Roast profiles

This builds on the basic stuff from http://coffeenavigated.net/roasting-coffee/

On this page I have collected examples of how to control the roast on the Bullet R1 roaster and look at the bean temperature curves from different roasts.

Details in the roast process of a batch are called the Roast profile.

kaffetaarnet-abdullah-sh  bullet-sort-hvid

 

.

.

My point is not to promote one way of roasting but to give you “a map” of different ways of doing it – to give inspiration. You have to try out for your self: what works in your setting, with your beans and to your taste.


CONTENT

Reading the Bullet Roasting profile graphs
A start profile
Difference in first and second batch
John-Paul McCarthy’s strategy
Bob Werby; two updates, he started roasting without computer
Roasting to start of 2nd crack – Peter van Wijk
Into 2nd Crack – Davin Gegolick
Control points during the roast , Therese
Good tasting coffee with the same bean
Working on a nice declining ROR curve
– Kaffetårnet, Denmark
– Jed, USA
Roasting small batches (like 100 grams), by Klaus
Repeating a roast – Playback

 

Reading the Bullet Roasting profile graphs

The straight blue curve is the bean temperature (BT) measurement.

The jumping blue curve is ROR = Rate Of Rise – thats the rise in degrees of the bean temperature probe per minute. Its calculated during the roast. Typical peaking around 2-3 minutes.

kaffetaarnet-abdullah-750g-185c-okt2016

Forget about the red curve. The IR sensor to measure Drum Temperature works best when there is no beans in the drum. The purpose is only in the preheat phase.

At the bottom you can see a red/orange/yellow line – thats the power setting during the roast. The blue line is fan setting.

Remember that bean probes do not measure alike. Some get first crack (FC) start around 170 C, others around 200 C. This is due to technical differences in the probes. So look out for this when you compare with your own roaster.

A start profile

Here I will recommend a simple start profile for new users of the Bullet. So you have somewhere to start – and can modify from there. You can’t give a bulletproof profile. You have to try out for yourself in your own setting.

Preheat temperature – depending on batch size
400 grams use 170°C
500 grams use 175°C
600 grams use 180°C
7-800 grams use 190°C
1000 grams use 195-200°C

Use power P6-p9 depending on batch size. And fan on F2.

Look at the color change of the beans. Note the time for the first sign of light brown. At this point start to turn the power down; like one step every minute – ending at P2 before First Crack. If you do a 4-500 grams batch you can stay at fan F2, with larger batches turn it up to F3.

If the point of yellow/ the first sign of brown starts around 3 minutes it is a fast roast -> that might be good for a light roast of a hard bean. But if you want it slower then use less heat.

If the point of yellow/ the first sign of brown starts around 5 minutes it’s a slower roast -> might be just fine for a lower density beans, like brazil. But if you want a faster roast use more heat to begin with.

If the point of yellow/ the first sign of brown starts around 10 minutes I would think it was too slow. But taste for yourself.

Next step in your roast -> listen for First Crack. For a medium-dark roast you can stop 2.5 or 3 minutes from start of first crack.

Or go darker and listen for Second Crack.

Difference in first and second batch

It is well known in coffee roasting that the first batch responds differently to the following batches on the same program / setting of power and fan.

This is due to the roaster getting heated up from running one batch. With the first batch the roaster will absorb some of the heat applied.

Here Davin from Canada have run batch 1 and 2 exactly alike: same bean, same batch size, same program – but resulting in different 1st crack start. The first batch reached FC at 8:16 minutes. The second batch at 7:51 min.

First_batch_second_batch_Davin

So if you want to repeat a roast profile using the Playback function from a previous roast – then bear in mind if it’s the first batch or a following.

Use extra heat for a first batch to compensate.

Steffen Lav has this strategy for the first batch, when he do a playback of a previous roast:

“First observe if the red DT curves are on top of each other. Then when the roast is going: if Turning Point (TP) is the same – if not increase or decrease heat immediately to compensate. If TP is the same it is normally a matter of compensating with more or less airflow – if that doesn’t help then heat compensate.

However, observe that heat will changes quite slowly to new P settings whereas airflow change BT quicker. So do not overcompensate with the P setting.”

