Rehydrating beans - any experience?

[nrdlnd]

I have worked a lot with the rehydrated beans. I didn't have more of the Guatemala Asoproguate washed 2019 but I had a Guatemala Antigua washed 2021 that is rather similar. When I measured moisture it was rather similar maybe a little more moisture. The moisture is around 11.5% to 11.8%. I have used 95g beans and 5g H2O sometimes a few 1/10th of grams more. The batch moisture has been around 16.5%.

I've had some problems to get the developing phase ROR right and there were a tendency to a crash and flick at the end. I thought the PID-logic overreacted so I tried a power zone at the end of the roast. After many iterations the ROR looked rather ok but it didn't taste ok. There were a bitter aftertaste that I think you could call "baked" even if the ROR didn't look so bad. So I went back to the PID-control. I've found out that what's happening before 1st crack very much affects what's happening during the developing phase. If you have too much heat before that phase there is a risk for a crash and a flick. Now with my last roast I had less boost before 1C but continued during developing (during the cracking phase when the beans give away a lot of moisture) and started the negative phase 3 later when the beans go exothermic. This is the last log:

220219_GuatemalaAntiguaRehydr_DLightR2fanmod_1.2.png (100.03 KiB) Viewed 6560 times

L became 1.2 and it is a light roast but not very light (the lightest I've made is 0.99-1.0). I stopped the roast at DTR 20.5% but I could have stopped it earlier to get a lighter roast. I always cut the beans in halves to see if they are evenly roasted inside and it was (as it seems to always be with this profile). The original D-Light profile from Damian has a very high fan speed. I have lowered it a little in the beginning and more at the end to lower the power demand. The fan speed is still high compared to most other profiles though. It's not possible to see in the log as there is no scale. The max needed power is moderate at 1.38 kW. I think the ROR looks nice!

I haven't tasted this roast yet. I have tasted the other roasts. I haven't made a cupping session as I'm "Post Covid" and I can't taste the finer nuances. I have instead made brews with the best drip coffee method I have and I can taste if the brew has a nice and balanced acidity and sweetness. I'm sensitive for bitter taste. I have mostly enjoyed the lighter rehydrated roasts. The lightest roasts have been with rehydrated beans. The beans that haven't been rehydrated tend to become more medium light with a L around 1.7 at the same DTR around 20%-21%.They were more bitter but not uncomfortable. They tasted quite ok but wasn't as clean and balanced as the rehydrated lighter roasts in my opinion. Rehydrated beans come out lighter at the same DTR. More hot steam make the roast more efficient.

This is the latest incarnation of my modified profile:

DLightR2fanmod.kpro

(4.06 KiB) Downloaded 261 times

If you try it you may have to modify it for your bean. I think it can work well also for non rehydrated beans. The default L is 2.0 and that is a good start to be able to mark the 1:st crack. I don't think you should go higher than that level. The very first single crack tends to come rather early with rehydrated beans. I wait until there come more in succession before I mark 1st crack. I'm using my mobile to record the cracks as a visual help: https://kaffelogic.com/community/viewto ... ?f=5&t=326.

[nrdlnd]

Hi,
Maybe someone is interested what happens with another bean?
This time a heirloom Ethiopian I've never tried before. A Limmu Kossa/Sakka grown in high altitude around 2000m. Fully washed. This is not a single origin. The coffee is sold via The Ethiopian Commodity Exchange. This means the coffee comes from many small farmers. It's grade 2 and the size of the beans are varying. Anyway the beans are smaller (most of them!) than the Guatemala Antigua. My reasoning was that that I could roast a bigger batch 100g instead of 95g + water for rehydrating. I'm not quite sure this was an OK decision. I measured the moisture of the beans and added 5g water to the 100g beans. Anyway this is what happened when I used the same profile as before:

220310_LimmuKossa⁄SakkaRehydr_DLightR2fanmod_1.2.png (78.83 KiB) Viewed 6483 times

The end of the roast is worse than with the Antigua. There is a tendency to a crash and flick. The roast became very light (Cinnamon) and that is good in my world! I tasted it 12 hours after the roast and it tasted rather good. I don't think it was very baked. I decided to make a new try. I let the zone 3 start later and my thought was to give some more heat during cracking so I let the +1 zone 2 extend until zone 3 started.

