Development time as a ratio of roast time by Scott Rao - Page 10

Discuss roast levels and profiles for espresso, equipment for roasting coffee.
User avatar
[creative nickname]
Posts: 1832
Joined: 11 years ago

#91: Post by [creative nickname] »

I got turned on to this by Tom Chips, who posted a profile for this year's crop of the Yrgacheffe Aricha DP that blew my mind. I had enjoyed my more traditionally profile roasts of that green, but when I followed his lead and started slow, keeping a pretty flat-line profile and stopping in the midst of first crack, it produced a result that was just an avalanche of sweet fruit. I had thought it would have to be under-developed, but the slow start seemed to cure that. It's a tool I am happy to have in the toolbox for coffees where I'm willing to sacrifice some body and Maillard flavors in exchange for enhanced sweetness and acidity.
LMWDP #435

User avatar
TomC
Team HB
Posts: 10557
Joined: 13 years ago

#92: Post by TomC »

bigabeano wrote:Likewise, I've been in the roasteries of some very high-profile US roasters, where they had no idea that ALL of their bean data was massively inaccurate, beyond the usual bit of thermometric lag. For example, one successful, cultish roaster had a 6mm-wide probe in a UG22 that was bent so its tip was 1/2cm from the face plate of the machine and the probe was covered in a 2mm thick layer of creosote. Their first crack temperature reading was off by about 20 degrees F and they didn't know their probe was a problem.

I've seen this firsthand here in the Bay Area. The probe was so thick it looked more like a bent finger bent over back towards the faceplate.

But to the bigger purpose of large scale production roasting, it's becoming more and more like McDonalds. Train the person to respond to a flashing light and require as little individual control and input as possible. It leads to more consistency, which is something larger roasters are more bound to than hobbyists who like to play with coffee profiles and see how many ways they can stretch a green coffee. Personally, as long as the probe is consistent (only in these specific settings) I wouldn't care if it read 20° off, or 100°, more of the data points that large commercial roasters are going by are based on sensory perception anyway.

But, and to agree with the broader context of your point, a novice roaster, one who's learning what he or she can draw out from a particular coffee, not bound by constraints of following a production schedule and/or possibly someone else's profile, then that person should want not only precise and accurate thermometry, but also know how and why and where to place them. Without knowing the reasons why, you might as well just stick to organoleptic/sensory input alone.
Join us and support Artisan Roasting Software=https://artisan-scope.org/donate/

osanco
Posts: 121
Joined: 11 years ago

#93: Post by osanco »

[creative nickname] wrote:I got turned on to this by Tom Chips, who posted a profile for this year's crop of the Yrgacheffe Aricha DP that blew my mind. I had enjoyed my more traditionally profile roasts of that green, but when I followed his lead and started slow, keeping a pretty flat-line profile and stopping in the midst of first crack, it produced a result that was just an avalanche of sweet fruit. I had thought it would have to be under-developed, but the slow start seemed to cure that. It's a tool I am happy to have in the toolbox for coffees where I'm willing to sacrifice some body and Maillard flavors in exchange for enhanced sweetness and acidity.
I have a full bag of some wonderful Yirgecheffe from Cafe Imports. Since I have at least 100lbs to experiment with, I'll forgo my usual violence against innocent coffee, laissez-faire, throw it against the wall to see what sticks approach and actually try to accomplish a couple flat line and declining ROR profiles and see what happens.

All I have to do is find time to do nothing but roast for a couple of hours... :oops:

User avatar
bean2friends
Posts: 687
Joined: 14 years ago

#94: Post by bean2friends »

A slight tangent - that you guys are probably on to - still it's part of my DNA so I feel obligated to cite it. I spent most of my working career in the steel industry where the most highly paid operators - like the Blast Furnace Blower took great pride in the art of their profession. Every thing and everybody's success depended on the Blower knowing what to do and when - anticipate temperatures, adjust the wind, knowing just when to tap the furnace (read drop the charge). Then along came W. Edwards Deming, who helped the Japanese learn and implement real quality control. His lessons became the new norm in steel making. You must measure everything. Instead of feel, it became science and technology. Of course the Blower is still a responsible job, but now the Blower has tools to use to make his decisions. There is a record to refer to to see where things went right and things went wrong.