 

John-Paul McCarthy’s strategy

John-Paul from Coalbox Coffee in Canada wrote this about his experience so far (July 2017):

As I have become more accustomed to the Bullet roaster and how the beans respond I’ve also begun honing my methodologies:

– 190 C preheat as standard for 800 grams
– Begin roast at P9, DS5 until yellow point
– Achieving consistent yellow point at 4 min
– At yellow point I increase drum speed each minute until FC to reduce tipping potential. And increase convective heat transfer.
– As I adjust drum speed I reduce power and increase fan speed toward FC. I’m still tweaking this to reduce the exothermic bump at FC. But I’m aiming for a P3 and F4 around FC point.
– Achieving consistent 7 min FC
– Achieving consistent 9 min roast duration
– 22-23% development (DTR)

Note: I’m still using the original chaff filter. So I’ll have to revisit my methods once the new filter it arrives.

Lately I’ve been regularly roasting two bean varieties:
– Fraijanes, Guatemala, SHB 1700+
– San Agustin Loxicha, Oaxaca, Mexico

800g batches are very consistent with the Guatemalan bean.

The Oaxaca beans were great initially but it became progressively more difficult to roast consistently from batch to batch at 800g. I couldn’t maintain a high ROR (above +20 C delta) through to yellowing. My yellow target is 4 min but I was getting 6 min with the 800g dose with roasts stretching to 12 min plus to Full City. It didn’t feel right.

Even at P9 it was as if the beans would suck all the heat out of the Bullet once my bean temp hit 100 C. This made if very difficult to reduce the exothermic spike that occurs during FC. The results tasted fine but I found that post-degassing aroma development was quite poor.

I seem to have hit a sweet spot when lowering the Oaxaca dose to 700g. The reason might be that the Oaxaca bean has a greater moisture content. Again it’s really a matter of continual analysis, calculated experimentation and determining what works best for a given bean variety.

Bob Werby, Virginia, USA
– with 2 updates

First: Roasting with out a connected computer, december 2016

Bob have roasted 100 kilo coffee in 5 months and found it all to have tasted fine.
He doesn’t have a computer connected to the Bullet.

Mostly he roast 1.000 grams batches. Otherwise it’s 700-800 grams. He aims at total roast time in the 12 minute range

With a preheat at 200°C the 1 kg beans get FC start just before 9 minutes.
With preheat at 170°C then FC starts as late as 13:30 min. His bean probe gets 1st crack (FC) at 187 – 199°C. End of 1st crack at 208-209°c

When to stop the roast
Bob stops the roast 20-40 seconds after 1st crack has ended, which with his bean probe is around 212-214°C. If he is unsure when FC did end, he stops at 2:30 after FC start.

He says: “There is a little bit of subjectivity in calling end of FC. My rule is that once I get a 3 second pause between cracks, I call it the end. Otherwise a few weird beans could cause the end of first crack to go on forever. Then I wait 20-30 seconds before dropping.”

Update july 2017 Roasting with a computer

We all modify our way of roasting along the way. Here an update from Bob Werby:

Earlier I did a P9/F2/D3 run from start to finish only reducing power at the start of first crack while running 1000 to 1020 gram batches.

Now I know that D3 is too low for drum speed. I liked that setup because the milestones – Temperature at the Start of First Crack, End of First crack, and drop temperatures matched what I was used to from my previous roaster.

In retrospect that really has little relevance. My roast times at 1000 grams and without computer control were lasting in the 15:00 to 17:00 minute range with first crack at 200°c to 202°c in 14:00 to 15:00 minutes while ending at 30 seconds after the end of first crack was about 208-211°C.

With computer
After transitioning to using a laptop to monitor the roasts I realized that I was probably obtaining a roast development under 20%.

I experimented with ways to improve that but found that I could not push enough energy into the beans and maintain a decreasing ROR easily with the large 1000 gram batches.

I experimented with multiple runs at 900 grams, 850 grams, 800 grams and 750, 680 grams, 620 grams, and 600 grams and found that it was much easier to to get a declining ROR, a 10:30 total roast time and a 22-25% development with 600-700 gram batches.

New strategy
My new protocol is P8/D8/F1 to start, then P7/F2 at dry, then F3 at 160°c and stay there until drop.

My first crack times are now in the 7:30-8:00 range @ 178°c-180°c
with end of first crack at 10 minutes and 190-194°c.
Drop is at 10:30 at 192-195°c.

