Halfpenny Brewery, Lechlade

Discussion

Important Texts

As stated above, these recipes are primarily modeled on evidence in Judith Bennett's recent book.

Also important in the crafting of this ale were several Elizabethan-era references. William Harrison, in 1577, provided a detailed description of how his wife brews beer [Misc-4]. Gervase Markham, in 1615, published The English Housewife, which contains not only an entire chapter devoted to brewing and the keeping of wine, but also an entire chapter on how to malt grain [Markham]. Kenelme Digbie, in 1669, wrote The Closet ... Opened..., a book of recipes, many of which are for various alcoholic beverages [Digbie].

Other useful modern texts include Cindy Renfrow's A Sip Through Time, a collection of recipes and other useful information [Renfrow]. Also worth reading is Beer, A History of Suds... by Gregg Smith. Though most of this book is devoted to the history of Beer in America, the first few chapters have some valuable information for the medievalist.

About Medieval English Ale

In England in the middle ages, particularly before the Plague (which first reached England in 1348), the most common drink of the day was ale. Ale, during this time, was a drink made from malted grains, water, and fermented with yeast. Malted grain would be crushed; boiling (or at least very hot) water would be added and the mixture allowed to work; finally the liquid was drained off, cooled and fermented. The ale might have been spiced, but it would not have had hops as an ingredient.

Beer, on the other hand, was made from malted grains, water, hops, and fermented with yeast. Hops added a measure of bitterness to the beer, and also helped preserve it. We will see below that the successful addition of hops required a change in the process that had a profound effect on the resulting product: after the liquid was drained off, it was boiled again with the hops.

Bennett has found a record of a Richard Somer who was selling Flemish ale (i.e. beer) in Norwich in the late 13th century. This was, however, a special case; the next record of the selling of beer that Bennett found was not until some hundred years later, with merchants along the eastern and southern coasts of England starting to import and sell beer in the 1370s [Bennett, p. 79]. It should be emphasized, however, that this was still somewhat of a special case -- beer was not a wide-spread drink until the late 16th century [Bennett, p. 81; Smith, pp. 25--26]. And even then, ale was a widely popular drink. Again quoting Bennett:

... As John Grove put it in 1630:

WINE: I, generous wine, am for the Court
BEER: The City calls for Beer
ALE: But ale, bonny ale, like a lord of the soil in the Country shall domineer.

[Bennett, p. 81]

Fresh Ale

In medieval England ale was served fresh, still (or very recently) fermenting, as opposed to stale, or done fermenting and cleared.

Since ale was basic to the diet of ordinary people, each household required a large and steady supply; a household of five people might require about 1 1/4 gallons a day, or about 8 3/4 gallons a week. Yet ale was both time consuming to produce and fast to sour, lasting for only a few days. 17
[Bennett, p. 19]

17. At Elmley Castle (Worcestershire) in 1446, for example, alebrewers were proscribed from selling ale more than four days old. Warren O. Ault, Open-field husbandry and the village community, Transactions of the American Philosophical Society, n.s., 55, part 7 (1965), item 147, pp. 77-78
[Bennett, p. 190 (text of endnote 17)]

From this evidence, we can conclude that any production method that allows the ale to keep for a long time is probably not what we are after. Likewise, any ingredient or production method that forces us to age the ale for a long time to become drinkable is not right either. Earlier in the medieval period, the ale-brewers were making smaller batches in their own houses, rather than brewing in industrial quantities. They simply did not have either the space or the resources to age the ale for long periods of time before selling it.

Expensive Ale

Ale was somewhat expensive to both produce and buy. In 1310, an unskilled laborer was paid at most 1 1/2 d. (pence) per day of work [Bennet p. 24]. We will see below (under "grains") that in the Oxford market in 1310, a bushel of wheat cost between 10 and 12 pence per bushel, or 6 2/3 to 8 days of unskilled work, with barley going for 7.5 d/bu, or 5 days, and oats going for 4 d/bu, or 2 2/3 days. A bushel of grain would yield between 7 1/2 and 10 gallons of ale [Bennet p. 23], which would be sold for 1 1/4 to 1 1/2 d. per gallon.

Now aside from the economics of profitable sale, discussed below, it is worth pointing out that both the grain, and consequently the ale, was expensive for the laborers. In short, a gallon of ale cost roughly a day's work. The combination of modern large-scale farming and production techniques, and increased wages, make grain in the 20th century United States roughly 1/10th as expensive as it is was in 13th C. England.

