Coffee Data Science
Caffeine Across Roast Level
A review of some cool data
Recently, a study was published on caffeine and extraction from different levels of coffee roasts. I pulled the data to re-plot and took a look.
The study sought to examine caffeine content across coffee roast, and the popular idea has been that darker coffee roasts have less caffeine than lighter roasts. This study found more caffeine in lighter roasts for their specific brew method, but darker roasts had a higher caffeine content for a similar extraction rate.
One challenge is to get a similar extraction rates to compare caffeine per extraction yield content.
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).
Coffee Roasts
The authors made 5 roasts by two bean variants (Natural and Washed) on the Aillio Bullet with a batch size of 500g, and the development was based on the time the roast was dropped:
- First Crack (FC)
- FC+1 minute
- FC+3 minutes
- FC+5 minutes
- FC+7 minutes
They measured a few things, so we had some interesting data points to explore for both whole bean and ground. There is a fun trend between bean drop temperature and bean color, but it depends on natural or washed.
We can also compare ground color to whole bean color, and this trend seems linear until the end when the roast gets a lot darker.
They also measured post-roast weight, and they measured the weight 10 days later at brew time. This allowed them to measure the weight loss from de-gassing. The washed coffee degassed faster than the natural coffee.
Brew
They used an Aeropress to do an immersion brew for 1, 2, or 10 minutes. At the end, they used a force stand to do a fast 30 second draw down at ~1 bar pressure.
For coffee grind, they used a sifter to try to control the grind distribution caused by the change in how the coffee particles were ground due to roast. This has some bias in how much fines are excluded, which could be problematic. However, this has not been determined yet.
Looking at Extraction Yield (EY), the darker roasts have less extraction. This is odd because typically in espresso or pourover, darker roasts extract more. So there is some artifact of immersion brew. I plot these as the coffee color instead of the development time because the color is more helpful in understanding the roast development for different coffees, batch sizes, and roasters. On the Agtron scale, a darker roast has a lower number.
Looking at 1 minute and 10 minutes for washed vs natural, we can see close trends, but there is some weirdness especially for the 36 Agtron Natural.
Looking at caffeine, longer immersion gives more caffeine similar to EY, but the Natural roast has similar levels of caffeine regardless of roast. For the washed coffee, caffeine seemed to have a slight pattern.
Comparing natural and washed at 1 and 10 minutes shows an interesting trend for naturals being higher at the longer brew time.
We can normalize the caffeine by EY, and then the real trend occurs. All the data points to darker coffee having a higher caffeine content extracted per EY. This might be an artifact of brew method, and as an espresso drinker, I would be curious how espresso performs.
Comparing washed and natural, they have similar trends.
I have collected data for espresso on EY across a few roasts levels. Looking at this espresso data, the darker roast (48 Agtron) extracts higher in the same output ratio as the others.
I wonder in espresso if a higher extraction yield (EY) would also mean a higher caffeine extracted, and I am curious how much the brew method impacted the results.
This study showed darker coffee extracted high caffeine relative to coffee solubles than a lighter coffee on immersion brewing. However, darker coffees had a lower EY in this brew setup. This study helps push the knowledge in the area of coffee extraction vs caffeine extraction.
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
