HomeRoast Digest

Topic: Profiling by bean type (22 msgs / 765 lines)
1) From: Vicki Smith
This is a long post. I probably should apologize now and get it over 
with ;).
Early this morning, I ended up rereading Tom's tip sheet on the IR2 to 
answer a question here. In it, Tom talks about different profiles for 
beans grown at different altitudes. In that same article 
http://sweetmarias.com/hearthware.iRoasttipsheet.html)he says that he 
uses the same profiles most of the time when he is roasting for cupping.
The article has been haunting me all day.
Now, I'm wondering how other people approach profiling. Do certain kinds 
of beans, certain origins, things like DP vs WP
influence or determine the way you roast? I'm not thinking roast levels 
as much as I am thinking about things like:
* slow or fast buildup to 1st
* time elapsed between beginning and end of first
* time between the end of 1st and the beginning of 2nd (if you go that far)
What I'm asking may be unclear, so I'll post a bit of what Tom wrote, in 
the hopes that folks understand what I am trying to ask. I know the 
details of time and temp are specific to the iRoast 2 and may be/are NA 
for other roasting methods.
 From the tip sheet:
Here's the roast curve that I use for almost all samples to get a City 
or City plus roast (which is considered a cupping roast I know and is 
probably lighter than most folks like):
Total roast time: 9:30 min
     * Stage 1: 350 F for 2:00 min
     * Stage 2: 400 F for 3:00 min
     * Stage 3: 460 F for 4:30 min
Island Coffees: Coffees from Jamaica, Hawaii, and to a lesser degree 
Puerto Rico, have a lower bean density because these island coffees do 
not have the altitudes of such origins as Kenya. They benefit from a 
lower initial temperature during the warmup time. Here is the program I 
am currently using for a City roast:
     * Stage 1: 350 degrees f for 3 min. ( = 310 f onboard thermometer 
     * Stage 2: 400 degrees f for 3 min. ( = 360 f onboard thermometer 
     * Stage 3: 450 degrees f for 3 min. ( = 375-390 f onboard 
thermometer reading) (It might be wise to make the last stage 3-4 
minutes, then simply stop the roast manually when you reach the "degree 
of roast" you like - this is always the best method to target an exact 
degree of roast.)
Brazil Coffees for Espresso: Like the Island coffees, Brazils come from 
lower elevations. They benefit from a slower warmup and a longer overall 
roast. For Northern Italian Espresso I like this program, and you can 
use it for other light espresso roasts too:
     * Stage 1: 350 degrees f for 3 min. ( = 310 f onboard thermometer 
     * Stage 2: 460 degrees f for 3 min. ( = 370 f onboard thermometer 
     * Stage 3: 470 degrees f for 4 min. ( = 390 f onboard thermometer 

2) From: Brett Mason
Geez you CAnadians are up late!
us Iowans too...
  Cheers, Brett
On 1/13/07, Vicki Smith  wrote:

3) From: Brett Mason
So you're asking if people profile according to bean type,
then asking if people profile according to elevation, and finally
if people profile according to individual bean?
Also, some people don't profile at all, but go for "just good beans" ...
  I am in this last group.
I know there are people out here who are working on the science and careful
management of home roasting...
  Sound off folks!
For me, FC+ by sight, smell and sound is all I know...
On 1/13/07, Vicki Smith  wrote:

4) From: Ken Mary
My profiles are in 3 classes, Brazils, dp, and wp.
Tom is right on when he says Brazils like a slow warmup, but I discovered
this only recently. From a cold start I use only 65 to 75% of my roaster's
power, and maintain this setting throughout the roast. The heatup is slow,
but during first crack and beyond this is faster than my other profiles and
maintains brightness. The semiwashed Sumatras may also fall into this class,
but I have not done enough roasts to decide. Like other dp beans, there is
no specific drying step.
For dry process beans, I use full heat at startup and skip the drying phase.
Enough drying happens during the ramp to first. At the appropriate bean
temperature, the heat is reduced so that the ramp through first to the
finish is what I want, usually 5 to 8C per minute. I wonder if peaberries
should be included here, since the first crack is usually very weak.
Aged beans are a subset of dp. No drying, and the finish ramp is about 3C
per minute.
Wet process beans start with full heat but have a drying step at 125C bean
temp for 2 to 6 minutes. This is trial and error depending on the strength
of first crack. A long noisy first means not enough drying. The finish is
the same as for dp beans.
Most of my roasts stop at city to city+ level, a few at full city- just
before the first snap of second, determined by bean temperature.
A lot of this may change after my next roaster mod that will enable
preheating and charging with the drum in place.

