CoffeariumBeta
  1. Plant
  2. Farm
  3. Process
  4. Trade
  5. Roast
  6. Brew
  7. Cup
Post-harvest processing · Step 4

Fermentation

The processing step in which yeasts and bacteria break down the mucilage and sugars around the coffee seed – how it runs, how to monitor it, when to stop it and what it does to the cup.

Source-checkedby Coffearium verification pass· Last source-checked 8 Oct 2026· Prepared with AI assistance· 15 sources (15 tier A)· 5 contested· Updated 8 Oct 2026
Reading depth
Contents
  • In washed coffee, pulped seeds rest in a tank – with or without water – until the mucilage has broken down and can be washed off; Codex gives 12–36 hours, reviews 12–48 hours and 6–72 hours.1,15,2
  • Microbes use up the sugars and make lactic acid, acetic acid, ethanol and mannitol, so the pH falls – from about 5.5–6.5 to about 4.0–4.6 in field and trial fermentations.3,6
  • Temperature sets the pace. In one trial with Castillo coffee, mucilage breakdown took 17.7 hours in spontaneous fermentation but 42.1 hours with the tank held at 15 °C.8
  • The end point is judged by touch, by pH or by measuring how the fermenting mass settles; washing late raises the risk of vinegary and "fermented" cup defects.2,6,9
  • Natural coffee ferments in the fruit as it dries; honey coffees can be made by stopping tank fermentation early.5,10

What fermentation does#

After pulping, the seeds are still covered in sticky mucilage. In the washed process, they are left to ferment until microorganisms have broken the mucilage down and it can be washed off.1 In wet fermentation, lactic acid bacteria dominate the microbial community.5 How the mucilage itself is taken apart is covered under mucilage.

Fermentation is not only a cleaning step. The microbes consume sugars and make acids, alcohols and aroma compounds, and some of these end up inside the seed.2,5 Length and conditions of fermentation therefore change the cup.3,4

Where it happens#

  • Dry-tank fermentation – pulped coffee is left in a tank without added water. In a Colombian trial, no water was added to the 30-litre fermentation tanks.8
  • Underwater (submerged) fermentation – the pulped coffee is covered with water.1,2 In an Ecuadorian trial, about 150 kg of pulped coffee was submerged in clean water in a concrete tank.4 Some farms add only a little water: in one Colombian study, about 4 litres were added to 10 kg of cherries after pulping.7
  • During natural drying – whole cherries ferment on the drying yard. Acetic acid bacteria and yeasts dominate, and their products, such as gluconic acid and sugar alcohols, build up in the drying outer layers of the fruit.5 Cherries left in bags or heaps for days heat up and ferment quickly, which Codex links with quality loss and a higher risk of ochratoxin A.1
  • Honey coffees – fermentation can be stopped early so that only part of the mucilage is removed before drying. A Cenicafé trial found a 22–77% probability of defective cups in such honey coffees.10
  • Controlled variants – starter cultures, controlled temperature and sealed, oxygen-poor tanks are used to steer fermentation.2,8,11 Anaerobic fermentation and carbonic maceration are covered in their own entries.

What changes in the tank#

  1. Sugars rise, then fall. Sucrose, glucose and fructose are released from the mucilage and then consumed.3 Microbes use sucrose first, then glucose and fructose, then citric and malic acid.4
  2. Acids and alcohols appear. Lactic acid, acetic acid, mannitol and ethanol accumulate, mostly after the first day.3
  3. pH falls. In a Nicaraguan field study, pH started at 5.5–5.7 and fell sharply to about 4.6 as fermentation neared completion.6 In a Yunnan trial, it fell from about 6.0–6.5 to about 4.0.3
  4. The mucilage lets go. The texture changes from slimy to gritty.2 The pectin chemistry behind this is covered under mucilage.

