Drying ramp length; should beans still be green?

[DavidMLewis]

I remember someone here saying that one should extend the drying ramp, i.e. the section from 225 F to 295 F, so that the beans are tan rather than green at the end of it. In the computer-controlled Hottop, if I drop in at a relatively high temperature (420 F air), then hold the element on until I get to a bean temp of 208 F, which carries it quickly up to 225 F, it takes much longer than the usually recommended three minutes 225-295 to get the beans not to be a bright green at that point. To what does that color change correlate, and is it in fact important? The reason I ask is that I've always been puzzled by Jim Schulman's insistence that the ramp from 295 to 380 should be as quick as possible unless you were trying to accentuate woody flavors; in my roasts if I run that too fast I don't get any flavor at all. I was beginning to wonder if I should be slowing down the drying section instead.

Best,
David

[another_jim]

I may need to do a retraction on this.

What I say applies to air roasters, and to a good extent to well ventilated drum roasters, at least according to the pros I talked to. However, the rules may be quite different for the smaller, mostly unventilated drum roasters like the Hottop.

I was visiting Ken Fox, who uses a Burns knock off 1 pound sample roaster that is very similar in configuration to the Hottop and most BBQ drums (no airflow, perforated drum, direct heat underneath). His roast have improved quite dramatically in the last year. He now runs a very fast but even ramp from his drop in to the first crack, taking around 8 to 9 minutes (basically lower drop in but a lot more heat compared to before). The roast finishes in another 3 to 4. However, the original Hottop profile had a flat section early, and took around 5 to 6 minutes to hit 300F bean temp, and only then went to maximum heat and a full speed ramp to the first crack.

So, as far as I can tell, people's practices here aren't quite adding up to a consistent set of guidelines.

[cannonfodder]

With my HotTop, I let the roaster heat beyond the 'ready to go' beep. I drop my roasts at 200-250F depending on the volume I am roasting.

[Ken Fox]

Beans have to finish their roasting more or less evenly done. Exactly how you get to that point is going to depend upon your roaster and how it transfers heat to the beans. Being at an even point (among the beans) when you begin first crack is an intellectually satisfying idea that is probably not possible for many or even most roasters. The key is to finish with the beans at about the same level of roast throughout the bunch of beans you are roasting.

Exactly how you get to this point is going to be entirely roaster dependent. I can't think of a way of arriving at this point consistently without trying a whole bunch of different things, including roaster charge weights, the temperature of the roaster when beans are introduced, and how you apply heat during the roasting process. It is very helpful to have some way of monitoring your roast temperatures, in a way that is repeatable. I am talking about thermometry here.

Improvements in your roasting results are apt to occur in stairstep fashion, e.g. you will progress incrementally over a period of time. The only way to know how any given modification effects your results is to taste them, then try to repeat what you thought you did and to taste again. That is how you get better at this.

IF you want to get to be good at this, it is going to take a lot of time. You don't have to be the best roaster out there in order to have decent results. You will have to decide how much effort and expense you want to put into this process, and that will determine to a great extent how successful you will be.

ken

[mrgnomer]

I've got a Hottop programmable and was wondering about the drying ramps too. Currently I'm roasting with profiles that are closer to the iRoast2 I used to use: steep steady ramp for about 8-10 min to 1st crack, level off for about a 2 min rest after 1st and ease into touching 2nd. I start off with sort of a drying stage for the first two segments programmed to max temp and about 2min 30sec each but after that the segments are 1 min for max temp up to the 7th. I'm more interested, though, in keeping the heating element at max up to the 6th segment. The roasts are bright, I find. Good if you like lighter roasts that preserve a bean's varietal character, I think.

[Jasonian]

So much of this is dependent on a few factors.

The freshness of the green coffee crop (i.e. - has it stabilized at 10-12% moisture yet?).

The origin of the green coffee crop.

The altitude the coffee was grown.

And then you get to the equipment being used.

For higher density coffees, I like to stretch out the drying stage for quite a long time. 5 minutes or more in some situations.

For lower density coffees, it really depends on the origin. Brasils tend to like a very slow even ramp all the way from start to finish.

It helps to bee keen about the point in the roast where the coffee ceases to be endothermic, and begins to exhibit exothermic properties.

By "slow even ramp" I don't mean of supplied BTU's, but of overall bean mass temperature.

It takes a light touch.

Can you be more specific about the coffee you're using, and what flavors you're experiencing in the cup?

