Steeping the perfect read...
Quality takes time. One bean at a time.
Every coffee ferments — the variable is how. Understanding the microbial chemistry behind washed, natural, and anaerobic processing explains why the same estate can produce a clean chocolate cup and a grape-forward lot from the same harvest. This article maps the mechanism to the label, with data from ICB's catalog of 1,238 Indian coffees.
Most descriptions of coffee processing work backward from the cup: washed coffees taste clean, naturals taste fruity, anaerobic lots taste tropical or wine-like. Those associations are accurate. What they leave out is the mechanism that produced them. That mechanism is fermentation — microbial activity that converts mucilage sugars into chemical compounds roasting later transforms into cup character. Each processing label is shorthand for a fermentation environment, not a flavour dial.
The most common misconception about fermentation in coffee is that it only applies to experimental or natural-process lots. It does not.
Washed coffee ferments in water for 8 to 72 hours immediately after the skin and pulp are stripped — this is how the mucilage layer breaks down before the coffee can be washed clean and dried. Natural coffees ferment inside the intact cherry as it dries for 15 to 25 days on the drying bed, with microbial activity occurring in the enclosed fruit environment. Anaerobic lots ferment in sealed, oxygen-deprived tanks for 36 to 72 hours. The difference between these methods is not whether fermentation happens but the environment in which it happens.
In India, washed fermentation typically runs 12 to 36 hours. Estates at cooler high altitudes — Bababudangiris, the Shevaroy Hills — can sustain longer windows because lower temperatures slow microbial activity and give more control over the endpoint. At warmer low-altitude sites, shorter windows reduce the risk of the fermentation overshooting.
Note: The Coffee Board of India's traditional grade names encode the fermentation environment directly. 'Plantation A/B' signals washed — short aerobic fermentation, mucilage removed before drying. 'Cherry AB/PB' signals natural — extended in-cherry fermentation during drying. These terms predate specialty vocabulary but describe the same processing distinction.
The microorganisms doing the work during fermentation fall into three primary groups: yeasts (primarily Saccharomyces and Pichia species), lactic acid bacteria (LAB, primarily Lactobacillus and Leuconostoc), and acetic acid bacteria (AAB, primarily Acetobacter and Gluconobacter). Each group metabolises mucilage sugars differently and leaves behind distinct compounds.
Yeasts produce esters — organic molecules with characteristic fruity aromas. Isoamyl acetate reads as banana; ethyl butyrate as pineapple; ethyl hexanoate as green apple. Which species dominate, and under what oxygen and temperature conditions, determines the ester mix. Low-oxygen environments favour yeast and push ester output significantly upward — research suggests anaerobic conditions boost ester production 300 to 400 percent compared to conventional washed fermentation. That is the direct chemical basis of the tropical-fruit character anaerobic coffees consistently produce. The four-fold shift also explains why anaerobic cups read as categorically different from naturals rather than simply more intense — the ester mix changes under low-oxygen conditions, not just the concentration.
LAB produce organic acids, primarily lactic acid, which at moderate concentrations contributes a smooth, yoghurt-like brightness. This acid diffuses into the bean during fermentation and shapes perceived acidity — brighter and more defined than the washed baseline without becoming sharp. Longer fermentation windows, managed carefully, let LAB build lactic acid in ways that add complexity while retaining the structural clarity washed processing is known for.
AAB produce acetic acid. At low concentrations it reads as clean and mildly sweet. Above roughly 2 mg/ml, it tips into vinegar — the defining signal of over-fermentation in a washed lot. The balance between these three groups shifts with oxygen availability, temperature, duration, and cherry condition.
The precursor-roasting link: These microbial byproducts do not create flavour directly. They alter the bean's chemical matrix — building up specific amino acids, modifying sugar pools, depositing organic acids into the parchment layer. Roasting acts on this altered matrix through Maillard reactions (amino acids reacting with sugars) and Strecker degradation to produce the hundreds of aromatic compounds that make up cup character. A coffee fermented for 12 hours and one fermented for 72 hours arrive at the roaster carrying different precursor profiles. Same farm, same roaster, same machine — different cup. The fermentation's effect is baked in before roasting starts.
Oxygen level, duration, temperature, and whether the cherry skin stays intact — these four variables define a fermentation environment. Different configurations produce different microbial balances and therefore different compound profiles.
