Coffee Farming/Roasting

Small-Scale Roasting Recipes

This is a collection of tried and tested recipes to roast coffee in small quantities using different home equipment. Since you will rarely have the exact same equipment and circumstances, you will have to find your own perfect settings. These recipes can still be a good starting point.

HIghland Arabica Roasting: Pan on Ethanol or Gas Stove, French Roast

Environment

Using an ethanol fueled, pressureless burner (model Origo 1500). Any propane / butane / natural gas / biogas gas stove can also be used, with slight modifications to the process. The beans are Highland Arabica Coffee from Nepal, namely the first batch of the 2017 harvest of this product.

Result

A "perfect" French roast at 23.43% roasting loss (French roast is said to have 23.2% roasting loss – see). This is the darkest reasonably possible roast.

Process

1. Use a Origo 1500 ethanol stove with a full tank.
2. Use a cast iron flat pan, stir with a long spoon or spatula (to stay away from the heat).
3. Put in 200 g of green beans.
4. Start the stove and heat on 100% power ("4 of 4") until 140 °C bean temperature. (That should be a bit after the start of the first crack, around 7:30 min.)
5. Lower stove power to 68% ("2.7 of 4").
6. Finish at 220 °C bean temperature. (That should be the middle to end of the second crack, around 25:15 min.)
7. Cool by stirring with a spoon in a colander until hand warm.

Explanations

For these beans, temperature seems to be a much better indicator than time, color or sounds. That is because these beans have much of the silverskin on them, are of mixed size (resulting in loosely defined first and second cracking), and also due the difficulties to achieve the same heat uptake (depends on wind speed etc.) and to set an ethanol or gas stove to the same output. (Power output depends on ambient air pressure for gas, outdoor temperature for ethanol, and filling degree of the tank).

To determine temperature, use a handheld "gun type" IR thermometer on a small (ca. 1 cm diameter) section of beans right after turning them over, so with their hot side up. Check the distance to diameter ratio of your thermometer; for example if yours has a 8:1 ratio, this means measuring from a distance of 8 cm to get a sampling area of 1 cm diameter. Use 3-4 measurements, some from the center and some more (but not too much) from the sides, and make an approximate average to know the "real" bean temperature.     

Highland Arabica Roasting: Pan on Electric Stovetop, Espresso Roast

Environment

Using an electrically heated glass ceramics stove (normal home appliance), namely one of the two larger cooking fields. The beans are Highland Arabica Coffee from Nepal, namely the first batch of the 2017 harvest of this product.

Result

Medium to dark roast, good for a typical Italian espresso preparation in a stovetop espresso maker. Could be roasted even a bit darker for that purpose. Due to the dark roast, the resulting coffee does not have much of the beans' own flavour (here: hazelnut / roasted hazelnut / chocolate). But that is by intention.

Process

1. Take a wok type pan and pre-heat it to about 80-100 °C.
2. Put in 100 g of green coffee beans.
3. Set stove to 89% power ("8 out of 9") and start a timer.
4. Start stirring the pan with a wooden spoon.
5. At the beginning of the first crack, lower the stove power to 72% power ("6.5 out of 9").
6. At 13:15 min, pour the beans into a metal colander. This should be in the middle of the second crack.
7. Cool the beans by shaking in the metal colander until they are hand warm.

Explanations

Pre-heating the pan is to simulate how the pan will be for the second and following batch anyway, in order to achieve consistent results. As always with stovetop roasting, you have to stir continuously throughout the whole heating process to prevent the beans at the bottom to be burned on their underside.

With this coffee, it is challenging to judge the roast process. Regarding colour, much of the silverskin is sticking to the beans, creating the false impression of bright spots and an uneven roast. The best bet is to judge just from the colour of the dark parts of beans. Regarding the cracks, the first and second crack are not sharply separated as there are beans of many different sizes present. Expect cracking to slowly increase (first crack), then become less active but still audible, then to become more active again (second crack).

The 100 g batch size is good for tests, but it is probably possible to also do 150 g or even 250 g in the same way all at once. Roasting time may increase of course.