5) From: Edward Bourgeois
Being a rather newbie(130 roasts) I still learn from every roast. I  monitor
and record my bean mass temp. by the min.. As the roast progresses the
degrees/min. rise is what I monitor equipment wise and use to adjust heat
level.  I record start and finish temps of first crack and start of second
if going that far. The main differences I've found due to bean type is the
amount of slowing down of temp/min rise into first and the span between
first and second.

6) From: Tom Ulmer
Profiling the bean roast will benefit or detriment the flavors in the cup.
The examples from the tip sheets are good places to jump off. A general rule
of thumb is that beans grown at a higher elevation are denser and are not as
prone to produce ashen flavors when heat is aggressively applied. Wet versus
dry processed does not influence my initial profile. Origin, the size and
weight of the bean are the biggest indicators of the initial profile. I then
develop a profile to accentuate and another to flatten the flavors of the
bean from there.
I never buy any bean in less than five pound quantity and I roast only for
my household. This gives me room to maneuver, however I may roast the entire
five pounds and never be satisfied with the profile. Sometimes I think I hit
it spot-on and Jan remarks that I should roast that chocolate-tasting bean
again - I am the only one who knows it is the same bean just a different
week and a tweak of the profile and whether it was intentional.
I roast 4 different bean origins at a time and a double volume of one
(usually a Brasilian) to ensure that I get a press pot of each as a single
origin, some opportunity blends, and weekend espresso. Quite frequently I am
full of myself and believe that I am profiling for these circumstances and
when it comes together it works quite well as personal congratulations for a
job well done.

7) From: Peter Zulkowski
This profiling thing still has me confused!
Hey, I turn my roaster on full acceleration (1475W Turbo Oven)  with 650 
Gr of beans and the temperature creeps up at about a degree per second 
until 350F or so,
and that is if I preheat the empty roaster to 450F.
This is with Harrar Horse.
Different results with different beans.
Different results with different amounts of the same bean :)
So I remove the TO and measure the temp of the beans in the roaster with 
a hand held Infra Red meter... the IR monitor reads within a few degrees 
of the thermocouple output. Now I know that the TC and the IR give the 
approximate temperature of the beans themselves.
At least at 350F.
The temperature rises more slowly as the temp gets higher... to a 
certain point.
With such low power, profiling is sketchy at best, unless you limit the 
amount of beans you roast.
OTOH, even with unlimited power, the beans will absorb only so much heat 
per  minute in a 450F environment.
diff beans absorb differently also.
To be real, you should also consider differences in the place where the 
roasting is done... wind, humidity, temperature...
Batch after batch, outside, things change.
That being said, when one sets ones roaster to 350F for 3 minutes, let's 
say; does that mean that the beans themselves are at 350F for 3 minutes?
Or are they just exposed to that environment for that amount of time?
Where are you measuring the temperature?
If you have the temp probe outside the drum that is entirely different 
from having it within the bean mass as in a Popper or TO setup.
Also, inside of the roaster, drum environment is different from HG/DB 
environment, is diff from SC/TO environment, is diff from HG/BM 
environment is diff from Rosto, is diff from IRII.
Even fluid bed roasters have there differences dependent upon where the 
thermometer or thermocouple is placed.
Sound so complicated I wonder how we can roast at all.
Sure is fun though,
and the coffee is great besides :)
Brett Mason wrote:

8) From: Edward Bourgeois
Peter I also use a GG TO top. same watts. My pot is 12"x5 1/2"  Just did a
trial batch tonight with no temp adjustments just full heat. Usually I cut
temp a bit at end of first cr. I only preheated to 350 where I normally
preheat to 450.so the first few mins. were slower to get going than normal.
Batch was 660gr.Nice Costa Rica   Here's a min by min
1min- 183
2        222
3       250
4        277
5        296
6        314
7        332
8        347
9        363
10      377
11      388
12      402  start first
13      408
14       412
15       422
16        429
17        442  dump just before hint of 2nd      if pre heated to 450 and
slowed at about 428 batch would have finished about 14 1/2min.http://coffee-roasting.blogspot.com/