Medieval English Ale and modern Homebrewing

Another aspect of this project is that it is an all-grain brewing. Many brewers within the SCA make beer in about the same way any other modern hobbyist home-brewer does: they buy commercial malt extract, which is a concentrated or dried malt-sugar solution. This is combined with water, and the result is boiled. Hops, dried hop flowers, may be added, or the malt extract may have been hopped in the factory that produced it. This then is cooled, yeast is added, and it is fermented.

All-grain brewing is somewhat more ambitious, since the sugars must be extracted from the malt by the homebrewer. It is, however, more controllable. The flavor of the beer can be changed substantially by the choice of type, quantity, and roasts of malted barley, as well as other grains. The flavor and "body" can also be effected by the method of mashing -- process by which the starch in the grain is converted to sugars -- as well as the mash temperatures. I personally find all-grain brewing to be preferable to extract brewing -- I like the added control that mashing affords, and it is much closer to how the process is done on a commercial level today, as well as historically. It is also cheaper; the grain for an all-grain batch costs about half of what the extract would to make an equivalently sized batch. This doesn't come for free, however. All-grain brewing requires some very large and fairly expensive equipment.

For the SCA brewer, all-grain brewing has another advantage: one can use the techniques used by brewers in the few late-period recipes that we have. These techniques are somewhat different than those used today; for a full description please see my thoughts on Elizabethan Homebrewing.

For this project, however, there was no choice. One cannot buy malt extract that is anything like the wort produced by this brewing; these batches are sufficiently unlike anything that can be purchased, so it cannot be "simulated".

One interesting irony of this project is that it matches the observation by Bennett, that beer brewing requires more equipment than ale brewing [Bennett, pp. 86--87]. In creating 2 1/2 gallons of ale, the only specialized piece of equipment I used was a large cooler, which I used as a lauter tun. In retrospect, I could have easily made a 5 gallon batch of this ale with the same equipment, save for a larger fermenter. I did not need the large pot, nor the burner to put this pot on, that I use for making 5 gallon batches of beer.

Materials

Grains

In Elizabethan times, as today, the most common grain in ale and beer was malted barley. Wheat was also fairly common, often in smaller quantities, as are oats. In medieval England, however, the preferences were somewhat different. Bennett reports several extremely useful clues on this. Looking again at the Clare information:

Our most direct evidence of domestic brewing comes from elite households. In 1333--34, the household of Elizabeth de Burgh, Lady of Clare, brewed about 8 quarters of barley and dredge each week, each quarter yielding about 60 gallons of ale. Brewing varied by the season of the year, with vast amounts produced in December (when more than 3,500 gallons were brewed) and quite restricted production in February (only 810 gallons). The members of the Clare household drank strong ale throughout the year, imbibing with particular gusto during the celebrations of Christmas and the New Year.
[Bennett, p. 18]

Dredge is a combination of oats and barley [Bennett, p. 17]. A quarter is a unit of dry measure of approximately 290 liters [Bennett, p. xv]; there are 8 bushels to the quarter, making a quarter approx. 36 1/4 liters. (A modern U.S. bushel is 35.24 liters, and a modern U.K. bushel is 36.37 liters [AmHer].)

So the Clare household ale was made using between 2/3 and 3/4 malted barley, with the remainder being oats.

Consider, for example, the possible profits of brewers in Oxford during the early years of the fourteenth century. In the late autumn of 1310, a jury gathered to review grain prices and to set the ale prices accordingly. Stating that wheat had recently sold in the Oxford market for 8s., 7s.4d., or 6s.8d. a quarter (depending on quality); barley for 5s.; and oats for 2s.8d., it set prices for good ale at 1 1/4 d. in cuva and 1 1/2 d. in doleo. The first price was the more standard of the two. What profits could a brewer have expected from buying grain at these prices and then selling ale made from it at 1 1/4d. the gallon? For a brewing of 3 bushels of malt, the range of possible costs (shown in the four left-hand columns of figure 2.1) and receipts (shown in the four right-hand columns) was very broad.

The first two estimates of cost assume that an Oxford brewer would have brewed malt made from equal amounts of wheat, oats, and barley. ... The next two estimates of cost assume a different ratio of grains in the malt, one similar to that employed at St. Paul's: two-thirds oats and one-sixth each for barley and wheat. ...
[Bennett, pp. 22-23]

(The figure on [Bennett, p. 23] shows production costs ranging from roughly 20 to roughly 30 pence (d.), and income ranging from 24 1/2 d. at a concentration rate of 6 1/2 gallons per bushel to 45 d. at 21 gallons per bushel.

