Updated: Mar 5, 2020
We’ve celebrated our third month in business and are staring our fourth right straight square in the face. During this time, I’ve had to overcome the learning curve of roasting coffee on a five burner gas grill converted into a coffee roaster while nailing down the roast development necessary to make great coffee.
This learning curve has been helped considerably by Shane Lewis at RK Drums who I can’t speak more highly of. Shane has been instrumental in helping me understand the how to roast coffee on our somewhat less-than-conventional and often quite finicky propane grill intended for burning burgers and charring chicken breasts.
The grill we use for roasting seems to have a mind of its own. One day it will roast perfectly on the numbers while others it simply doesn’t cooperate by accelerating or decelerating the roasts by minutes instead of seconds. All that to say, I probably speak for the minority of RK Drum users who have these sort of problems as many can calibrate their grills and roasters to within seconds of the RK profile chart. (I blame all of this, by the way, on Texas weather. It is in no way, shape, form, or fashion my fault nor the fault of the equipment.)
Nonetheless, I’ve come across some recent discoveries that may assist other RK Drum roasters in creating a more consistent product from batch to batch while eliminating some of the less desirable properties our specific grill tends to deliver.
Time and Temperature
The essential function of all coffee roasting is time and temperature. Pro roasters will incinerate me (and probably Shane) for saying that internal bean temperatures are not necessary for producing an excellent, quality, and consistent roast. Nevertheless Shane, along with the company’s creator Ron Kyle, developed a simple time and temperature chart for their drums where under a given temperature and green coffee load a coffee should reach first and second crack at a certain time.
A snippet of the 12 pound drum roast profile chart from RK Drums. You can find the full chart as well as the charts for all of the RK Drums here.
However I, along with many other roasters, find that the time and temperature chart is a guideline and doesn’t always align perfectly what my temperamental roaster wants to do in any given day. With that said, I did some research and developed a system that seems to work really well for our machine and our coffees.
While I can’t with any confidence express that I’m an expert in roasting, roasting techniques, nor roasting green seeds beyond a brown, grindable, drinkable substance we call “coffee”, I’ve sufficiently studied some roasting techniques in order to better understand how professional roasters roast great coffee time and time again.
I stumbled across the Cat & Cloud podcast above with the Czar of modern coffee, Scott Rao, which really turned me on to exploring more technical aspects of our roast. While Rao’s “textbook” on coffee, The Coffee Roaster’s Companion, is one of the predominant texts for coffee roasters and enthusiasts, his article, Roast Development Time and the Prerequisite for Any Successful Roast, published on the Daily Coffee News really shines a clean albeit simple light onto the process of roast development.
In Rao’s article, he mentions that great, memorable coffee has two consistent characteristics that stand out from roast to roast. One, Rao says, is that coffee hits first crack at 75-80% of the roast and the coffee develops 20-25% of the total roast time. In layman’s terms, the time between roast start to first crack should consume 75-80% of the roast while the time after first crack starts should be 20-25% of the roast.
Rao’s Roast Development on an RK Drum
While Shane’s chart is a great starting place, my numbers almost never line up. Depending on how the roaster decides to heat on any given day and various ambient factors, I’ve had to adjust accordingly on our RK setup.
For roaster’s looking to improve and measure a roast on an RK Drum, the simple way of doing so is recording the time of first crack’s start* and using some simple division to calculate an approximate development time and when to pull a roast to optimize origin character and roastiness. Using this method will negate any external factors that may be affecting your roast and allows the roaster using an RK Drum to make realtime calculations on their roast.
Our 12 lb. RK Drums setup on a 5 burner propane grill.
Let’s use an example from Shane’s chart in the paragraph’s above.
Take a look at the 4 lb. example in Shane’s chart. According to his chart, we should be hitting first crack at approximately 10 minutes. Assuming that we are hitting first crack at 10 minutes, using Rao’s methodology, we can divide our first crack time by .80 and .75 respectively to give us a range for when to pull our roast.
Using the formula 10 ÷ .80 and 10 ÷ .75 we get a range of 12:30 and 13:20. Meaning, our coffee has developed for 20% of the roast at 12:30 and 25% at 13:20. But if we look at Shane’s chart, a 4 lb. roast should hit 2nd crack at 12:00 which is a little too long for an evenly developed roast and at risk for hitting second crack. Ipso facto, I generally adjust my constants in my equation to .85 and .80 respectively.
Using the constants of .85 and .80, assuming we hit our first crack at 10:00 as per Shane’s chart, we can create the following formulas for our development range: 10 ÷ .85 and 10 ÷ .80 giving us a range of 11:45 and 12:30. With that said, we can be watching to pull our roast within the timeframe of 11:45 – 12:30 total roast time. Using this method and assuming you hit first crack on schedule, pulling your coffee in the calculated timeframe will give you an evenly and adequately developed coffee without getting into second crack.
*For the longest time, I marked the start of first crack when first crack started “rolling”. In other words – when the beans were popping and cracking in rapid succession. While measuring first crack’s start is somewhat arbitrary, Rao suggests measuring first crack, “the moment [you] hear more than one or two isolated cracks.” If you measure differently, you may need to adjust your constants accordingly.
Personal Preferences in Our Roasts
While second crack isn’t inherently a bad thing in the least, we prefer to keep our coffees just shy of second crack in that developmental “sweet spot” where origin character and the sweetness of the roast complement each other nicely. In fact, Shane coaches a technique he dubbed “10 in 10” where the coffee is pulled from the machine once the “operator” hears 10 snaps of second crack in any 10 second window. This gives the coffee a very balanced combination of origin and roastiness, but I have found it a little on the roasty side for our personal tastes and the tastes of our customers seeking something towards the lighter side of the roast spectrum. (Again, this is a personal preference. 10 in 10 will make an excellent coffee, and no one will complain about it.)
I used Shane’s 10 and 10 method (once I got the hang of it), but found that my personal senses never quite acclimated to catching the sounds of second crack in time leading me to either pull the coffee too early because I got nervous of over roasting or pulling it too late thinking I hadn’t heard 10 in 10. With the addition of ambient motor noise, creaks and groans from the rotisserie spinning in the supports, and the coffee tumbling through the drum, I had trouble discerning those more faint second cracks whereby this method became my staple to determine when to cool my roasts.
Some Notes on Roast Development
While this formula works well for most applications and adapts to your roast for the day, it isn’t the end-all be-all for home roasters. There is certainly room for variation, but this formula is adaptable to any roast. If your roaster is running fast, the formula adapts accordingly. If the roaster is running slow, that is accounted for as well.
Roast development really doesn’t have to be rocket science, but far too often we make it just that. The problem I’ve always run into is mastering the art of “guesstimation”. I’ve always played guessing games with the development of our coffee in the roaster and made assumptions as to when and how well it’s developed. With this method, I’ve seen far more consistent roasts and far more consistency in the cup.
As for Shane and RK, I can’t say enough good about the quality of the equipment, the personal service, and dedication to perfection that Shane displays even months after my initial purchase. Shane even called me on a Sunday afternoon prior to a family obligation! Now that’s service of the best kind!
I’m interested to know how others have been able applied this or similar tactics to help predict their roasts on RK equipment, so please leave comments down below. If needed, I can always publish a followup piece that explores additional methodology.
Updated February 12, 2017 at 4:58 P.M.