Now that things have cooled down. . . .

another_jim wrote:So far we have:

.
.

• Diagnostics How to identify and fix roasting problems.
  .

could be the best bang for everybody's buck. . . .that is, if we have to pick only one, or prioritize the list.

Jim,

I hope the article is still in the works, because i refer often to the front page of this thread. I've really only had a roaster capable of profiling for the last half year, after i instrumented my Hottop. In that brief time, I've begun to ask questions about (and could use more rumination/illumination on) the following topics:

1. bean specifics - i don't know if this needs to be broken down by geographic region, or by bean characteristics (water percentage, hardness, etc) or both. Obviously this would focus on how different beans roast differently.

2. roast stage specifics - specific sections of the roast that incur common flaws or are more important than other sections. Kinda like "drying/first ramp/second ramp" but with more detail. For example, my wondo bonko roasts go very "exothermic" post first crack, while my la maravila seem to almost go endothermic. Is the first or second part of the first ramp more critical? and exactly how flat can I make the second ramp without stalling the roast?

Maybe what i'm looking for is the connection between the two topics. How, and more importantly why, does each coffee roast differently? If there was one "golden" profile that fit every coffee (and possibly every machine) roasting coffee would almost be trivial, once we got the proper instrumentation to repeat that profile. Instead, it's defining the "best" for a bean that may change even as it sits in storage for 3 months.

I'm not Jim, nor do I look/sound/talk like him, but I've been struggling with some of the same questions.

#1 : I noticed a consistent variation between my stash of Sidamo and a recently procured Sulawesi. Not only was the Sulawesi a larger bean, but upon cutting open the green bean, it also appeared to be less dense. I have no idea whether this is related to growing altitude, varietal, or some combination of earth, wind and fire, but the results in my roaster speak for themselves. At the same given load (340g), same charge temperature (400F), the Sidamo consistently requires me to use higher heat in order to reach the "dry/yellow" stage by a set time (4:00), whereas the Sulawesi uses less heat to reach the same stage at approximately the same time.

Now, my roasting device may not be as tweakable AND measurement-friendly, a la Ed's "DreamRoast", so there could be several reasons why this is happening, but I know my roaster and the stash of Sidamo well enough to recognize that, when minimizing variables to the best of my abilities, it's consistent. I noticed a similar issue with an order of Costa Rican beans, but at the time I didn't think much of it. I've got about 700g left, so it may warrant another few tests.

#2 : In poring over H-B for roasting info, I came across enough posts from members whose opinions are well-regarded that I started putting a crib sheet together. The closest that I've found to "specific" information regarding common flaws was condensed into the following:

- grassy vs dusty/leathery (too short/long drying phase)
- bready/cheerios vs. no toasty flavors (short/long ramp to 1C)
- too acidic vs. flat/caramelly (short/long finishing ramp)

Can you explain what you mean by endothermic vs. exothermic, as it relates to the wonko/la maravila? I have my own understanding of those terms, and want to make sure that we're on the same page.

Also, the Roasting 101s: Understanding Degrees of Roast has some good information about what happens during 1C:

another_jim wrote:After the halfway point of the first crack, the cellulose cell walls of the beans enter a "glass phase," in other words, they become a gooey liquid that expands as the remaining gasses inside the cells expand. If your roaster cannot rech that temerpature, then you end up with very hard, small beans.

another_jim wrote:Bean expansion happens when the cellulose is above the temperature where it goes soft, roughly 410F, and below the temperature where it starts breaking down, roughly 430F. This is independent of baking the beans at lower temperatures, or scorching and tipping them with higher environmental temperatures.

good luck,
-s.

There's two issues, one technical, one taste based

The technical one is the changing thermal properties of the bean at the first crack, when they lose their water. I never really appreciated how dramatically this affects drum roasts. In an air roaster, all the heat is lost and replaced from moment to moment, so if the bean's thermal properties change, its very easy to adjust. Drum roasters store heat, and people have talked about the beans going exothermic or running away at the first crack. I suspect the mechanism is that the beans release superheated steam at the first crack, i.e. a substantial amount of heat. In an air roaster, it vents; in a drum, it accumulates. Whether the steam release is "exothermic" depends on the quibbles in the definition: is a boiler explosion exothermic or just the release of stored heat? Every bean is a little exploding boiler at first crack.

In any case, my new and improved roast simulations estimate a 10C to 30C rise in the temperature surrounding the beans after the first crack. The amount depends on the bean being roasted and how long it took to get the first crack (i.e how much they had been dried before).

In practice, this means what every drum roaster has known for ages, you need to drop the heat when the first crack comes, or the roast runs away from you. Moreover, you need to drop the environmental temperature too, since the drum measurement usually doesn't register this rise.

