Rwanda Inzovu Espresso Roast



I purchased these beans here.

We have settled upon the following IKAWA roast recipe. Tasting was done 7 days after roasting.

Tasting Notes: A light and delicate bean with good florals. A fruity freshness with notes of apple, vanilla, black tea and plum. The apple really does come through with the roast above in my opinion.

Origin: Rwanda

Estate: Various

Varietal: Mibiriza/Bourbon

Grade: Conventional

Certification: n/a

Processing: Wet Processed


Hey, thanks for the profile! I am still currently trying to figure out this Tanzania Peaberry (and my updated roaster) and I’ve almost thoroughly exhausted though my experiences from the old roaster. I do believe I have some Rwandan beans around here somewhere I can dig up and test, but they are a bit “home aged” so to say. Would you be able to post how you approach this style of roasting? What do you feel like is a good temp/time range for the various segments and how much development/time does this style take after 1C. Are you getting into 2C with these roasts? Any and all input is welcome, I’ve started threads on profiling if you want to add onto them or post a new thread, thanks in advance!


My method is simple. I tend to buy a kilogram bag of green beans. Then I look for something similar, typically using the preparation method and bean type as the key influences. Also, often the bean’s seller will give you clues as to where you should be taking the roast. If there is nothing obviously close, why not start with the default roast? So I have a series of Kilner Mini JARS labelled 1-5. I will try my starting roast (sometimes looking up roasting tips in a Coffee Bean guide). Then I listen for the 1st crack and note down that time. That should usually be about 3 mins in. Now, at the end of the default roast, look at the beans, and try chewing one. There is a lot of information in the crunch to tell you if its right. If they look under developed, try a 2 at a hotter temperature, and the a roast 3 at a longer time at a lower temperature. I am not a huge city roast fan and don’t like ashy coffee. I will create five variants in all. Then leave them in the jars for a week. Then my boys and my wife have a tasting session. We usually have a best candidate at that point, and I then repeat the process using that winning recipe as the starting point. I will increase and decrease the temperature by about 5%, and do the same to the duration. We then wait another week, and try these variants. Usually at this point, it’s pretty decent. I love peaberry beans but found that they are a bit more delicate. This is allegedly due to due to their difference in size and density to regular beans, and so the heat is transferred throughout the bean differently. I find low and slow is usually better. Also, my personal tip … get a really bright LED torch and clean the glass lid of the IKAWA. I leave the torch pointing down and find I can then watch the colour change in real time. So, I would find the temperature and time for first crack, then drop the temperature down by 20-30 degrees, and vary the duration - low and slow is the starting point.

I can see a Kilo of Kenyan Kisii Peaberries for sale. What is bothering me is I have purchased them before, and my recipe is missing from the app. I’ll buy a kilo and let you know in time. When you say your beans are “home aged” - are you talking several years? Green coffee tends to last well.

As for the second crack, I tend to rarely get there in my final roast profiles as that’s what we tend to like. I don’t think it’s a good idea for peaberry either … but roasting is an experimental and experiential process and as much art as science.


That’s great, thanks for sharing your insights. While I do think there can be a lot of variations in preferences for different roasts, I’ve found peaberries both dry and wet processed do well roasted darker than regular beans, but wet processed coffee’s regardless of variety, tend to burn out if roasted to the same temp spec as a dry processed and need a lower finishing temp. From what I’ve seen, peaberries tend to show the fruit roasted much darker in color. A 3 min to first crack is interesting given the overall time of your profiles, those kind of profiles still elude my understanding when trying to compare to drum roasts. So far these Tanzania PB WP did ok with a Rwandan style, I have yet to try your Rwandan profile on them but they also did well with the NFLT10 found in the Washed Ethiopian thread.


Dear Deven,

My first salvo at cracking the Kenyan Peaberry has begun.


I did some academic type research on the known issues before, There is an interesting paper here which describes roasting the peaberry in a device logically identical to the IKAWA.

They suggested that 5 minutes at 250 centigrade is optimum.

I noticed on my first attempt that the relative high density was resulting in poor movement of the beans, and so increased this quite a bit for the following roasts.

I stuck around the 250-260 range as this was working well, and found I could hit first crack with a nice gentle ramping up at about 5 minutes. Then I would typically hold for five minutes.