In retrospect I believe I am producing better roasts and much more easily reproduceable ones under computer monitoring than I was before.

I also feel that the long period of roasting without a computer gave me an opportunity to get a really good grounding in the smells, sounds, and visuals of the roaster without being a slave to the electronics. I would not place too much faith in my probe temperatures as I did replace my probe during this time.

Also as you may recall, I do not significantly use temperatuers to roast. I do it by start of first crack and time of end of first crack dropping at 30 seconds after the end of first crack to obtain a City+ roast.

All the people the who routinely have had my coffee both before and after computer assistance still comment that they prefer my coffee to others. The feedback was extremely positive before and it is extremly positive now. I guess that is a testimony to the ability of the Bullet to produce great results within a very wide range of parameters. For one thing, I have never burnt a roast under either method – manual or computer. I also believe the 30 seconds + or – after end of first crack is a pretty foolproof system for getting reliable results. Another thing I must add is that I only roast very high quality (and more costly) beans with high ratings. On the few occassion where I have roasted bargain beans, I have had poor results.

Update october 2017
– John Francis calls this the “Werby protocol”

Bob Werby writes:
My new normal roast is the following for high altitude beans which is all I roast:

700 grams is normal charge amount;
Preheat either 195°C or 200°C;
P9/F1/D8 to start;
P8 at 150°C (my dry point);
P7/F2 at 160°C;
F3 at 170°C;
F4 at 175°C;
then P8 from 180°C to drop.

I have also run some while keeping P9 unitl 160°C and then going to P7/F2.

After I upgraded my 110volt socket to a full 20 amp service I found an improvement in voltage so that I could roast even 1000 gram charges under 12 minutes total. For that I have used:
Preheat 200°C;
P9/F1/D8 untill 160°C (skipping any change at 150°C dry) when I go to P8/F2;
then at 170°C I go to F3;
then at 175°C I go to P7/F4,
and at 180°C go to P6 – until drop at about 10 minutes total.

Roasting 1000 grams of Kenya gave FC start at 8:15 min. FC end at 10:40. And then drop at 11:15 min.

The exact same settings worked differently on different beans and required no changes on my part. I used the exact same settings for 1000 gram batches of Ethiopian beans on the same day and first crack started a minute+ later.

I do not think I will roast less than 700 grams ever again as the roast moves too quickly for me.”

 

beans-smoke-sh2

Roasting to start of 2nd crack, Peter van Wijk

Peter van Wijk writes:
I usually roast up to 2 crack. I like the chocolat and nutty tastes.
I find it difficult to get a nice declining ROR all up to 2crack. Delevopment % is always between 30 and 40%.
Strange thing is, almost all profiles give a good taste.

This roast is 700 grams of Cuba, preheated to 185C:

PetervanWijk-to-2nd-crack

Daryl commented:
“Same here. Prefer drifting right up to the edge of 2nd for my espressos. Always tastes good no matter how you get there.”

Peter van Wijk replys:
“I agree. Some roasts that have a bad looking profile, like having a little crash and then a raising ROR, that I expected to have to throw away, turned out to taste OK. Total roast time of around 14 minutes tasted usually better than 9 or 10 minutes roasts. Also letting the beans age for one or two weeks does a lot of good. Talking about espresso of course.”

Daryl answers:
“Yeah that’s my big mistake, not letting them rest long enough before drinking them.”

Roasting into 2nd crack

Davin Gegolick do this profile for customers who wants very dark roasts

DavinGegolick_2ndcrack

Adjusting profile by control points
Therese, Denmark

Why
The current roast doesnt necessarily respond the same way to power and fan settings as a previous roast. And the roasting proces is slow to respond on changes so you have to make changes in advance – its like sailing a ship.

So I wanted to find control points along the roast that tells me where the roast is headed – so I can adjust the heat in time.

I first wrote about a strategy in November 2016. But I have since continued to develop my approach.

Update October 2017
After having been focusing for a year on the bean temperature profile and ROR – I have now moved it to a secondary position. Because it didn’t really help me make good coffee. And because what happens in the bean probe is not exactly the same as what happens in the beans, see Bean probe sensitivity.

But I still use the BT and ROR as input to how I manage the roast.

