Saison Brewery Visits - some general thoughts

After visiting a couple of different Belgian saison breweries over the past month or two (and some others over the past couple of years) I thought I'd give a bit of insight into the different modern takes on this style. And this also gives me a chance to put up pictures I have from some of the different leading saison breweries. This is a bit more of an overview and the more specific details on what I took away from different breweries will probably be saved for posts dedicated to those breweries.

As I mentioned in my about me section, the saison beer style is held together more by spirit and feel than by the sort of specific technical parameters you'd find in most beer styles. So while the strength, color, presence of spices, hop levels, etc. in different saisons may vary significantly, all the top commercial examples are dry, refreshing, yeast-driven beers with exceptionally low finishing gravities. Here is some of what I've seen from what I consider different leading Belgian commercial brewers of this style.

The Brasserie a Vapeur steam engine (from 1896).

Brewhouse facilities
Saisons, sometimes called farmhouse ales, are often romanticized as being rustic beers. While this is certainly true for the history of the style, not all commercial breweries (including many of the best) fit into this rustic ideal. Some of the breweries located in the center of villages/towns fit perfectly within the farmhouse feel/image while others out in the countryside have a modern setup. Some breweries (most notably Brasserie a Vapeur) are using some equipment over 100 years old while others, such as De Glazen Toren, have a beautiful new brewing facility that they acquired within the last 10 years or so. So while the style may hold a farmhouse connection to some, the farmhouse connection is certainly not ubiquitous.

The De Glazen Toren brewing setup.

Ground malt waiting for the next brew day at De Glazen Toren

Saisons are generally yeast-driven beers so the fermentation is going to control the final product. Some breweries are using pure cultures for fermentation and some are using mixed cultures, both to beautiful results (though I'm not sure all the breweries with mixed cultures are intending to do so). Some brewers, such as Dany Prignon at Fantome, will at times choose different yeasts and bacteria to add for some of their different brews. So while the fermentation (and the fermenting organisms) may vary greatly, the fermentation tanks are the only consistently reasonably modern equipment I've seen in all the saison breweries I've visited. I suppose that this is unsurprising given the advantages that conical fermenters and clean-in-place systems provide and because while a brewery only needs one mash tun and boil kettle, they will need many fermenters. There was still some reasonably variability between some of the breweries I visited but the variability between fermenters within a given brewery surprised me more than the varying fermentation setups at different breweries. Especially when considering the equipment differences on the hot end.

Spices and 'feel'
My favorite more typical saisons are either dry and rather hoppy (such as Dupont, Thiriez, and De Glazen Toren) or much more mild with the hops and spiced with a blend such that no single spice character seems to dominate the beer (such as Fantome and Brasserie a Vapeur). Achieving the appropriate balance, character complimentary to the yeast, and choosing some original spices is important here. In my mind Dany at Fantome is the master. He doesn't keep a written log of what he does but he's used so many different spices over the years that he has a feel for what levels of which spices are appropriate for the balance.

This sort of intuition or feel that I observed with Dany is something that stood out in the saison  brewers (and for that matter lambic producers as well) that I visited. It goes beyond recipe formulation though, as I saw many parts of the process controlled by feel or quick observation rather than measurement. And in general I think this is a big difference between many commercial brewers and most home brewers. Home brewers often jump around to very different styles, brew less frequently, and have more regularly changing equipment, environment, and process than commercial brewers. And even when the commercial brewers are regularly brewing something new or something they haven't done in a year or so (again taking Dany at Fantome as a perfect example), the feel they have for the style and how to use their ingredients makes a big difference.

Mashing in at Brasserie a Vapeur. The water to grist ratio was adjusted based on the feel of the mixture coming out.

