Drying, Sorting, and Green Coffee

Phenomenon

In the corner of a roastery sit a few jute sacks of green coffee. Each is printed with an origin, a grade and a lot number; some come with a spec sheet that reads "moisture 10.8%", "water activity 0.55", "screen 17/18".

Pour out a handful: grey-green to pale yellow-green, very hard, smelling of hay, beans or fresh grass.

A freshly picked coffee cherry left at room temperature will ferment and mold within days (Chapter 9). Yet these green beans have crossed an ocean and sat in a warehouse for months, and they can still be roasted. What does a seed have to satisfy to manage that?

Variables

Between the drying yard and the roaster lie roughly these stages:

Two quantities are worth separating from the start:

• Moisture content: how much water is in the bean, as a percentage of weight;
• Water activity: how "free" that water is — whether microbes and chemical reactions can use it — on a scale from 0 to 1.

Experiment

Thought experiment: split one lot of washed parchment coffee in two, and dry both halves to the same final moisture content.

• A: dried on raised beds over a dozen days or so, coming down slowly;
• B: put in a mechanical dryer at a very high temperature, done in two days.

Write down your prediction before reading on

Both halves have the same moisture content. Are they the same green coffee? After a period of storage, which one might run into problems sooner?

Your prediction

Save prediction

Stored only in this browser on this device. Never uploaded.

Second experiment: split a lot of sound green coffee in two and store it for a year.

• C: in ordinary jute sacks, in a hot, humid port warehouse;
• D: in sealed inner bags, in a cool, dry warehouse.

Write down your prediction before reading on

After a year, which half will roast up "flatter" and less lively? Might one of them grow mold?

Your prediction

Save prediction

Stored only in this browser on this device. Never uploaded.

Observation

First experiment: the same moisture content does not mean the same beans. Researchers who looked under the microscope at parchment coffee dried at different temperatures found that seeds dried at 40 °C kept their cell structure intact, while in those dried at 60 °C the cell membranes ruptured and oil and other cell contents leaked out Confidence: [Strong Evidence]. Drying conditions also change the sugar content of washed green coffee Confidence: [Strong Evidence]. Green coffee dried too fast or too hot tends to age faster and go "flat" in the cup sooner Confidence: [Rule of Thumb].

Second experiment: one study stored differently processed green coffees under standard conditions for two years. With long storage, cups developed a typical "flattening", plus off-notes from lipid oxidation; the seeds lost viability within six months, except washed coffee kept in its parchment, which stayed viable for more than a year Confidence: [Strong Evidence]. Another study found that sealed big bags (with or without injected CO₂) kept green coffee's quality largely intact over 12 months Confidence: [Strong Evidence].

As for mold: in the lab, green coffee needs a water activity of at least about 0.80 (roughly 14% moisture) before molds can produce ochratoxin A Confidence: [Strong Evidence]. Sound green coffee sits well below that, so shipping and storage are not usually a major contamination route — but if the coffee gets damp, that route opens again.

So what counts as "sound"? The International Coffee Organization's quality programme asks that exported green coffee have a moisture content no lower than 8% and no higher than 12.5% (measured by the ISO 6673 method) Confidence: [Common Practice]; it is a voluntary programme. Specialty Coffee Association standards require specialty-grade green coffee to have a water activity below 0.70, and grading uses a 350 g sample Confidence: [Common Practice].

Model

Drying removes usable water

What microbes and most chemical reactions need is not "water" but "water they can use". The lower the water activity, the harder it is for microbes to grow, and the slower many chemical reactions run Confidence: [Established Mechanism]. Moisture content and water activity are related but not the same: at the same moisture content, water activity can differ between beans and at different temperatures.

So the goal of drying is not "as dry as possible" but to land in a window:

• too wet: risk of mold, fermentation and toxins (Chapter 9);
• too dry: the beans become brittle and break more during hulling; coffee is sold by weight, so lost water is lost money;
• too fast, too hot: cell structure is damaged and the beans age faster;
• uneven: some beans in the lot over-dried, some under-dried, and they will roast unevenly too.

Many producers rest parchment coffee or dried cherries for a while after drying and before hulling, letting moisture even out between beans and within each bean Confidence: [Common Practice].

