Brewing Historical Porter with Brett

For many years I have been on a quest to find out all I can about porter. This has meant hours and hours of study in places like the National Brewing Library in England, the British Library in London, and the Beinecke Rare Book Library at Yale. All of which has yielded filing cabinets full of notes about the origin of the beer, who made it, what it was made from, where it was made, and so on. But it never answered that most important of questions — what did porter taste like in the 18th century?

Aside from the difficulties of reproducing raw materials from that time, there is the problem of knowing what happened to porter during storage. In those times porter was stored in wooden vats for months, or even years, before it was sold. So it might well have developed various flavors through the action of bacteria or “wild” yeasts present in the wood, or even in the beer itself before storage. It has been suggested by one British researcher that the high hop alpha-acid levels in porter would have inhibited the bacteria, such as those producing lactic acid, and the major effects of storage on flavor come from yeasts other than Saccharomyces species.

There is no direct evidence of this, but there is some good secondary evidence. In the early part of the 20th century a Dane, N.H. Claussen identified a yeast species as being responsible for the secondary fermentation in British Stock Ales. He called it Brettanomyces, and it was later named B. claussenii in his honor, as other variants of the species were later discovered, notably in Belgian “spontaneously-fermented” beers.

Stock Ales were high alcohol beers stored in wooden vats for long periods, during which they underwent a slow secondary fermentation, as the Brettanomyces yeast consumed polysaccharide molecules which Saccharomyces species could not ferment. It was considered that this was what gave British stock ales their characteristic taste, later found to be largely due to high levels of fatty acid ethyl esters, notably ethyl acetate and ethyl lactate, which were produced by this yeast. One writer has suggested that the relatively high levels of these esters may have given the beer a narcotic effect, making it appear stronger than its alcohol level would suggest. Claussen’s work proved there were at least two different yeasts at work in these beers, which lead many English brewers to refute Hansen’s suggestion of using single culture yeasts, a practice which soon became common in Continental Europe. Ironically, not much later British brewers phased out Stock Ales almost entirely, and turned to brewing only “running” beers, which were shipped out of the brewery and drunk before there was time for the slow Brettanomyces secondary fermentation to occur.

Therefore, with perhaps a little stretch it seems that Brettanomyces claussenii might well have made a significant contribution to the flavor of vatted 18th century porters. So it seemed to me that I needed to experiment with Brettanomyces in a porter. For the record, three varieties of Brettanomyces are available to us:

• B.claussenii (White Labs WLP645); this is reckoned to give a relatively mild and subtle Brett character, with the accent on aroma rather than flavor, the aroma being estery/fruity, reminiscent of pineapple according to some.
• B.bruxellensis (White Labs WLP 650, Wyeast 5112); classed as a stronger Brett character than the former, giving the notorious horse blanket flavor, and quoted as being used for secondary fermentation of Orval Trappist ale as well as in lambic beers.
• B.lambicus (White Labs WLP 653, Wyeast 5526); produces the most distinctive Brett flavors and aromas, sometimes described as “like cherry pie;” it is found in lambic beers and Belgian sour brown ales.

All of these, particularly the last two, also produce some acidity, and all of them take upwards of three months for full flavor development, although as might be expected the rate of secondary fermentation increases with increasing temperature.

The first point about these yeast strains is that, leaving the Belgians aside, most brewers consider them to be undesirable wild yeasts. Also, once a Brett strain becomes established in a brewery it is very difficult to eradicate, which is probably why such yeast are suitable for the Belgian “spontaneous fermentation” approach. Therefore, if you plan to be primarily a brewer of regular ales, stouts, and lagers and to only occasionally dabble with Brett-fermented beers, you MUST be scrupulous about cleanliness and sanitation (which you should be anyway). I have spoken to many British brewers who were aghast at the idea of using more than one Saccharomyces species in their brewery, let alone a Brett species. But many American craft brewers use more than one yeast variety and quite a number have used Brett yeasts without cross-contamination problems. If you use only stainless steel and glass your normal cleaning and sanitizing techniques should ensure you have no contamination problems.