Malting is a process which both preserves a grain for longer-term storage than is possible in the raw form, and activates enzymes in the grain that help turn starches into sugars. After the grain is harvested, it is moistened and allowed to sprout and grow for a few days, then dried in an oven at low to moderate temperatures, and finally threshed and stored.

Traditionally the sprouting would be carried out on a large floor [Markham, pp. 182-185], often in the attic of the malt house. As the grain sprouts, it generates heat, which must be allowed to escape so that the malting grain does not cook itself. The young plants also require carbon dioxide to continue growing. To facilitate this, the malt is turned (scoop it up, flip it over) at regular intervals. A few British maltsters are still producing floor-malt today, though it is rare and expensive.

Historically, kilning was often carried out in what is essentially a large wood oven or smoker. The malt would be spread out on a false-floor made of hair-cloth, straw mat, or other suitable material, on top of some type of loose material, in order to allow the hot exhaust from the kiln's oven to evenly penetrate the grain. Then a wood fire would be built in the oven, and the malt baked for several hours, and occasionally turned to prevent burning [Markham, pp. 186-190].

Modern kilning is quite different. Today the malt is roasted in a drum with a water spray to control temperature, patented by D. Wheeler in 1817 [Harrison]. Because of this spray, malt can be roasted much, much darker than was possible in an oven. The water spray also allows lighter malts to be made much more consistently, and without picking up any flavor from the fuel used to fire the kiln.

Often, to better approximate amber and brown malts, I have roasted some portion of Pale malt in an oven. For the first batch of weak ale, which was an initial test of a recipe and several techniques, I did not go to the extent of roasting the malt, but used only straight Pale malt. Baird brand malt is roasted slightly more than many other English Pale malts; it is practically a Mild, rather than Pale roast. Mild is generally roasted just slightly more by the maltster.

For second batch, however, I oven-roasted a small portion of the malt. Roasting some of the malt better simulates the oven-kilned nature of medieval malts - particularly the unevenness of the malt kiln. Since this also adds a nice flavor to the resulting ale, it seemed like the right choice for the second batch.

I wanted to roast the malt some, but not too much. So I selected a portion of the malt and baked it in the oven, first for half an hour at 225 degrees F., then an additional half hour at 300 degrees F. This gave a nice amber roast to some of the grain. (It is important not to do this to to all of the malt, since the roasting procedure can destroy the enzymes that are used in the mash to convert unfermentable starches into fermentable sugars - see below in the techniques section.)

The information about the brothers at St. Paul's, brewing in 1340--41 [Bennett, p. 191; end-note text 21], says that oat malt was the predominate ingredient. Unfortunately, malted oats are currently unavailable commercially. So I was forced to compromise by using the Clare household mix of 3/4 barley, 1/4 oats, and further compromise by using unmalted oats. As a result, the flavor won't be quite right, but it will be closer than by completely ignoring this evidence and using entirely barley malt.

Yeast

Though there have been some reports of successfully culturing yeast from bottles of beer found in ship-wrecks, none of these cultures have become commercially available to the homebrewer. Failing authentic yeast, I opted to combine two strains of modern yeast of the types used to make modern English ales.

Both Digbie and Markham recommend making a yeast starter in order to have a sufficient quantity of yeast to attack the large size batches they are making. Digbie advises [Digbie, pp. 99]:

... This quantity (of a hogshead) will require better then a quart of the best Ale- barm, which you must put to it thus. Put it to about three quarts of wort, and stir it, to make it work well. When the barm has risen quick scum it off and put to the rest of the wort by degrees. The remaining Liquor (that is the three quarts) will have drawn into it all the heavy dregs of the barm, and you may put it to the Ale of the second running, but not to this. Put the barm you have scummed off (which will be at least a quart) to about two gallons of the wort, and stir it to make that rise and work. Then put two Gallons more to it. Doing thus several times, till all be mingled, which will require a whole day to do. Cover it close, and let it work, till it be at it's height, and begin to fall, which may require ten or twelve hours, or more. Watch this well, least it sink too much, for then it will be dead. Then scum off the thickest part of the barm, and run your Ale into the hogshead, ...

Markham recommends something similar, though not as complex a technique. He says to combine some of your wort (presumably cool enough) with some barm (yeast), and let these work while the main batch is cooling. Then when the main batch is cool, stir up this starter well and mix it in. The cooling time can be quite substantial, much more than an hour or two.

Using a starter is good practice in modern, as well as medieval brewing. Starting with a large quantity of yeast will reduce the effects of wild yeasts and other microorganisms by overwhelming them by sheer number, and by eating up all the available sugar. This will help keep the ale from catching unwanted infections, since there will be less for undesirable microbes to feed on.