Sorting separates seeds that are not alike

After hulling, a lot of green coffee is a mix of seeds of different sizes, weights and colors — and some problem seeds. Sorting separates them:

• by size: using screens with different hole sizes, conventionally numbered in 64ths of an inch, so a screen 18 has holes 18/64 inch across Confidence: [Common Practice];
• by density: using a vibrating gravity table to separate light and hollow seeds;
• by color: using optical sorters or human hands to pick out off-colored seeds.

Defects can come from anywhere along the chain:

Grading standards usually count defects by category, with more serious defects weighted more heavily Confidence: [Common Practice]. Sorting removes many problems, but not all of them: some defects cannot be seen from outside.

A grade name is not a flavor name

Many origin grade names are, at heart, about size or defect count: Kenya's AA and Colombia's Supremo are mainly size grades; Ethiopia's Grade 1, Grade 2 and so on are based mainly on defects and cup evaluation Confidence: [Common Practice]. Each country's system is different. Larger beans are easier to roast evenly, but "large" by itself is not "delicious".

Green coffee keeps changing, slowly

Drying makes green coffee stable, but it does not stop it changing:

• it starts out alive, and green coffee with its parchment removed mostly loses viability within six months Confidence: [Strong Evidence];
• the exact cause of flavor "flattening" during storage is still being studied; one line of research links it to slow Maillard reactions using up aroma precursors during storage Confidence: [Open Question];
• heat and humidity speed all of this up.

So green coffee has its own "freshness": when the trade talks about "new crop" and "past crop", it means how long ago the coffee was harvested and processed Confidence: [Common Practice]. That is a second clock, separate from roasted-coffee freshness (Chapter 17).

Transfer

In the Processing Mill scene you can watch the final drying stage of all three pathways and compare how much water each has to lose. When reading a green-coffee spec sheet, or a roasted bag:

Storage after roasting is a question for Chapter 18.

Misconceptions

"Green coffee is dry, so it keeps forever." Green coffee is far more stable than roasted coffee, but it still ages: it loses viability, its aroma precursors decline, its lipids oxidize, and the cup goes "flat". Heat and humidity speed all this up, and dampness can lead to mold.

"The faster it dries, the better." Drying too fast or too hot damages the seed's cell structure and makes green coffee age faster. What drying needs is a suitable pace, and evenness.

"The drier, the better." Over-dried beans are more brittle and break more during hulling, and they lose value when sold by weight. The target is a window, not "as low as possible".

"High grades like AA or Supremo always taste better." These grades mainly describe size, not flavor. Smaller beans can be excellent.

Real Lab

There is no registered lab card for this chapter. Here is an activity for looking closely at green coffee.

You need: about 100 g of green coffee (some home-roasting suppliers and roasters sell it), a sheet of white paper, a kitchen scale and an airtight jar. Even better if you can get a washed and a natural from the same origin.

1. Look at the color: pour the beans onto white paper and look in daylight. Washed green coffee often leans blue-green or grey-green, naturals more yellow-green or yellow-brown — but that is only a common tendency Confidence: [Rule of Thumb].
2. Smell: green coffee smells of hay, grass or beans, completely unlike roasted coffee. Is there anything musty or fermented?
3. Feel: press with a fingernail. Properly dried green coffee is very hard and will not take a mark. Do not bite it.
4. Sort: go through every bean and separate out the broken ones, ones with small holes (insect damage), ones that are clearly black or pale, odd shapes, and any stones or debris. Count each group.
5. Weigh (optional): take 50 g, put half in an open bowl in the kitchen and half in the airtight jar, and weigh again after a week. If your scale reads to 0.1 g, you may see a small change — green coffee exchanges moisture with the air. A very small change is normal.

If you cannot get green coffee, sort roasted beans instead: pour a bag onto white paper and look for beans that are clearly paler than the rest, with wrinkled surfaces. The trade calls these pale beans that will not take color "quakers", and they mostly come from unripe cherries Confidence: [Common Practice]; they often bring a raw, peanut-like taste.

Check Your Model

Predict. At the peak of harvest, a farm runs out of drying space. To free it up, they bag parchment coffee that is not yet fully dry and stack it in a warehouse for a few days before finishing drying. What do you expect to happen?

Diagnose. A roaster notices that one lot from one farm tasted great when it arrived in March, but by December the same roast profile gives a flat, woody cup with a hint of cardboard. What is the most likely cause, and what could they check next?

Explain. In your own words, explain why moisture content alone is not enough to judge whether a lot of green coffee is in good shape.