Kenya and Ethiopian Coffees
Kenyan and Ethiopian coffees are recurring Roest reference points because they often expose problems in heat transfer, first-crack handling, development, sorting, and rest. This page summarizes community practice for roasting these origins on Roest, with emphasis on how they behave differently, what targets have worked, and how to diagnose common cup outcomes.
Starting Profiles and Adjustment Map
Use this section as the canonical starting point for Kenya and Ethiopia. Detailed mechanics for airflow, development, weight loss, and between-batch handling are covered in Airflow and Fan Settings, Development Time and Drop Decisions, Weight Loss Targets, and Cooling and Between-Batch Protocol. curated Treat the batch sizes in this table as profile references, not as model-capacity guidance: use only batch weights within your specific Roest model’s rated range; 150–200g sample-roaster profiles are not appropriate for smaller S/L models that are not rated for that load, and 100–200g profiles should not be directly transferred to a P3000 without using P3000-specific guidance.
| Coffee type | Starting batch | First-crack / timing target | Heat and airflow approach | Development / drop | Weight-loss reference | What to watch |
|---|---|---|---|---|---|---|
| Kenya, 100–130g fast Roest style | 100–130g | Many successful Kenya profiles target first crack around the 5–6 minute range; one documented TW-style Kenya profile showed FC at 04:55 and drop at 05:54, while another Kenya profile showed FC at 05:31 with 9.9% loss 2 sources. | Keep enough energy into crack, but avoid a large uncontrolled ET rise entering development. One Kenya approach targets 237–242°C max ET and uses max airflow after crack; another later approach favors lower air around 35–40% before crack because it appeared to cook the inside more, then higher air at crack for cleanup 2 sources. | Start with 45–60s development unless the cup clearly needs more. Several Kenya discussions cluster around 50–60s, while very long development is associated with savory notes 2 sources. | Treat 9.5–10.5% as a light reference range, not a pass/fail rule. Later Kiandu examples were successful at 9.2–9.5% loss, while earlier Kenya discussions considered 10.2% too little 2 sources. | The target visual is inflated, big, rounder beans. A flat face or small, stressed-looking beans suggests energy/pressure formation was too low or slow source. |
| Kenya, slower / larger-batch style | 150–200g | For 200g batches, one target was about 3:40 to yellow, 2:30 Maillard, and 45–60s development; another broader timing framework for many beans was 4:30–5:00 to yellow, 3:15–3:30 Maillard, and 1min+ development 2 sources. | Larger batches often require a different profile rather than a direct scale-up. Some 150g Kenya batches were described as diluted and mellow compared with 100g batches, and 180–200g batches may need much higher inlet temperatures to reach FC before 6:30 2 sources. | Adjust development to taste. If a slower profile produces florality but misses berry/fruit, it may be too long and too low rather than simply underdeveloped source. | Do not compare weight loss across batch sizes without tasting and checking inside/outside color. | See Batch Size Scaling before treating 150–200g as equivalent to 100–130g. |
| Washed Ethiopia | 100–150g | Washed Ethiopia has often been roasted faster than other beans to preserve depth and complexity. One Ethiopia profile stayed around 234°C ET at FC and climbed to 236°C after; another reference placed FC at 200.5–201°C BT and held to max 203°C for about 30s source. | Small washed Ethiopian beans can need early power and may punish soak/slow approaches. Faster roasting helped improve evenness in one comparison, while slow Ethiopia roasts were associated with loss of depth/complexity 2 sources. | Start around 30–50s development, then cup after rest. Under-30s development on a natural geisha went underdeveloped, while washed examples sometimes needed more relative development than naturals 2 sources. | A common Ethiopia reference was 11–11.2% loss; 10.5–11.5% was still described as light, while 9.5–10% was often considered under for Ethiopia 2 sources. | Tiny, dense beans may roast unevenly, with small beans darkening more than normal-sized beans, especially in longer 7–9 minute roasts source. |
| Natural Ethiopia | 100–175g | Several successful natural Ethiopia examples sit in the 5–6 minute roast-time range rather than 8–10 minutes. One natural Ethiopia tasted good only at 10.7–11% loss in 5–6 minutes; at 9.5–10% it was under and flavor was not yet there source. | Fast/light natural Ethiopia can be clean and fruity, but fermented/sugary small beans may need careful heat management. One 74158 Wollega Natural target used 3:30 to dry / 150°C BT check source. | Keep development short enough to preserve aromatics, but not so short that the cup is empty, sour, or green. | Natural Ethiopia examples ranged around 10.5–10.9% loss for 10.4% moisture greens source. | Watch for quakers and process defects before blaming the roast. |
