Charge and roast time experiment

Thanks to Jono from Rosetta roastery, for inviting me to participate in this experiment. Although the sample size was small, both in roast sample size and number of samples, some valuable lessons where learnt.

Experiment: To roast the coffee 6 ways only changing the charge temp (CT), keeping flame at the same throughout and once First crack (FC) is reached have dev time ratio of (DTR) 10% of the roast. Mostly Drop Temp (DT) – was around 3°C later.

Sample size: 200g roasted. Two coffees were tasted, then the one we liked was identified as the favourite roast or middle roast:

Shortest roast (SR)  200C / 392F 7:45 210C / 410F  49sec (10%)  213C / 415F
Middle Roast / Favourite (MR)  187C / 367F 9:05 205C/ 401F  57sec (10%)  208C / 406F
Longest roast (LR)  177C / 350F 9:45 201C / 394  62sec (10%)  204C / 399F


Of all the coffees tasted the LR and MR where sweeter and more complex. The MR was the most complex with stronger Jasmine aroma, “natural” type red fruit, citrus, juicy mouthfeel, pleasant acidity. The SR tasted green, pithy, underrated.

Jono’s conclusions: Temp far less important than allowing sufficient overall time for development to take place. Too much time leads to obliteration of acidity, but too little the (FC at 07:45) leads to underdevelopment.

Obviously, batch size will have very real implications on dev time and development levels, but safe to say that a much hotter but shorter roast can lead to far lower levels of development than a far cool but longer roast. Hinting at total roast time being more important that a more aggressive temp profile….

However too many variables. Too small a data set.


While Jono and I are okay with sharing this knowledge now, we reserve the right to change our opinions at any time in the future.

Coffee used: Rwanda – washed

1 thought on “Charge and roast time experiment”

Leave a Reply

Your email address will not be published. Required fields are marked *