and they can be almost undrinkable before aging into the end product that they are looking for. So for now I'm not too worried about this and I'll let them age for a bit to see if it ages out.

Edit: Thanks to Dan at MTF for reminding me about a great old MTF thread regarding the same sort of (presumed) phenolic brett-derived bitterness that I am noticing in my beers.

Some quick plans to brew for future blending:

Carboys in storage while I was away.
Fruit forward - Many of my beers weren't as fruit-forward as I'd like. I find that North American sours can be too fruit-forward for my taste and lacking some of the complexity that I love in lambic so I'm glad they weren't all super fruity, but I would like a fruit forward component or two to blend in with my beers that are, at least at this point, more phenolic. So I may try to come up with a blend of dregs/cultures for a future brew that I think will promote more fruityness in one or two carboys for use in blending.

New ambient microbes - I'm now living in a new location with different conditions so I'm interested to see what the microbes around here are like. My new place is a bit drier and more exposed rather than the more wooded valley conditions of my previous place. There are also fruit trees in the back yard compared to old oaks where I was living before. Given how variable different lambic microbes are in the relatively small region of traditional lambic production, I expect to see some fair differences in my current location compared to my previous. So I'll probably put out some more spontaneous starters to see what is around here and I am also looking forward to trying some fully spontaneous beers as the weather cools.

Fruit - I have some fruit from this year's fruit season frozen and waiting for aging on some beers. So when I do a more detailed tasting of the beers that I think may be ready for blending and try some blends I will pay attention to how the fruits I have may compliment them. I've got sour cherries, apricots and rhubarb (I know it's not a fruit but it is fruit-like in flavor contribution). I am already thinking that the apricots could be a nice addition to a blend with the sour saison in it given the light apricot/stonefruit character of that beer.

Finally I plan to re-try at the Rodenbach Vin de Céréale-inspired brew. I'll probably change the malt bill to a paler base with some higher color malt and may try to get it on used oak earlier in the process.

Otherwise I expect I'll carry on as normal with these mixed-culture beers. For hopping I think I'll stick with around 500 g/HL (~0.67 oz/gal) for now. The lower hopped beers were ones that I didn't like as much and I didn't notice any clear problems from higher hopping. I'll probably re-brew the rye base that I still have a bit of around and maybe something darker like a duivelsbier but otherwise keep it pretty close to the classic lambic tradition. Though I'm not calling my brews lambics (and mostly so far they are not spontaneous), that is the sort of final end product I am looking for in terms of balance, complexity, and general flavor/aroma characteristics.

This write up and tasting has somewhat critical but that is because I want to keep a high standard for my target end result and I don't want to settle for something that is only good. The beers are all in the direction I am looking for, so with more time, some minor tweaking, and a greater diversity of components I think I am on track to create what I want.

Ok, that's plenty for one post. If you want more details on recipes of certain components and/or how some are tasting feel free to ask! There was already so much in this post that I didn't want to expand it further but I'm happy to share it.

Saturday, July 30, 2016

Duivelsbier of Halle

Duivelsbier of Halle - Den Duvel zit in dat bier of het kan anders niet zijn!

If you were to ask the average enjoyer of Belgian beers what a Duivelsbier was, they'd probably give some response to the effect of 'Oh, like Duvel? And those other golden strong beers with devil-themed names? Yeah, a golden-colored, strong but light-bodied beer...'. And who can blame them. I would have given the same sort of answer not too long ago. A quick flip through Michael Jackson's Great Beers of Belgium even breaks down beers following this, with Duvel and other devil-themed beers all grouped together (and there's quite a few of them). But while Duvel is perhaps the most successful of these Belgian beers with names relating to satan, it certainly didn't invent the idea. Duvel's website traces the lore of the name to a fateful passing comment by a drinker in 1923. This blog post deals with an earlier devil's beer, as presented by Frantz Stockmans (who it seems really knew his beer) in Le Petit Journal du Brasseur in 1912, but tracing a history back centuries before that. So, on to the topic of the day - Duivelsbier of Halle.