Many homebrewers will often hear the advice that they should repeat the same recipe to get a hold of their process, and I think this is really important. But for those homebrewers out there who would get bored with this and lose what they enjoy in the hobby, try something like what Dany is doing. Find a style of beer you like making and/or drinking. It could be pale ales, it could be saisons, it could be stouts, whatever it is you really like. And brew some similar batches in a row. You can change parts of the recipe or parts of the process but it will still give you an idea about what different ingredients and techniques will contribute to your final beer. Whenever I get a new pack/vial of yeast (especially considering that yeast pitches can be one of the more expensive parts of brewing), I grow it up and split my starter for multiple batches in a similar family or style. Then I'll brew these batches over the next month or so. Keeping this approach of brewing similar styles with similar process and/or ingredients for a few batches in a row helps get a hold of the influence of my ingredients and process. There is no substitute for actual brewing, the closer together the batches are the better, to gain a feel for your system and an intuition for how different ingredients and processes will influence you beer.

Closing news - note: edited Jan 2015
And I'll close here with an exciting piece of saison-related news. Historically, some saisons would likely have had a character similar to younger lambic from the extended conditioning period and the non-microbiologically pure fermentation. And there is some evidence of a portion of lambic being blended into saison (for a good quick history of older saisons check out the Yvan de Baets chapter in Phil Markowski's Farmhouse Ales). This would certainly aid in the refreshing qualities of the beer and this sort of idea can be seen in some commercial beers today. Most notably Cuvee de Ranke, which is a 70/30 blend of a hoppy pale beer and lambic from Girardin. Edit: the info on the De Ranke website describes this a bit differently, and perhaps the base beer isn't as hoppy as the shelton website would lead one to believe. I've had it on tap in Belgium where it tasted rather hoppy and not very sour, lending credit to the Shelton description, but it may have been a case of a mistake in which beer was put on tap.

Edit fall 2015: An earlier version of this post included information that Jean at Cantillon was interested in some saison-related trial projects. I interpreted this to mean actually doing a bit of brewing with pitched yeast, but I believe I was wrong on that. I've heard no further mention of this so perhaps it isn't going anywhere. It would still be awesome, but I think focus is going elsewhere like toward the new facility expansion and this idea is on the back burner and/or shelved. Or, I think more likely, I was really excited by the idea and didn't fully understand what he meant by it. Jean is involved in saison projects by providing his lambic for blending with saisons produced elsewhere (such as a De La Senne and Hill Farmstead) and other beers (like with Agullons and Birra del Borgo (before the buyout - ed May 2016)), so this is more likely what was meant by his interest in saison-related projects. I expect he will continue to be involved like this and this is probably what he meant when he said he was interested in projects inspired by historic saisons. And I think my enthusiasm let my imagination run a bit. I also got a bit of a feel that this was a bit more of a longer term dream/goal, so it may be possible that he will expand these collaborations. I'll update if I learn of any tangible developments.

Edit May 2016: While I was certainly to quick to post about Cantillon experiments, I was also too quick to edit back that post. I won't say much more here and will leave any further discussion of this to the people involved if/when the time comes. On the whole the original information was valid - that Jean has an interest in continuing to produce quality traditional lambic as well as see how his lambic could interact with other beers/worts.

Alright, that's all. I'll plan to write up some specific brewery visits but thought that this general comparison would be a good way to get some ideas out and set those specific visits up.

Notes from specific saison breweries:
Brasserie Thiriez
Brasserie a Vapeur

Old Ale w/ Brett - Tasting notes

Again, sorry for the delay in getting this post up. I spent some time visiting breweries in Belgium and didn't get a chance to get the post up before the trip. Also, sorry for such a long post with no images. But the visits in Belgium have given me a good amount of photos for future posts. Anyway, on to the post.

This is the tasting and blending notes for my Old ale with Brett c recipe. As a recap, I bottled three different treatments: the non-brett base alone, a blend of 67% non-brett base with 33% brett aged, and the brett aged portion alone. I added White Labs Brett. c. to a portion in the secondary and all three treatments had the same 'clean' (no brett) primary fermentation. The clean and blended bottling was done at the beginning May 2013 (original brewday was 19 October 2012) and the 100% brett aged was bottled at the end of August 2013. As a reminder, the 3 gallons of clean portion were primed to a low level (1.8 volumes) appropriate for the style and after bottling about a gallon and blending in the gallon of brett aged old ale, no additional sugar was added. The 100% brett aged portion was primed to 1.8 volumes.

