Roast and Learn Together - March/April 2015

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
Team HB

#1: Post by TomC »

Some exciting news for March. Marshall Hance has offered to not only lead the discussion (on a stellar washed Ethiopian Gedeo Kochore) he's also going to help us source it thru his Shrub account. He's passing on the savings he gets as a bulk buyer to us, and we are getting access to a green coffee from a source that might otherwise alude us. So a big thank you to Marshall for that.

There is one important stipulations to note: this coffee is in short supply. He only has under 40 pounds available to us (37.5 to be exact). He has created a link where we can buy the green coffee from him, he's already set up for shipping and distro, so the prices are quite good and he's not profiting on this green distribution. Participants for March are limited to no more than 5 pounds of green. Ideally, for the hobbyists who have Quests, Huky's, Hottop's, they'd buy a limit of 3 or less pounds. Folks with USRC's and Mini500's, if they need a bit more than 3 pounds, so be it. The coffee will sell out quickly, so the participation this March will be limited to about 8 or so members.

I'm really looking forward to this topic. I've already tried it, and it's very, very good. Very intense florals and wonderful developed sugary sweetness.

Here's the link to buy the green. Note, Marshall will not process any orders asking for more than 5 pounds.

* Link removed, greens sold out.
Join us and support Artisan Roasting Software=

User avatar

#2: Post by endlesscycles »

How's it going, y'all? For those who have already purchased greens, thank you. I'll be receiving them via UPS 1ZRV98360391713917 Thursday, and will repack and send out either that day or Friday. You'll receive USPS tracking info at that time.

A few guidelines, as per the usual each month.
1) Clearly state charge weight and roaster used as well as the roast date on the first sentence or in the graphic.
2) Try to update one personal post, meaning if your profile graphic is shared in a post and you're commenting on your impression of that particular roast profile, then the cupping/brewing notes belong in that post along with a date, not a scattered new one that makes the thread hard to read and follow.


Yirga Cheffe's are among my favorite coffees, and I've been having great luck explicitly following Rao's suggestions when roasting them. I'd like to encourage those who haven't tried his roasting philosophy (fully, not partially) to do so side by side with your typical approach to such a coffee. Doing so might require throwing everything you "know" about roasting out the window, since he's pretty explicit in what to do every step of the way.

To summarize:
Start with high heat. Assuming you can shed heat down to reasonable roasting temps in just a few minutes, go ahead and max it all the way out.
Continue with high heat for the first 3 minutes. As long as you have good agitation and airflow, this won't be a problem.
Use heat to control the roast vs. air and heat. I suggest "medium" airflow throughout, whatever that means to you.
Only decrease heat input through the roast, never raise it. Going gently and only as needed (see below) tends to work out best.
Never allow the ROR to rise or go flat, only decrease. No exceptions. Going flat for even 30sec should be avoided. 1min will show defects.
From the start of 1C, add 1/4 to 1/3 of preceding time to drop for 20-25% RDT. I think shooting for 22.5% is probably best to start.

It's my experience that looking at the temperature and clock isn't as helpful as looking at the ROR curve, since that's the central idea to the philosophy. I wouldn't worry about final temperature much, either. Done right a 25% RDT roast within normal roasting times is between very and over developed, if not baked. Following the rules, it's unusual to get above 410F/210C with a 392F/200C 1C. Don't worry about it.
-Marshall Hance
Asheville, NC

User avatar

#3: Post by Chert »

Thanks for leading the charge, Marshall and for the anticipatory post. For initial roasting in our new cooperative roaster this will be a treat. We have some grade 2 Yirgacheffe from Atlas importers that I will follow with the same effort as you outline while awaiting the package from you. I think that will be a great comparison for a re-newbie like me.
LMWDP #198


#4: Post by Roasty »

I'm pumped for this, sounds great and will definitely follow that game plan and compare it to my usual.

I think I just bought the last pound btw.

User avatar
TomC (original poster)
Team HB

#5: Post by TomC (original poster) »

Good coffee sells fast. Fortunately, Mill City Roasters happened to have the same green, so more folks can participate. Check out their listing in the Marketplace section.
Join us and support Artisan Roasting Software=

User avatar

#6: Post by NoStream »

I roasted up my first two batches last night. (Thanks again to Dave / Mill City for the sourcing the greens.)