Halle (Hal in French) is a city southwest of Brussels in Belgian lambic country. It is Flemish speaking, and is one of the larger towns in the area (population ~38,000 today). Jesuits had a reasonable presence in the city from at least the 1600s. Beer travelers might be familiar with Halle as the easiest way to get from Brussels to Beersel (for 3F or Oud Beersel) by train involves changing trains in Halle. Being in lambic country, lambic producers and cafes were once reasonably common in Halle, but those days are past. With them, it appears, went duivelsbier. One brewery (Boon) still makes one (see this quick blurb on lambic.info, a site you should all pay attention to if you like lambic - there is more English language lambic information there than anywhere else by a huge margin, see also this Dutch-language wikipedia page). I've not had it, but from reading online reviews and the Dutch page it seems it probably doesn't reflect historic Duivelsbier. Interestingly Vander Linden (a now-closed lambic producer in Halle) was producing the beer up until 2001, something I wasn't aware of. So while presenting the existence of duivelsbier may not be news to some, especially Belgians, hopefully some of the production methods and technical specs of circa 1900 duivelsbier are.

That's a promising start to an article! (ignore the reproduction prohibited part).
"Legend: Like Brussels lambic and other beers, the Duivelsbier of Halle has its history."

So what is Duivelsbier? Overview and Legends: It is a dark, strong beer made from spontaneous fermentation. According to this article in Le Petit Journal du Brasseur, this beer first came around somewhere near the end of the 17th/beginning of the 18th centuries. Lore (so take this as you will) has it that the Jesuits were brewing and one day they were without yeast for the beer they had just made so they put the wort into their barrels without adding yeast, planning to blend it with their next brew some time later. But it started to ferment on its own before that time, so they left it in the barrels. A year later they took a sample and found they had a vinous beer that, when sweetened, was quite enjoyable. And so the beer was born. The name Duivelsbier came shortly after that (again, from the lore). Perhaps the author and his sources, all residents of Halle, might not be unbiased in this little dig at the visitors from Brussels, but the story goes that some Brusseliers came through and, being unused to beers of that strength, drank a bit too much. On their way back home, when they were stopped by the mayor, they blamed the beer, saying "The devil is in that beer, it can be no other way!"(Thanks to Kevin of Belgian Beer Geek for translation help! The original Flemish quote is at the top of the post). As the story goes, the name stuck and, either way, this dark strong beer of spontaneous fermentation became well-loved in Halle.

Vander Linden's Duivelsbier (photo taken from a breweriana auction).
Making Duivelsbier: The production for the beer is basically what one might do for lambic (excepting maybe malt color). The grist is 50-55% raw wheat with the remainder as malted barley. Based on the end color of the beer, I think it is reasonable to assume that the barley used here isn't as pale as what one might find in modern lambic. Perhaps something like a Vienna, or maybe even something like Munich, is more accurate. This is guesswork as the text just says malted barley (and the text doesn't specify multiple different malts, so I'm assuming one diastatic malt with some color). The text also says the beer is brown/amber in color and that the krausen from active primary is black/brown, suggesting that the wort is fairly dark and color doesn't come only from candi sugar added later. The OG is in the range of 1.060-1.070, with the average beer probably clustered around the middle of this. This is comparable to lambic of the time, though perhaps slightly stronger than average/on the high side (the text, at least, argues that duivelsbier was the strongest of the beers of spontaneous fermentation, coming in at ~7.5-8% abv).

Though exact production methods were variable depending on the producer, the author favors a turbid mash approach, variations of which were common for many different Belgian beers (even non-spontaneous beers). The mash had 4 temperature steps, reached by infusions, of 45 C (113 F), 55 C (131 F), 65 C (149 F), and 75 C (167 F) and with turbid runnings withdrawn to a second boiler at the 55 C, 65 C and 75 C steps. The turbid runnings are boiled throughout this mashing and then returned to the mash tun to be clarified by passing through the grain bed before being transferred into the main boiler. Much like with lambic of the time (see these first and second earlier posts on lambic in the mid 1800s), later runnings of the mash were sent to the now-empty second boiler to make a lower gravity beer (called Mars, again like lambic).