This was the first time I had tried to blend two batches just before bottling (aside from a couple of random bottles here and there) and as I eluded to in the first post it proved to be a bit trickier than I had anticipated. I had a great plan. This was to bottle 1 gallon of the clean old ale (3 gallons total in a rather full 3 gallon carboy) and then siphon in one of the gallon jugs of brett old ale, mix gently, and bottle the blend. My plan for siphoning and starting the siphon is where I had a bit of trouble. I wanted to use my racking/bottling cane, full of wort, to start the siphon of the brett portion. I though this was a great idea to minimize mid-bottling sanitaiton steps and air exposure. The issue came with the bottling wand tip. I didn't want to put it all the way to the bottom of the 1 gallon jug as that would stir up some sediment, and the angle and wand length were not enough for me to push it on the side of either the carboy or the 1 gallon jug to depress the tip and get flow.

The deliberation for a new plan mid-process and removal of bungs while trying things out allowed for more air exposure and also resulted in the less than ideal solution of quickly pinching off the tubing, removing the bottling wand, resanitizing the end of the tubing and starting the siphon. Also with my 3 gallon carboy on the counter for bottling I was without a good place to put the 1 gallon jug to get and keep a siphon going. So I had to end up holding it up on my shoulder for the duration of the siphoning which was not especially pleasant and also almost certainly resulted in more jostling and more sediment resuspension and transfer. Lesson learned for next time.

I've tried to detail what I thought were my mistakes pretty well to help others avoid them in the future. While it didn't go as smoothly as I had hoped, it still worked out pretty well. So here were my 4 take away messages from this experience to make it go better next time.

1) Have multiple racking canes sanitized and ready to go from the start. Maybe you won't use them all but it's better to have them there than need to adjust one or resanitize mid-process. With only one racking cane on hand I had a tricky time here.

2) Have the gravity feed setup in place for all steps before starting. You'll potentially find yourself with rather full hands with the racking canes and other miscellaneous stuff so moving carboys mid-process can be tricky.

3) As mentioned in the step 2, your hands will often be full. Have everything you need (sanitizer spray bottle, caps/bungs, racking canes, CO2 bottle for purging, etc.) close by. And maybe have some options for holding racking canes that don't result in either a lost siphon or non-sanitized equipment that you will need to use again. I had no place to put down my racking cane without losing the siphon and that made moving things around tricky.

4) Have some help. Bottling alone can be fine but with a new process involving multiple vessels, possibly multiple sets of tubing and more scattered equipment, it is really good to have another set of hands around. That definitely saved me.

On to the tasting notes. Both the clean portion and the blend were tasted at the end of August 2013 (3.5 months in the bottle). The 100% brett aged portion was tasted in the middle of September 2013 (only 3 weeks in the bottle). Perhaps what stuck out most to me was an acidity in the brett treatments. Where did the acidity come from? It is definitely possible that it came from bacteria, but I don't think that is the case because the brett portions weren't sour before bottling and both went sour in the same way after bottling (meaning they would have needed to pick up the same sort of bacterial contamination at their different bottling times). Plus the sourness did not seem to increase with added time in the bottle, there were no bottle pellicles, and there certainly wasn't an excess of CO2 production. I think it came from brett acetic acid production due to oxygen pickup. While brett on its own doesn't really produce much sourness in fermentation, if you give it extra oxygen after the growth period it can produce a lot of acetic acid. So with the bit of O2 I introduced with bottling I got a bit of acetic acid, which was not enough to be especially unpleasant and production stopped when the O2 ran out.

Clean Old Ale:

Aroma - Rich candy sweetness and caramel dominate, the more earthy English base malt comes through nicely, virtually no hop presence as would be expected for the style and the recipe, mild alcohol.

Appearance - very low tan head, fair retention, peach-amber color with a fine haze

Taste - Rich and earthy maltyness, fig, grape, and a bit of oxidized apple character, caramel, the Marris Otter base comes through nicely, light alcohol presence but overall smooth for it's strength. The rich English malt character dominates with a some pleasant supporting fruit in the background.

