Ethiopia Lemon Candy

Many, Many Roast Profiles.

We roasted this coffee in 20 different ways before choosing our final roast profile. It is very surprising how different this coffee can taste with different roast profiles! To run these tests, we roasted at out “Production” batch size of 600 grams. This means that when we find a profile we like, we can replicate it with complete confidence and accuracy. Nearly all roasters use a small sample roaster and roast around 50-200 grams to understand how that coffee can taste, then they roast a couple of batches on their big production roaster to find the final profile. This is a disadvantage because they would ever be able to roast 19 batches to “sample” a coffee on their production roaster because it would not be profitable for the company to waste that much green coffee, let’s say 600-1400 pounds, or $3,000-$14,000 depending on roaster size and green cost/quality to test 19 roasts on their production roaster.

Another advantage for us is we don’t have to translate our sample roasts to our production roasts since we use the same roaster and batch size for both our sample roasts and production roasts. When Roast Masters have to translate profiles between two completely different roasting machines, it is wildly inconsistent and flawed. It is like developing a recipe to cook bread in your home oven, then having to cook 10 loafs in a giant commercial oven all at once using the heat exact recipe of time and heat (and complicate this even more with other coffee roasting variables!). This is why we believe the best roasters are the ones who also have the most experience sample roasting and translating profiles.

You can save a lot of time this way by working with a roaster that’s currently 15-30x bigger than ours, but that kind of workflow calls for massive quality control measures that we have a hard time seeing in a majority of roasters.

1st Test Roasts. More developed.

After roasting 6 profiles that tasted subpar to our favorite Nordic-Style roasters, we decided to roast lighter. We were shocked because our roast weight loss was 10.7-11.7%, which was much less than our Mexico roast at 13.2%. Because we had more weight loss from this Ethiopian green and the fact that our end temps of both our infrared and bean probe were similar to the Mexico roast, we assumed the Ethiopian profiles would have been much lighter than the Mexico profile we had used previously. We were wrong. This is a completely different bean of smaller size and less moisture.

2nd Test Roasts. Less developed.

We roasted 7 new profiles (V7-V13) that were very different and experimental with the goal to bring out more clarity, acidity, and bright sweetness. Our new roasts had a weight loss between 8.8-11.2% (most hovering around 10%-10.3%).

These roasts really told us the diversity of flavors this very high-quality green can present! With one roast profile (V13) we roasted very fast and tried to replicate a flavor profile you might experience from a very light roaster like Sey in New York. With another roast profile (V10) we found some of these same flavors, though they tasted sweeter and with a lot of texture, similar to a roaster like The Barn in Berlin.

POUR OVER NOTES:

V10: Lemon candy, white and purple florals, expressed orange peel, bright and clean sweetness but also with a dense body/sweetness that pairs together to remind us of panela sugar, milk chocolate undertones, a big syrupy mouthfeel with a lingering finish.

V13: Lemon Candy, white florals, clean sweetness like simple syrup, pink florals and subtle bright tropical pink fruits, vanilla, enhanced aromatics, a juicy mouthfeel with a slightly lingering finish.

Pursuit of Perfection

We decided to go a step further and explore both profiles in greater detail. We roasted the “Sey” profile with more development, hoping to bring out even more sweetness and texture. We also roasted “The Barn” profile with less development, hoping to bring out more florals and acidic clarity while hoping to retain the same or similar sweetness. These tests (V14-V19) were obviously to find a middle ground between these profiles. It was purely a pursuit of perfection.

Reality Sets In

After roasting these 6 new profiles, we stored them in plastic Ziplock baggies since we ran out of our old first edition pink bean bags. Unknowingly, these bags are not great for coffee preservation. We noticed that the coffees flavors aren’t nearly as vibrant and dimensional as they should be. Luckily, we use a pour over brewing method that can present acidic clarity, even in older or oxidized coffees. Our favorite roast out of the bunch, V18 was also luckily one of the few we stored a small sample in a mason jar. Once we brewed that coffee, we knew it was the one.

Curiosity In Storage Methods

 After that last storage hiccup effecting coffee quality, I read this article called “Investigations on the hot air roasting of coffee beans” by Stefan, Schenker published in 2000 as a Doctoral Thesis

submitted to the Swiss Federal Institute of Technology Zurich. It states, “High temperature roasted beans exhibit a greater bean volume, a larger cumulated pore volume and larger cell wall micropores than low temperature roasted coffee of identical degree of roast. These properties are assumed to increase the undesired mass transfer and to accelerate the staling process.”

This made me question the effects of roasters like UG Series vintage Probats or other roasters with thick cast iron drums. When roasted lightly, they impart a flavor profile that is both sweeter and more textured than you can get out of other roasters. We noticed our Aillio Bullet roaster also imparted the same flavor characteristics based on how we like to roast with it: always high heat and mainly change fan speed to control the roast in an ideally hotter and more stable environment on this electric roaster.

In order to test this theory, I thought it would be wise to store whole bean coffee in different environments that have different amounts of oxygen to see how staling effects our specific roast. I then thought it would be interesting to also pre-grind coffee right after roasting and store them in a container, ideally thinking the gasses from being ground would create an inert gas layer quickly (instead of over time like whole bean does, if much at all) which may help to preserve the coffee in an even more oxygen-reduced environment, leading to brighter acidic flavors with enhanced clarity and depth.

Roast 20 (V20) was our first test. We pre-ground 68g (for our batch brew recipe) right after roasting and stored those doses in glass containers. After 2 days of resting pre ground, we brewed one dose on batch brew and it seriously blew us away! The flavor clarity was immaculate and the notes were the most vibrant we had ever seen from the nearly 100 brews we’d done to explore this coffee thru out the profile roasting process.

For V20, we also stored equal portions of unground whole beans inside mason jars (large and small), and vacuum-sealed bags (using a straw to suck out oxygen instead of our full pressure vacuum). We divided these so we could freeze half the beans right after roasting and the other half to rest them. This way we can tell exactly what storage methods work best, and why. We also wanted to test all of these storage methods being ground right before brewing, and also being ground and stored 2 days before brewing.

After this test, we were sure we’d have a much clearer picture on what we could do to package the tastiest coffees possible to ensure you have the maximum flavor potential of your coffee!

Storage Method Results

Its amazing how you can do all of that work and have a test that doesn’t tell you much… We found our pre-ground coffees tasted INCREDIBLE 2 days after roast and 7 days after roast. All of the other storage methods used were lacking in acidity and vibrancy. We honestly got a lot of muted, chocolatey notes that weren’t very exciting. After going through all of the coffee by day 8, we found out that ONLY the pre-ground coffees were extracting nearly 1% higher. We compared that against some other roast that had been rested for 28 days and the extraction lined up with the pre-ground coffee.

Our loose theory is that the “muted” flavored cups are being effected by carbonic acid which is covering acidity. Also, the lower extraction combined with out “high extraction” batch brew method may have also helped a massive impact on our results.

The Final Profile

After that terrible test, for the past few weeks we have been roasting and brewing the absolute best cups we’ve ever had with this coffee…consistently!! Even brewing 4 days off roast espresso and 10 days off roast espresso are both tasting phenomenal! Maybe the green has finally opened up after being in its vacuum sealed state or our luck is as vibrant as these cups have been!

The final profile (V20-CJ), which is included below, is very similar to many others we have roasted. This goes to show visual color is not the most important thing, nor are end temps, “traceability” and ROR curves. I believe ambient temperature (and weather) may impact roasting more than what often shows on your temperature readings. The flavor of the cup is what, oddly but surely, truly matters most.