[popeye]

I've got an iroast2 and 118-121v here in downtown san diego, and i find that heat transfer occurs way too quickly. I think my iroast happens to be on the "supercharged" side of the quality control spectrum too. Anyway, it's helped me realize that temperature is only one variable. And it helps if i look at temperature through heat flow (some sort of hazy integral relationship comes to mind), because it's really the heat the beans receive that raises their temperature. In my case, this helps to pinpoint airflow as the other major variable to work with. so i play with temp and airflow.

I say all this because any temperature ramp you do is also dependent on the heat transfer characteristics of your roaster. I think this is pointed out best by the air/drum roaster split. And it makes me wonder how good of an approximation of internal bean temps our external bean measurements are.

(disclaimer: the above contains definite errors, as i am definitely not an engineer).

[TimEggers]

I was visiting Ken Fox, who uses a Burns knock off 1 pound sample roaster that is very similar in configuration to the Hottop and most BBQ drums (no airflow, perforated drum, direct heat underneath). His roast have improved quite dramatically in the last year. He now runs a very fast but even ramp from his drop in to the first crack, taking around 8 to 9 minutes (basically lower drop in but a lot more heat compared to before). The roast finishes in another 3 to 4. However, the original Hottop profile had a flat section early, and took around 5 to 6 minutes to hit 300F bean temp, and only then went to maximum heat and a full speed ramp to the first crack.

I've found this in my RK Drum setup too. My roasts became far better when I cranked the heat pre-first crack. My probe reads close to 700F in the grill (outside of the drum), which seems high but the coffee is turning out so damn good.

First crack comes right at 10 or 11 minutes at which point I reduce the burners to low (ambient temperature steadily falls to 480F approx. in about 3 minutes) with second crack coming at 13:30 to 15:00 minutes. I've noticed the beans temperature seems to ramp faster (higher thermal inertia)? And if I don't reduce the heat the roast races out of control.

It's like my roasting has gone from a racecar steadily racing the track to a drag car pounding the pavement. But the coffee is turning out so damn well. Even color (bean to bean and surface to center). Excellent aroma (in fact the best ever), amazing brightness yet with the body I'm accustomed too.

This came from reading Jim Schulman comments too.

[another_jim]

In my air roasts, I like to go slow below 300F, then as fast as possible to the first crack. I think with the unventilated drums, this may be a non starter, there's just too much thermal mass to accelerate that hard. Tim's and Ken's strategy of max heat early may be the best bet for these. You can still control the ramp by timing the drop in precisely.

I'm glad people think they learn from me. I have a sneaking feeling that most of what I say is wrong; but that people do learn how to experiment, rather than settling on what the manual says or on what they've done before. Maybe that lesson is a lot more important than giving out the right information.

[Rainman]

You're right, Jim- I think we do learn quite a bit from you, if nothing else to be more willing to experiment away from product literature. I'll be interested to see how this Ambex roaster compares, but my sc/to results have been pretty consistently awesome by my own (and my wife's) taste. I found by talking w/ the Ambex people, that their roasters have some air flow participating in the roast process-- the stated rate of transfer of heat from convection to conduction are appx 4:5 in their equipment. I don't know if that's an optimal number using drum technology, but to choke off air flow (eg. a dirty roaster) produces worse results-- go figure. I'm going to guess that the amount of time I use now going from drop in to the beginning of 1st crack is at least a bit faster (because it's all convection- correct me if I'm wrong?) than what the drum will require for the same bean and roast degree.

I had an email discussion going back and forth w/ a roaster manufacturer a while back when I was looking into importing a similar roaster as Ken's (from Pinhalense in Brazil). He offered a suggestion to drill holes in the back of the drum once I got it to double their number- he kinda made this tip sound like a trade secret... Anyway- this would effectively double the amount of ventillation, suggesting that this is an area where the traditional Jabez/Burns designs suffer. At some point, convection heat transfer is fairly important-- I have no idea how the RK users manage that.. Tim?

Ray

[Ken Fox]

You're right, Jim- I think we do learn quite a bit from you, if nothing else to be more willing to experiment away from product literature. I'll be interested to see how this Ambex roaster compares, but my sc/to results have been pretty consistently awesome by my own (and my wife's) taste. I found by talking w/ the Ambex people, that their roasters have some air flow participating in the roast process-- the stated rate of transfer of heat from convection to conduction are appx 4:5. I don't know if that's an optimal number using drum technology, but to choke off air flow (eg. a dirty roaster) produces worse results-- go figure. I'm going to guess that the amount of time I use now going from drop in to the beginning of 1st crack is at least a bit faster (because it's all convection- correct me if I'm wrong?) than what the drum will require for the same bean and roast degree.

