Anaerobic Fermentation in East African Lots
I've been cupping anaerobically processed coffees from East Africa for about three years now. The first batch came from a small washing station in Nyamasheke, Rwanda. The cup was divisive. Half of our team thought it was exceptional. The other half said it tasted like cough syrup.
That experience set the tone for everything that followed.
Background: Controlled Fermentation Arrives in Africa
Anaerobic processing gained traction in Latin America around 2015. Colombian producers led early experimentation. The technique involves sealing coffee cherries or mucilage-covered beans in airtight tanks, displacing oxygen with CO2, and allowing fermentation to proceed in a controlled environment. Temperature, pH, and duration become variables that producers can manipulate.
African producers were slower to adopt the method. Infrastructure limitations played a role. Many washing stations in Ethiopia and Kenya operate with equipment dating back to the 1970s. Installing stainless steel fermentation tanks with pressure release valves requires capital that most cooperatives simply do not have.
The shift began around 2019. A handful of washing stations in Sidama and Yirgacheffe installed their first sealed tanks. By 2022, anaerobic lots from Ethiopia were appearing regularly on cupping tables at Nordic Approach and Collaborative Coffee Source. I tasted my first Ethiopian anaerobic from Halo Hartume that year. The lot scored 89.5 on our internal scale.
What I Found in Three Growing Seasons
Our lab tracked 47 anaerobic lots from Rwanda, Burundi, Ethiopia, and Kenya between the 2022 and 2024 harvest seasons. We recorded fermentation parameters where producers shared data. We cupped each lot blind, using SCA protocol with a five-person panel.
The results were inconsistent. That is the most honest summary I can give.
Rwandan lots showed the highest standard deviation in cupping scores. The range spanned from 82.25 to 91.0. Ethiopian lots clustered more tightly, mostly between 86 and 89. Kenyan producers sent us only six anaerobic samples across three years. Two were excellent. Two were flawed. Two were unremarkable.
Fermentation duration varied wildly. Some producers ran 72-hour cycles. Others extended to 200 hours or beyond. The correlation between duration and cup quality was weak. A 96-hour Burundian lot outscored a 180-hour lot from the same washing station using identical cherry selection.
pH monitoring data, where available, told a more useful story. Lots that maintained pH between 3.8 and 4.2 at the end of fermentation consistently scored above 86. Lots that dropped below 3.5 showed acetic acid notes that our panel flagged as defects.
The Halo Hartume Lots: A Closer Look
I want to spend some time on the Halo Hartume washing station because we received the most complete data from their operation. The station manager, Testi Bekele, documented fermentation conditions across four separate lots in 2023.
All four lots used the same cherry selection: ripe Wolisho and Dega cultivars from surrounding smallholders. All four went into identical 1,000-liter tanks. The variables were duration (72, 96, 120, and 144 hours) and one temperature intervention (the 120-hour lot received ice at hour 60 to slow fermentation).
Our blind cupping produced the following scores:
The 144-hour lot had noticeable alcoholic fermentation character. One panelist wrote "overripe banana peel" in their notes. Another wrote "nail polish." The 96-hour lot had the cleanest expression of the Wolisho terroir. Stone fruit. Bergamot. A floral finish that lingered.
I spoke with Testi over WhatsApp in January. She said the 144-hour duration was experimental. She does not plan to repeat it. The station's standard protocol for anaerobic lots is now 90 to 100 hours with twice-daily pH checks.
Equipment and Infrastructure Gaps
The data quality problem in African anaerobic processing is significant. Many washing stations do not own pH meters. Temperature logging requires equipment that costs more than some stations earn in a month. Fermentation decisions are made by smell and feel, passed down through experience.
This is not a criticism. Kenyan producers have made exceptional washed coffees for decades using concrete tanks and intuition. But anaerobic processing rewards precision. Small deviations compound over multi-day fermentation cycles.
I visited the Kangocho washing station in Kirinyaga County, Kenya, last March. The manager showed me their new anaerobic setup: two 500-liter plastic drums with hand-welded lids. No pressure release. No temperature control. They were producing anaerobic lots by sealing the drums and waiting.
Kangocho Station Assessment
The coffee I cupped from those drums scored 84.5. It was clean. It had pleasant acidity. But it did not justify the anaerobic label or the price premium attached to it. I told the manager this directly. He agreed. He said they were still learning.
Pricing and Market Pressures
Anaerobic lots from East Africa command FOB prices between $8 and $25 per kilogram, depending on cup score and buyer relationship. Standard washed lots from the same stations sell for $3 to $6.
This price differential creates pressure. Producers see anaerobic as a path to higher income. Buyers see it as a way to differentiate their offerings. The incentives push both parties toward production volumes that may exceed current processing expertise.
I have tasted at least a dozen African anaerobic lots in the past year that should not have been sold as specialty coffee. Fermentation defects. Unclean cups. Inconsistent profiles across bags from the same lot. The market is absorbing these coffees because demand for novelty is high and quality control is difficult at origin.
This is not sustainable. At some point, roasters will grow skeptical of the anaerobic label. Producers who invested in equipment will face declining premiums. The washing stations doing excellent work will be penalized alongside those cutting corners.
What I Think Works
I am not opposed to anaerobic processing in Africa. Some of the best coffees I have tasted in the past two years came from sealed tanks in Sidama and Nyamasheke. The technique can enhance clarity, extend flavor complexity, and highlight cultivar characteristics that traditional washed processing obscures.
But the technique requires investment beyond equipment. Training matters. Data collection matters. Willingness to discard failed lots matters.
The producers getting consistent results share certain practices:
- They select cherries at narrow Brix ranges, usually between 22 and 24
- They monitor pH at least twice daily
- They release pressure manually rather than relying on automatic valves
- They keep detailed records and adjust parameters between lots
Halo Hartume does all of this. So does the Buf Café collective in Burundi. The Kenyan washing stations I visited do not. The gap shows in the cup.
Next Steps
We are continuing to track anaerobic lots through the 2025 harvest. Our focus is shifting toward yeast inoculation, which several Ethiopian producers are now testing. Early samples suggest inoculated lots ferment more predictably, but I do not have enough data to make claims yet.
I am also planning a return visit to Kirinyaga in June. The Kangocho station purchased a pH meter and temperature logger after my March visit. I want to see if the equipment changes their outcomes.
This piece is part of an ongoing series. The next installment will cover carbonic maceration in Colombian and Ecuadorian lots. Expect that in late summer.
If you are a producer or importer with anaerobic lot data you are willing to share, my contact information is on the About page. I am particularly interested in lots with complete fermentation logs and cupping scores from multiple sources.