In the beginning of the roast I look at these parameters to see how quick the roast got started:

Temperature of turning point (TP)
The ROR peak level
Time when bean temperature (BT) reaches 115 C.

If the roast is going too slow for what I want → I will use more heat (P). If it goes too fast → I will use less heat.

What happens in the beginning of the roast very much determines the direction of the rest of the roast – and more so the larger the batch is. It is very difficult to change direction later on.

From there on I note the physical change: color and start of First Crack.

Time of yellow point – and note temp and ROR
Time of medium brown – and note temp and ROR
Time of start dark brown – and note temp and ROR
Time of First Crack start – and note temp and ROR

I have in mind approximately how fast I want to reach First Crack, the length of the middle phase (=from yellow to first crack) and ROR level around first crack. I adjust the heat according to how the roast is progressing and how I know the roast behave in the different stages.

To stop the roast I go by time: counting seconds from first crack start.

I don’t worry about getting the exact same BT and ROR curves as an earlier good roast of the same bean. I don’t worry about have a perfectly nice declining ROR curve.

 

From November 2016:

Principle

Find control points along the roast that tells you where the roast is headed – so you can adjust the heat in time.

When the roast has reached Bean temp at 115°C at a certain time, you know if its ahead or behind plan. And so forth at different bean temp = control points.

When you look at the ROR at each point, you know how quickly it moves. That predicts where the roast is headed, like the time to reach FC.

How
Analyze your batches. Group them in successful and unsuccessful. Then you get values for where you want the roast to be at the control points during the roast.

Make your own guide to control points. Our bean probes are different – so you can’t use my list if your FC onset isn’t around 170 C.

udsnit-regneark-analyse

 

My control point guide – update December2016

I find high grown beans for a light roast is best when FC start comes around 8-9 minutes. Where as darker roast and if the bean is lower grown like Brazil, its better with FC start around 9-11 minutes. I typical roast 400 grams and for filter.

Light roast with 1st crack start at 8-9 minutes
For 400 grams preheat 175°C. Low power (like P5) and F2 the first minutes, then P7 for some minutes. After yellowing increase fan to F3. Then decrease power gradually. But all the way adjusting settings to aim for these control points:

Turning point: if less than 98°C the roast is behind/to slow
ROR peak: 15-17
BT 115°C nice between 2:10 and 2:30 minutes
_Note time for Yellowing (typical 3:20-4 min)

BT 140°C achieved around 4-4:20 min (at 4:15-20 and ROR at 11 → FC start at 9 min)
BT 150°C around 5 min … should have ROR round 9-11
BT 160°C round 6-7 min … ROR should be at least 6, but less than 10.

_Note time for FC start (typical at 165-175°C)
Around FC onset aim for lower ROR, something like 3-6°C per minute

 

Darker roast with FC start around 10 minutes (need to explore more)

Turning point: fine with 92-98°C
ROR peak: 12-17
BT 115°C around 2:30-3 minutes
_Note time for Yellowing (typical 4-5 min)

BT 140°C achieved around 4:20-6 min … ROR 5-10
BT 150°C around 7-8 min … ROR round 5
BT 160°C round 8-10 min … ROR round 5

Considerations: maybe try a slower start and then higher ROR at 140-160 C ?

 

When to stop the roast
For a light roast: Depending on the bean stop the roast 50 sec to 1:20 min after FC start(for a light roast).
For a medium dark or espresso: 2,5-3,5 minutes after FC start … I just learned to smell to the beans after FC onset (with the trier) … and note how it changes .. and decide from this.

beans-trier-sh1What is a good roast to Therese ?
“To me most important to achieve is
1) avoid bad taste (bitterness, cardbord, burns and so on),
2) get big aroma … In light roast I get a certain kind of grand aroma, that I can’t associate with a  specific food/flower/whatever … But that is high up … and stays as a pleasant aftertast for a long time in the mouth.”

I mostly brew filter. But sometimes espresso.

Update may 2017

New strategy

A high heat start → higher preheat than earlier

To prevent to big a drop in ROR after the initial peak (at somewhere around 1-3 minutes), increase heat at this point

Already at 150°C; aim for ROR level around 3-5 at FC start (typical around 170°C with my bean probe) by slowly decreasing Power

Settings
Preheat for 400 grams: 185°C – but if its the first batch 190°C

Start at P6. At turning point go to P7. During the ROR peak rise to P8. Exactly how this is done depends on Turning point temperature, the ROR-level of the peak and the time for reaching 115°C and 125°C

Fan at P2 until yellowing point or just as browning starts then F3.