| Geisha / delicate African-adjacent profiles | Usually 100–150g | Geishas in the discussions often triangulated around 6–7 minutes total on Roest, though specific lots varied. A Panama Geisha reference was around 7:00 total, with 25+ cracks and 45–60s development source. | Avoid blindly shortening drying or time-to-crack. In one geisha comparison, shorter drying around 3:30 versus 4:20 produced vegetable/bell-pepper/tomato-water aromas despite similar drop temperatures source. | 40–60s is a common exploration range; some users found 50s too much for a particular geisha and considered 40–45s for more florals source. | Use weight loss and color together. A Janson Geisha example at 145g showed FC 08:14, drop 09:44, 15.41% development, and 12.41% loss source. | First-crack detection may be unreliable on some geishas; manual judgment may be required source. |
Adjustment map
| Symptom | Likely direction | Adjustment |
|---|---|---|
| Kenya exterior dark but inside raw or cinnamon-colored | Too much exterior heat / RoR too high entering FC, with insufficient internal development | Reduce the aggressiveness of the hot start or landing, and inspect inside/outside color before changing only development time. This defect was repeatedly described as roasted outside but not inside 2 sources. |
| Kenya becomes savory, wet-chimney, roasty, or too dark after rest | Too much roastiness, long development, or over-roasted exterior | Shorten development, lower end heat, or roast lighter; do not judge only immediately after roasting because Kenya can become browner and more savory with rest 2 sources. |
| Kenya is vegetal, hay-like, green, cardboard, woody, or “Schweppes”/baggy | Underdevelopment, poor internal heat transfer, or insufficient cleanup around crack | Increase usable development or improve heat before crack; higher air at crack improved appearance, cleanliness, astringency, and baggy taste in one Kenya workflow 2 sources. |
| Kenya has florals but lacks berries/fruits | Too long and too low, or insufficient energy before development | Try a lower-energy but still sufficiently developed profile rather than simply extending the roast. Patrick’s 150g / 75% fan example was described as too long and too low when berry/fruit was missing source. |
| Kenya crashes harder around FC after controls change | Sudden inlet increase plus fan decrease can worsen the crash | Hold inlet steady while tapering fan up, or taper inlet down and fan up together source. |
| Ethiopia tastes green, woody, or empty | Too little early heat, too low RoR into crack, or low moisture making development difficult | Add early heat rather than soaking indefinitely; low moisture was described as making taste harder to develop 2 sources. |
| Ethiopia tastes green-tea bitter, dry, papery, or baggy | Profile shape or green sorting issue | Avoid high-ET-start / low-ET-finish profiles for washed Ethiopia when they create green-tea/fruit bitterness; also sort aggressively when defects are visible 2 sources. |
| Natural Ethiopia becomes roasty, astringent, or muted | Too slow, too dark, or too much late heat for the lot | Move toward the natural-Ethiopia starting lane above and cup after rest before making large changes. |
| Cup is promising but harsh/fizzy at one day | Too fresh to judge | Rest before deciding the profile failed. See Resting and Degassing. |
Kenya Behavior on Roest
Kenya often demands enough energy to build internal pressure and sweetness, but it also punishes uncontrolled exterior heat. A well-executed Kenya should look inflated and rounder, while beans with a flat face or stressed/dark small beans point toward poor pressure formation or uneven heat transfer 2 sources.
First crack can be active and sensitive. One Kenya target was 25–35 cracks, with 40+ considered undesirable in that workflow, while low crack counts were attributed to too little preheating, low voltage, or cold conditions source. Crack counting is not a universal roast-quality measure, but it is useful when comparing the same coffee on the same machine.
Kenya can show desirable currant, blackberry, blueberry, rosehip, tomato, rhubarb, orange, maple, brown sugar, and floral notes. Tomato is not automatically a defect: some tasters actively want ripe tomato in Kenya, while others treat tomato-like notes as roast-approach dependent rather than inherently good or bad source. The practical distinction is between ripe, sweet, tangy tomato complexity and vegetal/bell-pepper/eggplant/tomato-water underdevelopment.