Knowing when to stop#

The end point is when the mucilage is degraded and the beans are washable.1 Producers use four kinds of test:

  • Touch – a handful of beans is rubbed between the fingers; when the mucilage is gone they feel rough and gritty instead of slimy.2,9
  • Stick – a stick is pushed into the fermenting mass; the hole it leaves is read as a sign of the end point.9
  • pH – the sharp pH drop near the end can predict completion and help prevent over-fermentation.6 One review reports that fermentation was complete within about 2 hours of the mass reaching pH 5.2
  • Volume – Cenicafé’s Fermaestro measures how far a sample of the mass settles as degraded mucilage drains away.9

Effect on the cup#

Well-run fermentation can raise quality. In controlled trials in Yunnan and Ecuador, longer fermentation gave a fruitier cup with more acidity.3,4 Run too long, it spoils the coffee: over-fermentation affects taste and flavour,12 and Cenicafé names vinegary and “fermented” as the cup defects tied to this stage.9

Kinetics#

The published curves differ between sites, but the pattern repeats: sugars peak and fall, acids rise late, and pH drops.

pH. In Nicaragua, seven farm batches started at pH 5.5–5.7 and fell sharply to about 4.6 near completion, while glucose fell throughout most batches and ethanol or lactic acid rose sharply near the end.6 In Yunnan (1,300 m), pH fell from about 6.0–6.5 to about 4.0 in every variant, with the major drop after 36 hours.3 In Ecuador, pH was about 4.5 after 16 hours and 4.0 after 36 hours.5 In Nariño, Colombia (1,959 m), lactic acid rose to 0.37 g/L over 48 hours and pH fell from 5.2 to 4.2.7 In a Colombian dry-tank trial, the mass started at pH 5.53–5.67 depending on the variety (5.61 for Castillo); spontaneous fermentation ended below pH 3.5, while fermentation held at 15 °C levelled off near pH 4.0.8

Sugars and °Brix. In Nariño, the pulp started at 5.3 °Brix, with 0.82 g/L glucose and 1.16 g/L fructose; °Brix and sugars rose during the first 12 hours of fermentation.7 In Yunnan, sucrose peaked within 8 hours and glucose and fructose after 36 hours.3

Temperature. The Yunnan fermentations started at about 15 °C, followed the day–night cycle and ended at about 11–13 °C.3 In the Colombian dry-tank trial, at an ambient average of 26 °C (23–31 °C), the spontaneously fermenting mass warmed to close to 26 °C by the end.8

33.544.555.566.57pHNicaragua, 7 farm batches5.5–5.7 → about 4.66 Yunnan, China6.0–6.5 → about 4.03 Ecuador4.5 at 16 h → 4.0 at 36 h (no start value given)5 Nariño, Colombia5.2 → 4.2 at 48 h7 Colombia(Castillo), spontaneous5.61 → below 3.58 Colombia(Castillo), held at 15 °C5.61 → about 4.08
Data table
pHSource
Nicaragua, 7 farm batches5.5–5.7 → about 4.66
Yunnan, China6.0–6.5 → about 4.03
Ecuador4.5 at 16 h → 4.0 at 36 h (no start value given)5
Nariño, Colombia5.2 → 4.2 at 48 h7
Colombia (Castillo), spontaneous5.61 → below 3.58
Colombia (Castillo), held at 15 °C5.61 → about 4.08
pH early and late in wet fermentation in five studies. Each line runs from the last to the first value reported (for Ecuador, 36 h to 16 h, as no starting value is given); intermediate values are not shown. The Colombian rows start at the Castillo mean (5.61); the paper gives the end values for all three varieties together, and the spontaneous fermentation, which ended below pH 3.5, is drawn at 3.5.3,5,6,7,8

Metabolites#

In the fermentation water. At the end of fermentation of depulped coffee in Yunnan, the water held 3.0–3.2 mg/mL mannitol, 1.7–1.8 mg/mL lactic acid and 0.7–0.9 mg/mL acetic acid.3 After 64 hours in Ecuador, the most abundant compounds in the water were lactic acid (8.2 mg/mL) and 5-ketogluconic acid (5.7 mg/mL), with acetic acid at 2.4 mg/mL, mannitol at 1.9 mg/mL and ethanol at 1.8 mg/mL.4

In the beans. After 64 hours, the Ecuadorian beans contained 4.6 g/kg ethanol, 4.5 g/kg acetic acid, 3.6 g/kg mannitol and 2.2 g/kg lactic acid.4 Lactic acid, acetic acid and mannitol made in the mucilage were also found in the endosperm in an earlier study.5

Sugars leave the seed. In the same Ecuadorian study, fructose, glucose and sucrose in the endosperm fell significantly during fermentation; the authors link this to the seed itself consuming sugar anaerobically under water, and a further drop during soaking to osmotic loss into the water.5 Drying experiments point the same way: glucose and fructose in washed coffee fell within the first day of drying, with no process water present.13 Both mechanisms are discussed under washed process.

