Brewing Experimental Beers **************************************************************** By Jeremy Bergsman jeremybb@leland.stanford.edu http://www-leland.stanford.edu/~jeremybb/beerstuff/beerpage.html **************************************************************** An effective way to learn the effect of a certain ingredient or process on the beer it is used to produce is to make a beer with and a beer without this ingredient or process. When there are more than one alternative these can all be tested together, often reducing the work involved and yielding more informative results. Examples of things that one might want to test as a homebrewer are: The flavor and/or aroma of the various types of hops The flavor, color, extract, etc. of the various specialty malts The effect of boil time on hop utilization The effect of boil time on hop flavor and aroma The flavors and fermentation characteristics of the various yeast strains. The effect of various adjuncts The effect of different fermentation temperatures Here are some things to keep in mind when planning these types of experiments: You should always have a "control", that is, a beer that does not have or use the thing being tested. In some cases this means plus and minus the thing, for example with and without a specialty malt. In other cases, such as the effect of yeast strain, the alternative to a given strain would be another strain, preferably a very well known one, or, even, better, several batches with several different strains. A control gives you something to compare to and reveals when an experiment has gone wrong. Let us say that there happens to be a high level of mold in the air the day you do an experiment to test the effects of yeast strain. If you only made one batch and it came out tasting moldy, you might falsely conclude it was the yeast that caused that flavor, rather than the contaminating mold. If you have a control beer made with a known yeast (known not to produce moldy-tasting beer), you can check (to some degree) whether all beers that day would have been moldy, or whether in fact that yeast strain was bad. If you had made several beers and they all tasted moldy, you would also be safe in assuming that it was not something about the yeasts that caused the effect. The various versions and controls must be made as similarly as possible. Using the example above, if the control were made on a different day or in a different room, it might not turn out moldy and you might falsely conclude that the yeast being tested was at fault for the moldy test batch. This leads to one of the major problems of these experiments: making several batches of similar beer. Many of these experiments allow you to use one wort: for example the different yeast strain experiment discussed above. These are the easiest to do and to do well: one wort will be divided up into similar fermenters (fermenter geometry may have an effect on many beer characteristics) and pitched with different yeasts or moved to environments of various temperatures, etc. Even in an experiment where separate worts are required, for example in the testing of various specialty malts, it is probably best to make one large base wort and add different small worts to parts of this. See below for details on this idea. If worts must be boiled separately consider the use of hop oil extract and isomerized alpha acid extracts to make these characteristics identical among the batches. By choosing a small batch size one can take steps to minimize the amount of wort needed. One gallon is a practical minimum size. Even this size presents problems of loss and oxidation during liquid handling. A carbon dioxide-driven racking system is very useful for preventing oxidation in all, but especially in small batches. Also important to keep in mind is that one gallon will only produce 9 bottles which may not be sufficient for tasting. See below for hints on one gallon batches. So what if you have scaled back to one gallon batches and still need a lot of wort? If you have a friendly brewpub nearby you might be able to convince them to donate/sell some wort from one of their pale ales. This will be a drop in the bucket to them and they might be interested in the results. Otherwise malt extract is the way to go. As long as you can reproducibly dilute the malt extract this should produce a uniform wort. If buying in bulk, dry malt extract is easier to weigh accurately than the liquid form. One can also supplement the extract with samples from one mash or one steeping of crystal and/or dark malts. Why complicate things with a partial mash? Recipe design is more important than one might guess for these experiments. My initial experiments assumed that the background beer should be nearly as bland as possible, to most fully reveal the effect of the experimental variable. This has a few problems however. First, it can make the experiment hard to interpret if an inappropriate base beer is used. For example the fermentation characteristics of a yeast might be different on worts of different OG's or sugar profiles. Second, very bland beer exposes subtle flaws that detract from evaluating the effect of the experimental variable. This has happened to me. In a test of the flavors of different hops my base beer was so bland that small amounts of DMS in one batch and of diacetyl in another came to the front of the flavor and made it difficult to accurately asses the hop flavor. The difficulty of producing multiple batches increases the chance of these types of flaws appearing. Keep the recipe simple, but not really far from that of a real beer. By the same token, you may want to use more of an ingredient or exagerate a process (for example use a very high fermentation temperature) to make the difference easier to detect, but this may well prevent you from getting a handle on how the experimental variable would affect a real beer. Excessive hops for example add unusual flavors that are not detectable at normal hopping rates. Why make the recipe different from a real beer at all? Maybe you don't have to but consider the experiment. Do you want to know the effect of boil time on hop utilization? Multiple additions a single batch are going to make analyzing and drawing conclusions from an experiment nearly impossible. This experiment must be done with only one addition to each batch, at varying times. If you plan to do a very large experiment you will need assistants. Having led groups making beers like this I would make a few recommendations. First you must find people who will commit to come because too many people is as bad as too few so you want to invite the exact amount that you will need. You should have all the jobs worked out in advance. You should assign yourself only the task of administration/supervision and record keeping. This will turn out to be a full time job to which you will need to restrict yourself. If you start to do any of the beer production work mistakes will probably be made. You must be prepared to direct people rather than let them choose their own course or else things can quickly get very disorganized. Hints for one gallon fermentations: I have found that one gallon glass jugs for cider and the like are easy to find and accept the same #7 stopper as most carboys. If these are used there is the problem that headspace must be allowed in the primary fermentation and you are now down to a .7 gallon batch. I have overcome this problem for some versions of these experiments as follows: use whatever specialty ingredient is being tested at four times its desired final concentration and make .25 of a gallon of this. (For example if you wanted to know what the effect of one pound of a specialty malt was in a five gallon batch, you would make 1 lb/5 gallons = .2 pounds in .25 gallons.) In addition to all of the .25 gallon specialty malt worts, a large amount of base wort is made. The actual amount needed is .75 gallons per test plus 1.5-2 gallons. .5 gallons of this is added to each .25 gallon wort, and .75 is added alone to a one gallon carboy. The remainder is fermented separately in any size carboy. This remainder is used to top up the .75 gallon batches at transfer. This same idea could be used for a hop flavor/aroma experiment with hop teas or dry hops added to the various batches, but for yeast it will not work: what do you ferment the topping up beer with? Hint for mini mashes: keep your eyes open for things that can easily be converted into a mini lauter tun. I use a 32 oz shampoo bottle (HDPE). I cut off the bottom, perforated it, and shoved this back in through the new hole in the bottom until it met the top. Upside down this is a nice lauter tun for .5-.8 lbs of grain. The original top is even a valve that can be opened and closed. Another hint for mini mashes: steeping most grains is not the same as mashing them. Buy some "Koji" or amylase enzyme from your homebrewing supply store. Use this according to its directions to mash specialty malts by themselves.