Yeast Culturing, and lack thereof

Yeast was cultured in the medieval period, but not to the extent that we would think of today. Back then many people were brewing, and yeast was freely shared amongst brewers. Since yeast that produced good ale and beer was valued and shared, good yeast was cultivated.

Now, however, not only is good yeast treasured, it is also cultured in laboratory conditions, usually starting from a single cell culture. Since yeast reproduces by budding, most cultured yeasts today are all alike to the point of being clones of each other [Ligas].

Since these were small batches, I simply rehydrated yeast from dried packets. For the first batch, in order to slightly simulate the multiple strains of yeast that would make up medieval ale barm, I used two different strains of yeast. For the fourth batch, I've gone back to this mixture, since it gives by far the best results so far.

Experiments in yeast mixing

After presenting the first batch at a local event, and posting the recipe for the first batch to the historic home-brewers' (internet) mailing list [hist-brew], I received two interesting comments regarding the yeast mixture. Marc Bloom (AKA "Red") commented that if I was interested in yeast diversity, in addition to ale yeast I might consider adding bread yeast. He also said that he remembered reading that one (modern) English brewery uses a very unusual yeast strain, available to consumers as Fleischmann's brand bread yeast [Bloom]. So I added this to the mix for the second batch.

In addition, Al Korzonas wrote:

Finally, I'm willing to bet that all beers made more than 200 years ago had a significant amount of bacterial and Brettanomyces sourness. For authenticity, I'd pitch the dregs from a bottle of unfiltered Lambic (like Cantillon, Boon Marriage Parfait or Lindeman's Cuvee Rene) for the Brettanomyces and lactobacillus that was most certainly a part of any beer made more than 200 years ago. Pitch it early in the ferment for more character, late for less. You can expect to have a rather long, lingering fermentation when you add Brettanomyces because it is pretty slow and because it will eat many carbohydrates that Saccharomyces simply won't eat. If [you] plan to bottle, a year is not too long to wait for the Brett and lactos to finish their work.
[Korzonas - letter]

Following this advice, I also added the dregs from a bottle of Cantillon Gueuze to the yeast mix for the second batch (even though these ales are meant to be drunk much earlier than a year old.

As noted above, the second batch developed a strong nasty taste that was very much like one of the minor tastes in good Lambic. So I left the Lambic dregs out of the third batch. The third batch did, however, have the Fleischmann's bread yeast as part of the mixture. This third batch quickly developed a very clean, very bad smell and taste that was reminiscent of paint thinner.

So it seems that it was not necessarily the Lambic that gave the off tastes; instead that Fleischmann's bread yeast is unsuitable for brewing ale. As I have yet to try adding the Lambic dregs to a mixture of otherwise known-good yeasts, I'm reserving judgement on this move until I try the next logical experiment.

Water

Many breweries in England generally had fairly hard water, notably including Burton-on-Trent [Noonan Scotch Ale]. Fortunately, so does Pittsburgh, so I did not feel that I needed to alter the mineral content of the local water for these brews.

Oak

In the 16th C., beers and ales were usually tunned in casks, often made of oak. Also, in some of the referenced later period techniques, notably Digbie's Scotch Ale recipe [Digbie; Renfrow p. 11], some of the processing of the liquids were done in oak casks.

On the other hand, according to Bennett, there is evidence of brewing in vessels other than oak:

When Denise Marlere died in February 1401, she left behind a thriving brewing business in the town of Bridgewater. She bequeathed the bulk of her business to her servant Rose: half of a tenement, all of her brewing vessels with a furnace, three sacks full of malt, a cup, a brass pot, a pan, a goblet bound with silver, a chafing dish, two silver spoons, and some other carefully specified goods. She also left brewing utensils to other heirs, giving a leaden vat each to her parish church, her parish priest, and two local monasteries, and leaving to her daughter, Isabel, two more leaden vats, a brass 3 gallon pot, a pan, a mortar and pestle, and the proceeds of one brewing.
[Bennett, p. 14]

Clearly Denise Marlere was brewing in lead vessels. I'm not interested in low-level lead poisoning, so I avoided this type of vessel. Still, oak is not the only alternative, and one could reasonably brew a recreation of a medieval ale in, say, an open stainless steel vat with a clear conscience.

For the first batch, I chose oak. To simulate some of the oak flavor that may have been picked up by fermenting in an oak vessel, I chose to boil a small amount of oak chips (available at better homebrewing shops) in a bit of water, and then added some of this water to the wort. Some care should be taken not to over do this, as the oak can have a surprisingly aggressive taste.