After yellowing point, around 150°C decrease power step by step. I adjust this according to:

Aim for ROR around 10 at 150°C . And at 160°C ROR at 8 and power stepped down to P4.

At First Crack start have power at P2 and have ROR around 5. During FC the ROR has a tendency to drop. Don’t worry too much about it. But try and keep ROR between 2 and 4. If ROR is to high try with fan at F5 or even F6.

If ROR is higher than 5 after FC start, then do a shorter time to drop.

 Here a 400 grams roast of an Ethiopean bean (Homa)

Homa 25juni

 

Good tasting coffee with the same bean

Martin Kjeldsen and me (Therese) lives in different parts of Denmark, but we got the same bean – and the same roaster, the Bullet. The bean is from Uganda from Mount Elgon (on the border to Kenya). A real good bean.

We both have roasted the bean several times. The examples below is where we both were suprice of the taste. Remarkable good.

Martins was roasted 2 minutes from FC start to 192ºC. Mine was roasted 2:45 min and to 175ºC. But our bean probes do not measure alike. Notice the difference in FC start: Martin at 182ºC. Mine at 167ºC and 169ºC.

Martins roast, 500 grams, preheat 185ºC
Martin used it mainly for filter brew, but it also works as a lighter espresso.

kjeldsen-mt-elgon-2min20sek-okt2016

Therese’s roast, 400 grams

mt-elgon-great-2016-11-21-05-29-00Taste: no burn flavour, no bitternes, nice round and good aroma.

So even though the ROR curve dont look alike for Martin and my roast profile, its still taste remarkably good.

I roasted it again 9 days later
First the the curves were very close. But after 5,5 minutes the ROR levels differed. Here both curves:

Two roast of the same bean Mt Elgon nov2016

The higher ROR level at second batch gave an earlier FC; at 7 minutes … whereas the first was at 8:45 min.

mt-elgon-notgood-2016-11-30-06-39-48

This coffee didn’t taste good to me. It had bitterness and burned taste. And a weak aroma = boring coffee.

What could be causing this difference in taste ?

Alike elements
The development-% was around 24% for both (the time after FC related to total roast time). And weightloss was about the same: 14,3 and 14,5%

Development time (DT) was 2:44 minutes for the good one. The bad with more burned taste you would expect to be roasted longer, but no, its roasted a bit shorter: the DT was only 2:19 minutes.

3 possible explanations for the burnt taste

(1) A higher temperature rise in the bad one: 12°Celcius (and BT ended on 181°C) … where as the good one only rose 6-8°C (to end-BT at 175°C).

(2) When entering First Crack the bad one had a ROR around 10°C pr min. Whereas the good one was around 5. I have heard a recommandation around 5°C per minute – and not as high as 10°C when entering FC and for the rest of the roast.

(3) The american roaster Rob Hoos talks about the importance of the middle phase: from yellow point to FC start (see his book “Modulating the flavor profile of coffee“). He calls it the Maillard phase. In these two roasts the drying phase up untill yellow point are not that far apart: 4:00 and 3:45. But the lenght of this middle phase is 3 minutes for the bad one, and 5 minutes for the good one.

A roast consultant told me he prefers 3 minutes, so thats not criminal in itself. But maybe it suits this particular bean better with a slower roast; a longer middle phase and lower ROR levels during FC.

 

Working on a nice falling ROR curve

Read basics about the ROR curve here.

For many of the Bullet users it’s new to get a ROR curve. And the challenge is to control the heat and fan to get a nice declining ROR curve.

.

fil-24-10-2016-17-09-13October 2016, Kaffetårnet

Kaffetårnet is a small cafe in central Copenhagen, Denmark, with a Bullet roaster.

Here 750 grams of beans, preheat at 185 C. Settings:

→ first no power, P0 (thinking to avoid scorching of the beans)
→ when BT reaches 100 C, then P9
→ then stepping down: one step every 10 degrees rise in BT (bean temp) and stepping up Fan every 10 degrees.