I'm really pleased to be working with a washed Ethiopian coffee. They're my very favorites, low in supply at this time of year, and I struggle to retain their floral aromatics in my roasts. (Does anyone have suggestions for getting these aromatics to be prominent?)

I attempted to Marshall / Endlesscycles' suggestions. I'm using moderate airflow throughout dry and ramp (5 on an early fan Quest) and then maxing out in development, charging really hot (>480 F BT on my Quest), using a good bit of heat early, and hitting ~20% development time. (I shoot for 1:45 for my roaster.) I'm dropping these at 402-406 BT, which is a true city roast, with the last snaps of 1c occurring after drop.

Here we are...

Rao-sts 1-2, 3/9/2015, 12.27 & 12.53% WL respectively

Cupping, two-days post, 3/12 1:18, 1.3 TDS, 20% ext.

The first profile above was quite nice, sweet, tartaric acidity, a bit of jasmine (though it always shows up more on the cupping table than in brewed coffee, in my experience). The second was dramatically worse. There was a hollow, baked, and savory quality. I think there are two suspects - one, slightly flattened RoR ("plateauing RoR") and, two, an excessively long Maillard phase in an attempt to achieve declining RoR and possibly develop more florals.

V60, 3/14/2015, five days post.

I brewed up the first profile above. 1:16.5, 202 F, ~1.4 TDS, ~21% ext.

This was really nice - very sweet, with prominent tartaric and citric acidity. There was a bit of a bitter cacao finish and a mild astringency, which indicates to me that I went slightly overboard with the hot charge temp. I'll try 490 F MET rather than 500 F. The roast retained the florals. I brewed it for half a dozen people, all of whom enjoyed it. A friend of mine whose palate I trust referred to it as "explosively floral."

Two things I've learned that can inhibit florals in my Quest are excessive airflow and baking. I think I used to intuitively imagine that aggressive heat application would decrease florality, but that doesn't really seem to be the case. Rather, over-aggressive heat application early just resulted in a slight bitter edge.


I also roasted three more batches as an airflow experiment. Profiles were the same as above, but with different airflow so I won't post profiles. (I also accidentally overwrote the best one. Ugh.)
My Quest has the older style fan which allows lower airflow settings. I tried 3-3-max, 4-4-max, and 5-5-max for dry-ramp-development respectively. Although minor differences in development RoR wound up confounding the results a bit, I did find that 5-5-max traded off a bit of aroma intensity, especially florals, for a slight increase in sweetness. I preferred the slightly higher complexity and aroma of 4-4-max to the slightly higher sweetness of 5-5-max.
I remain unconvinced that flat RoR in dry or early-mid ramp (pre-caramelization) has any negative effect on cup quality.


Here I charged slightly cooler to avoid scorching (but still really hot), coasted a bit with the Quest's minimal thermal mass, moved into ramp pretty aggressively, but still long enough to generate aromatics, and dropped with 1:45 of development, 20% or just less than that. Moderate airflow through dry and ramp seems to offer the best results.

Two profiles...

Airflow at 4 throughout dry and ramp, gradual increase to maximum, only maxing out at the end of development.

Same early airflow, faster maxing out of airflow in development. I am planning on implementing at least some airflow and amperage tracking for a better dataset.

Tasting, one day post, a few V60-02s 12:200 and 20:330, 201 F, 7.2 on the EK, with other palates

Tasted again three days post on a v60-01, 12:210, 201 F, 7.2 EK, 1.29 TDS, ~20.7% ext.

The first was cleaner, more honeyed, and more preferred. The second has a slight bitter, roasty edge. The flavors are relatively similar, but there's a chemical bitterness. Both are quite sweet. My best guess is that the slightly more aggressive MET charge in the second profile was at fault. Another hypothesis would be that the ramp got too hot, but that seems less likely as I've taken ramp up to 520 F before without any roastiness. Any other ideas?

Another profile, 3/20
I'm now using the events feature in Artisan, though events aren't always marked in 1c because there's a whole lot to think about.

From V60-02's 3/21, 201-202 F, 7.0 EK
Super sweet citrus, tropical fruit (mango), floral, milk chocolate
Thoughts from others - tangerine, strawberry, white grape, really sweet, one of my best roasts, etc.