Aerobic washed. Depulped beans ferment in water or open tanks with oxygen present. LAB and AAB are active alongside yeast, but the short window (12 to 36 hours in India) limits how far any group can run. Organic acids — primarily lactic and citric — dominate the flavour contribution. Across 292 washed coffees in ICB's catalog, the top flavour notes are nuts (55 lots), dark chocolate (51), chocolate (47), and citrus (44).
In-cherry natural. The whole cherry dries intact for 15 to 25 days. Fermentation occurs inside the skin, which acts as a membrane, with yeast working the fruit sugars slowly over weeks. Sugars and esters migrate gradually into the bean. The flavour outcome is more variable than washed. Drying conditions, turning frequency, and cherry ripeness all contribute, and results skew fruity and textured. Among 142 natural coffees in ICB's catalog, top notes are fruity (20 lots), berry/fresh (17), dark chocolate (15), and grape (12).
Sealed anaerobic. Cherries or depulped beans ferment in sealed, oxygen-deprived tanks, sometimes with CO₂ injected to clear residual air. The low-oxygen environment strongly favours yeast, driving ester production up. Duration is typically 36 to 72 hours for Indian lots. Among 111 anaerobic coffees in ICB's catalog, top notes are grape (15 lots), raisin (13), dark chocolate (13), peach (10), and caramel (10).
Carbonic maceration. Whole cherries are sealed under a CO₂ atmosphere — the most extreme anaerobic variant. The intact skin provides a secondary fermentation layer, and the CO₂ environment slows oxidation while microbial activity continues inside the fruit. The compound profile combines high ester output with preserved intact-fruit character. Among 28 carbonic maceration coffees in ICB's catalog, top notes are berry/fresh (6 lots), grape (4), plum (3), and cherry (3).
Fermentation-forward processes — anaerobic, experimental, carbonic maceration, and double-fermented — account for 241 of 1,238 coffees in ICB's directory: 19.5% of the catalog. Almost all are in stock (239 of 241 available). A decade ago this segment was effectively absent from Indian specialty listings.
Regional concentration is tight. Anaerobic lots cluster in Chikmagalur (34 lots), Kodagu/Coorg (11 lots), and Sakleshpur (8 lots). Carbonic maceration sits primarily in Kodagu/Coorg (9 lots) and Sakleshpur (6 lots). That Western Ghats concentration reflects access to estate-level curing infrastructure — sealed fermentation tanks, temperature monitoring equipment, CO₂ supply — that smallholder cooperatives generally cannot replicate. Experimental fermentation in India is estate-led, not cooperative-led.
Among roasters stocking anaerobic lots, Coffeeverse carries the largest share (14 lots), followed by Fraction 9 Coffee (10 lots), RiverSide Coffee (7 lots), and Blue Tokai and Hill Groove Coffee (5 lots each). The concentration in sourcing-active roasters with direct-farm relationships suggests anaerobic has moved past early-adopter status into a repeatable segment of the Indian specialty catalog.
Note on the 'experimental' label: ICB carries 80 coffees tagged 'experimental'. The category covers lots that don't fit neatly into named processes — extended-fermented washed lots simply labelled 'washed', co-fermented lots, lactic-process lots, and others. The actual fermentation-forward share of the Indian catalog is likely higher than 19.5%; the experimental tag catches what the named-process taxonomy misses.
Two distinct approaches have emerged in Indian experimental fermentation, shaped by different ideas about what controlled processing should mean.
Ratnagiri Estate in Chikmagalur, operated by Ashok Patre, runs the most extensively documented precision program. The 117-hectare farm produces 60% specialty-grade output and, according to Cafe Imports' documentation, applies 15 distinct fermentation protocols across a single season: multiple washed variants, carbonic maceration naturals, anaerobic naturals, yeast-fermented naturals, aerobic fermented honeys, and a signature carbonic macerated washed. One documented lot placed cherries at 25.9° Brix into bio-fermenters, inoculated them with lab-grown yeast and bacterial cultures, and CO₂-purged the tanks for 96 hours. Another — the carbonic macerated washed — harvested cherries at 23 to 27° Brix, fermented them in-cherry anaerobically for 48 hours, then moved the depulped beans to CO₂-purged stainless steel tanks for a further 90 to 120 hours. ICB lists 72 Ratnagiri lots across all process types, sourced by roasters including Subko and Coffeeverse.
