Another problem in roasting that may be addressed by profile variation: insufficient glass phase

In the Roasting 101 article, another_jim describes the glass phase thusly:

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.

From this we can extrapolate that, regardless of roaster type, if you have the symptom of hard, small beans, one of the causes may be insufficient roaster temperature - the roaster generates enough heat to initiate first crack (AKA 1C) (~400°F BT, gathered from various internet sites, including Sweet Maria's visual pictorial guide to the roasting process), but whether by user control or design, doesn't generate enough heat to complete 1C.

To tie this back to the 3 ramps that have been discussed, this would be a problem with the finishing ramp.

-s.

Another consideration for profiling: initial bean mass

This will depend on your roaster's abilities (see a trend here?). I think that the 3 ramps can also apply here, but think that the absolute timing for a 125g roast in an unmodified air popcorn popper may NOT be the same as the timing for a 5lb BBQ drum roaster (i.e. the 9 minute roast in a Poppery vs. an 18 minute roast in an RK drum).

Time ratio may be a better measurement than absolutes in this case.

hmm...

Perhaps we should christen this thread the H-B "Three R's" of roasting

-s.

To expand on the temperature side, I've been wondering about which temperature of the three (bean surface temp, roaster environment temp, exhaust temp) would be most beneficial in profile tweaking, and it seems that again, it will vary with your roaster. What matters more is that, whichever temperature you're using, you're getting repeatable results.

In an earlier post, I gave more weight to temperature than to sight or smell. This was not intended to dismiss the other two, and here is where all three can be used to help triangulate your location within the roast, if you will.

e.g. If you're using a dogbowl a la JimG, and you notice that the beans are turning yellow, but the temp readout is 350°F, then shoot for that EVERY time. If you're smelling wet hay at 382, ditto. 1C at 445°F? Repeat it. When you can hit that note again and again, you can work backwards and tweak (as much as your roast method allows), adding a minute here or taking away a minute there, or adding 5°F per minute per ramp, etc..

-s.

coffee.me wrote:Who's that noob ?

Might have been me as I mentioned any prospect of roasting 101 a week before and then also asked for one of his profiles... since I have recently finished an air roaster based upon a popper and expected a certain level of transferability where BT is controlled directly.

Thanks for lots of good material so far: agreed that BT is the most important measure and it should be comparable across different roasters with allowances for air/coffee ratio during a roast and also perhaps the depth and frequency of temp cycling of individual beans moving from heat source to bean mass within a roaster during a roast. There must be some way to characterise a roaster... If, like most roasters, there is no direct control over BT then complex recipes to manipulate the controls are required to achieve a given profile. Starts to get messy.

On accidents, certainly learn a lot and perhaps reporting of these and their effects can help others? Gives one hints where to look. I have found doing prohibited stuff such as dropping and oscillating temps are not always damaging.

When documenting a BT profile I use a shorthand say 110-4-150-2-190-1-200-6-230 where by their size, times and temps are obvious (and it is obvious whether F or C by value) add :30 etc to times as well if required, still unambiguous. Just a suggestion. Since my roaster just does what it is told, I only need to note and deviation of PV from SV during a roast.

Still very early days for me - only ~40 roasts done, lots of A/B comps where only 1 thing is changed. Very slow as I can't drink that much coffee.

Eric

Ken,

I don't claim to understand your frustration, but I'd count myself among the frustrated "coffee engineer" types. If I grok your post, it's that none of this matters if it can't be applied to the results in the cup. Then again, it could also be interpreted as the HB version of "You kids and your gizmos, get offa my lawn!"

If your intent is the former, I'd completely agree, as I perceive this thread (and others like it) to be similar in goal, and encourage this continued pursuit. The point you make about "Anyone who has roasted coffee for a long enough period will have produced very good roast results by accident" needs to be broken down:

"Very good roast results" are what I(and presumably, the rest of the less experienced roasters who keep asking these questions) am trying to achieve, based on the limitations of my own experience and roasting method. "by accident" is what I'm hoping to avoid through this process of profile tweaking and tasting the results I'm hoping that the most direct result of this thread will be a set of (somewhat) standardized parameters or metrics that can then be applied to a given roasting method in the quest for a consistent and tasty result.

Given your own years experience, would you care to chime in on the information that's been presented thus far?

-s.

Sherman.

I think Ken's point is obvious: some of the posts have been confusing the gas pedal with the road map. It's supposed to be a thread on how to profile, not on how to set up a roaster so you can profile. I mostly agree with Ken's point; but sometimes the two issues are hard to separate, since one has to use whatever controls are available to fix whatever problems one is having.

