charting the diff

November 10, 2005 – 1:11 am

seeming proof that ambient temperature has everything to do with a desirable roast curve. notice the very rapid warm-up period when the room temperature starts at 25 C (77 degrees F), denoted by the green bars. the night before, roasting the same amount of coffee in the same electrical outlet with the same roaster settings but much cooler ambient temperature, you get a much longer warm-up period but a much more continuous roast curve throughout — no flat spots in the middle of stages, where the warm-roasted beans are presumably stalling or baking.

the roast shown in green, by the way, ended up burning. “duh,” you say. but the signs of full city came much quicker in the late stages, taking me by surprise and costing me precious seconds before i recognized them and hit the cool button (notice how close the readings are at around 5:20, then how wide the disparity becomes just before cool-down). had i been paying closer attention to the actual temp levels instead of mindlessly punching them into the spreadsheet, i would have noticed the readings soaring well past 260 C (500 F) — a sure sign the heat was becoming a bit too volatile for a controlled finish. so not only is a cooler beginning temp necessary for a gradual, continuous roast curve, but it helps foster a more controlled and precise finish. again, that might seem obvious, except that the total roast times were exactly the same. so the colder start-up doesn’t necessarily make a roast take longer, it just makes the roaster work harder to deliver a final product. i love winter.

p.s. of course, the idea isn’t to find a cold spot in which to roast. the point, eventually, is to be able to adjust my heat settings in order to reverse-simulate the beneficial aspects of the cold, i.e. a measured roast curve and a paced, exact finish.