Adding selected yeast or bacterial cultures — rather than relying on ambient wild microbes — reduces batch-to-batch variation and lets producers hit consistent flavour targets across harvests. Ratnagiri's "culture washed" label refers specifically to this controlled-inoculation protocol on their washed lots.
Black Baza Coffee takes a different view. Working with tribal smallholder farmers in Andhra Pradesh and Odisha, they use locally foraged cofermentation agents — wild ginger and kombucha cultures from the same ecological zone as the farms — rather than lab-grown strains and stainless-steel vessels. In 2025, more than a dozen microlots used these coferments. The approach treats fermentation as a function of the landscape's biology rather than a controllable precision variable. Their documented lactic-process line achieves a cream-bright cup through deliberate LAB inoculation, producing an acidity character distinct from both standard washed and fruit-forward naturals.
Both approaches — precision inoculation and ecological cofermentation — extend Indian specialty's flavour range beyond what conventional washed and natural protocols allow. The underlying mechanism is the same; what differs is the philosophy of how to guide it.
Over-fermentation occurs when acetic acid accumulates beyond the clean-sweet threshold — above roughly 2 mg/ml — or when other bacterial byproducts form from uncontrolled overgrowth. The cup reads as sharp vinegar, pickle brine, ethyl acetate (nail varnish or solvent), or rotting fruit. Four causes account for most defects: the fermentation window extending past the intended endpoint due to temperature spikes or inadequate monitoring; contaminated tanks or water; over-ripe or damaged cherries with compromised skin that allow uncontrolled bacterial entry; and delays between harvesting and pulping that let ambient fermentation start before the intended process does.
Identifying over-fermentation: Intentional fermentation character — tropical, wine-like, or fruity notes from well-managed anaerobic or natural lots — is distinct from defect sourness. The marker is clarity: well-fermented coffee has definition even when fruit-intense, with a clean finish. Over-fermented coffee feels murky or chemically sharp, with astringency that worsens as the cup cools. ICB's catalog records a 'Sour/Fermented' sensory category appearing in 19 natural coffees versus 5 washed — a signal that natural process carries higher defect exposure, particularly when drying management is inconsistent.
India's conditions add pressure here. High humidity, warm temperatures at lower altitudes, and variable monsoon weather mean washed fermentation running 12 hours past its intended endpoint can produce defect notes in the cup. Temperature monitoring has become standard practice at estates running controlled experimental fermentation for exactly this reason.
Reading flavour notes on an Indian specialty coffee bag is more interpretable once you can decode the fermentation environment the label implies.
Washed / Plantation A / Plantation B signals short aerobic fermentation — 8 to 36 hours in India, mucilage removed before drying. Mixed microbial activity (yeast, LAB, AAB) constrained by duration. Cup character: clean, defined acidity, transparency to variety and terroir.
Natural / Cherry AB / Cherry PB signals extended in-cherry fermentation during drying — 15 to 25 days of slow aerobic microbial activity inside the intact fruit. Cup character: fruity, textured, variable depending on drying management.
Anaerobic specifies the oxygen environment during fermentation — sealed, oxygen-deprived — not whether the cherry skin is intact. Duration for Indian lots is typically 36 to 72 hours. Yeast dominates; ester production is elevated. Cup character: wine-forward, tropical, grape or stone-fruit intensity not found in conventional processing.
Carbonic maceration is the extreme end of anaerobic: whole cherries sealed under CO₂, the intact skin providing a secondary fermentation layer. Cup character: berry-bright, defined acidity alongside fruit intensity.
Culture washed / Inoculated means specific yeast or LAB strains were added to the fermentation vessel rather than relying on ambient wild microbes. Ratnagiri Estate uses this approach to hit consistent flavour profiles across harvests.
Extended fermentation / Double fermented covers washed lots with deliberately longer windows (48 hours or more) or two sequential fermentation stages. The acid profile is typically more complex than standard washed, sometimes with more body, while retaining the process's underlying clarity.
Lactic process means LAB were deliberately encouraged — through inoculation or controlled conditions — to produce a cream-bright, clean acidity distinct from the sharpness of acetic-dominated profiles.
The proliferation of sub-labels — culture washed, double fermented, lactic process — is a sign of category maturation: producers naming their methods precisely because roasters and buyers have started asking.
Humanizer changes: Double "That" anaphora in ester paragraph fixed ("That four-fold shift" → "The four-fold shift"); em dash parenthetical pair in natural process paragraph broken into two sentences for cleaner rhythm.