Let me give an example:

The green coffee coop recently sold a Yemen called Anesi. When I visited Ken in the spring for the freezing experiments, I had a chance to try it, and found it a very clean, berry and chocolate shot. I was kicking myself that I hadn't bought any. Turns out that other buyers were not nearly so happy with the coffee, and were getting dried up, leathery shots instead.

(Now that I've scored my five pounds,) I can tell the people struggling that there is nothing particularly mysterious about this problem; it means that the beans are being overdried, and need a faster drying period, and possibly ramp to the first. However, what I can't tell people is how to go about doing that with their roasting equipment.

Now it appears the people having these problems primarily use the Behmor. They may know perfectly well that over drying is the problem, but not know how to fix it on this machine. To solve this sort of problem, we don't need general discussions on how to instrument roasters, but specific tips for specific equipment ( e.g drop the bean weight by X% and use profile Y).

If roaster "A" and roaster "B" can each be configured to drive the bean mass temp (BT) along the same prescribed time/temperature path, to what extent do the beans care, or even know, whether they were roasted in A vs B?

Are they smart enough to know how they were heated?

Or do they only care about the rate at which they changed temperature?

If the variations in air flow and MET from roaster to roaster were relatively less important than BT profile, then a family of device-independent profiles could be developed. Implementing the profiles on a particular roaster then becomes a separate exercise.

Jim

hopefully this will help get this thread back on the right path.....prior to this thread i'd only heard of the glass phase in passing (probably came up in a thread somewhere here), but i never really thought about it.

jim's comment/explanation (as quoted by sherman in this thread) made me wonder about it. i'm really interested in how to achieve this phase correctly in my roasts. also, what is the significance of getting it right when it comes to the final flavor?

up till now, i shoot for a 4-5 min ramp to 300 (though i've never really gotten lower than 5 mins with my roaster), then give it a bit more heat to gain momentum into 1C. when 1C starts i dial back the heat to avoid losing control of the roast. what i'm wondering, is if it might be a smart idea to boost the heat halfway through first to power the glass phase, but not so much as to let the temp go out of control. or maybe its better to not cut the heat until first is slowing?

another_jim wrote:... To solve this sort of problem, we don't need general discussions on how to instrument roasters, but specific tips for specific equipment ( e.g drop the bean weight by X% and use profile Y).

One of the things that I have always wanted to see on this site is some detailed reviews not just of espresso machines, but of some of the marketed home-roasters, like the behmor, hottop, etc. This would give the reviewer the space necessary to really explore these machines, and their capabilities, while allowing those of us who need the reviews to either be able to use our products from a more informed standpoint, and allow those who are looking to buy a manufactured home-roaster the ability to make an informed decision.

It seems that the original intention of this thread has been trampled by rants and musings that has distracted it from its very important and admirable goal: to teach the basics of roasting coffee beans. The purpose of a forum is not for a bunch of moody old know-it-alls who have "been there, done that" to discourage those who are just beginning their journeys based on their own inability to remember that everyones learning process in the coffee world progresses by steps taken with faulty, deficient equipment. But in learning how to master these pieces of equipment at each level of our journeys, or at least max out their potential, we are increasing our knowledge and competence by leaps and bounds. I do not doubt that I could get a better shot of espresso out of, say, peacecup's Saeco machine than I could out of a Brewtus owned by someone who didnt know or care what they were doing and simply had the money to buy top-of-the-line. This is where it seems Ken misses the value of this thread. The whole point of a forum at all is to allow the mixing of the knows and the know-nots for the benefit of the whole (as occasionally the know-nots, who are usually less afraid to be honest and appear ignorant, make observations that albeit rarely teach the knowledgeable something too).

There should be no fear over producing ANOTHER thread explaining the basics to those pesky beginners who dont even know that without the best equipment, their endeavors will be profitless. There are now many excellent beginners guides to espresso and how to pull shots, etc, simply because these guidelines had been repeated so many times on forums before (such as alt.coffee) and were over time tempered and reduced into a set of very usable criteria upon which everyone who tries to make good espresso has used as their starting point. This process has not been given enough time yet with roasting, I fear. But that is exactly why this thread is so darn important.

There is an ancient Greek quote (I cannot remember whose right now - not one of the obvious philosophers though), which, paraphrased, says "Once there was a classroom in which a student upset at the uselessness of what he was learning asked his teacher what possible value there was in learning the material. The teacher turned to his assistant and said, 'give that boy some coins, since he has to profit from what he learns." Yes, Ken, we certainly will not be able to get professional roaster results with home devices. That does not in any way diminish the value of learning with home-devices, whether it involves battles with microchips or not.