All of these discussions about coffee roasting remind me of blast furnaces. It's not a blessing or a curse, just a happy coincidence I think.

pngboy
Posts: 137
Joined: 12 years ago

#95: Post by pngboy »

If I look back @ my last couple years of roast logs on my quest much of my higher grown washed beans have a declining ROR and more of lower grown or dry processed have a flatter ROR profile.

osanco
Posts: 121
Joined: 11 years ago

#96: Post by osanco »

Just for the sake of sharing: I have continued to experiment with declining ROR roasts and I think I may be getting the hang of it.

To make this work, I have had to lower my drop temp (on my roaster) to 325F. I'm turning my fan off and keeping my burner lit at the minimum (TJ-067) .5 kpa. After the turning point, I'm running the fan at a minimum (again, on a TJ-067 to about 30%) and bumping up my heat to a ROR of around 30F per minute.

If you think of the math, it starts to make sense. About 1:30 to the turn at 175F. About 3:30 at 30F per minute = about a 5 minute drying time. (Sorry, Scott :D )

From there, I'm reducing the gas as necessary to hit first crack at about 385F at 10:15 and exit first crack at a ROR of about 10F per minute.

One thing I find interesting, after Scott pointed it out, is seeing the slight cooling associated with the onset of first crack. I never noticed that before, but focusing on ROR it's clear and I can anticipate and compensate for it.

For the record, I'm logging with a 1048 Phidget and have replaced the stock BT logging probe with a custom built, very fast bead type thermocouple.

Again, doing the math: 385F + 10F per minute for 3-3.5 minutes yields a (TJ-067) Full city or full city +.

I'm roasting 1 kg charges on a gas roaster with a ton of headroom. YMMV

What's different about this for me is generally sweeter, more aromatic coffee.

I'm curious about how this might work with an electric roaster. Sometime this fall, I'll compare the electric and gas side by side to see what happens.

I'd like to confess again, that I'm not a particularly serious roaster. All of this may be old hat to some, but this method is producing very new results for me and the process may be teaching me something.

Old dogs, new tricks etc.

compellingrich
Posts: 60
Joined: 10 years ago

#97: Post by compellingrich »

I could talk about this post all day. Osanco I have an electric roaster (a fluid bed, furthermore), and depending on the density/processing my drying stage is usually around 5 mins. (I basically exaggerate a classic Diedrich profile.) I increase heat from yellow to 3/4-through 1st-C, then decrease/cut off entirely based on where I'm looking to end up. And fwiw my "most-successful" roasts (strongest feedback, personal preference, highest Coffee Review ratings) have all been 15-minutes plus. Although I do Rao-profile certain origins/densities/varietals, I'll debate low-and-slow roasting for others alllll day :)

osanco
Posts: 121
Joined: 11 years ago

#98: Post by osanco »

Sure, I get that and have done much the same, although probably not as well, with both gas fluid bed and electric drum roasters.

That long slow roast probably maximises sweetness, minimizes acidity, and increases solubility and body. All good stuff.

I'm also guessing that for many beans, this style potentially reduces aromatics, which I suspect Scott especially prizes.

Different strokes, etc.

I did another declining ROR, 1 kg roast today with a nice Tanzanian Peaberry. I actually got a nice crack at first crack (a first with this bean and no pun intended) and dropped slightly closer to Scott's ratio.

So far, cupping these roasts has yielded some of what I have been paying a couple bucks a pound extra for, but missing, all these years. I think I'm getting some of the floral and aromatics and nuance that I usually only read about in the cupping notes.

