Stretching Out The Roast
Stretching Out The Roast
by Christopher Schooley
What are you looking for in a particular coffee? Do you want a sweeter coffee? A brighter one? Are you roasting for espresso and would like more body and lower acidity? All of these characteristics can be altered through roast development, specifically by altering the length of time of certain parts of the roast. In the first part of this article, we'll look at stretching out the time after the end of 1st Crack and see if it has a tastable impact on the perceived acidity due to the breakdown of acids and compounds. In part 2 we'll stretch out the 1st Crack itself and note how it affects the perception of body and mouthfeel due to the breakdown of particular carbohydrates. We'll see sweetness greatly affected by the amplification of the interaction of body and acidity, as well as prolonged caramelization. For the 3rd part of the article, I stretched out the drying stage of the roast, when the bean is losing moisture in the form of steam but there is not yet any expansion of the cells. Will this have an affect on the finished cup and the perceived body/acidity/sweetness?
Stretching the Time Between the Cracks
I took this idea of stretching the time between the cracks and did some experimenting. This is generally the point at which under or over development occurs, though many are moving away from using that terminology because it can be so subjective. In four roasts, the time between the end of 1st crack and the beginning of 2nd crack is lengthened, and the roast stopped at the same point each time. Then by tasting and comparing the results, I arrive at some conclusions about what roast brings out the characteristics of the coffee I enjoy more.
I did four roasts of the same coffee, each time stopping at the first sound of 2nd crack. Each batch had roughly 30 seconds more time between the cracks than the previous batch. In tasting the results, I'm looking more at the effect on certain characteristics of the coffee and not the quality of the green coffee itself.
Because everybody will have a different roasting situation even with the same equipment, this article is less about how to stretch the roast and more about the effect in the cup. For this article I used a Probat PRE 1Z single barrel electric sample roaster. This machine can produce the same roast over and over again; with just some minor adjustments to airflow you can really shape the roast profile. The roaster does not have a bean probe, so my parameters were the physical changes to the beans themselves in relation to time. Here are the roast particulars.
Start temp: 320 degrees F
Yellow Stage: 5:15
1st Crack starts: 7:30
1st Crack ends: 8:39
Start: 318 degrees F
1st C: 7:36
1st C end: 8:42
Start: 320 degrees F
1st C: 7:38
1st C end: 8:46
Start: 320 degrees F
1st C: 7:33
1st C end: 8:43
The coffee I used was from the Coko Cooperative in Rwanda. This is an ideal coffee for this type of experiment because of it's a fantastic bourbon varietal with characteristics of cocoa and cola sweetness, orange blossom floral attributes and clean mandarin orange acidity. Delicate features with sustained sweetness, balanced body, and clean finish.
I cupped these roasts myself along with one untrained cupper and then again with a panel of 8 people, some trained and some untrained cuppers. I asked the panel to look specifically at which coffee had the most Brightness, the most Sweetness, the most Body. The panel did not know what they were tasting.
First I'll post my impressions from the intitial cupping:
- Roast A: wet aroma, bright and lively, short finish, front loaded, very sweet and bright on front of the palate. most malty sweetness Sweetness: yes, malt, sensed in the front of my palate; Body: condensed; Acidity: agressive, front-middle
- Roast B: long sweetness with a peak in the middle, more shape to body. more defined body Sweet: lasting through finish, candy; Body: syrupy; Acidity: middle
- Roast C: flabbier body wise, lots of sweetness in finish with brightness near the rear of the palate. front of palate is open which maybe lends to the flabbiness. Sweet: yes, a little cocoa sweet/bitter in finish, fruit sweet; Body: broad, thinner syrup; Acidity: middle, yet slightly muted compared to B
- Roast D: wet aroma muted brightness, the body and acidity seem to be more integrated, but with less dynamic in the cup. Not devoid of sweetness or acidity though acidity is muted in comparison. most bitter. more caramelized with slight roasty note in front Sweet: center, more caramelized, bitter cocoa; Body: flatter; Acidity: stretched throughout palate
In the panel cupping, the findings were:
- Roast A was the sweetest for half the panel
- Roast B was the brightest and was the favorite for most
- Roast C was the sweetest for half the panel, and also had noticeably more body
- Roast D least sweet, least bright, but more body although it was flat. The untrained cuppers noted that this was the most balanced cup.
