I roasted my first batch of decaf beans last night, which was a little interesting. The profile I was running has a setting of two minutes at 335, two minutes at 385, and then 6 minutes set at 435. In most cases, I halt the roast just on the verge of second crack, about 1.5 to 2 minutes into stage three.
To digress, for normal beans, what this typically means is that after about 4 minutes, the chaff really starts building up on the exhaust side and bean circulation slows to a crawl. When this happens, bean circulation drops and roasting chamber temperature tends to rise, as the chamber becomes more of an oven and less of a fluid bed. At the end of the day, the bottom pile of beans start roasting much hotter and start entering second crack with momentum, while the top beans still have not reached Full City. While I’ve yet to punch a hole in my roasting chamber in order to run in a thermocouple, my best theory is that we’re looking at temperature exposures at the heater exhaust in the 485 degree range, with a 20 or more degree differential to the top of the bean stack. I can try reduce the amount of beans I’m roasting to reduce the amount of chaff I’m producing (actual bean volume does not seem to be my limiting factor), but after a certain point, this becomes academic as I am no longer producing enough coffee for more than a few pulls of espresso. I’m considering modifying the top vent for the unit by removing the protective screen since I vent outside through dryer ducting … but again, I’m not real excited about just mauling the design of the unit. Getting a variac for voltage consistency (I’ve seen enough weirdness in general to reason our voltage is suspect) and setting lower temperatures for the roasting profile (recognizing that with backpressure they will rise substantially higher than set) is probably the best approach. End of digression.
With the decaf beans, there is no chaff, so the exhaust backpressure does not build up as the chaff slowly clogs the system, so I get great bean circulation and airflow all through the roast cycle. This means things are a lot more controlled, the roast is much more consistent, and things are behaving like they are supposed to. The problem with that is that a finishing roast profile at 435 ends up not producing the same oven effect I get in the normal roast. I have read some suggestions of putting some chaff from a regular roast in the collector to jam it up, but this seems counterintuitive, as I feel it is a positive thing that the beans are actually circulating and behaving as designed. It does not make sense to me to break a correct process to work like a more familiar incorrect process. I see software developed this way and processes engineered this way all the time (well, I had to work a slap on the development community in here somehow), and it makes me cringe. “That’s ridiculous!” “But … that’s the way it’s always been done!”
In any event, I jacked up the final roast preset to 460 and 8 minutes, and I hit second crack with great circulation and the slightest bit of momentum dead on at six minutes. I pulled my first home-roasted decaf shot this morning with the second batch (I suppose I should try the baked batch to keep me honest; maybe it’s the best ever), and it was still better than any regular shot I’ve purchased in a commercial espresso shop (keeping in mind that mostly consists of Starbucks, Caribou, Border’s, Espresso Royale, and the little indy shops in C-U; this is not exactly a stellar cast). It didn’t have the same richness, the crema was somewhat weaker, the sweetness wasn’t there … but considering this was my first pull of the morning, I’m pleased.