9) From: Jeremy DeFranco
---This is a long post. I probably should apologize now and get it over
with ;).
Please don't apologize, this is a great topic, and I will reply with
likewise long post!
---Now, I'm wondering how other people approach profiling. Do certain kinds
of beans, certain origins, things like DP vs WP
influence or determine the way you roast?
Certainly, all these factors have a great impact on the way a particular
bean responds to the roasting process. An easy way to divide your beans into
separate profiles is by bean density- Hard, Medium Hard, Soft.
Thisarticle discusses this
method in some more detail.
goes into further detail about how to set up a profile for Hard
beans. As far as processing goes, the same author mentions that processing
method greatly affects roast characteristic. The reason being that method of
processing has a profound affect on how the bean pore structure "sets up" as
it dries out. Unfortunately, he doesn't go into detail about this, and I
would greatly appreciate if someone on this list could explain this
phenomenon. The author does hint, however, that dry process is better suited
to a lower heat profile than a wet process might (wet process would take
better to higher heat profile than a dry-process). Any reasoning anyone has
for this would be greatly appreciated : ). So to answer your question, yes
bean type determines the way I roast (although I haven't quite got it all
figured out yet : ) ).
---I'm not thinking roast levels
as much as I am thinking about things like:
* slow or fast buildup to 1st
* time elapsed between beginning and end of first
* time between the end of 1st and the beginning of 2nd (if you go that far)
      I have not been roasting long enough to have any good answers to these
questions, so someone with some more experience would be better suited to
answer these questions. I will offer my 0.02, though, and what I have read
from experienced roasters so far.
      What seems to be one consensus is that time between beginning of 1st
crack and end of roast (no matter what degree it be) be atleast 3 minutes
long. This seems to give the best flavor in more experienced roasters'
opinions (exceptions likely apply). There was a recent post that questioned
max time between beginning of 1st and end of roast, and the general
consensus seemed to be about 6 minutes, although I'm sure there are
exceptions to this, as well. Extending roast too long runs the risk of
depleting the beans of precious volatiles, while not long enough runs risk
of having too much cholergenic acid, and hence a bitter taste (the amount of
cholergenic acid is indirectly proportional to total time of roast).
     Time between beginning and end of 1st can be important as well. Too
quick means you are most likely letting exothermic reactions take over the
roast, and your roast will run out of control. A smooth roast rofile will be
lost. Too long a time may mean you are stalling the roast, and not letting
vital chemical reactions take place, leading to a "baked" or flat cup. So,
an easy answer would not be any specific time period, per say, but rather a
first crack that is properly controlled. This leaves lots of room for
manipulation... You could do a 1st crack that does not have extremely rapid
pops, but not 1 pop every 20 seconds either at peak rate of 1st. OR you
could  do a 1st crack that maintains the slope you had at warmup (i.e. a
faster progressing 1st crack that is still controlled). Although there are
always exceptions- some beans will have barely audible 1st cracks,
     Slow or fast buildup to 1st is a much more controversial topic, it
seems. This is also a topic I have no clue or experience with as of yet, and
it seems hard to find any consensus w/ this info. Some people favor an ~ 45
degree slope to 1st, some like to do a slowly increasing slope to 1st. Some
people always do a drying phase. Some only do drying phase for beans with
moisture content over 13%. Some do warmup phase. Some do not. It would be
interesting to hear theories on this. I hope to know better with more
experience, reading, and tasting. What does seem to be agreed upon, though,
is that you do not want to heat the beans so rapidly as to carmelize the
outside while the inside is just getting warmed up, and you don't want to go
so slow, as to bake the beans. All area in-between seems grey and
controversial. I would love to hear opinions and experience.
     Final thing to keep in mind would be that what is for sure is that any
different profile that takes the same bean to an identical endpoint (City,
FC, etc.) will create a dramatically different flavor profile. It might also
be best not to characterize each profile as "good" or "bad", but rather
unique in its own regard. Like with roasters or brewers, each is different,
and comes along with its own idiosynchrasies.
     Here are some other good articles worth a read:
     http://groups.google.com/group/alt.coffee/browse_frm/thread/4ad50880e24e13cd/251f56bb9248b8a0?lnk=gst&q=roasting&rnum=5&hl=en#251f56bb9248b8a0     -http://groups.google.com/group/alt.coffee/browse_frm/thread/1c8d61f9d164fcf5/233bb1182d4d21b8?lnk=gst&q=roasting&rnum&hl=en#233bb1182d4d21b8     -http://groups.google.com/group/alt.coffee/browse_frm/thread/4c5030a590ad400/fd767a4ae327694a?lnk=gst&q=roasting&rnum9&hl=en#fd767a4ae327694a     http://www.sweetmarias.com/roast.carlstaub.html     http://www.masterroaster.com/articles/cuppers.html">http://www.bootcoffee.com/articles.html     -http://groups.google.com/group/alt.coffee/browse_frm/thread/4ad50880e24e13cd/251f56bb9248b8a0?lnk=gst&q=roasting&rnum=5&hl=en#251f56bb9248b8a0     -http://groups.google.com/group/alt.coffee/browse_frm/thread/1c8d61f9d164fcf5/233bb1182d4d21b8?lnk=gst&q=roasting&rnum&hl=en#233bb1182d4d21b8     -http://groups.google.com/group/alt.coffee/browse_frm/thread/4c5030a590ad400/fd767a4ae327694a?lnk=gst&q=roasting&rnum9&hl=en#fd767a4ae327694a     http://www.sweetmarias.com/roast.carlstaub.html     http://www.masterroaster.com/articles/cuppers.html

