Wednesday, August 11, 2010

Recipe Formulation with Joe Walts

Joe Walts Brewing in Oshkosh
If you’re the homebrewing type who likes to put together your own recipes, here’s something for you. This is Joe Walts’ magnum opus on recipe formulation. A close read of what follows will definitely improve your ability to build your own beer.  Now for the bad news: Joe Walts is leaving us. After almost a year of making beer for Fox River Brewing in Oshkosh and Appleton, he’s moving on to work for Ale Asylum in Madison. We’re going to miss him in Oshkosh. Joe, you’ve been a great help to me and I appreciate all you’ve added to this blog. Thanks.

Recipe formulation is one of my favorite parts of the brewing process, and I was pretty excited when Lee asked me to write about it. I was tempted to geek out and highlight the fundamental calculations required to create beer recipes, but it would be a very long post and it wouldn't tell you anything new - provided you're willing to check out a presentation that I gave to the Madison Homebrewers and Tasters Guild last November. The file is named 'MHTG_Recipes.pdf' and you can download it at here.

When I make a beer recipe, I enter my recipe parameters into a spreadsheet. If you visit this RePublic link, you can download two different spreadsheets that will work for homebrew recipes. The one named Recipe_School.xls corresponds with the calculations in the MHTG presentation. The other, named Recipe_Gallons.xls is more detailed and acknowledges the complex interactions of ingredients (e.g. water chemistry, iso-alpha acid saturation and the affects of yeast starters on wort properties), but it's not necessary to make good beer or to learn about recipe formulation.

The first step of recipe formulation is to decide what type of beer to make. Until you have an intuitive feel for how ingredient quantities will affect the final outcome, I think the best place to start is with the BJCP guidelines. It should be mentioned that style guidelines are great at maximizing the objectivity of beer competitions, but they're not always accurate at describing commercial beers, especially those in other countries. That said, they're great for outlining the base specifications of beers within general stylistic ranges. For example, let's say that we want to brew a Witbier. According to the guidelines the beer should have an original gravity between 1.044 and 1.052, its bitterness should be between 10 and 20 IBUs, its color should be between 2 and 4 SRM, its grainbill should consist of 50% pale base malt and 50% unmalted wheat, and that Witbiers usually contain coriander and orange peel.

Here's where the fun starts. From drinking commercial Witbiers, I know they're usually pretty sweet. For a summer thirst quencher, I want something a little drier than a typical Wit. How do we address that? Disregard whatever you know about how Witbiers are mashed and perform a single infusion mash at 149 degf. Another method would be to replace the unmalted wheat with malted wheat, since malted wheat will contribute more amylase enzymes to the mash. It would be a shame to lose the unique flavor that unmalted wheat provides, though, so we'll stick with the BJCP grainbill. What about orange peel and coriander? They're nice spices and all, but they're totally played. Let's make something that could pair well with Mexican food. How does replacing the orange peel and coriander with lime peel and cilantro sound? Whether it sounds good or bad, you won't know the truth until you try it. Since Mexican food is usually somewhat spicy, and hops pair well with spice, we should hop the beer at the higher end of the bitterness range. 20 IBUs sounds good. Replacing the standard European hops with something like Amarillo should add a layer of citrus that works well with the lime and cilantro. As far as the processes go, though, we'll still be brewing something pretty similar to a BJCP Witbier.

In both of my recipe spreadsheets, white cells are your chosen variables and blue cells are the calculated results. From this point forward, I'll assume we'll build the recipe in the Recipe_School spreadsheet. Here's the information we'll need to know:
  • Target serving volume
  • Target original gravity
  • Target hop bitterness
  • Target boil time
  • Yeast strain and expected apparent attenuation
  • Maximum brewhouse efficiency (the percentage of available mash sugars that your brewing system can extract for a low-gravity beer)
  • Typical boil evaporation rate of your brewery
  • Typical losses in your brewery due to mash tun, kettle and fermenter geometries
  • Color, typical yield and extract contribution of each fermentable ingredient
  • Target mash temperature
  • Target water-to-grain ratio
  • Dry grain temperature prior to mashing
  • Hop varieties, alpha acid levels, form (pellet or whole), time added relative to the boil and usage rates of flavor/aroma hops (in lbs/barrel)
  • Other ingredient usages
  • Yeast type (dry or liquid) and beer type (ale, lager or hybrid)
Let's assume a 5-gallon batch and choose a mid-range original gravity of 1.048. We already know that we'll be hopping the beer to 20 IBUs. Boiling for an hour is usually a good compromise between protein coagulation, alpha acid isomerization and DME volatilization vs. not evaporating too much water (homebrew systems have much higher surface-to-volume ratios than commercial systems and therefore evaporate much higher percentages of kettle water). The obvious yeast choice for a Witbier is Wyeast Belgian Witbier or Wyeast Belgian Wit Ale. According to Kristen England, both products are the Hoegaarden strain. Since we'll be mashing at a low temperature and we all know to oxygenate the hell out of our worts, we can assume the high end of the manufacturers' attenuation ranges. Wyeast claims 76% and White Labs claims 78%; we'll use 76% because it sucks to end up with less ABV than you plan for.