On the middle sample, I increased the time duration to about 8 minutes, and due to the angle, it’s not that clear, but the volume was significantly increased in the middle sample. All roasts use exactly 50g +/- 0.2g

I haven’t done a brew with any of them, as I leave for 7 days. However, based upon my crunch and chew tests and knowing what good beans tend to taste and feel like in the mouth, I think it’s one of the latter 3 roasts that will win out. DW, DS1, DS2 and DS3 will form a tasting panel next weekend. See here for what those acronyms mean, but at the moment, DD1 is away, DD2 only drinks coffee with confectionary on top and stirred in, and DS4 is ambivalent , hence that’s my tasting panel.

My research into these beans led me to discover that PEABERRIES also have 13% more caffeine in them which also has an impact on the sex life of mice (see here). There is some weird stuff you learn when you go researching. In a past life, I developed and launched the UK’s first caffeinated vodka (search for Alice Extraordinary Vodka - there are many traces left like this review), and that’s where I learned a lot about caffeine and its effect on taste - and oddly enough, some other ingredients that have linkages to the libido. I was then taught a very hard lesson about the costs of launching and developing a spirits brand (mostly due to the costs of duty payable when giving away samples and the oligopoly markets you have to pay to sell your products through).

If this caffeine level thing is also the case, I will look to reduce the water temperature in my espresso by a degree or two, and use slightly finer grind to create a bit more back pressure and maybe a slightly smaller does. The lower temperature tends to result in less bitterness coming through.

For those interested, I sourced the beans here. The description from the seller is:

  • Origin: Kenya PB
  • Estate: Kisii Estate
  • Varietal: Heirloom Typica, SL28
  • Grade: PB
  • Certification: Conventional
  • Harvest Year: 2018/19
  • Processing: Wet Processed
  • Tasting Notes: Great floral aroma, strong notes of prominant stone fruit, cane sugar and molasses.
  • Roasting Notes: Will darken quickly, but follow it through to just past first crack. This gives it time to round off a great bodied coffee.

An update next week then. Regards, Cliff.


Hey Cliff! What I am about to say may come off as somewhat a poo poo your parade but please do not think that my intention is so. I am 150% looking forward to your results and enjoy your enthusiasm immensely. But I wish to help you by correcting some of the things you mentioned, not as a deterrent or because of my god-like complex but something to help your Google-fu, so by continuing your research you can expedite your understanding (if you wish) and ultimately yours (and I) roasting for the the most exceptional cups of coffee.

So first of all, those glass jars you store your beans in are top notch. I am a sucker for glass and those types of jars are my favorite.

How do I put this…typically a normal fluid bed is like a spouted bed (think of water fountains like the Bellagio) however, we are using a type of centrifugal fluid bed that is based on a vortex generator. This type of heat transference is much more efficient and precise than a typical fluid bed because we have a forced laminar flow in a vorticity (the shape the air moves in a upward spiral movement) that transfers heat much more efficient than a typical spouted (fluid bed) or a basic laminar flow. All of what was previously said; is because of the shape of the air (and the equal movement of the beans) the heat of the air has a much more even heating of the beans. So we can be more precise for all beans at the same time, not only in what way we apply the heat (fan speed/bean height) but also durations and timing. Vis a vis, if you find a profile that resonates with the family, tuning the different segments of the profile based on your experience and tastes, usually will lead to an optimized eye popping profile, if the profile itself isn’t already your ideal. Though like your DD2, I started out on Starbucks white chocolate mochas. I still love them, and roasting a bean for that type of drink is still no walk in the park. But maybe, just maybe one day you may roast something that dissuades her from her evile ways.

For me, the understanding of what profiles achieve what results, is the goal for me. But of course, the extent of our satisfaction and tweaking of the profile (what does what) is solely up to you. Sometimes (often in the past), once roasted enough, references and experience with a type of bean, a base profile usually was close to the ideal and so I could play and experiment a lot to see what else was (if/if not) there. Ah…but please forgive my rambling on. Currently, I’m working on a Rwandan Peaberry and my typical “A” profile does not live up to its name, I did adjust the profile and listed them below, the original is based on a Sumatran that was traditionally processed and the second profile is adjusted for a Rwandan wet processed Peaberry. I elevated the initial heat and lowered the ending temp, because of the processing, and shortened the profile (density). Maybe this was the correct thing to do, I will know in a week or so. Initially, I always brew once the day after roasting but the 2nd and 3rd brews of that roast I also do anywhere from 3-14days, I do have many roasts on going simultaneously, so it is a bit random.

As a healthcare pro, I can’t say I enjoy that you are mixing alcohol and caffeine, but Cest la vie. Yes certainly libido is effected, you may want to try eating the rind (white part) of a watermelon. But don’t blame me for the results. Also caffeine is highly soluble, however I doubt you’d be able to taste it’s bitterness. The bitterness is mostly coming from the high heat intensity from the roasting profile forming other things. Lowering the temp will only dissolve less of the things you created when roasting.

