Roasting Science

0 of 58 lessons complete (0%)

Roaster Types

RS 1.09 Pros and Cons of Fluid-Bed Roasting

Fluid-bed roasters are efficient at transferring heat to the beans, and they allow for the fastest roasts of any type of roasting machine. The large volumes of hot air lift and mix the beans, making a lot of heat available to them. . Because the beans do not touch any surfaces for long, there is virtually no risk of bean-surface burning, allowing the use of high air temperatures. Furthermore, according to Goldfine (1944), high airflow blows away the layer of air immediately surrounding each bean, which would otherwise insulate the beans and reduce heat transfer.

Fast roasting reduces production time, but it also has distinct effects on the structure of the roasted coffee and the resulting flavour in the cup, which we will discuss fully in Part 2 of this course. While high air-to-bean ratios allow faster roasts, excessive airflow may strip away some of the aroma compounds or dry the beans too quickly, resulting in flat or dull coffee (Schenker 2000). With a fluid-bed design, the airflow is determined by the speed needed for fluidisation, so it’s not possible for a roaster to reduce the airflow to alter the roast profile.

A fluid-bed roaster with a 50 kg-capacity at Zarraffa’s Coffee in Queensland, Australia. Photo by Coffea, published under a Creative Commons licensce

Thanks to the short roasting times and efficient heat transfer, fluid-bed roasters are ideal for industrial-scale roasting. The limited contact between beans and surfaces means the roasting temperature can be rapidly changed, even in large roasters, by simply changing the incoming air temperature. Fluid-bed roasters can roast fragments and broken beans that may get stuck in or fall through small gaps in drum roaster designs (Eggers and Pietsch 2001).

Fluid-bed roasters have fewer moving parts than drum roasters. Without the need to turn the drum or mixing paddles, fluid-bed machines are generally reliable and require less maintenance (Eggers and Pietsch 2001).