This addition of oak was criticized by Al Korzonas, who noted a very important point about oak in brewing: American Oak, which is what is often sold in brewing stores, has much more flavor than European Oak. He notes that several old English brewing books recommend against using American oak casks because of the flavor they will impart [Korzonas - oak]. I would recommend anyone who is considering using oak for a "more authentic taste" to read the article Mr. Korzonas wrote on the subject first.

Following this advice, I removed oak from the ingredient list for the second batch.

Quantities

I decided to use proportions of 3/4 malted barley and 1/4 unmalted oats (as a compromise to what is available), hard water, and ale yeast. But in what quantities? Again reviewing the Bennett quote from pp. 18:

Our most direct evidence of domestic brewing comes from elite households. In 1333--34, the household of Elizabeth de Burgh, Lady of Clare, brewed about 8 quarters of barley and dredge each week, each quarter yielding about 60 gallons of ale. Brewing varied by the season of the year, with vast amounts produced in December (when more than 3,500 gallons were brewed) and quite restricted production in February (only 810 gallons). The members of the Clare household drank strong ale throughout the year, imbibing with particular gusto during the celebrations of Christmas and the New Year.
[Bennett, p. 18]

The first thing I did was to reduce these units to something I could better deal with. I started by measuring my malt. Rather than depend on someone else's values, I carefully measured out 7 U.S. pints of Hugh Baird Pale malt, and weighed out this quantity at 4 lbs 6 oz. This works out to 1.32 lbs per liter of volume of grain.

The American Heritage dictionary lists 1 U.S. (modern) bushel at 2150.42 cubic inches, or 35.24 liters, and a U.K. standard bushel at 2219.36 cubic inches, or 36.37 liters [AmHer, "bushel"]. A bushel is 1/8th of a quarter; Bennett lists a quarter as being approximately 290 liters [Bennett, p. xv]. She also notes that a gallon in the medieval period in England was approximately the same size as a modern U.S. gallon, rather than the modern U.K. gallon (which is 25% larger).

So the Clare household recipe proportions were some 290 liters (dry measure) of grain, some 383 lbs., producing 60 gallons of ale. This works out to about 6.4 lbs. of grain per gallon of ale.

Not only are these large quantities of total grain used, they are surprisingly high quantities of grain per unit of water -- modern beer recipes generally run between 1 and 1 1/2 lbs. of grain per gallon. Clearly the medieval process was either quite inefficient, or the resulting wort was quite thick. After working on this project, I have found that it seems to be some of each.

While a thick wort will produce a high alcohol level, which helps preserve the ale, it also takes longer to ferment out. For this ale, I wanted to brew something that would finish somewhat more quickly. Bennett has some evidence of a weaker ale being produced around the same time.

... In 1282, when Robert Sibille the younger was presented at the court of Kibworth Harcourt for selling his ale at too high a price, the stipulated price left him little room for profit. Having paid 2s. for 4 bushels of malt and required to sell 5 gallons of ale for 2d., he would have had to draw 60 gallons from his malt just to recoup his investment. His ale, in other words, would have been very weak indeed and his profits very low.
[Bennett, p. 21]

4 bushels of my malt weighs 191.4 lbs.; this works out to a rate of 3.2 lbs. per gallon. While still quite strong by modern standards, this is half the rate of the Clare household recipe.

It is worth pointing out at this point that, due to the processes used, the Clare household ale would probably not be twice as strong as Sibille's ale. It would undoubtedly be quite a bit stronger, though. The irony is that Sibille was probably getting a better efficiency, in terms of sugar extracted out of his grain, than the Clare household was, at least for the initial running of the grain.

So for the first batch, I opted for something under the maximum concentration that Sibille could have brewed without losing money (according to Bennett). I used 3.2 lbs. of grain to draw approximately 2 1/2 gallons of ale. For the second, I tried to match the Clare household proportions, intending to use 12.3 lbs. of grain to draw approximately 2 gallons of ale, but ended up getting less liquid out than I intended in the first running, so this ended up somewhat stronger than the original.

I will also note here a nice rule-of-thumb. (I like rules of thumb like this; they are easy to remember and simple to describe to others.) Looking at the original Clare proportions, and taking into consideration how much water would be absorbed by the grain (but not released in the running), the rule-of-thumb is: 7 measures of water, added to 5 measures of grain (by the same volume measure), yield about 4 measures of ale, more or less. And the combined water and grain mix in the mash will occupy at least 8 measures when mashing (possibly 9 measures).