In terms of taste, it's mainly a process of learning what flavors develop in what part of the roast. This is roughly the same for most beans, so changing profile for a given bean is based on deciding which of its flavors you want to bring out. For instance, an Yrgacheffe which has boring green tea roast flavors and wonderful fruit and flower origin flavors would get a roast that snubs the black tea (by a fast ramp to the first) and a that enhances the fruit (by slowing down during the FC, smelling the roast, and ending it the instant all the grassy aromas die down, or the caramel smell for a light espresso roast first appears). A Bourbon or Sumatra, on the other hand, will get a slower ramp to the first, so they can develop all those complex malty and oaky flavors, and then will be finished dark and fast enough to have rich caramel aromas while retaining a bit of acidity.

This is how I usually roast these different beans; but one can craft completely different roasts from the same origins when the bean is right, and you are feeling inspired.

So ultimately, profiling for taste comes down to cupping the bean, and then picking a profile that lets it shine. A good profile is more an interpretation of the bean than a scientific discovery; good roasting technique is must, but to get really good at this requires lots of experience. This is why the top pros, who cup all their coffees, do roasts that consistently outdo everyone else.

thanks Jim,

That goes with what i expected: once you get to a system that lets you profile, there's no substitute for experience. And i feel like i'm standing at that point now: it's neat how much it's really opening up. For the first time i actually need five pounds of a coffee i like so that i can work through 1-2 pounds developing the profile and then enjoy the roasts for a while. The problem is that i want to try so many coffees that sometimes i wonder if I don't give them a proper chance. Although the exceptional coffees seem to stand out early - even if the profile is not yet perfect, the potential is there, and you can spot it by the second or third roast. But roasting coffee is no longer about "set time, pick degree of roast, drink." It was a lot of fun developing my profile for la maravila, and really getting to know that coffee. And I barely even know it! I'm sure there are so many further tweaks to the profile that would bring out different aspects of the coffee, and i've just found one profile of many. I can't wait for next years crop.

Hi all,

I'm new here and new to roasting. If I can chime in, what would really help me at this point is an extremely basic parameter for each stage (drying, ramp to first crack, ramp after first crack) or at least an idea of what NOT to do (I apologize if this is already here somewhere). I realize it's impossible to make specific recommendations across various machines and bean types, but it seems like there are probably some absolutes. For example, the drying stage generally takes place below 300F (bean temp) and should take more than 30 seconds and less than 10 minutes. Most beans for light roasts stay in the drying range between 3-5 minutes.

I don't even know if any of that information is correct, it's just an example.
For me, the challenge so far has been to just make something that could be considered coffee. My early roasts have frequently alternated between very light brown, super tannic beans and ash, with the occasional success here and there.

Cheers

The best timing depends on what sort of roaster you use, do post on that. Drying periods are vital when you do light roasts using fast ramps to the the first crack; less so otherwise, since the beans have more of a chance to dry later in the roast.

I'm not sure I always accept the exo/endo-thermic terminology folks use, which refers to heat created/absorbed during a chemical reaction. It makes more sense to me if I think in terms of bean temperature increasing faster during first crack (assuming constant heat input.....) because I think there is less water vapor being created. Creating water vapor (boiling) takes far more energy than increasing the temperature of the vapor itself or the bean. So if one were to postulate that during the drying phase and approach to first crack that the energy input was heating up the bean and vaporizing water, you would expect that the bean temperature would begin increasing more quickly once all the water was vaporized. This rate of temperature increase then slows when the heat energy is absorbed while volitilizing other compounds in the bean at the end of first crack and into second crack.

After second crack, once the beans catch fire, you do indeed have an exothermic reaction!

P.S. - the thoughts above are not scientifically documented. I would be delighted to discover that there are indeed exothermic reactions taking place that primarily account for the temperature drop around first crack.

Wouldn't the release of vapor be considered exothermic?

As you stated the energy is stored in to heat moisture and when the vapor is released, the energy is released as well. This is an exothermic reaction. Is it not?

another_jim wrote:The best timing depends on what sort of roaster you use, do post on that. Drying periods are vital when you do light roasts using fast ramps to the the first crack; less so otherwise, since the beans have more of a chance to dry later in the roast.

I am using a home built air roaster with a heat gun element and I generally roast for espresso (ideally), if that helps at all.

Could I generalize and say that for light roasts, one would conduct an extended drying phase, then a fast ramp up to first crack, and then... stop? And for dark roasts one might have a steady climb to first crack, and then a slow ramp to either second crack or just before? Would you recommend shooting for first crack at a certain time for dark or light roasts?