For my experiment I used an 1850 recipe from Whitbread, one of the great London porter brewers. I didn’t use any of my 18th century recipes as you might have expected, because these call for the use of 100% brown malt. Modern brown malts are not made in the same way as those from that time, and cannot be mashed on their own. So for a brew of 5 gallons (19 L) my malt bill consisted of 80% pale malt, 5% black malt, and 15% brown malt, with OG 1.060 (14.7 °P). I opted to keep the bittering hops low, at an estimated 25 IBU, so as not to mask other tastes. I used Wyeast 1098 for the primary fermentation, which finished at SG 1.022 (5.5 °P) in two weeks time. I could have added the Brett in the primary, but I had already decided that I would split the batch into two and add the Brett to only one half. I had also decided I would use B.claussenii WLP645. My research had indicated that there was a difference of opinion as to whether or not it was necessary to make a starter with this yeast. I decided not to do so for two reasons, the first being that I would be pitching only half the brew with the Brett yeast, thus needing less than for a full batch. My second reason was that I was concerned about possible contamination of my brewery with the Brett, during the various operations needed to prepare a starter, and decided that it was not worth the risk at that point.

Since the beer was going to be in secondary for some months, it seemed to me that it would be best conducted in a stainless steel soda keg. That way the keg could be kept sealed and the beer brought into condition by the Brett fermentation, and there would be less chance of contamination with bacteria which might confuse the issue. At first I had thought that I might dedicate a keg to Brett fermentations, as I have some spare, having picked them up very cheaply. But these are 5-gallon kegs and I should only be putting 2.5 gallons in it, leaving a lot of head space with a consequent risk of oxidation masking other flavors. Instead I went for a 3-gallon keg (which is more expensive than any of the fives, and in shorter supply in my brewery!).

I racked half of the brew into the pre-cleaned and sanitized keg, and took it into the far corner of my basement, well away from the brewing area. I oxygenated the beer for about 30 seconds before adding the Brett culture, for two reasons. First, I wanted to give the culture a head start, as I was not going to prepare a starter, and second I didn’t want to oxygenate the beer after adding the culture, in order to simplify cleaning of the diffusion stone. Once the Brett culture was in I sealed the keg and I left it for six months. My basement is generally around 60–65 °F (15.6-18.3 °C) most of the year, so quite satisfactory for a slow secondary fermentation. The remainder of the beer I bottled, along with a little dextrose as priming.

So then came the moment of truth. I took two glasses, a bottle of the non-Brett beer, and the keg with the Brett-impregnated beer. To sample the latter I used a picnic tap, which even after cleaning I have since reserved solely for Brett beers. That may be overkill, as I could simply have replaced the tubing and cleaned and sanitized the tap and beer outfitting in the usual way. Then I filled the glasses each with one of the beers, and set about comparing them as closely as I could. The non-Brett beer was a sound robust porter, well-balanced, with a hint of the almost chewy effect brown malt typically gives, and background roasted notes from the black malt; overall it was a good beer, although a little thinner than I had expected. Then to the Brett-beer, which was not at first markedly different, with the above flavor notes still noticeable. It did have a more intense aroma, mainly fruity from what appeared to be various esters, but not what I should describe as “pineapple,” and I could detect no hint of “horse blanket.” It did taste a little fuller than the other beer, again probably due to fatty acid esters produced by the Brett. On the down side, there seemed to me to be a slight greasy aroma and flavor which I did not like; that might well be subjective, as I have handled quite a lot of pure ethyl lactate, and never really liked its aroma! There was also an apparent increase in acidity, making the beer quite tart, although not unbearably so; it was certainly less tart than a typical lambic beer, which I generally do not like.

However, I did feel that the Brett-fermented beer was not one that I would like to drink by the pint. So my next step was to blend the two beers, and I did this in the simplest way, at the point of serving. This was often done with porter in earlier days, but usually with an old and a young brew, and in my case I was using two old beers, one with and one without Brett treatment. I poured about one-tenth of a pint of the Brett beer into a pint glass and then added a half-pint of the other beer, and swirled them together a little before tasting. This was something of a revelation, with all the malt characteristics of the porter showing through as before, but now a little fuller on the palate, and just a little bite from the extra acidity adding an extra touch of interest.

Although my experiment did not perhaps result in a great revelation about 18th Century porter, I had never expected that it would do so. As I pointed out at the beginning, brewing authentic reconstructions of old beers is fraught with difficulties, even when recipes have been uncovered and deciphered. You have to try to solve the puzzle one step at a time, and this experiment is one more piece in place.

Issue: July-August 2010