Noah

JimG wrote:If roaster "A" and roaster "B" can each be configured to drive the bean mass temp (BT) along the same prescribed time/temperature path, to what extent do the beans care, or even know, whether they were roasted in A vs B?

Are they smart enough to know how they were heated?

Or do they only care about the rate at which they changed temperature?

If I have a choice, I always choose to roast stupid beans. They don't care how they've been heated.

But even with smart beans, it's a stretch to assume that measured bean mass temp (BT) can be compared between two roasters. Any measurement we make is merely an estimate. A slight change in the location of a BT probe in the same roaster can change the measurement significantly.

Since this is merely a brainstorming session I'll throw this in.
As Ken mentioned with bean quality, garbage in garbage out. No matter what you do you can't make a 70 point moo goo muck bean a 90 by how you roast it. I think something similar can be said about the quality and capabilities of the roaster. To put a 90 in and get a 90 out will not only depend on the knowledge of the Roaster but will also depend on the shortcomings of the roaster. How would we advise someone with a $100./lb green Panama Esmeralda Gesha and a unmodified popper or unmodified off the shelf roaster? I think H-B is a place for those with a deep respect for coffee and also respect for the tools we use. I do think some of the off the shelf roasters and alternatives such as poppers, bbqs, convection tops,etc. can be modified to be reasonably capable roasters with a reasonable amount of effort and expense. I do get frustrated when someone wants to get advanced level help with their roasting skills but is unwilling to put a few dollars and a little time into improving major problems with their roaster. Why struggle with the shortcomings of 30-50 degree temp.swings from a cheap thermostat in a turbo oven top when $30. for a good dimmer or a $100. for a good used variac and $10. for a on/off switch and a couple hours work will for the most part solve a major problem.

Thanks Ed,

I think your post puts the argument in good perspective. We are assuming that people on this forum are going to take some pains with their roasting. The only questions are what sort of pains are worthwhile taking, and how we can help each other out.

So here's another story that crystallized my answer to this:

Old fashioned, commercial, unventilated sample roasters like Ken's certainly can do great roasts, but the learning curve is somewhere between non-trivial and horrendous. The only reason to persevere is one knows for certain that one can finally do great roasts, since so many have in the past. At the opposite end of the commercial spectrum are the production roasters at Green Mountain, which I recently visited, ranging from a 60 pound stationary drum to a 5000 pound open bowl roaster, all made by Probat, all using heated air and mechanical stirring, all profilable for environmental temperature, airflow, and bean temperature (within the limits of heat transfer physics), all using the same control panels and displays, and all delivering identically cupping roasts for identical profiles.

Obviously, fine tuning a profile is a lot easier on something like Green Mountain's roasters than on an unventilated sample roaster. So the jaw dropping surprise is in their cupping room: for their sample roasting, they still use old style, unventilated Probat/Burns sample roasters (BTW, which Probat still makes and for which there are buyers' waiting lists). The initial profile for each coffee's production roast is based on these low tech sample roasts.

The high tech of the production roasters is not about developing profiles, but about ensuring the same roast independent of the amount being roasted or the roaster being used (It's mind boggling from an industrial engineering standpoint (or at least my 1980s one): Green Mountain manages to do everything from pound by pound mail order COE coffees to container loads of Keurig capsules out of the same roasting and distribution facility.)

So what's the moral of this story? The huge capital investment in high tech roasting is for getting QC, but not for developing roast profiles. For that, you sample roast, get to understand the coffee, and understand enough about your production roasters to know how to tweak their base profiles to get the best out that coffee.

(BTW, I'm not saying everyone must buy unventilated Probat/Burns style sample roasters; they are like qwerty keyboards, people in the industry know how to use them, but if one were inventing from scratch today, the sample roaster design would probably be quite different.)

And how does this bear on Ed's point on what sort of pains should we take and how can we help each other out?

I'm going to come back to the conventional wisdom. For the last 50 years, regardless of their size, their equipment, or their technologies, coffee roasters have developed profiles the same way. They do a standard light roast of the coffee, brew it, and taste it. Based on that, they decide whether the coffee is worth using, and if so, how to roast it. The "how to roast it" consistys of two things, how dark to roast, and how to tweak their standard roasting profile. The tweaks will be longer or shorter, or less heat, more heat at various stages of the roast.

