Coffee Data Science
How Important is QC in Coffee Roasting?
Aiming for consistency
Quality Control (QC) should be important for coffee roasting, and I hope it is. Like everything, I want to know how important it is. How much can roasts vary without an apparent change in flavor? This is the first experiment of a few experiments to look at that question. To start, I looked at repeating the same roast 4 times using the Roest and the same development time.
To evaluate, I look at post-roast metrics and espresso shot comparisons. As a reminder, my main love is espresso, so my questions revolve on tasting differences in espresso shots rather than on the cupping table or pourover.
My inspiration was from Scott Rao’s post about Prodigal’s QC Protocol: batch to batch roasts need to abide by these two criteria:
- +/- 2 Color points on the ground coffee color measurement
- +/- 0.1% Weight loss
I thought these were a bit tight, but in discussions with a few professional roasters, these numbers are realistic.
For smaller batches like the ones I regularly do on the Roest (120g green), the resulting roasted coffee weighs around 100g, so a 0.1% error in weight loss is the equivalent to 0.1g. That is a single bean, which is easy to lose in the roaster or when pulling the beans out no matter how careful I am. For 10 kg, 0.1% error is 10g, so clearly this is a spec that’s easier to manage for a larger roaster.
Coffee Roasts
I used 4 roasts of Guatemala Xinabajul Producers 2 from Sweet Maria’s. I aimed for similar drum temperatures at the start of the roast, but I think the first roast had a slightly lower start temperature on the drum. Luckily, the curve was very similar, just offset.
All the Rate of Rise (RoR) curves matched well except for the first one.
Similarly, the first crack curves had similar patterns.
Post-Roast Statistics
Here is where the going gets tough. For weight loss, even using 86.3% as the starting point, all the values are outside the +/-0.1% range. This is tough, but then again, this can help answer the question of whether this matters in the cup.
The moistures were similar.
Water activity had a bump, but I’m not sure what that means.
They all had a similar density, which is expected based on weight loss.
For color, if we take 54 as the baseline, all roasts are within +/- 2, but this is whole bean color.
For ground color, if we take 84 as the baseline, all four roasts are within +/- 2 in color.
After roasting, I add 4% moisture to the beans, let them sit for 24 hours, then use a vacuum chamber to suck out as much gas as I can for 1 minute. Finally, I let the beans rest in a vacuum jar for at least 1 week before use, and then I pull shots over a few days, so the taste should vary due to resting.
Tasting Equipment/Technique
Espresso Machine: Decent Espresso Machine, Thermal Pre-infusion
Coffee: Home Roasted Coffee, medium (First Crack + 1 Minute)
Pre-infusion: Long, ~25 seconds, 30 second ramp bloom, 0.5 ml/s flow during infusion
Filter Basket: 20 Wafo Spirit
Other Equipment: Acaia Pyxis Scale, DiFluid R2 TDS Meter
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).
Shots
For each roast, I had 4 espresso shots. These were all pulled within 24 hours of each other. I found some variance, but it seemed like noise. Each column is a roast within each shot number.
Looking at the roasts separately, I also didn’t see anything outside of the variance of each roast.
Double checking extraction yield (EY), there weren’t major differences in EY.
These tests were enlightening to see what appears in the espresso shot taste. My next tests will look minor differences in development time, and I will start expanding the range to see when a color change starts to be very noticeable.
If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.
Further readings of mine:
My Second Book: Advanced Espresso
My First Book: Engineering Better Espresso
