Ethiopian Coffee Dry Milling: Export Preparation Guide

From dried parchment to FOB-ready green beans: what every importer needs to understand about the final quality checkpoint before shipment.

Most attention in the Ethiopian coffee supply chain goes to origin: the region, the washing station, the processing method. But between a well-dried parchment lot leaving a cooperative in Yirgacheffe and that same coffee loading into a container at Djibouti, one more critical step takes place. Ethiopian coffee dry milling is where parchment becomes green beans, where grades are confirmed, and where the cumulative quality of every earlier decision either holds or falls apart.

For importers and roasters, understanding dry milling is not academic. Milling calibration directly determines defect count, screen size uniformity, and whether your pre-shipment sample accurately matches what lands at your warehouse. It also affects yield: the ratio of parchment input to green coffee output varies by 15 to 20 percent depending on processing method and milling settings. That spread is money.

This guide walks through the full dry milling process as it operates in Ethiopia, explains what each stage does to your coffee, and ends with a practical checklist of questions to ask any exporter before you sign a contract.

What Is Dry Milling and Why Does It Determine Export Quality?

Dry milling is the post-drying mechanical processing stage that converts dried parchment coffee (or whole dried cherries, in the case of naturals) into clean, sorted, graded green beans ready for bagging and export. It is distinct from wet milling, which takes place at washing stations immediately after harvest and involves the removal of the cherry skin and mucilage in the presence of water.

Where wet milling shapes fermentation character and clean-cup quality, dry milling determines physical lot uniformity. A batch of coffee exiting a dry mill should be homogenous in bean size, density, and color. That uniformity matters for two reasons: first, it is what grading systems measure, and grades determine price; second, uniform beans roast evenly. A mixed-screen lot roasts unevenly, generating a cup with both under- and over-developed notes regardless of how skilled the roaster is.

The Ethiopian Dry Mill Infrastructure

Where Ethiopian Dry Mills Operate

Ethiopian dry mills are concentrated on the outskirts of Addis Ababa. This is not accidental. Ethiopian regulations require dry mill facilities to have sufficient yard space for truck ingress and egress, and for full shipping containers to maneuver and load. The capital also sits near the central highway system connecting the growing regions (Sidama, Guji, Yirgacheffe, Hararghe, Jimma) to the Djibouti corridor, which handles nearly all Ethiopian coffee exports.

Mills are owned by a mix of private exporters and cooperative unions. Some private exporters own their own milling equipment and offer toll milling services to smaller exporters who do not. Toll milling fees in Ethiopia typically range from USD 0.10 to 0.18 per kilogram of green output, depending on processing method and service level (automated sorting only vs. automated plus hand sorting).

ECX-Route Coffee vs. Direct Sourcing (DSL) - Two Different Milling Paths

Ethiopian coffee reaches dry mills through two primary channels, and the milling path differs between them.

The 8 Stages of Ethiopian Coffee Dry Milling

Once parchment coffee arrives at a dry mill in Addis Ababa, it moves through the following sequence. Each stage can be adjusted or bypassed depending on the lot type, buyer specifications, and target grade.

Stage 1: Pre-Cleaning

Before any milling equipment touches the coffee, a pre-cleaner removes light impurities. This is typically a combination of three sub-steps: a pneumatic separator (airstream) that lifts loose dust, fiber, and broken parchment fragments away from heavier beans; a magnet that strips out metallic particles that could otherwise damage hulling equipment; and a destoner that uses vibration and airflow to separate small stones and clods of soil from the coffee stream. Together, these protect downstream machinery and remove foreign matter defects before they reach the grading stage.

