The genetic diversity of Ethiopia’s indigenous coffee stock isn’t the only thing that makes its coffees remarkable. A time-tested system of channel washing and then soaking (aka double-washing) the coffee parchment after a long fermentation (36 hours in this case) helps to sort by density and clean away any remaining fruit pulp and fermenting microbes. There are also some indications that this post-fermentation soak may improve quality by tricking the seed into the initial stages of germination. After all this, the coffee is dried and hand-sorted on raised beds.
The result in this case is a pristine grade 1 Ethiopian coffee, with a typically high density and low moisture content. The screen size is surprisingly small; nearly two-thirds of the lot passes through screen 16. It’s possible that as a result the coffee could appear unexpectedly dark while roasting, so be sure to keep a close eye on the roast and check Jen’s recommendations.
Richard and I brewed a Chemex of each roast, and despite nearly identical technique, somehow I ended up with a minute longer brew time. Anthony Rue, of Volta Coffee, Tea, and Chocolate in Gainesville, FL, happened to be in town and joined us to taste. While both roasts produced really enjoyable coffees, we were especially taken by Jen’s second roast, PR-0329, with its sugary sweetness paired with mango, nectarine, and a slightly dry bergamot finish. Anthony’s notes for PR-0328 included “magnolia,” a great descriptor I plan to use more often in the future when appropriate.
Sometimes what may look like an obvious flaw in your roast, may not result in a significant flavor discrepancy in the cup. These two roast profiles look nearly identical. The difference in total roast time in the drum, ratios of the three roasting stages, ColorTrack numbers, and even the end temperatures are are minimal. However, there is an obvious spike in the rate of rise on PR-0328 when I increased the gas at 4:14. In order to mitigate this drop in momentum on my second roast, I lowered my charge temperature and increased the gas by a half point. I still needed to increase the heat around the same time, but the rate of rise never dipped like it did in the first roast. This is because there was enough energy and I added heat at a time that moved seamlessly with the curve.
On the cupping table I expected PR-0328 to appear flawed, but instead it was the profile that I preferred because of its lively acidity and juicy mouthfeel. Honestly, both roasts were lovely and there was little that separated them on the table. Recently, a lot of roasters are implementing the roasting styles of Scott Rao who claims that the rate of rise should always decelerate throughout the roast. While I do believe this to be true with several coffees, I have also tasted several lovely coffees that do not comply with this technique. With this coffee, we hardly noticed a difference with the small spike. Personally, I will always rely on the cupping table over a roast curve to determine how successful a roast is.