Coffee Data Science
Positively Positive Pressure in Coffee Roasting
A little more data, and maybe some clarity
Through multiple explorations of positive pressure, I still am having trouble finding conclusive evidence that it is good for roasting. This has been confounded by multiple optimizations to the roast profile, so I thought I would explore it a little bit more. I wanted to add a few more roasts where the coffee color is closer to the baseline by ending at the same bean temperature.
Positive pressure is created by a high internal fan speed compared to the exhaust fan speed. In this case, I go from 3400 RPM to 4400 RPM on the Roest internal fan speed.
Post-Roast Metrics
Previously, I used the Sumatra and Ethiopia roasts, but I noted how the high internal fan speed came out a bit lighter in roast (higher Agtron color number). So I made two more sets with Burundi and Costa Rica beans where they ended at 212 C bean temperature. The metrics were closer across the board.
These two new roasts also used a different profile where the fan speed was low during the start of the roast for the drying phase (bean temperature < 150 C), so the internal pressure should have been even higher for that part of the roast.
Color specifically was within 2 points which is where I want it to be.
Metrics of Performance
I used two sets of metrics for evaluating the differences between techniques: Final Score and Coffee Extraction.
Final score is the average of a scorecard of 7 metrics (Sharp, Rich, Syrup, Sweet, Sour, Bitter, and Aftertaste). These scores were subjective, of course, but they were calibrated to my tastes and helped me improve my shots. There is some variation in the scores. My aim was to be consistent for each metric, but some times the granularity was difficult.
Total Dissolved Solids (TDS) is measured using a refractometer, and this number combined with the output weight of the shot and the input weight of the coffee is used to determine the percentage of coffee extracted into the cup, called Extraction Yield (EY).
Shot Data
I pulled 4 pairs of shots per roast set resulting in 16 shot pairs. The data shots most of the high internal fan speeds had improvement except for 4.
The low fan speed start didn’t affect these results either.
The average across taste components show the Sour and Bitter having the largest impact.
TDS/EY didn’t have a noticeable difference.
On a statistical level, one Sour, Bitter, and Aftertaste were significant according to a two-tailed paired t-test.
I considered doing more roasts because the p-value for Final Score was close, but I’m keeping in mind that I’m conducting a multitude of roast experiments, so I most likely have to repeat this experiment later to see if this still impacts the taste.
