The case of the thimble, a 5 gallon bucket and what any of this has to do with coffee roasting.
One of the most commonly misunderstood concepts is the heat required to roast a <insert LB’s of coffee here> pound mass of coffee. While we have come up with the following guidelines, some people don’t understand why we have arrived at those numbers. It’s true the numbers are slightly arbitrary in the sense that we don’t have measured science behind it, but they’re also based on a significant amount of testing, and are designed to give you plenty of extra heat\speed in roasting if you need it.
I recall at one point one guy saying… “Hey, my grill is 28,000 BTU and I have no problem getting to 650F, so why can’t I roast 8LB?” We in fact, recommend 45,000 BTU, so what’s wrong with his claim?
See, BTU is a measurement of heat output, not necessarily temperature output. Let me provide the following analogy to best illustrate the point.
“A candle produces approximately 100 BTU\hour.”
THE EXPERIMENT:
Imagine if you took a thimble, the little thing you use for sewing, placing it onto your finger. I use this as an example of a little tiny cup. Imagine if this cup were filled with water and placed over a common candle. Do you think this water in the thimble is capable of boiling?

What’s interesting about this experiment is that you have an analogy that’s easy for all of us to understand based on common everyday experience. I imagine in certain cases, the water might not boil, but I think most of us would agree the answer is “yes, the water in the thimble would boil”. After all, we’re talking about a very small amount of material that is heating up. This material, what we will call the “mass” is the thimble and the water.
“1 BTU is the amount of heat energy needed to raise 1LB of water by 1 degree Fahrenheit.”
Now imagine if you take 5 gallons of water and you place it over the same candle in the previous thimble experiment, do you think the water would boil?
“Do you think a 5 gallon bucket of water would boil placed over a candle?”
Most of us would say “no” this water will not boil in a 5 gallon bucket. Why do you suppose this is true? In both cases the candle, which burns an average 23722552 °F is the same in both experiments? So what’s happening here and what does it have to do with coffee roasting?
In the second experiment, the mass that we are heating up is the bucket and the 5 gallons of water. If the water heats up, it will heat up relatively little. What happens is that the entire mass (the bucket and the 5 gallons of water) is absorbing the relatively small amount of heat input from the candle, then what heat energy is absorbed dissipates into the environment before the water heats up to any appreciable degree. You can imagine an even more exaggerated experiment by trying to heat up a swimming pool with a candle.
So you see, as mass increases…. you need more and more heat input to be able to bring the mass to the desired temperature. How then does this translate when roasting coffee? Well, it’s exactly the same. Instead of a thimble or a bucket, we have a drum. Instead of water we have beans that contain water. The larger the mass grows (2LB drum up to 12LB drums) the more heat input you have to have. The flame temperature might be the same, but the capacity to input heat increases. This is the BTU number. Think of this as 50 candles under the 5 gallon bucket of water, and now we might have something to talk about.
In a real life example, we tried to get an 8LB drum to roast 8 pounds in a 28,000 BTU grill. This grill, when empty had no problem heating up to 650700F. But the moment you add a bunch of mass to the grill, (much like going to the 5 gallon bucket of water over the candle) that same grill with 8LB of coffee, struggled to get above 175F. You could roast 12LB in this scenario, but not 8LB.
We then later realized we needed to trap as much heat as possible and keep it in the grill. We used sheet metal to cap off every hole we could find in the grill. The result was that we barely achieved 600F by trapping every ounce of heat. The roast finished at 22 minutes, which is a pinch longer than our ideal. This could be considered a success, but this was literally with 100% gas on the burners. If one needed to go faster, the heat simply isn’t there. At 45,000 BTU you’ll have more than enough head room, even at 8LB to speed up the roast by increasing heat if you should need to.
This is a real life example of how BTU plays out when dealing with increased mass. The more mass you have (the larger your roast)..the more heat output (BTU) you need to finish the roast on time. Now this “on time” is key. In the above example of the 8LB roast, where we capped off all the holes, this was trapping enough heat to get the roast to finish on time. We could have left the vents uncovered andlost a bunch of heat, roasted at 350 for 45 minutes and finished the roast. The roast would have tasted terrible, but it would have finished. The trapping of the heat allows us to finish on time and speed up the whole reaction.
So in summary, More mass requires more BTU to get the coffee to finish with sufficiently short time. If you sacrifice any of these, the variables change and you can find greater degrees of success. For example, if you struggle with BTU, you can lower your mass…or lengthen your roast (not ideal), or you can supply more heat to compensate. Hopefully this sheds light on a complicated subject, which hopefully now makes a little more sense.
ShaneRK
Author: Shane Lewis
Shane is the founder of the green coffee exporter Old Spanish Main, Ltda. in Bogotá, Colombia, having practiced under the renowned grader & cupper Lelly Espitia. He is a cultivator of a #14 Colombian Cup of Excellence coffee. Shane currently designs & manufactures coffee, cacao & grain roasters with RK Drums.