Ethiopian Behavior on Roest
Ethiopian coffees are less uniform than the origin name suggests. Washed, natural, low-moisture, small-bean, high-density, and heavily fermented lots can behave very differently, and several discussions emphasize sorting, moisture, and bean size before profile tuning.
Washed Ethiopias can roast unevenly, especially when small beans darken faster than larger beans. A shorter, faster batch improved evenness in one comparison, while slow washed Ethiopia roasts were associated with lost complexity and depth 2 sources. Natural Ethiopias can be vivid and fruit-forward when kept fast and light, but the same process category can also carry quakers, unripe beans, ferment defects, or a baggy/astringent finish.
Moisture readings in Ethiopian greens in the discussions ranged from very low 9.6% Tagel Alamayehu to 10.3–11.0% across several 88 Graines lots 2 sources. Low moisture was associated with difficulty developing taste, so a low weight-loss number should be interpreted alongside cup quality, color, and green condition rather than used alone.
Green Selection and Sorting
Kenya AB lots received repeated positive attention. One experienced roaster thought buyers may have better chances with AB lots because many people jump first to AA, causing good AA lots to disappear quickly source. Good Kenya examples in the discussions include Kiandu AB, Kibugu AA, Mahiga AA, Gura, Kegwa, Kagamoini, and other washed SL-type lots, but evaluations varied by crop, supplier, and roast.
Ethiopia needs more sorting discipline. Several discussions reported many quakers or unripe beans in Ethiopian lots, including Tagel Alamayehu, Kabira, and some 88 Graines naturals; one ET2509 comparison removed 19 beans from a 15g dose, while ET2512 was described as clean, fruity, floral, and free of the baggy/astringent/unripe defect source. When a cup shows baggy, dry, astringent, unripe, or “Schweppes” notes, do not assume the roast profile is the only variable.
For broader buying decisions, supplier notes, moisture expectations, and defect sorting, defer to Green Coffee Selection.
Rest and Evaluation
Kenya and Ethiopia should not be judged only in the first few days. A common baseline is to taste at 10–14 days or later, and one contributor explicitly warned not to judge before that window source. Another rest observation was that coffee can taste good immediately, decline for 7–10 days, and then taste good again; if it still does not taste good after 10–14 days, the profile is probably at fault source.
Kenya can change more dramatically than many coffees. Examples include a Kenya profile that tasted bad in the first 2–3 weeks but became good after 40 days, a Kiandu AB at 26 days showing red berries, crisp red apple, and maple-syrup sweetness, and a Kenya KII that lost aroma and became savory after being opened before a 30-day rest target 2 sources. Ethiopia can also improve with longer rest; one Kaffa washed Ethiopia had herbal/vegetal notes at 2 weeks but was clean, sweet, peachy, and free of bag/astringent notes closer to 4 weeks source.
For profile comparison, cup multiple profiles side by side and smell grounds immediately after grinding and again after a short wait. One geisha comparison used at least four profiles and evaluated aroma immediately and at 2–5 minutes after grinding, revealing vegetable aromatics that were not obvious at first smell source.
Meaningful Conflicts and Evolution
CONFLICT (Temporal Evolution): Kenya weight loss. Earlier Kenya discussions treated 10.2% loss as too little for Kenya, while later successful Kiandu AB examples were praised at 9.2–9.5% loss. The practical resolution is to treat weight loss as a context marker, not a universal development target; inspect inside/outside color and cup after rest before deciding a Kenya is underdeveloped 2 sources.
CONFLICT (Unresolved): Kenya speed and development style. One Roest approach favors faster Kenya and washed Ethiopia profiles with first crack around 5–6 minutes for more vivid expression, while other trusted examples use larger batches or slower structures with longer phase timing and more development. The difference appears tied to batch size, fan/inlet setup, rest protocol, and desired sensory style rather than a single universal Kenya profile 2 sources.
CONFLICT (Unresolved): Tomato in Kenya. Some tasters consider ripe tomato desirable in Kenya, while others dislike tomato or see it as a roast-approach issue. The actionable distinction is sensory quality: sweet, ripe, tangy tomato can be part of a Kenya profile; bell pepper, vegetable soup, eggplant, tomato water, hay, or dry astringency should trigger troubleshooting 2 sources.