Acids that help and hurt. Low acetic acid gives a clean, sweet taste; above about 2 mg/mL it turns vinegary.2 Propionic and butanoic acids can give an onion flavour.2 The microbial groups behind these products are described under microorganisms.

Temperature and altitude#

Temperature controls the speed. In a Colombian trial with three varieties, more than 95% mucilage degradation took 16.8–17.7 hours in spontaneous fermentation, 17.5–20.0 hours at 30 °C and 41.3–42.1 hours at 15 °C; for Castillo the times were 17.7 ± 0.7, 20.0 ± 1.2 and 42.1 ± 1.5 hours (means with 95% confidence intervals).8 Only the 15 °C treatment differed significantly from spontaneous fermentation; holding the mass at 15 °C prolonged it by more than 24 hours.8 Average cup scores for the controlled fermentations were 82.7–83.8 SCA points (spontaneous controls 79.7–83.8), and quality did not differ significantly between controlled and spontaneous fermentation or between the two temperatures.8 This fits the wide range of reported times, which a 2023 review attributes to climate.2

Altitude also shifts the chemistry. In self-induced anaerobic fermentations of one variety from one Brazilian region, coffee from lower altitudes had more volatile alcohols, while coffee from higher altitudes had more esters, aldehydes and phenolics.11 Acetic, malic and citric acid were higher at the two upper sites (1,200 and 1,400 m), and lactic acid was highest at 1,400 m.11 At 48 hours the mass was at 21.6–25 °C according to the paper’s table (18–25 °C in its results text).11

Water, climate and soaking#

Dry tank or submerged. None of the sources opened for this entry compares dry-tank and submerged fermentation of the same coffee, or ties the choice between them to climate. What they do tie to climate is duration. A 2023 review gives 6–72 hours for submerged fermentation depending on the climate.2 In Kenya, fermentation takes 1 to 4 days depending on the weather, faster on warm days and slower on cold ones.14 Temperature itself is well documented: in a dry-tank trial, fermentation at 15 °C took more than 24 hours longer than spontaneous fermentation.8 Whether adding water slows fermentation by cooling or diluting the mass has not been measured head-to-head.

Soaking after fermentation. Some mills hold the washed parchment under clean water again, from overnight up to, rarely, 48 hours.15 Where fermentation came first, this is sometimes called secondary fermentation.15 In Kenya, soaking for about 12 hours after fermentation was found to improve colour and taste.14 In Yunnan, a 24-hour soak tempered the effect of longer fermentation and evened out green-coffee quality.3 In Ecuador, ethanol, fructose, glucose, lactic acid and mannitol in the beans dropped during a 24-hour soak.5 No primary source opened for this entry describes a two-stage (“double”) soak.

Reference protocol#

The table brings together the parameters reported in the sources. It shows the range found in practice and research, not a single correct recipe. No study in this entry compared dry-tank and submerged fermentation of the same coffee directly.