I'm not making any claims or arguments here. I'm only reporting what seems to be working for me.

dogjamboree
Posts: 258
Joined: 13 years ago

#99: Post by dogjamboree »

Thanks for your post Steve -- much appreciated. I've been meaning to try this experiment myself but have been either out of town, under the weather, or over the moon (ok, made that last one up). :)

User avatar
boar_d_laze
Posts: 2058
Joined: 17 years ago

#100: Post by boar_d_laze »

osanco wrote:That long slow roast probably maximises sweetness, minimizes acidity, and increases solubility and body.... I'm also guessing that for many beans, this style potentially reduces aromatics, which I suspect Scott especially prizes.
Yes. However, the loss of acid and floral notes, while noticeable compared to roasting to the sort of profile Rao is promoting, is not huge.

At least it isn't with low air-flow roasts in my Dalian Amazon.
I did another declining ROR, 1 kg roast today with a nice Tanzanian Peaberry. I actually got a nice crack at first crack (a first with this bean and no pun intended) and dropped slightly closer to Scott's ratio.
Why "slightly closer?" Why not try the scheme Scott's talking about?
So far, cupping these roasts has yielded some of what I have been paying a couple bucks a pound extra for, but missing, all these years. I think I'm getting some of the floral and aromatics and nuance that I usually only read about in the cupping notes.
If you think of Drying and Ramp as separate intervals you may find it easier to control the roast.

For instance, you can go low power and air through Drying, and increase both for Ramp, maintain a declining RoR, and hit 1stCs in the mid nines to low tens. That's a good elapsed time to enter 1stC because it allows you to hit the desired finish -- which is really just the range of C+ -- in a time which comports with the 20% - 25% range.

My experience is that by manipulating Drying and Ramp differently, I can balance the amounts of fruits and florals in the roast against the time they'll stay stable as the roast ages. Dry too fast and Ramp too slow, and their intensity is maximized, but they fade quickly. Dry too slow and Ramp too fast, and the varietal characteristics are comparatively flattened but the roast will retain its distinctive character for a long time as it ages.

Also, it might help to think of Development as two -- or even three -- discreet intervals. In order to maintain control of Development you have to allow for momentum and exothermia, and lower your heat and increase air in anticipation of 1stCs. You can't wait for the first snaps, you have to be at least 10F in front of them -- I typically go with 15F - 20F.

If you aren't pumping too many therms into the drum and are using enough air, when exothermia ends -- about half way through rolling first -- you'll see an inflection on the plot as RoR drops. You'll have to stay on top of the gas (or power) to make sure it doesn't drop too much. You're probably looking for an RoR of 6F - 10F from a few seconds after the inflection all the way through the rest of Development to Drop.

It's the entire period going from anticipating 1stCs all the way through to Drop where real time BT plotting really pays off.

If you want to try "one profile to rule them all" which will work with a roaster the size of yours, and follow the general tenets of the style Scott wrote about, try:

Charge at 300F, very low gas, low air; 6:00 (total) Drying (to 300F), after TP, use enough air to maintain negative pressure in the drum and enough gas to hit your time; 3:45 Ramp (to 1stCs), with slightly more air; 3:15 Development (to C+, about 10F shy of 2dCs), with quite a bit of air as a starting profile, with an eye to what I wrote about anticipating 1stCs, and tweak from there.

The real power of that profile is not its universal perfection, but that it's a plan which fits the numbers and which is easily adjusted. You can go from a 13min ET to 11:30 or 14:30 very easily. If you want a 25% (or any other desired percentage) Development ratio, it's easy enough to plan and execute. There's no reason to dance around with "almost."

Here are three more maxims to go along with Scott's Commandments:
  • I. It's better to use your roaster to cooperate with bean chemistry than fight it.
    II. Plot your ideal profile before roasting, and then execute it.
    III. Roast on purpose.
Rich
Drop a nickel in the pot Joe. Takin' it slow. Waiter, waiter, percolator