This suggests what sort of pains are worthwhile:

• It is worthwhile learning how to do a standard, fairly fast, light roast. If you only roast for espresso, this can be a little darker than if you also brew, but in either case it should be at the light and fast end of all your regular roasting. If your roasting equipment cannot produce such roasts without grassy or excessively sour flavors, then it is time to alter it or change it.
• It is worthwhile learning how to taste this standard light roast to answer the following two questions. Is this a coffee I like? Do I want it roasted darker, longer, brewed this way, or that way?
• It is worthwhile learning how to roast longer or shorter, with less or more heat, at various phases of the roast. If your roaster cannot do this, change it or fix it.

When I started home roasting, I really didn't want to get into all this rigmarole; and most people still don't. So how can we help each other out?

• We can recommend coffees we like. People will consider the source and try or not try the coffee.
• We can tell people how we roast it: a little lighter or darker, faster or slower, than usual. They can use this by making similar modifications to their own roaster setup. This means they will always need a roast setup that allows such modifications

.

I think some people here are trying to work out a smart technology replaces the traditional route. If a roaster can get the most out of each coffee from the get go; all the cupping and tweaking is unnecessary. One buys the coffee, tells the roasting machine where it comes from and how one wants to drink it. It does the rest. One can taste the shots, and buy more or not, assured that one tasted the very best possible shot. I'm not sure if this is possible; and to be perfectly honest, I don't really want it to be possible. Such a roaster could be connected to an equally smart super auto, and we'd be out of a hobby.

Jim
Absolutely +1
ps any pics or comments on that 5000lb bowl roaster? if so, would appreciate a pm

Here's my take and where I agree with a certain amount of what Jim and Ken have said. A specific profile is going to help if you're generally roasting the same beans to the same point all the time. Profiling leads to repeatability. But getting profiles from somebody else and using them in that way is like buying a term paper. You may get a passing grade, but you don't really learn anything. On the other hand, developing an understanding of what goes on in the roasting process will enable you to walk up to a new bean and get, at minimum, decent results the first time you roast it with better, and possibly great, results on succeeding roasts.

Now here's here I have problems with some of what's been stated in this thread. Warning: rant ahead. You can just stop here or you can proceed at your own peril.

In light of a number of statements that have been made here, I must be doing something horribly wrong, since I get excellent (and consistently repeatable, which is also an important part) results on my Behmor without 'twiddling' or trying to 'defeat the microchip', but simply by paying attention to what I and the beans are doing. When folks here tell me that I can't get good results from the apparently piece of crap home roaster that I purchased only because I obviously didn't know any better, I get a little miffed. I mean, I regularly compare the stuff that I roast to fresh beans that I get from top roasters, just as a reality check, and there's not a huge difference in quality between what I'm getting for myself and what I'm buying. It also makes me wonder how anybody ever managed to roast anything before all these electronic controls and fine temperature measurements and all that. Then I realize that people were turning out excellent roasts for centuries relying solely on the tools that they were born with - sight, smell, and hearing. The primary thing that finer metrics and piles of data give you is the possibility of greater consistency from roast to roast. Not necessarily a better roast, just a more consistent series of roasts.

When somebody here says, "Here are the tools that I have. How can I get the best results from them?" the answer, "Oh, those tools suck and you'll never get good results from them. Here's what any discerning person would use." isn't responding to the question that was asked. The answer should be either, "I don't know." or, "Here's what you can do to maximize results from what you have." Anything else isn't really helping the person who asked the question. And we're supposed to be helping each other here, right?

I'm going to dare a broadly sweeping statement: Commercial roasters' primary concern is to turn out a consistent roast as much as it is to turn out a great roast. Klatch wants the batch of Belle they roast next week to taste as much like the batch they roasted two months ago as they can because that's what people are paying for - a particular taste. As an example, look at the people who complain about how Black Cat isn't what it used to be since they changed the blend. Home roasters, on the other hand, appear to be more concerned with getting the best flavor they can out of the beans that they have in their stash. Oh, yeah, there's a certain concern that this year's Sidamo might not be quite the fruit bomb that last year's was, or the new crop of Brasilian has more or less body than the previous crop, but, overall, the concern is not to make it taste the same way it did last time, but to make it taste the best it's capable of tasting. Which, to me, means that home roasters are generally treating coffee more as a culinary product by making the most of the particular ingredients they have at that time, the same way a good cook will use what's in season to make a great dish that may be slightly different that it was two months ago because the ingredients, like all agricultural products, change with the seasons. So, as a home roaster, think of yourself as a chef, taking the best ingredients available at the time and maximizing them, which involves understanding the process. As long as the dish you end up with is great, it doesn't matter if it tastes the same as it did last time.