In an effort to become a better brewer, I’ve resolved to brew more often. I figure if I can brew at least once a month, I should be able to nail down a great-tasting beer that can be reproduced fairly accurately. The plan will be to brew two stylistically different beers, alternating between the two for each brew session. That way, I won’t be stuck choosing between drinking IPA and IPA for the next three months. I started this project last month with a rye IPA. The first iteration of that is currently in the keg but that’s all I’m going to say about that for now because today’s post is devoted to the Purple Box.

The Purple Box is the tentative name for a beer inspired by the Girl Scout cookies known as Samoas (I reject the Caramel deLites alternative). It is the first time I’ve generated a recipe almost entirely from scratch. Normally, I grab one of several books and/or issues of Brew Your Own magazine, find a recipe that sounds interesting and grab ingredients that match as closely as possible. Designing a beer from the ground up was a fun way to conjure a recipe and a great way to force myself to really think about what I wanted in the finished product and how the ingredients and their proportions will work to those ends.

Before I could even begin compiling an ingredients list, I had to determine exactly what kind of beer I wanted. This was easier than I thought because all I want is a beer that tastes like a Samoa. To do that, I needed to deconstruct what makes a Samoa taste like a Samoa, and then incorporate those elements into the beer. For those poor souls out there who have never had one, a Samoa is a vanilla-flavored shortbread cookie covered in coconut and caramel with chocolate drizzled all over it. It is one of mankind’s greatest inventions.

Proof that God is a troop leader.

Proof that God is a troop leader.

Many people (myself included) will immediately think this beer should be a porter or stout. That’s understandable. The chocolate aromas that come naturally from roasted barley seems like the perfect fit. But with along with those chocolate notes comes a rather strong roasted coffee flavor. This is something I decidedly do not want in my beer. My co-brewer suggested we go with a scotch-style since those beers have a great big malty base that should fit the idea of cookie-beer rather nicely. They also tend to have a higher concentration of crystal malts that will lend a nice caramel flavor.

Grain proportions were derived by doing a few quick GOOG searches to get a sense for what other people do. For the most part these are just guesses but if I cared more about things like color, I could probably have been a bit more deliberate. Here is what I came up with.

Updated 03/26/2013 @ 16:07 PDT
I was quickly corrected that the original inspiration for grain proportions were determined by referencing the scotch ale section in Ray Daniels’ indispensable Designing Great Beers: The Ultimate Guide to Brewing Classic Beer Styles (Brewers Publications, 1998). My memory is terrible.

1% chocolate malt
5% caramel
5% biscuit malt
10% flaked oats
10% crystal malt
69% base malt

Once proportions are determined, you have to figure out how much grain it will take to reach your desired gravity. Before I explain how that’s done, I’d like to warn the more math-averse readers that we’re about to get real computational, real quick. If that kind of thing would cause your eyes to glaze over and perhaps start to wander over to something like The Superficial, then by all means, skip the next section and proceed directly to the dangers of caramel.


In order to determine the proper amount of grains to mash, you first have to know what your goal is. This is defined by your desired final gravity and the total volume of beer. I expected to get 5½ gal. at a gravity of 1.060° Plato. This gives me a total number of 5.5 x 60 = 330 gravity units (GU).

Since we know the proportions of grains in the grain bill, determining how many GU for each grain type is easy.

1% chocolate malt: 0.01 x 330 = 3.3 GU
5% caramel: 0.05 x 330 = 16.5 GU
5% biscuit malt: 0.05 x 330 = 16.5 GU
10% flaked oats: 0.1 x 330 = 33 GU
10% crystal malt: 0.1 x 330 = 33 GU
69% base malt: 0.69 x 330 = 227.7 GU

The wonderful thing about mashing grains is that they are rather predictable. For nearly every kind of common (and most uncommon) types of grain, test after test has been done to determine the maximum amount of fermentable sugars that are extractable per pound of grain. For example, it is well-known that 2-row pale malt can yield no more than 38 points per pound per gallon. That means if you soak 1 lb. of 2-row in 1 gal. of water, the resulting liquid will have a maximum gravity of 1.038. An example of a table listing the extract potential for a variety of grains can be found here.

Using a similar table I calculated the amount of grains needed to extract the appropriate number of GU.

chocolate malt: 3.3 GU ÷ 28 points/lb./gal. = 0.1 lb.
caramel: 16.5 GU ÷ 46 p/lb./gal. = 0.4 lb.
biscuit malt: 16.5 GU ÷ 24.5 p/lb./gal. = 0.7 lb.
flaked oats: 33 GU ÷ 22.4 p/lb./gal. = 1.5 lb.
crystal malt (80L): 33 GU ÷ 23.8 p/lb./gal. = 1.3 lb.
2-row pale malt: 227.7 GU ÷ 26.6 p/lb./gal. = 8.6 lb.

The more astute readers will notice that my maximum yield numbers don’t match any table you can find on the Interwebs. This is because my system is not perfect and is very rare that I will be able to achieve maximum extraction from any given mash. As such, all yields are scaled down to 70% of the maximum, which is most likely still too generous given how inefficient my system is, but whatevs.


That said, we really wanted to make the caramel flavor sing, so we resolved to put actual caramelized sugar into the boil. This was definitely a learning experience for us and I would like to share the lessons we learned. Most importantly, table sugar has a melting point of 366.8°F. If you are caramelizing sugar, be extremely careful that it does not touch you at all. This stuff is like napalm in that if you get any on you, it won’t easily come off. And it will be second-degree-burn hot. You cannot take your eyes off this stuff for a second because if it starts boiling, you’ve already lost the battle.

Lesson the second: the ultimate form caramelized sugar takes is determined to a large extent by the temperature one allows it to set at. The first time around, we heated it until it was liquidy and then dumped it into a glass bowl where it promptly turned into a solid hunk of caramel glass. This is great if your looking to ruin your teeth; not so great if you are trying to get it into the kettle. We gave it another go and added some hot water after caramelizing the sugar. This allowed it to set at a lower temperature and resulted in a mixture that was more amenable to things like, you know, pouring.

I’m not a big fan of extracts, so for the coconut flavor I got two pounds of dried coconut flakes. I had no idea where to put these in the brew process. Some recipes suggested in a secondary fermenter, some said in the boil, some even said in the mash itself. So I decided to go with all three. We just spread the flakes out on a baking sheet and toasted at 325°F for about 5 or 10 minutes. Your kitchen will smell amazing.

I wasn’t too particular about my hop choices for this beer because I don’t really want any hop aromas poking through. Adding just enough bittering hops to balance out the sweetness should be enough. To save some cash, I chose to use what was available in my freezer—home-grown Willamette hops from last year’s harvest. For yeast I went with Wyeast 1056 (American Ale ) for no better reason than because I use it for just about everything. If it ain’t broke…

The final recipe ended up looking like this.

Grain Bill
9 lb. organic 2-row
2 lb. flaked oats
1½ lb. crystal malt (80L)
1 lb. Belgian biscuit (amber) malt
1 lb. toasted coconut flakes
¼ lb. chocolate malt

Single infusion mash @ 156°F for 60 min.

Boil Additions
7½ oz. caramelized sugar
1½ oz. Willamette hops @ 60 min.
1 tsp. Irish moss @ 15 min.
½ lb. toasted coconut flakes @ 10 min.

Wyeast 1056 American Ale

I’ll let you know how it goes in Part Two.