I had an email discussion going back and forth w/ a roaster manufacturer a while back when I was looking into importing a similar roaster as Ken's (from Pinhalense in Brazil). He offered a suggestion to drill holes in the back of the drum once I got it to double their number- he kinda made this tip sound like a trade secret... Anyway- this would effectively double the amount of ventillation, suggesting that this is an area where the traditional Jabez/Burns designs suffer. At some point, convection heat transfer is fairly important-- I have no idea how the RK users manage that.. Tim?

Ray

My impression is that increasing the ventilation through the drum of my sample roaster would result in enough added heat loss that my ability to control the roast would be severely diminished, likely requiring a substantial reduction in charge weight if I wanted to get the same roast parameters I now get. I live at altitude (almost 6000 feet/1850m) and the burner that was supplied with my roaster was completely inadequate. With Barry Jarrett's help, I obtained a hugely more powerful cast iron burner, that if memory serves produces more than twice as many BTUs (it's rated to 13,000 if I recall correctly).

Even with this larger burner, and especially in the winter when I heat my garage to only around 50-55F, there is not a lot of spare heat capacity if I want to hit first crack at 8-8.5 minutes with a 1 lb charge weight. I roast under a custom smoke hood with an adjustable fan. If I really crank up the fan, which sucks more smoke out of the large roaster tryer hole, the heat loss is very obvious on my roaster's thermometry.

Altitude is an issue for me, but as I've indicated, I'm using a burner that is at least 2x as powerful as the one that came with the roaster. Given design constraints, I don't think a more powerful burner would even fit in the space available.

All of this is going to depend on one's particular roaster and where it is used. Nonetheless, I'd be wary of making a non-reversible modification increasing airflow in this sort of roaster, as this might well force you to reduce charge weight or result in "baked" coffee results when you find that roast times are excessive.

ken

[Rainman]

Yeah, I wouldn't drill holes in the back of my drum without some serious investigation to see if that's a proper solution. In the end, I chickened out of that venture because it didn't leave me with a healthy sense of confidence that it would be worth the trouble (your advice helped in that direction, too). There's always JB-weld to plug those holes if anyone has already done it!

[another_jim]

A ventilated drum roaster means pushing air across the flame and heating it to roasting temperatures before it enters the drum.

Pushing cold air into the drum is probably not a good idea, since dropping the environmental temperature around the beans is, according to roasting guru, Carl Staub, a very bad idea. Large scale industrial roasters, where cost is no object also control the inflow air so it never drops.

Ed Needham modified his homemade BBQ roaster by putting a fan that blew cold air upward through the burners into the perforated drum. He also built a flue by punching a hole into the lid of the BBQ and attaching a stack. This sped up his roasts from around 20 to 25 minutes, weather depending, to a weather proof 12 to 15 minutes.

[Rainman]

Right- I'm sure introducing cold air into a drum (or really, any roaster for that matter), is a good recipe for stalling a roast. What I'm wondering is if different amounts of convection current (assume I mean heated air, not cold ambient air) in relation to conducted heat (from the drum, face of the roaster, etc.. anything that imparts heat to the beans from non-air sources) will alter the length of time to a finished roast (think of it as altering the proportions of convection to conductive heat). I think it does, but there's little discussion about that issue (maybe it's still not fully understood?). In my own experience (limited, I'll certainly admit)- the amount of time to the same degree of roast given different roasters seems to go like this: fluid-air bed < sc/to < hottop. I can't say that I've really compared these methods in head-to-head fashion, but that seems correct. Going beyond the simple explanation that most people acknowledge like, "air roasters notoriously roast faster than drum roasters, so you do what you can to slow them down", etc... I can't help but wonder if the relationship of air flow (hot air) to conducted heat is one variable that we haven't accounted for... much. What I'm thinking is that introducing more convection heat probably shortens the time you need to get to 1st crack without losing much of the origin flavors (if I'm going to buy an expensive drum roaster, being vastly different that my little sc/to setup, I at least don't want to lose that aspect of my roasts).

not sure if that makes sense to anyone else other than me. Am I obsessing about a triviality?