Boiling was 4-5 hours for the duivelsbier and 10-12 hours for the mars, comparable to lambic production. Unfortunately hopping rates aren't listed, but for now I'll guess that they are in the range of lambic given that the lactic acid levels are comparable between the two beers (though duivelsbier was a bit less acidic). After boiling the beer was sent to a coolship for open cooling, and then to barrels for fermentation. The beer sat in those barrels for roughly 24 months before it was ready. A bit of candi sugar (exact amount depends on the cafe doing it) was added to the matured beer shortly before serving. Under a microscope the microbial composition of a bottle of duivelsbier appears comparable to that of Brussels lambic, which I think is unsurprising given that Halle is central to the lambic region between many modern producers and duivelsbier was made by spontaneous fermentation. Parts of the text suggest that perhaps there was a bit less brettanomyces in duivelsbier, but I'm not personally inclined to put a lot of stock in that.

Characteristics of Duivelsbier
The beer is described as soft in body, carbonated and easy to drink in volumes greater than you intended to. It was clear and with a color of cognac. The taste profile is described as being between Brussels lambic and a good aged Oudenaarde beer (so something in the Oud Bruin or Flemish red/brown range). Of course keeping in mind that we are talking about these beers around 1900 so we can't really compare to modern examples, this gives us some idea of the color/malt character and the way this came across in the finished beer. The sweetening with candi sugar shortly before serving may push the beer more toward the faro direction than a straight lambic.

I've adapted information and tables in the text into this table of units (below) that are more easily understood. As usual, at times a mixture of units is used and there is a bit of information that doesn't seem to line up, but based on the whole of the information, here is my best guess at the most accurate values. The acid values are more clear and I have more quantitative confidence in them. Note also that some of the lambic/gueuze gravities and alcohol may not be totally representative of the range of lambic.

The acidity of the beer was generally in the realm of lambic, though it was a bit less, especially for acetic acid. The lactic acid levels of duivelsbier were about 85% of contemporaneous lambic and the acetic acid concentrations were a bit over a third of those in contemporaneous lambic. The acetic levels in historic duivelsbier were comparable to modern day flanders beers, so there was definitely a noticeable amount, but it seems that, given the age of duivelsbier, brewers did a good job of keeping O2 out of the beer to suppress acetic acid production compared to historic lambic. The FGs, OGs and ABVs were generally in the same ballpark for duivelsbier and lambic (FG in the mid single digits in SG, OG around 1.065, ABV 7.5-8.2%) though some lambics of the time were a bit lower strength than this. And of course most modern lambic is lower strength than this.

Duivelsbier in the modern world: I think duivelsbier is a good example of how much lambic and lambic-type brewing has changed over the last century. 100 years ago lambic was still doing comparatively well (though perhaps not as well as earlier in its history). But still, there were many producers and more 'styles' of un-fruited lambic than the one that we find now (almost every modern producer makes one lambic wort stream, though Boon does make a Meerts and perhaps their Marriage Parfait is also stronger). Historically with lambic there were two wort streams produced regularly - Meerts and Lambic, which were combined to produce faro. This was standard practice across many different breweries. In addition, we have this regional/specialty wort of duivelsbier and I wouldn't be surprised to learn that there were more worts like this elsewhere. And furthermore lambic was regularly supplied for blending into other beers to produce lambic-top fermentation blends. Some of these are coming back, which is good news. But anyway a lot was happening with lambic in terms of wort compositions and blending to yield a range of final products compared to now. Of course now we're doing pretty well for lambic with a wide range of fruiting and use of specific barrels for flavor addition. So really one was born as another faded away.

So, to bring this all back together to where we began, there is one surviving duivelsbier today. One lambic producer from Halle, the now closed Vander Linden, made it until 2001. Since 2003 Boon has produced the beer, though it seems that Boon's duivelsbier has more in common with something like an abbey brown than with the historic duivelsbier. While it is good that someone is keeping the name alive, and I'm sure the people of Halle are happy about keeping a beer of their heritage around, I would be really interested to see a beer resembling historic duivelsbier today.