Parameter Dry tank (no added water) Submerged or water added Source
Water None added, 30 L tanks8 About 150 kg of pulped coffee covered with clean water in a concrete tank4; about 4 L added to 10 kg of cherries after pulping7 Peñuela-Martínez et al., Zhang et al., de Oliveira Junqueira et al.
General duration – 12–36 h (Codex); 12–48 h (2016 review); 6–72 h depending on climate (2023 review) Codex1, Poltronieri & Rossi15, Elhalis et al.2
Duration in trials To > 95% degradation (Castillo): 17.7 h spontaneous, 20.0 h at 30 °C, 42.1 h at 15 °C 16 h standard, 64 h extended (Ecuador); 48 h (Nariño) Peñuela-Martínez et al.8, Zhang et al.4, de Oliveira Junqueira et al.7
Mass temperature Rose to close to 26 °C (spontaneous) at 23–31 °C ambient About 15 °C at start, 11–13 °C at end (Yunnan) Peñuela-Martínez et al.8, Zhang et al.3
pH at start 5.53–5.67 (Colombia, three varieties) 5.5–5.7 (Nicaragua); 6.0–6.5 (Yunnan); 5.2 (Nariño) Peñuela-Martínez et al.8, Jackels & Jackels6, Zhang et al.3, de Oliveira Junqueira et al.7
pH at end Below 3.5 spontaneous; about 4.0 at 15 °C About 4.6 (Nicaragua); about 4.0 (Yunnan, Ecuador); 4.2 (Nariño) Peñuela-Martínez et al.8, Jackels & Jackels6, Zhang et al.3, De Bruyn et al.5, de Oliveira Junqueira et al.7
pH rule Complete within about 2 h of reaching pH 5 – Elhalis et al.2
°Brix at start – 5.3 °Brix, rising in the first 12 h de Oliveira Junqueira et al.7
End-point criterion > 95% mucilage degradation; Fermaestro void above 85 mm, > 96% mucilage removed – Peñuela-Martínez et al.8, Cenicafé9
Soak after fermentation – About 12 h (Kenya); overnight to rarely 48 h; 24 h in trials FAO14, Poltronieri & Rossi15, Zhang et al.3, De Bruyn et al.5
Washing delay Risk rises from 3 h late (2013); 0–11% defect probability at 2–8 h late (2022) – Cenicafé9,10
01020304050hours to > 95% mucilage degradationSpontaneous, ambient(23–31 °C)17.7 ± 0.7 h8 Controlled at 30 °C20.0 ± 1.2 h8 Controlled at 15 °C42.1 ± 1.5 h8
Data table
hours to > 95% mucilage degradationSource
Spontaneous, ambient (23–31 °C)17.7 ± 0.7 h8
Controlled at 30 °C20.0 ± 1.2 h8
Controlled at 15 °C42.1 ± 1.5 h8
Hours to more than 95% mucilage degradation for Castillo coffee in a Colombian dry-tank trial (means of five repetitions with 95% confidence intervals). The 30 °C and spontaneous times did not differ significantly. Cenicafé1 and Tabi took 41.3–41.7 h at 15 °C and 16.8–17.3 h spontaneously; at 30 °C Tabi (17.5 h) was significantly faster than Castillo.8

Monitoring protocol#

A sourced routine for a single batch:

  1. Record the start – time, mass temperature and pH of the mass or water. Research trials log temperature and pH continuously.4
  2. Track pH – expect a slow fall followed by a sharp drop near the end.6 Once the mass reaches pH 5, check often: completion followed within about 2 hours in one study.2
  3. Track temperature – cool mass means slow fermentation; in one trial, fermentation at 15 °C took more than 24 hours longer than spontaneous fermentation.8
  4. Test the texture – rub beans between the fingers and look for the change from slimy to gritty.2,9
  5. Measure settling – with a Fermaestro, lift the device, drop it three times from about 3 cm and wash when the empty space reaches the 85 mm mark.9 In field tests, this put fermentation at about 11.5–20 hours, against farmer estimates of 10–72 hours.9
  6. Wash promptly – delay after the end point raises the risk of fermentation defects.9

Risks and defects#

  • Vinegary – too much acetic acid, above about 2 mg/mL, gives a vinegary taste.2 Cenicafé names “vinagre” among the fermentation defects.9
  • Onion – propionic and butanoic acids can give an onion flavour.2
  • “Fermented” – the “fermento” defect in the cup is tied to the mucilage fermentation stage.9 Holding coffee in the tank longer than necessary causes over-fermentation, which affects taste and flavour.12
  • Stinker beans – stinkers are caused by excessively long fermentation or by unclean water.12
  • Heaped cherries – cherries kept in bags or heaps heat up and ferment in an uncontrolled way, with quality loss and a higher ochratoxin A risk.1
  • Partial mucilage removal – honey coffees made by stopping fermentation early had a 22–77% probability of defective cups in one trial.10