Stage 2: Hulling

Hulling removes the parchment layer (endocarp) from washed and honey lots, or the entire dried cherry from natural lots. Ethiopian dry mills use one of three huller types:

• Disc hullers: Use rotating abrasive discs to strip parchment through friction. Effective for washed coffees with uniform moisture. Risk of bean fracture rises if moisture is below 10% or if discs are worn.
• Drum (barrel) hullers: Common for naturals. Rotating drum applies pressure to crack the dried cherry layer. Requires careful calibration to avoid chipping.
• Friction hullers: Use a rubber-lined rotor and stationary screen to abrade the parchment. Lower bean damage rate, more commonly found in newer facilities.

Hulling calibration is highly consequential. Over-aggressive huller settings generate chipped and broken beans. These count as primary defects under the Ethiopian Coffee and Tea Authority (ECTA) grading system and reduce the lot's grade. Experienced operators adjust settings based on lot moisture content and origin, since parchment thickness varies between regions and processing methods.

Stage 3: Polishing (Optional)

After hulling, some silver skin (silverskin or chaff) may remain on the bean surface. A polisher removes this residue through light abrasion. Polishing is not universal: some exporters skip it entirely, some buyers request it, and others actively prefer unpolished beans.

The debate centers on heat. Polishers generate friction, and if the process runs too long or the machine is not properly cooled, bean surface temperatures rise enough to affect volatile aromatics. Some importers sourcing delicate washed Yirgacheffe lots specify no polishing for precisely this reason. Others, particularly those sourcing natural Djimmah or commercial-grade lots, prefer polishing for a cleaner appearance. This is a specification decision that should be made in writing before milling begins.

Stage 4: Size Grading

After hulling, beans pass through a mechanical size grader: a series of vibrating trays, each perforated with holes of a specific diameter. Beans fall through successive screens until they reach a screen small enough to retain them. Screen size is measured in 64ths of an inch; a size 16 screen has holes 16/64" in diameter.

In Ethiopia, beans retained on size 14 and above are generally considered specialty grade. This threshold is referenced in ECTA grading protocols and aligns with the SCA minimum for specialty-eligible lots. Beans falling below size 14 are sorted into commercial or below-grade streams and priced accordingly.

Stage 5: Density Sorting

Size grading removes small beans, but it does not separate beans of the same size that differ in density. High-density beans have fully developed endosperm, lower moisture, and greater cup potential. Low-density beans of the same screen size are underdeveloped, lighter, and prone to quaker defects after roasting. A density (or gravity) table sorts them.

The density table operates on two simultaneous axes: it is tilted both horizontally and vertically, and the tray vibrates at a controlled frequency. The vibration causes beans to stratify: denser beans migrate toward one corner, lighter beans toward another. Adjustable partitions then divide the output stream into two or more density grades. The lightest fraction is separated as a lower-quality stream. Denser fractions are retained for the specialty lot.

Density sorting is where a standard G2 lot can sometimes be refined into a G1 by removing the lightest 8 to 12 percent of the screen-sized batch. However, this increases milling losses and raises effective cost per kilogram, so the decision depends on the price premium available at the higher grade.

Stage 6: Color Sorting

Color sorters scan every bean passing through the machine using high-resolution optical sensors or, in newer machines, near-infrared (NIR) laser technology. Beans that fall outside the specified color range are ejected by a burst of compressed air into a reject stream. Color sorters remove visually defective beans: full blacks, full sours, partial blacks, and discolored partials that density grading cannot distinguish because they may have the same weight as good beans.

Calibration sensitivity is a trade-off. A tightly calibrated color sorter rejects more defects but also rejects some good beans that happen to be slightly darker or lighter than the programmed baseline. This is a particular challenge for natural and anaerobic lots, where intentional fermentation colors may trigger false positives. Experienced mill operators running premium natural lots often soften the color sorter parameters and pass the lot through hand sorting for a final clean.

Stage 7: Hand Sorting

Hand sorting is the final defect-removal gate and the most labor-intensive stage. Workers positioned along a conveyor belt or sorting table visually inspect every bean passing in front of them, removing anything the machines missed: insect-damaged beans, partial blacks that escaped the color sorter, quakers, shells, and any remaining foreign matter.