So does anything like the old duivelsbier exist now? Or has has one been around anytime recently? Not exactly (to my knowledge) but there are certainly elements of overlap in terms of taste characteristics with some other beers. And I think we may have come pretty close with 3 Fonteinen's Straffe Winter (note that the beer was not marketed as relating to duivelsbier). The beer was a stronger (8%) spontaneously fermented beer brewed with darker malts (amber and Munich malts as well as lager malt) and wheat, and with candi sugar and the beer had faro-like characteristics). That seems to hit the main points of duivelsbier as listed by this text: a stronger, darker, lambic-like spontaneously fermented beer sweetened with candi sugar. So, it seems to me at least, that this had the potential to be pretty close! It's been far to long for me to try to recall of the characteristics of the 3F beer with any accuracy, and when I had it I didn't know about duivelsbier or have much appreciation for a good faro, but it would be interesting to try that again with this in mind. With the current state of lambic and it's growth, perhaps we'll see such a beer in the futrue...

Friday, July 15, 2016

A Visit to Seizoensbrouwerij Vandewalle

Returning to a series of beer travel-related posts I was working on a few months ago, here is a post about a visit in May to Seizoensbrouwerij Vandewalle. Seizoensbrouwerij Vandewalle was started in 2011 (first beers released in 2012) by Chris Vandewalle. Chris is the 10th generation in a line of brewers and he has a serious passion for both beer and his region. As is not uncommon for small Belgian breweries, the brewery is a second job for Chris, who works as a regional historian for his day job. Chris's passion for beer and history make him a great resource for learning about beer of the region and he is proud to bring brewing back to his community of Reninge which, although the current population is only ~1000 people, he reports was once home to half a dozen breweries.

Hop fields outside of Poperinge.
The brewery is located in the municipality of Lo-Reninge in the southwest corner of the West Flanders (West-Vlaanderen) province. It is not far from Poperinge, one of Belgium's hop growing centers. This region of Belgium was hit especially hard during WWI. Perhaps nearby towns such as Ypres are familiar to those who remember some WWI history. Between people fleeing the fighting to start new lives elsewhere (and not returning) and those killed during the war, this time saw the region lose nearly a generation's worth of people and knowledge. This included brewing knowledge and the types of beers being brewed. Chris is passionate about collecting what was left/lost and bringing that knowledge and history back, for general life and culture as well as beer.

Even without meeting him his passion for his region and beer traditions should come through clearly based on the line of beers he makes and the regional West Vlaams product logo proudly noting the use of local ingredients.

The Vandewalle beers: Oud Bruin, Bitter Blond, Kriek, and BB à Lambiek
The beers
Chris's beers all have some common threads, most notably is a certain edge. To say that his beers are unpolished gives absolutely the wrong impression, but let me explain. I don't mean that his beers are rough takes that need working out yet. His beers are well crafted and are made to be exactly what they are through recipe and process. But the beers have an edge to them. On purpose. They have not been rounded out or softened for broader appeal. So that is what I mean by unpolished - they retain every bit of grit and character that Chris intends. The blond is firmly bitter - more so than I think any other Belgian beer I've had (excluding maybe Belgian made IPAs, but even then it is more bitter than most of those). The Oud Bruin is not as sweet as those familiar with the more commercial examples would expect. And the Bitter Blond à Lambiek is forward in its brett character. So I think unpolished is an accurate description in this context, but perhaps it's better to say it this way - Chris brews beers with an edge.

Chris's walk in cooler with bitter blond and a selection of bottled beer.
Chris is currently making 4 beers. Given his location in hop growing country, it may not be surprising that the Bitter Blond was his first beer (first released in 2012). And although I didn't ask about production breakdown, it seems to be his main beer as well. The accurately-named  Bitter Blond is a firmly bitter beer with a pleasant landrace-type hop character. While the hops are grown in West Flanders, they are English varieties. It is common now to find English varieties grown in Belgium as the historic Belgian varieties had been replaced. Some small farmers are starting to grow older Belgian varieties again, but it will likely be some time before they are produced at a commercial level.