10) From: Vicki Smith
Thanks Jeremy. One of the bonuses we get as SM customers is that we know 
enough about the beans we buy to be able to factor in things like bean 
hardness as we develop profiles.
Jeremy DeFranco wrote:

11) From: Michael Dhabolt
Kudos.  A very nice job of chronicling.  Most all of the caveats in the
places I would have chosen.
Mike (just plain)

12) From: an iconoclast
On 1/13/07, Peter Zulkowski  wrote:
Hence, the reason I timed 3 roasts on one day only out of curiosity and
never, not once, have I ever kept any written data. I need to rest my brain
while I roast.  It's good for me NOT to be anal. If I have to think and
write while I roast, it will no longer be fun. I need it to be different
than my everyday life or it becomes a job.  I need Zen along with my coffee.
I know I factor in the variables listed above when I roast and have come up
with some sort of informal profiling:
I start with my BBQ side burner on high for the first roast.
I always roast about 2 lbs of beans (measured by eyeballing the level of the
beans in the mesh colander).
I always start with the temp at 900 degrees give or take 20 degrees up or
down depending on ambient temp and how windy it is.
I maintain settings until first crack is under way.
If roasting decaf, I'll reduce the temp on the HG sooner than with regular
beans, to better control the roast.
After first crack, I decrease temp on the HG to allow for decreased stirring
and letting the bean mass create it's own heat.
After several roasts of the same bean, I recall it's personality and modify
temp as needed.
After the first roast of the day, I pour the new batch of beans in the mesh
colander and turn the BBQ side burner on low with the HG off. This allows me
to bag up and label the previously roasted beans while warming up the next
batch. It saves time and fuel. Then I start back at the beginning of this
My roasts are not always identical.  I visit my husband several times during
the roast when  he does the roasting.  We share information we acquire while
roasting. We tweak things as needed. I try different roasts for different
beans. Even if I did keep written stats, my favorite coffees would run out
before I could use a profile consistently. Then I'd be frustrated which is
very un-Zenlike. My coffee ALWAYS tastes better than any coffee I've
purchased in the past....so I can't lose.
Take care,
Sweet Maria's list searchable archiveshttp://themeyers.org/HomeRoast/index.htm

13) From: Vicki Smith
One of the reasons I was so interested in learning some more-or-less 
generalizable things about types of beans (by hardness, for example) is 
specifically because I almost never buy more than #5 of any given bean. 
If there is some sort of starting place for a given sort of bean, I'm 
more likely to get a good result the first time out, or at least before 
I run out of the shtuff.
an iconoclast wrote:

14) From: Ken Mary
Try my Brazil profile, the chocolate is very strong even just out of the
roaster. The concept is described earlier in this thread. Some bean
temperatures and times:
3 min  45C/113F
6     100C/212F
10    150C/302F
15    190C/374F
16.5  208C/406F first begins
17.5  212C/414F first ends
19    222C/432F city+ level
The initial warmup may not be optimized. But the results were so good that I
have not changed it.