220311_LimmuKossa⁄SakkaRehydr_DLightR2fanmod_1.3.png (76.5 KiB) Viewed 6483 times

Well maybe a little better but still a tendency to crash and flick. I tasted the roast 4 hours after roasting and it tasted ok but the 1st roast that had rested longer tasted better. I have had earlier experience that if there are too much heat before 1st crack it can induce a crash and flick at the end of the development. At the same time I think it's difficult to predict what the PID-regulation will do. So I made another try. I stopped the +1 zone2 earlier before 1st crack and then started the -3 zone3 later. This is what happened:

220311_LimmuKossa⁄SakkaRehydr2_DLightR2fanmod_1.3.png (78.09 KiB) Viewed 6483 times

This time it looks better. The beans are a little darker but still light. The end temperature became only a little higher (a half degree). The needed power was a little higher this time and I don't know why. To make it even better maybe make zone3 more negative -4 instead of -3? Important it's not stalling. Edit: Other possible changes is to let the second zone2 end even a little earlier and also extend the zone1 longer (and thereby shortening the zone2 even more. Maybe easier to control with the lower 95g bean weight I had before.

This is the profile with the changes in the screen shot above. It has the same name so you may want to rename it if you have downloaded the earlier version before.

DLightR2fanmod.kpro

(4.15 KiB) Downloaded 278 times

What is your experience working with zones? I'm trying to learn from my mistakes and also when it goes well but it's not easy. Edit 13th of March: Tasted the coffee as filter coffee today after 2½ days rest: Balanced, sweet and juicy. I haven't compared the three roasts directly but this last roast is sweeter and more juicy. I have measured the roast degree and it's bordering very light but I will call it "on the lighter side of light". It's still well developed with a DTR around 20.4%.

[nrdlnd]

I have made further small changes to the profile now called "R3" and then roasted two different beans in sucession. The changes I made was extending the positive +2 zone 1 to 4:20 and end the positive +1 zone 2 earlier at 5:35 well before 1st crack. The negative -4 zone 3 is unaltered. My reasoning for having positive zones in the beginning is because I have rehydrated the beans to about 16% moisture and that more heat is needed because of that. The reason for having a negative zone at the end is that ROR should decline steadily to avoid "baking". Both batches are 95g beans. The first Sidamo Wendo had a moisture of 10.1% and it got 7g of water 12 hours before the roast. The other bean a Guatemala Antigua had a moisture of 11.4% and got 5g of water. Both beans are fully washed and high altitude. The Antigua has more even size and they are a little bigger. The Sidamo Wendo is Gr2.

220322_SidamoWendo_DLightR3fanmod_Rehydrated_1.4.png (79.6 KiB) Viewed 6418 times

It became a light roast with L=1.4. DTR 20%. At the end of the roast 1.4 kW needed power. Evenly roasted. End weight 83.0g. The ROR-curve isn't bad. Maybe zone3 -3 instead of -4? A power zone at the end because of the alternating power could also be an alternative but my experience is that it's difficult to dial in.
The Antigua is even better:

220322_GuatemalaAntigua_DLightR3fanmod_1.3.png (77.83 KiB) Viewed 6418 times

This is also a light rost with L=1.3. DTR 20%. Evenly roasted. End weight 82.6g.
I think it's interesting that the profile worked well at the first time with two different beans even if both were washed and high altitude. They had to be rehydrated differently as the Sidamo Wendo was drier.
This is the last profile:

DLightR3fanmod.kpro

(4.31 KiB) Downloaded 323 times

EDIT: I have tasted the roasts as pour over. Both roasts were light roasts but the Sidamo was lighter bordering to very light even if "L" was a little higher. I tasted both roasts the day after and also a couple of days later. The Sidamo was good already the day after as these beans have a high sweetness. The Antigua needs more rest as it's a bit too acidic for my taste but otherwise nice. None of these beans are from selected lots but they are of good quality.
EDIT 220401: I have tried the last iteration of the profile with two new batches and different beans (both high altitude one washed and one natural) but changed Zone 3 to -3 instead of -4. I think it was better. The log of the natural looked very good. When you look at the pictures of my logs the ROR axis is X5 (the Y-axis) and makes it easier to see ups and downs. If you have lower X on the Y-axis of the ROR curve it will look "better" but you will get less information about eventually needed changes. I've kept 30 sec smoothing of logs as I think it's easier to see the bigger changes of the curve over time and I think they are more important and also possible to adjust. You can change settings in "Options" and "General options". Important to have the same settings when you compare logs. EDIT: About the needed power. In my case it's around 1.4 kW at the end of the roast. I have a 220V machine but my voltage is 230V. I'm roasting in the kitchen and the temperature there is around 23 degrees even during winter times. It's more than 1.4 kW as available power is dependent on available voltage AND the temperature in the environment. I have still not enough power for D-Roast for example as the fan speed is too high at the end of that profile. That's why I've modified the fan speed. If you have worse conditions than I have it's possible to lower the fan speed even more at the end of the roast.