Barrels of Oud Bruin.
While the Bitter Blond doesn't try to classify itself as such, this beer fits well with modern saisons. The first, and most obvious way, is in the pale, dry and firmly bitter characteristics of the beer along with the yeast character (though the hops play the dominant role in this beer). Additionally the motivation for the beer fits the lore of saison as a beer of farmworkers. Agriculture still plays a major role in the regional Lo-Reninge economy, and historically as well as in the modern day one of the major activities was growing (and then harvesting) grasses. This can be seen if you time a trip right, although now the cutting is done by industrial farm equipment. Historically this would have been by hand, and as Chris explained the field workers would have needed a refreshing bitter beer to quench their thirst. Chris has made his Bitter Blond in this spirit.

Chris's second beer was his Oud Bruin (also 2012). Sour beers were once common in the region, though you wouldn't know it by looking at many of the nearby modern breweries. But in line with the modern remaining Flanders red-brown beers, the wider region was formerly known to produce darker acidic beers. Like others, Chris's Oud Bruin is a blend of an aged beer (a brown which spends about 1 year in oak) with a younger brown beer. The beer is drier than many others on the market with a great acid balance and a touch of sweeter/mellowing character from the younger beer. This is a great modern example of these mixed-fermentation brown beers which were once ubiquitous in the region, especially as sweet beer showing only hints of age and mixed fermentation can be over-represented.

Kriek base in barrels awaiting the summer's cherry harvest.
The third beer is Krieken Rood, a kriek made using locally grown cherries. The beer ages in oak with the whole cherries (and as Chris confirmed with a laugh, those cherries are slowly and meticulously removed from the barrel by hand). The beer is brewed in February, where it waits in oak for the cherries. In July when the cherries are in season they are added to the beer and they remain with the beer until the following February, when the beer is bottled and the cycle begins again. It was just released when I stopped by in the first week of May, meaning roughly 2-3 months of bottle conditioning.

This kriek is unlike any other that I've had (note that kriek just means sour cherry and, as such, does not have any inherent tie to a given beer style). The acidity of Chris's kriek is mellow, but present, and the focus is more on a fuller/sweeter fruit (but it's not a sweet beer!) and almondy/woody character. Part of this likely comes from the varietal he is using, which his website lists as Nordkrieken from Veurne. Unfortunately I have no other experience with these cherries either on their own or in other beers, so I can't speak much to the character of these cherries. The time in oak barrels which are less neutral (more on this below) than what many other Belgian kriek producers are using and the longer contact time between the whole fruit and the beer likely play into the uniqueness of this beer as well. Here are some photos of the cherries from the 2016 harvest, coincidentally posted to the Seizoensbrouwerij Vandewalle facebook page the same day that I published this post.

The fourth and newest addition to the lineup is Chris's Bitter Blond à Lambiek. The first blends were released in 2015, though bottles list blending dates as early as 2013 so this beer has been a part of the Vandewalle plan for some time. The beer is made from blending his Bitter Blond with commercial lambic (coming from Oud Beersel) at a ratio of roughly 5% lambic. The blending occurs right before bottling. After this the beer spends at least 8 months in the bottle to condition, though I've not seen/had a bottle that was less 12-18 months old. Likely due to the substantial hop presence, the beer doesn't develop a lot of additional acidity from the extended conditioning; however the Brettanomyces certainly makes its presence known. Amos at Browne & Bitter talks about this beer and the Bitter Blond a bit more in his contemporary Bière de Coupage post.

The brewery and brewing process
Chris is brewing 450 L batches on a Braumeister system and he is employing some non-conventional process to make beer which (fittingly) breaks from the general Belgian norm. One of the most striking things about Chris's process, and something that he stresses when talking about brewing, is the time that the beers are given. For example the Bitter Blond, his beer which is released the youngest, spends 3-4 months (mostly in cold storage) between brew day and bottling. And the Bitter Blond à Lambiek spends at least 8 months bottle conditioning before release. Brewing at the scale and running the brewery that he is, Chris can fully let the beer dictate when it was ready and give it all the time it needs to get there. He firmly feels that this is best for his beer, and by tasting the products I agree with him.