15) From: Vicki Smith
Ken, I forget what you roast in. I know that a slow warm up to 1st makes 
sense for Brazilian beans, but I would have worried about baking if it 
took anywhere near 16.5 minutes for my ~1 lb. bread machine roasts.
I also have never succeeded in a 1st that only lasts a minute (assuming 
we are not counting the odd outliers here)
Can you say more, please?
Ken Mary wrote:

16) From: Tom Ulmer
Thanks Ken I will.
This is a few minutes longer than my typical profile for Brasils and a
gentler ramp. The chocolate fan in my house will be happy.
My reference temperature is more of an indication of the drum =
temperature in
the early stages of the roast but here's a general profile that I use =
brings out more of a tangy sweet flavor:
Load 12 ounces by volume.
Let the beans come up to the reference temperature of approximately =
and linger for about 8-9 minutes. Increase the heat to bring the beans =
first at another 4 minutes at about 400°F (it seems as though our =
is quite close on this). I then hold the temperature there for another =
minutes and a small bump in heat will usually start second. Dump and =

17) From: Ken Mary
Hi Vicky,
I roast in a homebuilt drum in a toaster oven. So it is easy to do slow
warmups. There is no baking since the profile spends about 8 minutes in the
250 to 400F range.
I count the very first pop and last pop since they are important. My first
cracks normally range from 30 to 90 seconds.
The Brazil profile has a delayed first crack, since a normal first begins at
202 to 204C. This delay comes from the very fast ramp into first, 12C/21F
per minute. Normal ramps to first are 5 to 8C per minute. My t/c measures
bean surface temperature, but the crack is caused by the internal bean
temperature. This seems to indicate that the bean surface is about 5C hotter
than the interior due to this profile.
If you have an overly long first crack then either your ramp into first is
too slow or your mixing is not efficient enough to keep the majority of
beans at the same temperature.

18) From: Vicki Smith
I'd be interested to now if most folks have a 30-90 second first crack. 
My bread machine stirs enough so mixing should not be a problem. I can 
get a faster 1st by moving my heat gun closer to the beans, but a 1st 
crack of 2 minutes did not seem to be an issue, and with some beans, it 
is all of that.
Now, I'm wondering....
Thanks for the response, Ken.
Ken Mary wrote:

19) From: Edward Bourgeois
Vicki    I consider the evenness of the greens to start and the evenness of
the roasting up to first crack as being the factors for the length of FC. If
the goal is an even roast then the short time of FC would achieve that.

20) From: Vicki Smith
I have noticed that some SO roasts--for example the Brazil Yellow 
Bourbon I'm enjoying today--seem to have a short 1st crack (and are very 
even at least visually), which is what I would expect. But when I am 
roasting Harars (for example), the 1st does not go as quickly, nor do I 
expect it to.
I'm thinking this is in line with what you are saying, Ed.  I generally 
don't roast South American beans very often, and lean toward Yirg, 
Harar, etc. I believe that is why I am used to seeing a longer 1st with 
my 1 pound+ bread machine roasts, especially if we are talking about 
including the outliers, both the early and the ones at the end, when we 
talk about the elapsed time for 1st crack.
Edward Bourgeois wrote:

21) From: Ed Needham
Different beans will benefit from a roast that is specific to that bean 
type.  A true statement.
A high grown 'hard' bean will roast differently than a less dense island 
'low' grown bean.  A dry processed bean will roast differently than a water 
processed bean.  The varietal distinctions will also affect the roast 
profile.  personal tastes will affect the roast profile.  Type of roaster 
will affect the roast profile.
Unless you are roasting bag after bag of beans for a commercial venture and 
need the ultimate in consistency, determining a repeatable roast profile 
will almost be impossible for a homeroaster.  The best we can do is shoot 
from experience with a particular bean type and listen to the roasting 
experiences of others and do the best we can do.  By the time we figure out 
what the best profile is, we are out of beans and the next order comes in 
slightly different, requiring more sampling, testing, cupping, comparing, 
and then we are out again.
Luckily, the process is an enjoyable one, so I continue to try to find the 
ultimate profiles for my favorite beans.
Fun facts to know and tell:
--Coffee grows best at about 20 degrees either side of the equator.
--The closer to the equator coffee is grown, the higher on the mountain it 
can grow.
--Higher grown beans are more dense because they grow slower in the cooler, 
thin air environment.
Ed Needham®
"to absurdity and beyond!"http://www.homeroaster.com(include [FRIEND] in subject line to get through my SPAM filters)

22) From: Ed Needham
Being a newbie myself (30+ years of homeroasting) I still learn from every 
Ed Needham®
"to absurdity and beyond!"http://www.homeroaster.com(include [FRIEND] in subject line to get through my SPAM filters)

HomeRoast Digest