From Killenso to the World: Why a Third-Generation Ethiopian Coffee Family Chose a Chinese 10TPH Coffee Bean Cleaning Line
A third-generation Ethiopian coffee family in West Guji needed to scale cleaning without sacrificing specialty-grade quality. A turnkey 10TPH coffee cleaning line went from first search to full operation in roughly three months, covering layout design, shipping, installation, and commissioning.
Killenso, in Ethiopia's West Guji zone, produces some of the country's most sought-after washed arabica. The family that farms there has been in coffee for three generations - and in 2025 they did something the previous two never had to: they bought an industrial cleaning line, and they bought it from China.
A family business at an awkward size
Third-generation producers in Guji face a specific squeeze. They are too large to sell cherry to the local cooperative and too small - by international standards - to justify the kind of turnkey plant a multinational exporter would install. Their coffee was good enough to attract specialty buyers in Europe and Japan, but the final cleaning, grading and colour sorting was still being done by hand on tables.
Hand sorting is not automatically bad; for very small lots it is still the gentlest option. It does, however, put a hard ceiling on volume. During the peak of the season the family was turning away cherry they had already contracted, simply because they could not clean it fast enough.
What they were afraid of
The hesitation was not about capacity. It was about whether a machine could handle speciality coffee without damaging it. Three concerns came up repeatedly in our first calls:
Breakage. Aggressive conveying and over-long drop heights crack beans. Cracked beans lose cup score and get downgraded.
Over-sorting. A colour sorter tuned too aggressively ejects good beans along with the defects, and in speciality coffee those ejected beans are the most expensive ones in the building.
Service distance. West Guji is a long drive from Addis Ababa. If the line stopped, no technician was going to arrive the same afternoon.
The answer: a 10TPH line built around gentle handling
We supplied a complete 10TPH coffee cleaning line, laid out in 3D against the family's existing warehouse before a single machine was built. The process is conventional - the engineering that protects quality is in the details:
Stage | Equipment | Purpose |
|---|---|---|
1. Feeding | Low-speed bucket elevator with cushioned intake | Lifts beans without impact cracking |
2. Pre-cleaning | Air-screen pre-cleaner | Removes dust, floaters and light chaff |
3. De-stoning | Density de-stoner | Removes stones picked up on drying beds |
4. Gravity separation | Specific gravity separator | Drops immature and insect-damaged beans |
5. Colour sorting | Removes sour, mouldy, faded and foreign beans | |
6. Grading | Deck grader with screen set for speciality calibres | Sorts by size for lot consistency |
7. Dust control | Central cyclone and filter system | Keeps the hall breathable and beans clean |
Breakage was addressed mechanically: the elevator runs at low speed, every transfer point was designed with a short drop, and contact surfaces were lined where beans slide. Over-sorting was addressed in software. The sorter was commissioned with a conservative reject threshold calibrated against the family's own hand-sorted reference sample - the machine had to match what their best sorters did by hand, not exceed it. The client keeps that reference sample and can re-calibrate against it at the start of every season.
For the service-distance problem, every line ships with a remote diagnostics module. Most faults can be diagnosed from our side, and the spares kit delivered with the plant covers the parts that actually wear. See the colour sorter range for the platform this is built on.
Three months from first search to running coffee
Week | Milestone |
|---|---|
1 - 2 | Requirement review, sample testing, 3D layout design signed off |
3 - 6 | Manufacturing, assembly and factory acceptance test with the client's own beans |
7 - 9 | Container loading, shipping to Djibouti, overland transport to West Guji |
10 - 11 | Installation, calibration against the hand-sorted reference sample |
12 | Commissioning, operator training, first commercial lot run |
The factory acceptance test was the step that convinced them. Their own beans - the actual lot they were worried about - were run through the line in our facility in China and the sorted output was shipped back for cupping before the machine was ever disassembled for export.
What changed
Throughput: from roughly 1.5 TPH by hand to 10 TPH sustained.
Breakage: held below 0.5%, verified against screen analysis after commissioning.
Lot consistency: grading by screen size produced uniform lots, which made the cupping scores more repeatable across a season.
Peak-season capacity: the family no longer turns away contracted cherry.
Labour: about 30 seasonal hand sorters were redeployed to drying-bed and agronomy work, where attention to detail pays more.
'My grandfather sorted by hand and my father sorted by hand. We did not want a machine that was faster but made the coffee worse. What convinced us was running our own beans through the line in China and cupping the result before it shipped.' - Third-generation producer, West Guji
Should a speciality producer mechanise?
Our honest answer: only once hand sorting has become your bottleneck, and only with a machine that can be calibrated against your own reference standard. A colour sorter is not a replacement for good agronomy or careful fermentation - it is the last step that protects the value created upstream. If you are at the point where you are declining contracted volume, or where buyer rejections are costing you more than a line would, it is worth a conversation and a sample test. Send us half a kilo of your beans and we will run them and send the sorted output back.