FC was reached at 9:38 min.

kaffetaarnet-abdullah-750g-185c-okt2016

August 2016, Jed from USA

This roast was made by Jed:
jed-v-zanden-11aug-nr1

Jed wanted to see how the ROR curve acted when nothing was ajusted during the roast. So power is set on P6 (see the orange barr in the buttom) and fan is set on F2 (see the blue barr). Batch size was 450 grams and preheat was 160 C (a bit low, for 450 grams I would use 175C).

The graph shows how ROR rises after 8 minutes. Too much. So you need to turne down the heat earlier on.

 

jed-v-zanden-11aug-nr2In this second roast Jed reduced the heat: first steps down from P6 to P5 at 6,5 min and so on.

It helps. The ROR curve doesnt rise as much as before. But many prefer the ROR levels even lower around First crack.

 

You can work further on the heat adjustment. Like using Kaffe-Thomas technique with stepping down from 5 minutes, one step every one minute. And you can work with the fan setting. Excatly how much depends on the you batch size, bean type and roasting conditions (room temp and humidity) and … and ….

Larger batch size
Here is a roast from Eric with 1.000 grams Sumatra beans – again with no change in settings (apart for the first minute). Notice how the ROR level rise in the last part of the roast.

 

Eric Heflin 1kilo maj2017

 

Roasting small batches

Aillio says minimum batch is 350 grams to have the right BT measurement. If smaller batch the BT probe will not be sufficiently covered. But you can still roast smaller batches. Just ignore the BT reading.

Klaus investigated this with 100 grams batches to do sample roast. He preheated to 140 C.
→ Started off at P5, F2 and d6.
→ At 3:30 minutes lowered power to P4. He got yellowing around 4:30.
→ At 5:30 lowering to P3 … and even later lowering to P2

Klaus roasted 8 batches this way and got a FC start between 7:30 and 9 minutes.

His point with the 8 batches in a row was to explore the development time. The first was roasted 50 seconds from First Crack start. The next 1:10 minute. And so forth jumping 20 seconds all the way up to 3:10 minutes:

klaus-kampmann-8rist-af-mt-elgon

How did they taste ? Klaus reports: If you seek a light roasted coffee with a lot of fruity notes the roast with DT around 01:10 minute was the best. Better aroma than the 50 sec and the 1:30. For a darker roasted coffee with more body and chocolaty notes the one wiht DT around 02.30 min was good.

 

Coffee tasting good despite the ROR curve looks

These coffee tasted great.
Despite the Development time ratio was 10,8% and not between 20 and 25%.
And despite the ROR curve was not evenly declining. But still it was rising either.

benga-hill-17juli-cut

Repeating a profile – Playback

In the software RoastTime, there is a Playback button when you choose a previous roast. This function can repeat the settings of Preheat, Power, Fan and drumspeed to a new roast. It doesn’t adjust according to BT during the roast. It simply repeats the shift in settings along the timeline = repeat the program.

How to:

→ In Off-mode press the “Playback” button, and the preheating will start
→ When the Bullet says “Charge” add the beans, and the program will run
→ You have to stop the roast manually (so have this planned, whether its time after FC start or at a certain temperature)
→ If you want to roast it longer than the playback profile, look out to turn up the heat – since it goes to zero when the original roast stopped.
→ When done, press “Stop Playback” (otherwise it will go on in the next roast).

You can override the program and manually do another setting. But the program will take over again at the next change of setting in the timeline.

Look at the ROR curve during the roast – and how it fit with the ROR curve of the profile you want to repeat. This tells you where the Bean Temperature is headed.

Note: it will make a difference whether its the first batch you roast or a following. Because when the roaster isn’t fully heated, it will absorb heat – this gives less heating of the beans. See graphs.

View 2 graphs together: use ctrl while picking the roasts from the list. Then they show up together.

fil-25-10-2016-21-44-19-1This was a roast from Martin that I tried to repeat with the same beans.

The BT curve is not exactly the same. Most likely due to small differences the BT probe. But the shape and ROR curve looks very alike.

.

.

 

Here Steffen from KaffeRist made 7 repeated roasts of the same bean – this was even on the same day, on the same roaster. But still small variations happens. Notice how first crack (FC) starts at different time and temperature.

Steffens 7 gentagne rist april2017