V60-01, 3/22
I'm sipping on it right now as a 12:210, 204 F, 7.4 EK V60-01, and I think it's even better. Going a couple degrees hotter really brings out the sweetness even more.

Espresso, EK43, 1.4 grind, 18.3 dose in a VST 18, 56.8 g out, 22 sec. 200f. 7 bars.
crisp, sweet, super ripe; bright; insane condensed jasmine sweet tea florality; a really great shot

This roast was pretty spot-on. Next roast will be similar, but I'll try applying more early heat instead of in ramp and see what happens. I did achieve very good development with as much / more heat toward the middle of the roast.


#7: Post by marista »

NoStream, i've been delighted to get my hands on some washed ethiopian aricha greens and was frankly delighted to smell florals on the tryer for the first time since i started roasting. :D i know this is not the same green as discussed, but this was my profile for achieving the florals. i'm also using a quest m3, and this was a 100g batch. weight loss for this batch was 13.3%.

my hypothesis is that you're really looking for only one sweet spot where finish temp and development time coincide. meaning to say, i think certain taste characteristics you're looking for might only exist at a certain finishing temperature.. provided it happens at the ideal amount of development time.

edit: do note that artisan did some funny marking on my graph, drying phase ended at 5:00 not 6:42 as the green bar says


#8: Post by rgrosz »

marista wrote:edit: do note that artisan did some funny marking on my graph, drying phase ended at 5:00 not 6:42 as the green bar says
Problem is that artisan marked your turning point at 6:42, which is clearly wrong.
LMWDP #556
Life is too short to drink bad wine - or bad coffee

User avatar

#9: Post by Andy »

Modified Poppery
150g charge

Received Gedeo in yesterday's mail from Mountain Air. Graphs of my first 2 roasts are below. Both are roasted to about 420F in 11:50. I plan to brew Saturday morning and add notes to this post.

#1: 3/5/3, end weight 89.3%

#2: 4/4/3, end weight 87.3%

Edit: Sat., 14 March. I just finished cupping these 2 roasts, first by standard cupping, then with Melitta pourover.
Fragrance: Some kind of sweet fruit candy dominates; maybe orange lifesavers.
Aroma: Mostly cardamom-like spice; also generic white-sugar sweetness, but no more lifesavers.
Palate: Black tea and lemon, soft acidity. No significant change after cooling 5 minutes.

These 2 batches were very similar, in fact I could not distinguish between them. This is lovely coffee IMO. It has all the best of what wet processed Yrga Cheffe has to offer: sweetness, delicate aroma of citrus and black tea, but not an acid bomb. Next I will try a very different profile just to see what happens, but I will be surprised if I'm able to improve on this.


#10: Post by 9Sbeans »

Just received 4 pounds of beans from Marshall, and immediately fired up three 227-g batches last night. All three batches were aiming for the "Rao" style (don't have better terminology, suggestions welcome). To simplify the test, I didn't play with the ventilation fan speed adjustments. The fan was set at my default 30%, and increased to 35% at the end of Drying. (For my batch #3, I had to reduce the fan back to 30% at the end of first crack to prevent stalling.)

To briefly explain my notations, my roaster can supply 0 ~ 600 mmAq of gas (propane) pressure, and I divide the gas pressure by 5 to get a % reading, 500 mmAq (100%) as the maximum. For example, P16 (16%) means 80 mmAq, and P40 (40%) means 200 mmAq. Since each roaster is controlled differently, the absolute value on one machine is meaningless for others. Therefore, these should be considered in a relative (increasing/decreasing) fashion. I break the overall roasting process into several segments, and use +/- to mark the timing of the gas adjustments.

After charging, there are:
1. Turning point
2. End of drying phase, start of ramp, RoR slows down
3. Later ramp phase, flat RoR
4. First crack start, RoR plunge
5. First crack end, RoR rebound

"+" means increasing the gas pressure from the previous segments, and "-" means the opposite. For example, "+ + - - -" means I increased the gas pressure at TP, increased gas again at the End of Drying, and subsequently decreased gas at the later Ramp phase, First crack, and after the end of the First crack. "303F Rao ++--- 411F" means charge @ 303F, declining RoR, drop @ 411F. Complicated? Yes.