Open questions#

  • Migration into the seed. Ethanol, acetic acid, mannitol and lactic acid have been measured in beans after fermentation.4,5 In dry processing, the data suggest a slow migration.5 How far these compounds penetrate, and how much survives roasting to shape flavour, is still debated.2
  • Dry tank versus submerged. The sources report both, but none in this entry compares them directly on the same coffee or links the choice to climate; only fermentation time is linked to weather.2,14
  • Tolerance for delay. Cenicafé’s own studies disagree on how quickly defects appear after the washing point (see Contested).9,10
  • Optimum length. Codex advises short fermentation,1 while trials found that longer fermentation improved fruitiness and acidity.3,4

Key studies#

Jackels & Jackels, 2005 – Chemical indicators of fermentation. Follows seven farm-scale batches in Nicaragua, measuring temperature, pH, glucose, ethanol and lactic acid. The sharp pH drop near completion is proposed as a practical way to time washing and avoid over-fermentation.

Journal of Food Science 70(5) · doi:10.1111/j.1365-2621.2005.tb09960.x

Zhang et al., 2019 – From cherry to cup in Ecuador. Compares 16-hour and 64-hour submerged fermentation of Typica, measuring microbes and metabolites in water and beans. Extended fermentation put more ethanol, acetic acid, mannitol and lactic acid into the beans and gave a significantly fruitier cup; acidity also scored higher, though less clearly.

Applied and Environmental Microbiology 85(6), e02635-18 · doi:10.1128/AEM.02635-18

Zhang et al., 2019 – Processing parameters in Yunnan. Tests fermentation length, mechanical demucilaging and soaking on Catimor at 1,300 m. pH fell from about 6.0–6.5 to about 4.0 with the main drop after 36 hours, and longer fermentation gave a fruitier, more acidic cup.

Frontiers in Microbiology 10, 2621 · doi:10.3389/fmicb.2019.02621

Peñuela-Martínez et al., 2023 – Controlled temperature. Ferments three Arabica varieties in dry tanks at 15 °C, 30 °C and ambient temperature. Cold fermentation took 41.3–42.1 hours against 16.8–17.7 hours for spontaneous fermentation and ended at a higher pH, yet cup quality did not differ significantly between treatments, with treatment averages of 82.7–83.8 SCA points.

Fermentation 9(11), 976 · doi:10.3390/fermentation9110976

Sanz-Uribe & Velásquez-Henao, 2022 – Incomplete and prolonged fermentation. Uses the Fermaestro to stop fermentation early for honey coffees or extend it by 2–8 hours past the washing point. Honey coffees carried a high risk of defects, while prolonged fermentations rarely produced defective cups.

Revista Cenicafé 73(1) · doi:10.38141/10778/73105

Contested
  • Typical fermentation time: the Codex code of practice gives 12 to 36 hours1, a 2016 review 12 to 48 hours15, while a 2023 review reports 6 to 72 hours depending on the climate.2
  • Short or long: Codex advises fermentation as short as possible1, while controlled trials in Yunnan and Ecuador found that extended fermentation gave a fruitier, more acidic cup.3,4
  • How much delay is tolerable after the washing point: Cenicafé reported in 2013 that washing 3 hours or more late raises the risk of fermentation defects9, while a 2022 Cenicafé trial that kept coffee in the tank 2 to 8 hours past the washing point found only a 0–11% probability of defective cups.10
  • Migration into the seed: fermentation metabolites have been measured in the beans4,5, but how far they penetrate and how much they shape flavour after roasting is still being researched.2
  • Sugar loss from the seed: the lower glucose and fructose of washed coffee has been attributed, at least in part, to metabolic processes rather than leaching into the process water13; a field study agrees for the tank stage but adds that sugars also left the beans by osmosis during a later soak.5
Share or cite this entry