[nrdlnd]

I roasted a single origin Finca El Socorro from Guatemala. High altitude and natural processed. This is a high quality bean. 95g beans and added 5.6g water 9 hours before the roast. DTR became 20.5% and end temperature 217.4 deg. L=1.4. I've measured the roast and according to the measurement a very light roast (Cinnamon). Taste as filter coffee one day after the roast and also after one week rest a nice and balanced acidity and sweetness. I'm still post Covid so I can't taste the finer nuances. I've ordered more of this bean.

220401_GuatemalaFincaElSocorro_DLightR3fanmod_1.png (80.82 KiB) Viewed 6311 times

This time zone 3 = -3. ROR looks nice but I think it's possible to try with less negative zone 3 and it may be possible to let the PID logic handle it without modification? EDIT: I've tried with -2 but -3 and -4 is better. I still think the extra heat in zone 1 and 2 is beneficial because of the higher humidity of the rehydrated beans.

This is how the log looks with 15 sec smoothing instead of 30 sec. I'm not sure it gives more information. I get about the same info with 30 sec smoothing.

220401_GuatemalaFincaElSocorro_DLightR3fanmod_1.4_15sec.png (85.08 KiB) Viewed 6304 times

Edit: This is the default setting X1 and 1 sec smoothing (excuse that the screen shot looks different in the last picture as there is more information with the side panel. It IS the same log but I marked the used power a little differently. That doesn't affect what I want to show:

220401_GuatemalaFincaElSocorro_DLightR3fanmod_1.4.png (89.6 KiB) Viewed 6291 times

My setting with X5 and more smoothing (in my case 30 sec) I think gives more information about the bigger changes that I can use the zone boosting function for. What do you think?

[nrdlnd]

Hi,
I'm still rehydrating my beans and I'm working with my adaptions of the original D-Light profile from Damian.
I've made a "smaller" modification of the original profile by making it a little steeper and I've also lowered the fan speed at the end of the roast. My intention with this is to shorten the development time. My hope with this change is to lessen some bitterness. I have made comparisons and the development time becomes shorter but "L" also becomes higher at the same DTR. The taste is also less bitter. The beans are still well developed inside-out. For some beans the former R3 profile is better but for other beans the new modified R4 profile. Weight loss is a little less with the R4 profile even if the roast tends to be darker. It has to do with the shorter developing time I think. I prefer my roasts rather light. The R3 profile tends sometimes to be too light even if they are fully developed. I'm using and comparing both variants with a new bean. Default "L" is still 2.0 but I recommend to start at L=2.5 if you mark 1st crack and check DTR and end temp.

EDIT 2022-06-16: I made a roast with a single origin bean from Guatemala and this time with the 'R3' profile. My notes is the following a couple of days ago: "95g+5.4g Finca el Socorro 2000m Natural. End weight 81.9g. Light. Evenly roasted inside/out. ROR looks better than R4 yesterday. Went a little higher in DTR and temp than I usually do with R3. Could go even higher. Taste after 23 days rest: Very good taste, balanced acidity, nyanced and fruity. Some chocolate. No bitterness. Good aftertaste. Comment: DTR was 22% and L=1.4. Measured Tonino=110. This light roast needed more than 3 weeks rest to give its best! Some of my 'R3' roasts have been very light (Tonino around 120='Cinnamon') but I liked this better but it's still a light roast. What I think is important is to test a profile at different roast levels (different 'L') and also especially for lighter roasts taste them after different rest time. My wife commented that she very much liked this roast and this was without her usual milk!

DLightR4fanmod.kpro

(3.03 KiB) Downloaded 306 times