With the panel, we discussed how the perceived acidity moved back through the palate with each successive roast. In the shortest roast, the acidity was front and center. In the middle two roasts, there was generally more brightness in the middle of the palate with the third roast having a brighter finish. In the fourth roast the acidity was very muted and the perception of the coffee was rather flat. There is an interesting geography to this idea if you look at it that way. Where we perceive acidity on the palate had a rather large influence on our perception of body. Where the brightness was more in the middle of the palate, the roasts seemed to have a rounder and open body, with more flavors throughout the palate, while the shortest roast with the aggressive up-front acidity had very thin body through the finish.
In terms of sweetness, the general rule is the more caramelization the less sweetness. But in these tests what was noticeable was not just the level, but the type of sweetness. In the shortest roast the sweetness was more malty while the second roast had more candy sweetness. The third roast had fruited sweetness and the fourth, more of bitter cocoa sweetness and began to show some carbon/roasty notes. Keep in mind that each roast was roasted to approximately the same level, they just took progressively longer times to get there.
What this line of testing shows is that altering the length of time between the first and second cracks can help shape the flavor profile of a coffee by featuring sweetness and acidity in different ways. There is some room to play here to get the coffee to express itself a little differently.
The panel discussed which roast would feature best in different brewing devices. The shorter two roasts would probably feature very well in manual pour over methods since the acidity would be expressed clearly. The first roast would probably be too sharp for a Roasted Coffee Pictorial Guide. ">french press or even auto-drip machine. The last two roasts would most likely be the ones to try as espresso.
Stretching the 1st Crack Itself
Now we look at the effect of stretching out the 1st crack itself and how that changes the sweetness, body, and acidity in the finished roast.
What is happening during 1st crack? The coffee expands and begins to release moisture, but more importantly it is when caramelization begins. This is important mostly because if caramelization stalls you can develop baked flavors in the cup - harsh cereal notes with straw and sometimes corn-like flavors. This makes the attempt to stretch out this part of the roast a little tricky, since you want to make sure that it never dips below the temperatures where caramelization occurs - between 340° and 400° INTERNAL bean temperature. Depending on your probe and roasting environment, this could be a reading of around 380°-415°, but a good rule of thumb is to just be sure not to let the crack itself stop or let it get below the temperature at which the 1st crack began.
I chose once again to use a Bourbon coffee from Rwanda for this experiment. A coffee that has some delicate floral and citrus features, but also has a range of sweetness with a potential for a balanced body and clear acidity. The end point for each roast was approximately 30 seconds after the last audible crack. I manipulated each roast with adjustments to air flow. For the first roast I made no adjustment to airflow, allowing the first crack to roll right through. For the second I increased air flow about 15 seconds earlier, then on the third roast about 30 seconds earlier, and so on.
I did 5 roasts, with the 3rd roast actually stalling. Since I tasted the baked effects of the stalling on my first round of cupping, I did not include it in the panel cuppings.
The Particulars: Coffee: Rwanda Karengera
1st C: 8:05
1st C end: 9:12
1st C: 8:04
1st C end: 9:27
1st C: 7:52
1st C end: stalled - 9:40 ish
1st C: 7:27
1st C end: 9:55
1st C: 7:44
1st C end: 9:44
My tasting notes from these cups, excluding Roast 3:
Roast 1: Body: clean, condensed,Sweetness: mid palate, not so much in the finish, Acidity: tea like, dry, center
Roast 2: Body: sheer, Sweetness: nice through finish, Acidity: mild
Roast 4 : Body: more syrupy, Sweetness: very honey sweet through finish, Acidity: peak in the middle, even some tang at the sides of the palate
Roast 5: Body: broad, big, Sweetness: marshmallow, sugar sweetness, Acidity: not as bright as the other coffees, but some in the back and finish
The first panel that I put these coffees in front of was a panel of untrained/inexperienced cuppers. Roast 4 was the universal favorite. As you will note, roast 4 actually has a longer 1st crack than roast 5; almost 2:30 mins compared to 2 mins exactly. With more experienced cuppers, Roast 5 was the favorite. With these longer roasts, the mouthfeel was more full and tactile which created a longer and sweeter finish, and in the case of Roast 4 it also allowed for a more expressive acidity.
Roast 1 was their second favorite of both panels, with a few people stating that there were aspects to Roast 1 that they liked more than either longer Roast. It had the clearest and brightest acidity. It was sweet, but not through the short and dry finish. There was an overall condensed feel to the cup with everything right in the middle of the palate. This roasting style could lend itself rather nicely to a pour-over brewing method and I could see using a coffee roasted in this way being used in a component of a blend as an accent, perhaps even a blend with the same coffee roasted more like Roasts 4 and 5.
