Kenya Peaberry roasting advice


I’ve bought some green Kenya Peaberry and would appreciate advice on roasting please? I’m a beginner so words of not many syllables would be helpful!


My go to profile for most africans is extended Shakiso profile. Just make it a few minutes longer, and watch the roast, let it through the crack and then it depends on your preferences. Stop it manually 30s to 60+s after the first crack ended, maybe if you have enough beans make one drop 30s after 1st crack and second batch drop at 60s after crack (or if you prefere darker then maybe move those points further awayfrom the crack) give it a go and adjust according to what you taste …

Peaberry is different, so it might not be the best profile, but so far if I started an african bean with this profile it produced a nice roast. Except of one pesky rwanda that just did not like any profile i tried :wink:


My understanding is that Kenyan peaberry is very dense and can take a lot of heat. The Burundi profile might be a good start. Or something in between the Burundi and Shakiso.

Not roasted any myself so I’m guessing here.

As Pavel says, extend the profile you choose and cut when you think appropriate.

Let us know how it goes.


So I’ve roasted quite a bit of one type of Kenya Nyeri pea berry, and I keep being surprised by how quickly it roasts and how little heat is needed compared to what I’ve been told. I’m starting to suspect there is a bigger difference between drum and air roasting for pea berries than other beans. My theory is that PB’s, due to their shape, are less susceptible to conductive heat and more susceptible to convective heat than the average bean. So, in brief, the more aggressive approaches haven’t played out as well as say, something like kochere debo. Still have some tasting to do, but I might start with something like that and stop pretty soon after 1st crack.


Good to know - i have not yet roasted an african peabery, so … if I get some I will use some less agressive profile then.


That is very interesting and unexpected finding.

Have you found similar with normal Kenyan’s?


My suggestion is going easy on them because since the air roaster heats better from within to out, the denser the bean the less heat needed. As pavel made a good suggestion to drop @30s then 60s but I feel the window is more like 15s and 30s and dial it in in 5s increments.


As I said in the post, I am a complete beginner so need some help in identifying things like first crack etc. For info the beans are from the Tatu are and have been sun dried.
I do want to do more than simply buy the Ikawa beans and use the standard profiles but it’s a big leap from buying a bag of roasted beans!


Why then do the roasting notes on the Burundi say that a high level of initial heat is applied due to the density of the bean?


This is intentionally a bit simplistic but might be useful to you…

As you follow the Ikawa profile you will observe the beans changing colour from green through shades of green to mustard/yellow then beginning to brown then darker shades of brown. First crack is simply an audible popping or cracking noise that will be heard on the last leg of the profile. It usually starts with one pop and some seconds later more beans pop. The bean can’t handle the increasing internal pressure and explodes. After a while the cracking stops and cooling begins. You will smell the changes coming from the exhaust as well. Grassy to bready to more roast like as they crack.
The Ikawa profiles are very light and sometimes a few beans will not have cracked, so inspect the roasted beans and discard any that look very pale or uncracked. Very important with naturals as they tend to have a larger occurrence of this than washed.
If the profiles ran for longer then the beans would darken further and start a second period of cracking called second crack (not generally as loud). Darker roasts will go on past this point.
For filter generally speaking, roasts will finish between 1c and 2nd or on the first of the second cracks. Where and when is personal preference.

Why not extend the end points of a couple of Ikawa profiles and carefully observe the beans as they roast? You can stop a roast at any point by pressing the button on the machine and cooling will start.


You’d have to ask whoever created the profile…also a created profile is personal and someone created it.

Also don’t read stock profiles like they are King James Bible, not only is this roaster a small air roaster by the way it roasts is different too so there is nothing written in stone. Everything is up for critique and analysis. Try everything, trust nothing.

I’m just sharing that through my observations that this roaster roasts inside out.


How the roaster deals with density is obviously crucial to how we decide to roast a bean.

The 6 stock profiles seem to me to pretty much correlate altitude (assuming density correlates with altitude) with initial heat. It seems unlikely that is simply by chance. But it may be.

The Brazilian is started off the slowest. Your observation would, as I understand it, be the opposite case and have it start off the most aggressively because it is less dense?


Thanks for the helpful advice. I need to start playing a bit.


yes I agree, drum roasters created these profiles, I would guess the one when a negative sloping fan either worked with an air roaster before or just got lucky (like I did).

Yes my untested hypothesis would be to start aggressively or in other words give more power (harder/steeper slope) during the yellow if looking at my A profile. Though I haven’t worked with a not so dense low grown bean yet, I got an Ethiopian and Guat coming in 5lbs and a 1lb or Sumatra and prob won’t be picking up anything new (prob 3 months) until I finish them. Which I probably will then pick up Columbian and/or Brazil beans.So I guess I’m relying on you to risk or test your beans.


I’ve got a fair bit of the Brazilian left so happy to sacrifice for experimentation.
I have roasted it on a shortened version of your “A” profile and it turned out ok with more body than the stock profile.
Also on the Andino profile where it starts fast but I’ll need to amend it because the beans barely got to 1c.


My real qualm is not that a spike is bad but the end of the decrease should be when the beans are "turning around and there is no way without a bean temp probe to know when/where the right time is. That and going as high as 400*, my guess is for where the turn around is somewhere in the input temp of 350-375*F


Deven what exactly this turn around means for you on Ikawa? Because I always see a turning point to be a point when the temperature of drum roaster turns from declining to rising again, but, this is not the case with Ikawa, so, not sure how to translate it here.


Since the “turn around” point is when the bean temp probe starts coming back up, it is also when the beans are said to “start” the drying phase (actual bean temp of 212*F), so in relation to the IHR…it is when pale starts happening. Even though that is an oversimplification since the breakdown of chlorophyll aka color change doesn’t give the exact bean temp, just an approximation. Maybe one day I’ll have a tricked out IHR with a hygrometer to see an increase in humidity.


Since the Ikawa and beans start cold then the initial ramp is convective heat. Presumably the fall in RoR of input heat and decreasing fan speed will correlate to when conductive heat is now the on level terms or more significant . Perhaps this is analogus to the turn around of a drum roaster?

The problem being how do you where that point is?


The IHR technically starts at 150*F, I always do a pre warm up on my roaster. @450 for 8min with the least of amount of fan speed and time on cool down. Maybe for this next batch I will try not preheating the roaster and do a comparison. But I doubt I’ll like it, since I’ve done some testing so far…

The fall in ROR should correlate to the beginning in the changes of color…Conductive heat will never(?) be more significant than convective, the slow decrease in fan speed is due to the starting of conductive transference (+/-) of heat but at the same time taking the fact that the moisture loss is making the beans lighter too so I don’t want the beans to fly away or be pressed against the wall too much. (or maybe i do(?)…i dono yet) And also decreasing the pressure within the system to not strip the volatiles.

the “(?)” is because I could be proven wrong and I wouldn’t mind.

like I said… when pale starts.