Mouthfeel - Low fine carbonation (at a good level for the beer/style), medium/full body but no heavy sweetness, light alcohol warmth in the finish

Overall - Mellow and smooth, on the more subtle side with no aggressive character dominating. The alcohol gets more noticeable with warmth. In comparison this batch was more malty and much more subtle. Potentially my favorite of the three but the blend is a close second. We'll see how they develop.

Blended Brett Old Ale (67% clean, 33% brett aged):

Aroma - Tart and vinous, the brett is fruity oriented and its influence comes through strongly, the aroma of the blend is much more complex and less malt forward than the clean version, there is a light leather character that adds a nice complexity

Appearance - Pretty much the same as the clean version, peach/pink amber with a fine hazyness, very low head (lower than the clean version)

Taste - Much like in the aroma, fruity vinous and tart brett character come through strongly. The fruityness is grape and berry oriented. The malty and caramel character in the clean portion come through in the finish, and it is good to see that some of that is still there. The earthy malt character works well with the subtle leathery edges in the blend. There is defintiely a tartness here, and while it is interesting it is a bit more than I was shooting for and throws off the balance bit. The acidity is acetic oriented but is low enough to be ok to me. Generally I am not a fan of acetic sourness in beer but here it is sufficiently low to not be problematic or rough here.

Mouthfeel - Very low carbonation. The blend had no extra priming sugar added, so it had 2/3 the rate of the initial clean batch priming sugar. I expected the brett to slowly work through what is there in the clean majority of the blend but that doesn't seem to be the case as of this tasting. The body is a bit lighter than the clean portion, which is possibly helped out by the acidity. There is a mild acidic warmth in the finish with the alcohol.

Overall - The tartness is not drastically too much but it is a bit more assertive than I wanted. The blend is drier than the clean portion, which makes sense from the added attenuation of the Brettanomyces, and has a bit more bitterness and roughness in the finish. Overall the beer is smooth for it's strength which I'm happy with. I like the added complexity that the brett adds to the blend (like the fruityness and the mild leathery character) but I think next time I will decrease the brett portion a bit. A smoother blending process may also lower the rougher brett blend qualities as the brett aged portion on it's own was not really acidic (or acetic) before blending. If I can isolate the brett complexities without the acidity (or with only a hint) I think it will add quite a bit to the old ale.

100% Brett Aged (still the same 'clean' primary fermentation):

Aroma - The brett comes through strongly with earthy and musty/funky (not in a barnyard Brett. b. way) character along with a mild fruityness and mild acidity. The aroma of this treatment is not as acetic as the blend, likely reflecting oxygen exposure that the blend got during blending. A light alcohol similar to the clean portion comes through but it is mild and unobtrusive. The rich English malt and caramel from the clean portion come through well along with more of a toasty character.

Appearance - Very low coarse white head (though this is derived more from the pour than from carbonation). There is more haze than either of the other two portions but that may be due to the shorter time in the bottle. The color is amber without the pinker highlights of the other two treatments, but that may be influenced by the haze.

Taste - The taste is surprisingly tart, much more so than the aroma suggested. More earthy English malt character (like the clean portion rather than the blend) with dried dark fruit (prune and date) and some malty sweetness. The tart brett character dominates and the tartness enhances the perception of sweetness. The tartness fades away to a nice dry earthy/woody finish.

Mouthfeel - Medium body with very low practically non-existent carbonation. There is a moderate tartness with an acetic edge that is a bit rough but not terribly so.

Overall - The aroma is really nice but the acidity is too much. The flavor is nice other than the acidity and I think what I learned from brewing, blending and tasting these three treatments will help out next time. There are definitely ways that the brett added to the beer. Now to isolate those positive contributions while avoiding the not so good ones...

The only change I think I'll make for the next iteration of the base recipe is adding a bit of darker British crystal malt (135-165L) and increasing the level of chocolate a bit to darken the beer. Other than that I am happy with the base recipe. As mentioned above I would probably drop from 1/3 brett aged in the blend to something like 1/4 and I'll take care to avoid oxygen pickup while blending with what I learned from this first blending experience.