Coffee beans are sorted at various stages during processing. After harvest, they are sorted for ripeness with unripe or overripe cherries being rejected. This is especially the case when strip harvesting is used. Once the beans have been through processing they are again sorted to look for defective beans. Defective beans can affect the tastes of a whole batch of beans so it is important to detect them. Sorting beans can be done by hand, by a machine or electronically.
Typical bean defects are:
A quaker is an underdeveloped bean with a wrinkled surface. When it is roasted it will appear lighter than the others.
A green bean that is mostly black in colour due to disease, lack of water or nutrients during growth.
Dark green beans are generally darker than the rest and have a distinct rotten fish smell.
A discoloured bean that is different from the rest can be due to poor processing techniques such as over-drying.
A stinker is a defective bean that has a rotten smell when crushed. It glows under ultraviolet light and can be detected electronically.
Once beans have been sorted for defects, they are then graded and classified. There is no universal classification system and each country has its own standards for export.
Generally, the classifications are based on the following:
altitude
region
botanical variety
processing method
bean size, shape and colour
number of defects the appearance and quality after roasting.
The International Coffee Organisation (ICO) has four categories:
Arabica beans mainly from Colombia, Kenya and Tanzania which are graded by size.
Arabica beans from Mexico, Central and South America. These are graded by altitude, density, size and number of defects.
Arabica beans from Brazil and Ethiopia. These are graded by the number of imperfections.
Robusta from Africa and Southeast Asia. These are mostly graded by density, colour and number of defects.