Sources

  1. ACodex Alimentarius Commission (FAO/WHO) (2009). Code of Practice for the Prevention and Reduction of Ochratoxin A Contamination in Coffee (CAC/RCP 69-2009). Codex Alimentarius.
  2. AElhalis, H.; Cox, J.; Zhao, J. (2023). Coffee fermentation: Expedition from traditional to controlled process and perspectives for industrialization. Applied Food Research 3(1), 100253. doi:10.1016/j.afres.2022.100253 · other copies
  3. AZhang, S. J.; De Bruyn, F.; Pothakos, V.; Contreras, G. F.; Cai, Z.; Moccand, C.; Weckx, S.; De Vuyst, L. (2019). Influence of various processing parameters on the microbial community dynamics, metabolomic profiles, and cup quality during wet coffee processing. Frontiers in Microbiology 10, 2621. doi:10.3389/fmicb.2019.02621 · other copies
  4. AZhang, S. J.; De Bruyn, F.; Pothakos, V.; Torres, J.; Falconi, C.; Moccand, C.; Weckx, S.; De Vuyst, L. (2019). Following coffee production from cherries to cup: microbiological and metabolomic analysis of wet processing of Coffea arabica. Applied and Environmental Microbiology 85(6), e02635-18. doi:10.1128/AEM.02635-18 · other copies
  5. ADe Bruyn, F.; Zhang, S. J.; Pothakos, V.; Torres, J.; Lambot, C.; Moroni, A. V.; Callanan, M.; Sybesma, W.; Weckx, S.; De Vuyst, L. (2017). Exploring the impacts of postharvest processing on the microbiota and metabolite profiles during green coffee bean production. Applied and Environmental Microbiology 83(1), e02398-16. doi:10.1128/AEM.02398-16 · other copies
  6. AJackels, S. C.; Jackels, C. F. (2005). Characterization of the Coffee Mucilage Fermentation Process Using Chemical Indicators: A Field Study in Nicaragua. Journal of Food Science 70(5). doi:10.1111/j.1365-2621.2005.tb09960.x · other copies
  7. Ade Oliveira Junqueira, A. C.; de Melo Pereira, G. V.; Coral Medina, J. D.; Alvear, M. C. R.; Rosero, R.; de Carvalho Neto, D. P.; Enríquez, H. G.; Soccol, C. R. (2019). First description of bacterial and fungal communities in Colombian coffee beans fermentation analysed using Illumina-based amplicon sequencing. Scientific Reports 9, 8794. doi:10.1038/s41598-019-45002-8 · other copies
  8. APeñuela-Martínez, A. E.; García-Duque, J. F.; Sanz-Uribe, J. R. (2023). Characterization of Fermentations with Controlled Temperature with Three Varieties of Coffee (Coffea arabica L.). Fermentation 9(11), 976. doi:10.3390/fermentation9110976 · other copies
  9. APeñuela-Martínez, A. E.; Pabón, J.; Sanz-Uribe, J. R. (2013). Método Fermaestro: Para determinar la finalización de la fermentación del mucílago de café (Avances Técnicos 431). Cenicafé (Colombian Coffee Research Centre).
  10. ASanz-Uribe, J. R.; Velásquez-Henao, J. (2022). Producción de café con fermentaciones incompletas y fermentaciones prolongadas utilizando el Fermaestro®. Revista Cenicafé 73(1). doi:10.38141/10778/73105 · other copies
  11. AMartinez, S. J.; Simão, J. B. P.; Pylro, V. S.; Schwan, R. F. (2021). The altitude of coffee cultivation causes shifts in the microbial community assembly and biochemical compounds in natural induced anaerobic fermentations. Frontiers in Microbiology 12, 671395. doi:10.3389/fmicb.2021.671395 · other copies
  12. AMutua, J. (2000). Post Harvest Handling and Processing of Coffee in African Countries, 3.0 Post Harvest Handling and Processing of Green Coffee in African Countries. FAO.
  13. AKleinwächter, M.; Selmar, D. (2010). Influence of drying on the content of sugars in wet processed green Arabica coffees. Food Chemistry 119(2), 500–504. doi:10.1016/j.foodchem.2009.06.048 · other copies
  14. AMutua, J. (2000). Post Harvest Handling and Processing of Coffee in African Countries, Annex 5: Coffee Post Harvest Handling and processing in Kenya. FAO.
  15. APoltronieri, P.; Rossi, F. (2016). Challenges in Specialty Coffee Processing and Quality Assurance. Challenges 7(2), 19. doi:10.3390/challe7020019 · other copies

Tier A · authoritative · B · trade and expert · C · general. Supporting quotes for every source are in the data version of this entry. If a publisher’s site refuses access, try the doi link or “other copies”.

Suggest a change

On Fermentation. Tell us what you would change and why. A source that supports it helps most.