Happy roasting, I look forward to hearing what the panel says!


Ahh, the moral burden of being a healthcare professional. Oddly, we had conversations with a big pharma at the time, as it transpired that our recipe was almost the perfect treatment for patients presenting with strokes. There was something about alcohol, caffeine and riboflavin which made them very effective as a treatment - see research here and here. Anyway, my hopes of securing an alternative and more lucrative market came to none, but it just shows you. Also, the worst thing in classic Redbull (aside from the taste and with or without alcohol) was the dramatic increase in the risk of heart attacks and the irony was, I might actually have made the antidote :slight_smile: Still, that is all history. What amazes me is that coffee has such a good rep for health still, especially as it’s roasted - which is normally bad news on the carcinogen front, but it continues to be more of a saint than a sinner in research.

As for the other research … it was just a starting point for me. What I have done is gone longer, shorter, higher and lower, and we’ll see what happens.

As for caffeine being soluble and not that bitter … we’ll have to agree to differ. I have a pot of pure caffeine on a shelf here, and I know for a fact, it takes a lot of agitation to get it to dissolve (think minutes with a bamix - while salt/sugar dissolve almost instantly); and it genuinely is truly bitter.

As for those little pots … I have mixed feelings about them. On the plus side, they were in effect free and come with Heston’s Pate in Waitrose (it is very delicious, and to avoid a health lecture, if you are pregnant, don’t eat it as the levels of vitamin A are too high, and also don’t drink). The rubber seals naturally allow for venting, plus they are exactly the right size for a 50g IKAWA roast. The downside is, they are not light fast, so I put everything in a large biscuit tin.

The other thing you can do with the larger kilner glass jars is fast track cold brew. Here, what you do is put a good dose of coffee in, I do about 25% by vol, fill up with water filtered using the zero water thing; and then immerse in a sousvide water bath. See the process described here. This neatly links back to your comments on caffeine, as one of the defining characteristics of brewing this way is that caffeine is not extracted at this low temperature, so you end up with a super smooth brew with no bitterness and the maximum extraction of all the healthy bioflavonoids and other anti-oxidants.


Yes it is true that alone pure caffeine is bitter, but I would like to dispell the myth for you and others, that the bitterness is from the caffeine alone. Or that matter, caffeine with EGCG or Chorogenic acids, quinine and all the rest that alone are very bitter that come into brewed coffee at any roast degree. But importantly, that I can have a highly caffinated cup, even a pure robusta, that is not bitter but sweet to the taste, even a decaf can be sweet or bitter but it is because of the roaster (person). My point being that it matters significantly how the coffee is handled by the roaster. Not only in sourcing and the quality of the crop but once the person roasting receives the greens they sort through it and pull out any defects and then roasting the beans well so they go through all the proper stages and hit all the right marks to achieve a delicious roasted coffee. Though people enjoy strong bitterness in the cup, preference is important. I find it is necessary, as a part, for the balance to be achieved. I wrote the above to (in general) dissuade people from thinking that coffee can only be bitter, or there can only be one cup to rule them all. The best roast is the roast you prefer.

As I continue on my roasting journey, often times the myths propigated around coffee have led me put in lots of different (read: wasted) efforts like brewing methods to mitigate things like bitterness. All to really find out if I can roast the coffee well enough. I can brew with boiling water, using an espresso grind in my v60 while constantly whisking the slurry and still have a perfectly juicy, sweet cup of coffee. While I enjoy all the different brew varietals like espresso, immersion and percolation. I vary my brewing method not on mitigating the defects of my roast but things like the number of people I am serving, or what I’ll be doing while having the coffee or if I am using the coffee for a cocktail/ingredient or feel like adding milk and sugar, etc. If anything I am trying to brew as intensely or loosely as possible so that I can highlight even the faintest flaws of a roast. So that when I feel like plugging and playing, pouring and chugging a roast with no alterations, I could. I am not perfectly there yet, but I have come as far as to not only see the forest but also the trees.

I think the solubility (at room temp) for caffeine saturation is 2.17g/100ml of water. However sugar is something like 2g/1ml of water. So that is why you will see sugar dissolve so fast. When I said “highly soluble” I relatively mispoke, I meant it more from my experience rather that what typically most people would consider highly soluble like sugar. A funny thing about sugar is while saturation is 2.12g/ml the super saturation of sugar to water is more like 16:1 so it would only take 1g of water to liquify 33.92g of sugar at room temp.

As for your jars, an easy solution might be to wrap them in aluminum foil.