The second generation Roast Genie is nearing completion. This weekend, I roasted seven different 100 g batches of Guatemala Huehuetenango Cuilco to begin to get a feel for how the new roasting system is going to work.
The first thee batches were roasted Friday with different durations of roast time past the onset of first crack. For these I roasted at an initial set point of 510 F (environmental temperature at roast chamber exit) and then backed down to 490 F once 1st crack started.
Roast 1 – 3 min past onset of 1st crack (16.8% mass lost)
Roast 2 – 3 1/2 min past onset of 1st crack (15.0% mass lost)
Roast 3 – 4 min past onset of 1st crack (15.6% mass lost)
General observations during roasting:
First crack was very pronounced with this coffee. Onset was consistent from batch to batch starting at about 640 to 650 seconds into the roast. Darkening of the roast was not all that pronounced from the 3 min to the 4 min roast as it sometimes is with other beans. Although I don’t think the roast stalled, the frequency of cracks did not seem all that high. It seems like this bean can take quite a bit of heat before significant changes take place. The change in bean mass during roast was a little unexpected with the lightest roast having the highest loss in mass. This is probably due to the fact that I am roasting such small quantities that the measurement error is significant (+/- 2g measurement error = +/- 2% mass lost error).
None of these roasts were particularly impressive displaying a significant lack of body even after 36+ hours of rest. I perceive there to be significant sweetness and acidity in the cup but there is also an astringency in the finish and general lack of body that I find unpleasant. My suspicion is that these are under-roasted. I also noticed that the beans sound much harder or denser in the grinder than the Yemen beans I have recently been using. Those bean seem very soft and these much harder — at least in comparison. The Yemen beans developed in the roaster much faster than these seem to.
I completed the next four roasts today and attempted to adjust the profile to stretch out the roast a little in the hopes of allowing a little more body to develop. After reading this post on Stretchin’ Out the Roast at Sweet Marias and watching this video, I decided to extend the drying time by adding a 2 minute dwell at 350 F during initial heating. I roasted three batches with post first-crack times of 3 1/2 minutes but used three different post first-crack temperatures: 490 F (like roasts above), 500 F and 510 F. After observing less impact on this bean than I had expected. I roasted a fourth batch with a 3 minute dwell time at 350 F, 510 F post first-crack temperature and just about 5 minutes past initiation of first crack. For this roast, 2nd crack had just started.
Roast 4 – 3 1/2 min past onset of 1st crack @ 490F. Includes 2 min dwell @ 350 F drying time (16.8% mass lost).
Note that this roast profile shows 1st crack happening about 100s earlier than it should — for some reason there was a single bean crack at 580s or so but then nothing for another 100+ seconds.
Roast 5 – 3 1/2 min past onset of 1st crack @ 500 F. Includes 2 min dwell @ 350 F drying time (17.0% mass lost)
Roast 6 – 3 1/2 min past onset of 1st crack @ 510 F. Includes 2 min dwell @ 350 F drying time(17.0% mass lost)
Roast 7 – 5 min past onset of 1st crack @ 510 F. Includes 3 min dwell @ 350 F drying time (18.0% mass lost)
General observations during roasting
Again – I was rather surprised by how “robust” this bean seemed during roasting. Changes in roast time and temperature that I would have expected to dramatically impact other beans seemed to have much less impact on these. The 7th roast in particular was surprising in that the beans were not more significantly darkened. I have had trouble getting the results I wanted from Guatemala beans in the past and wonder now if more heat is required to get the desired roast levels. I will add tasting notes tomorrow after I have had a chance to begin sampling.
Thoughts on the Roast Genie system
Overall the 2nd generation system is working well but I have a ways to go to get all of the automated posting completed. The camera for taking pictures of beans after roasting still needs to be completed but the automated preparation of plots is working pretty well.
I had anticipated that having one pre-crack temperature setting and one post-crack temperature setting with manual control of the post-crack time would be sufficient control over the roast profile. After the experience here with this Guatemala, I think I may need to allow easier adjustment of these profiles.