The coffee that no one picked was Roast 2. I personally found it to be fairly mild with muted acidity and a sheer mouthfeel. Compared to every other roast, the cup was just generally flat. A more syrupy mouthfeel is related to the perception of particular carbohydrates that are released in greater levels with the stretching out of first crack. Stretching out the 1st crack did not have much effect on the sweetness of the coffee.
Stretching the Drying Stage
For part 3 I wanted to stretch out the drying stage of the roast, when the bean is losing moisture in the form of steam but there is not yet any expansion of the cells. There is some yellowing, but there is no Malliard reaction just yet. The Malliard reaction is the chemical reaction between a combination of sugars and amino acids which produce odors when cooking. In coffee, it produces those toasty aromas that you get before entering 1st Crack. Once the Malliard reaction begins, we start to enter the browning stage of the roast.
For this experiment I stretched the initial drying time, before any yellowing had occurred. I continued to use a Bourbon coffee from Rwanda so that there was consistency through all experiments.
Here are the roasts:
start temp: 300 degrees F
1st C: 7:47
end of 1st C: 9:06
Notes: air out to start, 50/50 at 1:30, all in @ 3:30, air half at 8:14, out at 8:45 secs,
start temp: 300 degrees F
1st C: 8:20
end of 1st: 9:40
Notes: air out to start, 50/50 @ 1:30, full in @ 4:30, air at 50-50 at 8:50, air fullis out at 9:20 for 10 seconds,
start temp: 300 degrees F
1st C: 7:05
end of 1st C: 8:30
notes: air out to start, air at 50-50 at 1:30, air full in at 2:30, air at 50/50 30 seconds after beginning of first crack, air full out 30 seconds after that for 10 secs
You can see in the logging that the variance between roasts is approximately 30 seconds, with Roast 1 being the control, or middle roast, Roast 2 the long roast, and Roast 3 the short roast. You will also note that the length of the 1st Crack, as well as the drop time (30 secs after the end of 1st Crack) are approximately the same for each roast.
For tasting, I again focused mostly on body, acidity, and sweetness; rather that specific flavor notes. Here are my cupping notes on these roasts:
body: 3D, bigger middle,
acidity: bright middle, and through finish, brightest acidity throughout the whole palate
sweetness: 2nd sweetness
notes: brightest dry fragrance and wet aroma, brightest break
body: longer, not as big in the middle, but present through the palate
acidity: muted, but present and balanced, light syrup
sweetness: more candy sweet, long sweetness through palate and finish, syrupy
notes: fairly muted on dry fragrance and wet aroma, muted break as well
body: slightly more condensed, short finish
acidity: tart, more forward acidity
sweet: sweet in front, but short finish
notes: more nutty on wet aroma, cinammon on the break, shortest finish
When tasting these coffees with the various panels, Roast 1 was the most common favorite, but most everyone did comment on the longer finish and sweetness in Roast 2. One common thread that I've seen in the panels for this whole series is that the people who prefered manual pour-over brewing methods prefered the shorter roasts in each experiment.The front loaded brightness and short dry finish has a certain appeal to them. In my opinion, I think these shorter roast profiles do fare well in pour-over brew methods, with the brightness prominent in the cup. But, in auto-drip machines and press pots, the dryness in the finish as well as aggressive acidity become a little too much.
The other interesting finding is that the results were almost identical to those in part 1 where we stretched out the time after 1st Crack. In both experiments, the longer roasts reduced acid, in a way similar to cocoa roasting. In the shortest roast, the brightness is all in the front of the palate; in the middle roast, the acidity is more centered, a more dynamic cup; and in the long roast brightness is in the finish but quite muted compared to the the other two roasts.
One interesting difference between this trial and the one in part 1 is that the long roast had a long sweet finish. In part 1 the long roast was pretty flat and the sweetness had begun to take on some cocoa bitterness due to prolonged caramelization. But here, in part 3, by stretching out the drying stage, acidity breaks down with no prolonged caramelization, so you are able to accentuate a pleasant sweetness. This is pretty key for roasting for espresso.
We've learned through this set of experiments that you can alter the perceived acidity by stretching out the time after the end of 1st Crack (Part 1), alter the perceived body and mouthfeel by stretching out the 1st Crack itself (Part 2), and alter the perceived acidity as well as sweetness by stretching out the drying stage of the roast (Part 3). The lesson here is this: if you really want to know the potential of any coffee, then it is smart to look at more than one roast of it.