The cooling tun and open fermenter.
Chris's beers are open cooled overnight. This open cooling is done outside in a dairy tank, which is then wheeled in to the brewing building where yeast is pitched and it ferments in the same vessel. This open cooling allows for the potential of mixed-microbe inoculation and when I taste the bitter blond I get the sense that there is something more than just sacch at play. Additionally, given Chris's focus on time, for many of his beers there is sufficient time for a mixed culture to express its different sides. The open fermentation in a relatively shallow wide pan certainly has an influence on the expression Chris gets from his yeast. Both of these processes - cooling outside in some sort of open vessel and then pitching yeast and fermenting inside in that same vessel - are accessible to home producers more easily than larger commercial producers. I may try some of this out when I'm back in the swing of brewing and the weather cools a bit.

More barrels at Vandewalle.
Chris is using oak barrels for the production of his Oud Bruin and Kriek, and additionally he is also aging a bit of Bitter Blond in oak for trials. It is interesting to note that all of Chris's barrels are new - their first use was his beer. He doesn't want used barrels (e.g wine barrels) as he doesn't want other microbes from whatever the barrels held previously influencing his fermentation. Now after 5 years of use he is happy with how his barrels are mellowing out and the characteristics they are giving now, though he is also expanding his barrel production and therefore not all his barrels are 5 years old yet.

Visiting - Public visits to the brewery are probably best set up for small/medium groups, and include 3 beers and some local snacks. Contact the brewery to set up a visit. Otherwise look for Vandewalle beers around Belgium. They can be found at some select good beer spots (Malt Attacks in Brussels, Mi Orge Mi Houblon in Arlon, and Bierhalle Deconinck in Vichte to name a few).

I'll close with a quote from Chris, which was accompanied by a firm pat of his stomach, about how he is brewing and what drives him to brew the way he does/the beers he does:

"A brewer does it his own way, following his belly!"

If you're interested in Belgian beer travel I've written up some general thoughts as well as specific insights on visiting lambic producers and saison producers (with links for posts of my visits to individual producers within those).

Tuesday, June 28, 2016

Categories of grisette and grisette strength

This post,which builds on some of the thoughts and label analysis presented in my 'What is Grisette?' post and my talk at Homebrew Con 2016, will dive into what information I have about the the general strength range that a historic grisette falls into. As usual there are still questions, but hopefully this helps to make the identity of grisette a bit more clear and helps you choose what strength to make your grisette. Also, as with the previous post about hopping grisettes, this builds off of an incomplete set of source material which relies heavily on Pelset's 1874 book. This text, while fairly thorough, has some inconsistencies, leaves some stuff out, isn't always clear, and in the end is only one source from one point in time for >100 years of beer brewed by many different brewers. The book is also on the earlier side of grisette info compared to other info out there (personal accounts from people who remember grisettes before they disappeared and labels, both of which would reflect mid 1900s grisette).
A Cat. 3 grisette and...

To start this off, I want to look at the more modern sources that today's Anglophone brewers might be familiar with and what they say about grisette in regard to strength. These are Farmhouse Ales (Markowski, 2004), Brewing with Wheat (Hieronymus, 2010), and (most likely via Farmhouse Ales rather than directly) Belgium by Beer, Beer by Belgium (Perrier-Robert and Fontaine, 1996).

-Farmhouse Ales cites Leon Voisin, retired brewer at the old Brasserie Voisin (see the saison label below) to say that grisettes would have been 3-5% beers. Shortly thereafter Belgium by Beer, Beer by Belgium is cited for an OG of 16.3-17.5 Plato.

a Cat. 1 saison from the same brewer
-Brewing with Wheat (citing Pelset, 1874) gives a gravity range of 10.2-11 Plato and mentions that there were three classes of historic grisette: young/ordinary grisettegrisette de saison, and grisette supérieure.

-Belgium by Beer, Beer by Belgium, drawing information form Pelset 1874, gives an OG of 6.5 to 7 degrees Belgian, which works out to 16.3-17.5 P or 1.067 to 1.072.