The grain-handling system I use at home - an adjustable hand-cranked mill and a 10-gallon Gatorade cooler with a false bottom - will give me brewhouse efficiencies around 92% for low-gravity beers. Most people aren't as geeky about water chemistry as I am (that's a good thing), though, so let's assume a maximum brewhouse efficiency of 85%. I'm not sure how much wort I lose to my mash tun geometry (I'm talking about wort below the pickup tube, not wort retained by spent grains), so I just call it a gallon to be super safe. My 9-gallon kettle and propane burner typically evaporates 1.2 gallons per hour and I typically leave about 0.65 gallons of wort behind in my kettle. My spigot is crappily high, though, so I bet you usually lose less wort than me. Let's assume a kettle loss of 0.5 gallons. My fermenter, an unusually-modified (not by me) 10-gallon cornie keg, loses another 0.65 gallons. Assuming you're using a glass carboy, let's lower that number to 0.3 gallons.

For a Belgian ale, using Belgian pilsner malt makes sense. Its color (according to a manufacturer) is 1.6L and, as the BJCP says, it will account for 50% of the total wort extract. Because the starches of flaked wheat are already gelatinized, it makes sense to use flaked wheat instead of raw wheat. Briess flaked wheat has a color of 2.0L and will account for the other 50% of the total wort extract. The presentation I mentioned earlier lists some typical ingredient yields. From that chart, we can assume a yield of 76% for Belgian pilsner malt and 64% for flaked wheat. Since most barley starches gelatinize at 149 degf, that's the lowest temperature you should mash at for single-infusion mashes. Otherwise, a lot of your enzymatic conversion will happen during the sparge. It'll work that way, but it'll be less efficient and completely unpredictable in terms of fermentability. If you want high fermentability, which we do for this beer, mash at 149-150 degf. If you want low fermentability, i.e. a lot of dextrins in the wort, shoot for mash temperatures of 154+ degf. Let's assume we'll be mashing this beer at 149 degf.

Typical water-to-grain ratios are around 1.2 qt/lb to 1.5 qt/lb. I usually mash at 1.5 qt/lb because thinner mashes increase fermentability, but mashing thicker is useful for high-gravity beers because your first runnings are stronger. Mashing thicker than that (numbers lower than 1.2 qt/lb) results in efficiency losses because there's not enough water to dissolve all of the grain sugars at once, which raises the gravity of the last runnings that are lost to the mash tun and spent grains. Thinner than 1.5 qt/lb can also result in efficiency losses because the low density of the first runnings causes the sparge water to mix with it to a greater degree (therefore also increasing the gravity of the last runnings). Since we're not making a high gravity beer, and we want the wort to be as fermentable as possible, let's mash at 1.5 qt/lb. Let's also assume your dry grain is 72 degf.

My general bittering strategy is to use the same varieties as my flavor/aroma hops when my target bitterness is below 20-30 IBUs. When my target bitterness is higher than that, I like to use a clean high-alpha bittering hop - such as Magnum - to keep the vegetative matter low. For massively-hopped beers such as double IPAs, where late hop additions contribute significant amounts of the total IBUs, using high-alpha hops for both bittering and the bulk of flavor/aroma additions becomes essential. Thankfully, we've chosen a simple recipe today. Let's use Amarillo for everything. Here are some rules-of-thumb that I like to use for flavor/aroma hopping rates (these are certainly not gospel):
  • 0 to 0.2 lbs/barrel for malty lagers and beers where yeast flavors dominate0 to 0.8 lbs/barrel for moderately-bitter beers
  • 1+ lbs/barrel for hop-focused beers like APAs and IPAs
Adding hops 20 minutes before the end of the boil is more typical for continental beers, though, so we'll stick with the tried-and-true and follow suit. According to HopUnion, the alpha acid content of Amarillo is typically around 8-11%. Until we actually buy the hops and know what we're getting, let's assume the middle value of 9.5%. Let's also assume we'll be using pellets because they stay fresh longer than whole hops and they're often easier to find.

According to some brewers on the forums, typical usage rates for orange peel are 0.35-0.75 oz per 5 gallons. With spices, it's better to be too low than too high. Let's use 1/2 oz of lime zest for our beer. I don't know anything about using cilantro in beer, so we'll just need to try it out and see how it works. Let's start with half the weight of lime zest and use 1/4 oz. I doubt there's such a thing as too little cilantro.

Finally, the limited variety of dry yeasts will force us to use a liquid yeast strain. Witbiers are solidly in the "ale" category with respect to fermentation characteristics. Here's what the spreadsheet spat out:
  • Belgian Pilsner malt = 4 lb 9 oz
  • Flaked what = 5 lb 7 oz
  • Total water = 8.7 gal
  • Mash water = 3.7 gal
  • Mash water strike temperature = 159 degf
  • Target pre-boil volume = 7 gal
  • Target pre-boil gravity = 1.039
  • Target post-boil volume = 5.8 gal
  • Expected final gravity = 1.012
  • Expected alcohol content = 4.7% abv
  • Amarillo hops for bittering (beginning of boil) = 0.5 oz
  • Amarillo hops for flavor/aroma (20 minutes from boil end) = 0.3 oz
  • Required yeast cells = 182 billion -> yeast starter volume = 1138 mL
My presentation lists a number of additional resources for creating recipes. Good luck and have fun!

1 comment:

  1. Thank you Joe. Best of luck and send some Asylum beers north!