In Ethiopia, hand sorting is performed predominantly by women, many of them temporary seasonal workers hired during the October through August export season. Sorting wages are typically paid per bag processed, incentivizing speed, which is why machine pre-sorting matters: workers sorting a heavily defective lot cannot sustain the same accuracy per bean as workers handling a mostly clean lot.

For Grade 1 specialty lots, most Ethiopian dry mills run the coffee through at least two passes under hand sorting: one pass after color sorting, and a final check pass before bagging. This double-pass protocol is standard for lots targeting SCA cupping scores of 85 and above.

Stage 8: Weighing, Bagging, and GrainPro Lining

Once sorted, green coffee is weighed into 60-kilogram lots and packed into jute bags. For specialty-grade Ethiopian coffee, the jute bag is lined with a GrainPro or similar hermetic liner before filling. GrainPro liners create a near-airtight microenvironment inside the bag, dramatically slowing the oxidation and moisture exchange that degrade green coffee freshness during multi-week ocean transit.

Bags are sewn, stacked, and held in a temperature-controlled warehouse until container loading. At loading, a certified sampler takes random samples from multiple bags per lot (typically 8 to 12 bags per 300-bag container) for the pre-shipment sample set. The container should only be sealed once the exporter and buyer have both approved the pre-shipment samples.

Washed vs. Natural: Different Milling Paths Through the Same Facility

Processing method changes what the dry mill actually encounters, and operators must adjust accordingly.

Washed coffees arrive at the dry mill as dried parchment with moisture content typically between 11.0 and 12.5 percent. The parchment shell is relatively thin and uniform, making hulling predictable. Size and density are more homogenous because the washed process removes the cherry before drying, and sorting was typically performed at the washing station. Color sorting is straightforward because the desired color range is narrow and well-defined.

Natural (dry-process) coffees arrive as hulled green beans, having been pre-hulled in the interior to remove the dried cherry, or they may arrive as whole dried cherries requiring a first-pass hulling at the Addis mill. Naturals have greater size variation, broader density distribution, and far wider color variance due to fermentation-influenced color development. Color-sorter sensitivity must be reduced, and hand-sorting passes are more intensive. Naturals typically generate 5 to 10 percent more milling waste per input kilogram than washed lots.

Yield and Losses: What Buyers Need to Know

Milling yield is the ratio of export-ready green coffee output to dried parchment or cherry input. It affects both cost and total available volume and is a number every importer should understand before placing an order.

These yield figures explain why FOB prices for washed specialty lots appear higher per kilogram than commercial naturals. When you calculate cost per kilogram of green output from parchment input, the gap narrows considerably. It also explains why exporters must purchase significantly more parchment coffee to fill a G1 specialty container than to fill a commercial G4 container of the same net weight.

Quality Control at the Dry Mill: The Pre-Shipment Sample

The pre-shipment sample (PSS) is the QC gate between dry mill and container loading. It is taken from random bags in the milled lot, typically after final hand sorting but before bags are sealed. The PSS serves two functions: it allows the exporter to confirm the lot meets the contracted specification, and it gives the buyer a final opportunity to approve or reject before the container is sealed and invoiced.

A properly executed PSS draw samples from multiple bags per lot. For a 320-bag (19.2 MT) container, a credible PSS uses samples from at least 10 to 15 percent of the bags, drawn randomly rather than from top layers only. The combined sample is split into at minimum three parts: one for the exporter's cupping lab, one for the buyer, and one retained as a sealed reference. In Ethiopia, ECTA also requires exporters to submit samples for official green coffee grading before any lot can leave the country.

What Importers Should Ask Their Exporter About Dry Milling

The dry milling process is largely invisible to buyers who work purely through samples and cupping scores. But asking the right questions before you commit to a contract can prevent expensive surprises. The following checklist covers the minimum information a buyer should confirm with any Ethiopian exporter before issuing a purchase order.