Excerpt from Pelset, 1874 mentioning gravity.
Some of this is in general agreement and some is quite different. To try to figure that out let's look at the original source material for Brewing with Wheat and Belgium by Beer, Beer by Belgium. This source is Pelset's 1874 book on grisette. Pelset gives the three classes of grisette mentioned above and gravities for the first two classes of grisette. For the third the author mentions that the later runnings were not mixed in and instead were used for a small beer. The section dealing with gravity, shown at left, gives the following:

Young/ordinary grisette brewed in winter: 6 degrees
Young/ordinary grisette brewed in summer: 6.25-6.5 degrees
Grisette de Saison: 7 degrees

Here's the source of confusion: these units are "degrees in an ordinary brewer's scale". At that time there were three primary units of degrees in use: Baumé, Balling, and Belgian. I don't have a strong handle on which would have been the most common in Belgium. It probably depends a good deal on where/when a brewer was trained, which would be influenced by scale (if they were more of an industrial brewery or a small local brewery). As mentioned previously in this series, the sources available are generally biased toward industry and grisette, if its roots are in mining, is a beer with an inherent tie to industrial activity. For reference, Lacambre (1851), one of the seminal books on Belgian brewing from the era (focusing on more large scale brewing activity), uses Baumé.
Results from an analysis of grisette labels, June 2016.

Given the gravities from Pelset (which are referenced to 10 degrees Réaumur, or 12.5 C), here are the equivalent gravities in OG at 20 C / 68 F:

6 Baumé = 1.042
6 Balling = 1.022
6 Belgian = 1.060

This is a big range, and only one of these three possible units is what Pelset meant. To try to determine which type of grisette would have been most common in the mid 1900s I looked through a collection of historical labels from breweries that closed before the mid 1990s, compiled and digitized by Jacques Triffin. The results from that analysis is in the tables at right.

Results from an analysis of saison labels, June 2016
While labels don't necessarily give clear information about recipes/characteristics, there is some goon insight hidden in there. First, some of the labels list a category of beer, which is the strength class of the beer (which to my understanding is for tax purposes, though it also conveys the strength to the consumer). Looking at modern beers, some Belgian beers with OGs in the realm of 1.050 still list category 1 on the label (see Girardin lambic). Categories like these are also used in Dutch brewing. In modern Dutch beer:

Category 1 beer would be 1.044-1.063
Category 2 beer would be 1.028-1.044
Category 3 beer would be 1.004-1.028.

Some historic and modern Belgian beers list category S (Supérieure, see Seizoensbrouwerij Vandewalle below as an example). In some historic labels this is listed along with Cat. 1, so this is either at or above Cat. 1. For the purposes of this analysis I have combined it with Cat. 1.

A cellar-matured saison. Keep it lying down and serve in a basket.
Brasserie Voisin's 'old' saison.
As you can see, the grisette labels cluster toward the lower end of the strength categories when the category is listed (given the caveat that not many list the strength) while saison clusters much more toward the higher end. The grisette labels were also more likely to mention something about being a table bière or a bière de menage (house beer) than saison. While grisette labels didn't mention anything about being old beer, this was not uncommon on the saison labels. Labels not listing some of this info are an unknown, but with the information given this supports that at the time period of these labels (many of the breweries in the collection closed in the mid 1900s) grisette was a lower strength beer that wasn't aged while saison was stronger and more likely to be aged. This is also true for paired labels (grisette and saison labels from the same brewery). Though with the range in both beers, it is clear that these beers were still heterogeneous and would vary in factors like strength and aging between different producers of the same time range.

Therefore, though there is a lot of good historical cultural insight in Belgium by Beer, I think they are mistaken with the OG of grisettes. That range is well out of line with the ABV range quoted in Farmhouse Ales and it also doesn't agree with the label analysis. Over the time period from the mid/late 1800s to the 1900s, gravities of Belgian beer were not dropping significantly (excepting during the wars) and in general gravities of average beers were rising. So it is unlikely that the gravity of grisette dropped significantly between Pelset's text in 1874 and the labels, mostly coming from the mid 1900s. The interpretation of degrees as degrees Belgian simply doesn't line up with the other sources.