Ray

[another_jim]

All this stuff is extremely well known by everyone except, it appears, copywriters

There is no intended conductive heat transfer to the beans -- beans are rounded and made of cellulose, so they will touch surfaces at a single point and scorch. Drum roasters are designed to keep the beans flying. If the drum rotates too slowly or is overloaded, you get conductive roasting, aka tipped beans

Unventilated drum roasters transfer heat from drum surface to the air in the drum, then to the beans. Some heat is also transferred by radiation. Home drums with perforated drums and a radiant heater, as well as some large scale roasters designed for dark roasting, do a lot of the heat transfer radiantly.

The amount of heat transferred per degree difference rises dramatically as the velocity of the beans through the air increases. In drums, this is done by keeping the beans flying, but is limited since if the drum rotates too fast, the centrifugal force will glue them to the sides.

It is also affected by the amount of air to the amount of beans (the air loses heat to the beans, more air, more heat to transfer). This is where air roasters gain their speed at low temperature roasting, they can pump more air at low source temperatures, while the amount of heat a drum can transfer to the air at a given temperature delta is fixed. However, one can get high speed low temperature roasts in a drum simply by lowering the charge, thereby raising the ratio of hot air to beans.

And that's enough of both hot air and beans for today.

[Rainman]

All this stuff is extremely well known by everyone except, it appears, copywriters... And that's enough of both hot air and beans for today.

Ah- thanks, Jim. I'm quite gullible as you're discovering, and copywriters often have their way with me. I'll typically trust anyone (at least once). I kinda like the hot air-- where'd we be without it?

Ray

[TimEggers]

My own experimentation may have relevance to what is being discussed here. I experimented with using a heat diffuser over the flames in my grill and with not using one. Without the diffuser the heat thermals really gets the air moving (I can feel that with my hand) and with the diffuser the thermals rise more slowly (but perhaps more evenly).

Today my preference is no heat diffuser, burners on high until pre-first crack. Jim is correct above that lowering the heat at just the right time is vital to establishing desirable post FC roast rates. Cut it too late and the roast flies out of control too soon and the roast could stall.

I've noticed too that without the diffuser chaff flies about (and out) my grill far more readily. I should note too that I removed the slide out grease trap under the grill this seems to allow more air to be sucked into the grill naturally by the thermals coming off the burners. It may be quibbling but I see a faster roast without the diffuser/drip tray in place but that may just be the specifics of the grill I'm currently using. Its internal volume of open space in comparison is large compared to the drum itself.

[Rainman]

Thanks, Tim- I know this is something not frequently discussed, but most commercial roasters allow for some sort of adjustment for airflow. The old lecture from my Dad keeps popping into my mind about how oxygen supports combustion, so I have no doubt air flow can be a limiter to a certain extent (however little adjustment some roasting setups have in that factor).

Ray

[mrgnomer]

I don't know much more except heat delivery for radiant drum roasting can be atleast two ways: direct contact when there's no air flow and the heat of the drum is doing the roasting and convection when you add air flow with a fan. I'd imagine the roasting is slower when it's just the drum supplying the heat and you can speed things up when you start circulating the heat with a fan. Just guessing. I might be wrong.

[e61brewski]

i've been playing with airflow a lot, and will cross-post some stuff from CG...

in particular, i've been using the two lids on the i-roast as rudimentary airflow controllers. the results, in terms of roast profiles and tastes in the cup, are fairly dramatic.

if, as various gurus have said, a crucial factor in final taste is not only the warm-up period but also the amount of time allowed to elapse between first and second crack, then wouldn't it be imperative for you to (a) find what that perfect time gap is for a particular coffee, and (b) be able to manage it predictably despite notoriously unstable factors that influence air roasting? like, er, winter? electric current vacillations?

adjusting airflow in real time seems to provide this ability to adjust on the fly, the way a deidrich user would.

this line of inquiry has been unraveling in a somewhat undisciplined fashion at the blog.

Post 1 - the lid juggle, explained
Post 2 - exothermy!
Post 3 - charts and such

feel free to wade in. it should be noted that i have NEVER before roasted anything good at some of the obscenely long times i've seen from some other home roasters -- 10 minutes and such. normally, dragging out an air roast that long (using the temp profiles i've seen) always gives me not-so-subtle notes of baked brick.

this is different. by controlling airflow, there would seem to be a way to properly dry/warm your beans -- and lengthen a roast -- without overly baking them. instead of simply playing with actual heater temperatures, in other words, play with the airflow as the roast develops and try delaying/speeding things based on actual temp readings.