Roast Approach: Swiss Water Process Rwanda Muhororo Natural

Today's coffee that we are going to approach is the Swiss Water processed lot of Rwanda Muhororo Natural.  The coffee comes from the Nyamasheke region of Rwanda, on south western edge of Lake Kivu.  The coffee is 100% Red Bourbon, characteristic of Rwandan production.  The coffee is natural processed.  Once the coffee had arrived in the United States, it was then taken up to the Swiss Water plant for decaffeination.  The decaffeinated coffee arrived back to the United States at the end of April, 2026.

Arrival QC:

The coffee arrived in February, 2026 to the United States, at which point part of the lot was taken up to the Swiss Water decaffeination plant in Vancouver, British Colombia.  When the coffee arrived to the United States, it had a humidity level of 11.8% and density of 1.15 g/ml.  After decaffeination, the coffee had a density of 1.20 g/ml and humidity of 11.6%.  The decaffeination process made the coffee more dense, but it retained a very similar moisture level.

The sample roast was preformed on an Ikawa with the roast profile that can be found on our website.

The Roaster:

For these roasts, I'll be using a Proaster 1.5 kilo drum roaster hooked up to natural gas.  Batch size is 400g, approximately 35% of roaster capacity.  I've used a small batch size on this roaster to get closer to mimicking the gas power that you might find on a larger production roaster.  It must always be mentioned that each roaster is different.  The type of probe, placement of probe, and batch size will all influence the way in which bean probe temperatures are registered in a roaster.  With that said, don't read too much into the exact temperatures, but look more closely at the theory behind the approach, time spent in different development periods, and rate of rise to get a better idea of how to translate this profile over to your own machine. 

The Goal:

When I started roasting this coffee, I began with roasting 4 different roast profiles.  I played around with heat application during various phases of the coffee roast.  I also roasted one of the profiles with 3 different charge temperatures to see how the coffee reacted to initial heat application and momentum.  On the cupping table, all of the coffees were astringent and lacked sweetness.  It was an astringency that I had never encountered before when using these same roast profiles.  The common cup notes for all of the coffees were lemon balm, red berries, lemon grass, and astringent.  I knew that there was more sweetness to this coffee, but I was struggling to unlock it in my little drum roaster. 

So I went back to the drawing board to try and address the sweetness and astringency issues.  I roasted the coffee three additional times.  Below is the roast profile that I found most enjoyable.  As many may know, this is one of the coffee components for our 2026 Hacea Roasting Championships.  I would like to take a moment to mention that I am approaching these coffees the same as all of the competitors.  I am not one of the judges, so this is certainly not prescriptive by any means for the competition.  My hope is that my experiences with this coffee will give you some guidance as you approach how you will roast it, and also potentially provide you with some troubleshooting help.

Hopefully this will provide you with the background you need to confidently approach this coffee in your own roasting adventures.

If you would like to learn more about my four phase approach to roast profiling, check out my “Roast Approach: Roast Strategy” blog.

The Approach: 

As mentioned earlier, my first approaches with this coffee were a run through of my standard roast profiles.  Usually when I run through these profiles, I am able to see a strong impact in sweetness and acidity depending on how I adjust heat application.  However, with this coffee the differences were very minimal.  They all had underdeveloped sweetness, a lemongrass acidity, and an astringent finish.  Something was going on with this coffee.  

Astringency in a roast can be a result of underdeveloped seed interior or possibly too much airflow in the roaster.  With the high density of the coffee seed due to decaffeination, I thought it was highly likely that the interior of the seed was not getting properly developed.  This also made sense with the tart lemongrass flavors I was experiencing in the cup.  But what about the sweetness? I theorized that the coffee may be drying out too fast and too early, leaving not enough available water molecules to be utilized during the sugar conversion process that happens during caramelization.  

To address these issues, I needed to more evenly heat the coffee seed and also maintain as much moisture as possible in the seed for caramelization.  For even heating, I utilized a "soak method".  The "soak method" involves reduced heat at the beginning of the roast, allowing the coffee seed to come up to temperature more slowly and create more even heating through the seed.  A potential drawback is that you can greatly reduce your thermal energy momentum going in to the Maillard reaction, leaving you with a muted acidity.  

For increasing the moisture in the drum, I looked at my airflow.  Reducing the airflow in the machine will keep the humidity inside the drum higher as water evaporation is happening at the beginning of the roast. 

I tried three new different roast profiles with this two-pronged approach.  On the first roast profile, I used a low charge temperature (380 F) with a starting gas application of 25%.  For gas application throughout the roast, I gradually stepped up the gas, then brought the gas back down at the end.  For the airflow, I kept the airflow restricted significantly longer than I normally do.  Usually at the end of the roast, I open up the airflow to 100% right before first crack.  You can see the roast curve, as well as the gas and airflow schedules below. 

For the second roast, I started with an even lower charge temperature and made an even slower heat application increase.  I wanted to really see what more drying time would do to this coffee.  How much "soak" is too much?  The gas and airflow schedule are provided below.  You can see they are almost identical, but with the second having a longer period of time in the drying phase.

For the last roast, I wanted to see what a "minimal soak" would do to the coffee.  I started the coffee with a charge temperature of 380 (similar to roast 1), however my initial gas application was 50% (where the first two were 25%).  I also made my heat application increases more quickly.  Because I was chasing sweetness in these roasts, I wanted to make sure that I was providing the coffee with enough time in the caramelization phase to build sugars and melanoidins (for body).  To counter the higher heat intensity at the beginning half of the roast, I dropped my heat application more drastically for the second half of the roast to stretch out time spent in this phase.  

On the cupping table, Roast 3 was the real standout.  It had a more pronounced and dynamic acidity, juicier body, and better sweetness than the other two.  Compared to the first set of roasts, this batch all had better sweetness, more pleasant acidity, and the astringency was gone.  I believe that because Roast 3 was able to gather more momentum into Maillard due to it's minimal soak, it had an acidity that was both more complex and more intense than the other two.  The roast with the longest soak (Roast 2) was "flat" compared to the others, however it did have more of the chocolatey notes come through.

The Takeaway:

Thinking about the chemical reactions taking place in the roasting process really helped me dial in this roast.  While we may know the moisture level of a coffee, rarely do we get any insight into how those water molecules are dispersed within the coffee seed.  How thick are cell walls?  How freely does the water move during the roasting process?  How quickly does the water migrate to the surface and evaporate?  I still don't have any of these answers, but the presence of astringency paired with lack of sweetness when roasting this coffee was definitely a red flag that something was going on with both the water molecules and density.  

When troubleshooting a difficult coffee, I like to think about the chemical reactions that are at the root of what I want to get out of the coffee.  If I want more acidity, I think about Maillard and what I can do to impact Maillard reaction.  If I want more sweetness, I think about caramelization and what might be going wrong there.  Roasting this coffee certainly made me think more about what was happening (or not happening) in my roaster drum, and in turn, made me a better roaster.  My biggest take away with this coffee is don't give up and don't settle.  This particular roast approach was almost a white flag of surrender to the monsters of astringency, but I took some time to think through what was going on and come back with a new approach and some much improved coffee.  Now all that is left is to ask for a bit of forgiveness for missing all of my publishing deadlines on this piece.

Happy roasting and best of skill to all of our competitors!