You’d never guess this coffee was honey-processed by the prep. Its clean, polished green appearance keeps a lid on the fact that the parchment was dried in mucilage. It’s a very dense, nicely sorted grade 2. Typically dry for a washed Ethiopia, besides the high density its standout feature is screen size. A steady 70% of the lot falls between screens 15 and 16 (and over 90% 14-17), which should make it a little easier than the average Ethiopia to track development and color changes during the roast.
A truly lovely coffee, performed well on both roasts. Our second roast, PR-0327, had clearer and more present florals on the cupping table compared to our first roast, PR-0326, which had more developed sugar browning flavors. These two roasts are a good example of how we would roast this coffee for a single origin espresso (PR-0326) and for filter brew (PR-0327). Although both have the same end temperature, their ColorTrack scores were very different, nearly 3 points apart.
When we look at the chart that displays the different stages of roasting between the two roasts, we can see that almost 65% of PR-0326 is Maillard reactions. PR-0327, on the other hand, is much less with only 56% of the total duration with Maillard reactions changing the color of the coffee. To achieve the shorter, lighter roast, we relied on a hotter drum using a high charge temperature and low gas, with a large boost of heat before the Maillard reactions begin. We kept the flame high through the finish because of the dense nature of this coffee and we wanted to get a relatively high end temperature in a short amount of time which gave us a lot of acid complexity in the cup.
Brewing a comparison of the two roasts in identical Chemex produced some interesting results. At a standard ratio and grind, the coffee in both cases responded very slowly to water. The resulting brews, even at 9 minutes, were tasty, but low in soluble material. Curious about how to work with this coffee, I asked Richard to try a Kalita and see if he could get the brew time under 4 minutes. Using a higher coffee to water ratio, Richard’s brew still took nearly 5 minutes, was high in solubles but still relatively low extraction percentage. Regardless, we tended to favor Jen’s lighter second roast, PR-0327, and Richard’s Kalita brew was seen as offering a bit more clarity and distinct floral notes than the Chemex. If you’re brewing this coffee in small batches, keep in mind you may have to wait a little longer than you’re accustomed to for similarly tasty coffees.