Taking degrees Belgian off the list then we are left with an OG of either 1.042 or 1.022. If we then assume that the historic Belgian category system is roughly in line with modern Belgian and Dutch levels (and perhaps this is a bit assumption, though I think it is not likely that the categories have grown significantly weaker from the mid 1900s to now), a category 3-2 beer would be somewhere in the range of the 1.020s to 1.030s, maybe into the 1.040s. This matches with both the remaining Balling and Baumé gravity scales, though these degrees fall toward the low and high extremes respectively based on the expected gravity range from label categories. If I had to choose one I'd probably lean toward 1.042 for a couple reasons. First, that matches the units of Lacambre, 1851. Second, degrees Balling would put grisette de saison in the mid 1.020s, which is too low for a bière de saison. And finally, interpreting the units as Baumé matches the interpretation given in Brewing with Wheat which likely came from Yvan de Baets, and he is pretty informed on these sorts of things.

A Cat. S beer from Vandewalle.
However Baumé doesn't quite line up with the labels listing a 3rd category beer. I'm not sure what to make of this, so perhaps it is safest to target a gravity range somewhere in between the two. Splitting the difference I think it is reasonable to base grisette in the 1.030s, possibly toward the middle to higher end to comfortably hit 3-5% abv on a beer with reasonable attenuation but without the benefit of prolonged mixed-fermentation to increase attenuation further. For this reason I tend to target gravities around 1.035 for the average (ordinary) grisette personally, but I think you could have some substance to an argument that beers which are both stronger and weaker could accurately be labeled grisette, depending on time period and grisette subcategory.

But what about grisette de saison and double/supérieure grisette? Brasserie De La Senne and Brasserie Thiriez brewed a grisette which Yvan de Baets told me was targeting the grisette de saison strength. This beer came out at 5.3% (DLS brew) and 5.5% (Thiriez brew), so guessing a bit on the FG this puts the beer somewhere around 1.045 OG. This fits with normal grisette being in the mid to upper 1.030s if grisette de saison is about 17% stronger that ordinary grisette, as stated in Pelset's text.

I don't have a lot to go on for double/supérieure grisette. Two distinct double grisette labels showed up in the label analysis but no strength classes were given on those labels. With grisette de saison in the mid 1.040s I think a double grisette has to be above 1.050, and probably not over 1.060, or at least not much over 1.060 given general beer strengths of the time (though this upper limit is purely a guess).

In closing, I want to say something about how this grisette strength info plays into the identity of grisette in the modern world. Two things are clear:

1) based on the labels, grisette was brewed at the same time as saison from some of the same brewers
2) It seems that most grisettes were lower strength, but that is not necessarily true. There were saisons brewed at the category 3 level and at least one grisette at cat. 2 (not to mention the double grisettes, which don't specify a strength but according to Pelset would be stronger than beers of roughly 1.045 OG).

Therefore it appears that there is something in the identity of grisette that, historically at least, made it different from saison more than just being lower strength. I think most grisettes were low strength beers, and if you were going to try to make a grisette it is probably best to start in that range (mid to upper 1.030s for a standard grisette by my estimation). Grisettes definitely clustered to lower strength than saisons in the label analysis so I do think that is part of the story (at least that grisette wasn't a beer requiring age and commonly brewed to as high an OG as saison). But there are other aspects that define grisette and help to differentiate it from historic saison. So far in my research those seem to be a requirement for malted wheat, generally lower hopping rates than saison (but still high enough to be noticeably hoppy), and beers designed to not be harsh/bitter/astringent when they are young (as both the grain and hopping rates of saison might contribute to).

The differences get murkier in the modern world, where saison is generally a softer beer than historic recipes suggest and most saisons are not mixed fermentation or are not aged long enough to give mixed fermentation time to express itself. But what I've seen so far supports grisette as a hoppy beer (but probably not astringently so) and does not support grisette as an acidic beer. Grisettes would not be sweet beers, as is the case from the one regular producer in modern Belgium, and from what I've seen it is best not to go for acidity when brewing grisette.

Other posts on grisette:
What is Grisette?
Grisette Recipe
Hopping Historical Grisettes