Roasting Info:

Bean: Ethiopia Yirga Cheffe Gedeo, Batch #1, 303F Rao ++--- 411F
Roaster: KapoK 500 sample roaster
Charge Mass: 227.0 g
Charge Temp: 302.9F
Dry/Ramp/Development: 4:41/3:24/2:15
FC-start temp: 376.5F
Finish Temp: 410.9F
Overall Roast Time: 10:21
Weight Loss: 13.74%

The dual-wall drum of my roaster has huge thermo mass, and in the first 3-min of the roasting it transfers its heat to the beans. Charging at a higher temperature, the BT will rise faster until the temperature difference between beans and drum reduces. After the fast energy transfer phase, the applied gas plays major role as the heat supplier. I gently increased gas twice at the TP and the 3-min mark without producing increased RoR. It entered first crack with momentum (33% power), and in theory I can let it run to even higher drop temperature while obeying Rao's rule. This is a proof of concept of "charge at lower temperature to reach a higher drop temperature".


Roasting Info:

Bean: Ethiopia Yirga Cheffe Gedeo, Batch #2, 360F Rao +---- 400F
Roaster: KapoK 500 sample roaster
Charge Mass: 227.1 g
Charge Temp: 359.6F
Dry/Ramp/Development: 3:32/3:36/2:06
FC-start temp: 381.9F
Finish Temp: 399.9F
Overall Roast Time: 9:15
Weight Loss: 13.17%

Charge at higher temperature, the Drying phase is short. I kept reducing heat supply in this profile. I got decent results from various beans by this roasting approach before, and hopefully it will be good this time.


Roasting Info:

Bean: Ethiopia Yirga Cheffe Gedeo, Batch #3, 297F Rao +--+- 405F
Roaster: KapoK 500 sample roaster
Charge Mass: 227.0 g
Charge Temp: 297.3F
Dry/Ramp/Development: 4:51/4:12/2:18
FC-start temp: 382.8F
Finish Temp: 404.6F
Overall Roast Time: 11:23
Weight Loss: 13.83%

Keep reducing heat is a simple approach to Rao's roasting style. However, if charging beans at lower temperature and kept reducing heat, it won't have enough momentum entering the first crack. Although both #2 & #3 are closely resembled to Rao's style, I constantly got better results from #2 than #3. In this batch #3, I briefly increased heat at the first crack to counter the expected RoR plunge. Not necessary in this Ethiopia beans.

If I stick my development time to 20~25% of overall roasting time, I would most likely end up with City+ roasting levels.


3/23 updated:

Ethiopia Yirgacheffe is one of my favorites. In fact, I grabbed 20# of Gedeo region Yirga from SM last year while it was available, and there are still about 10# remaining in my stash. Those could be old crop though, and hence I got additional 4# from Marshall for R&LT this month. I'm familiar with the Rao style, and regularly employ two roasting approaches as shown in my batches #2 (high charge temp) and #3 (low charge temp). Usually, when I open the container after several days rest, I can immediately tell the big, in-the-face flagrance and know that I get it. And this is the case of my batch #2. Exactly as Marshall has advocated for sometimes, charge high temperature, reduce heat, short overall roasting time, and Voila. It's already my SOP.

My batch #3 approaches (charging at lower temperature to get a longer drying phase and subsequently reducing heat) have never been as good as the #2 results. Indeed, when I try to compensate the RoR dip post-1stCrack, I may over react. When brewed by Aeropress, I can taste spicy pepper when it is hot in my Batch #3. The spicy taste is not presented in my other 5 batches, and I would say it's my roast defects. I'll abandon this profile in the future.

My biggest surprise is that, my batch #1 is really wonderful. It has the same in-the-face flagrance and hold up well. When making SO espresso, I like both my Batch #1 & #2, and my batch #1 was even better. The basal floral and flavors are the same, but #1 has fuller body and much more complexity. I don't know what makes the difference: simply the roasting degree (higher drop temperature), better development (13.7% vs. 13.2% weight loss), or everything in the roasting profiles counts. Usually when I employ my batch #2 approach, I can easily get superb results, but at a minor cost of faster fading. Yet my batch #1 holds up better than my batch #2. Fascinating.

I will apply my #1 profile on other origins and know if it's one times luck. :D