Sivetz Hot Air Fluid Bed Coffee Roasting

Comparing Drum Coffee Roasting to Hot Air Fluid Bed Coffee Bean Roasting (A reprint of Sivetz Coffee Company newsletter May, 2002)

In 1974, Michael Sivetz invented and patented an improved method for roasting coffee beans.

Books by Michael Sivetz

Jan 1, 1963 – Coffee Processing Technology Volume One and Two
Jan 1, 1973 – Coffee origin and use
Jan 1, 1979 – Coffee Technology
Jan 1, 1993 – A critique on causes of decline of coffee quality
Jan 1, 1993 – Coffee user’s guide

Timing of a Coffee Roast

A drum roaster has been and is still a poor way to transfer heat to a green coffee bean. For example, a Drum Roaster commercial batch of green beans can take 15, 20 or even more minutes to roast, whereas the same sized batch can be roasted in 8 to 12 minutes in the patented Sivetz roasters. This halving of roast time, alone, not only constitutes doubling productivity, but it produces a brighter, cleaner, improved tasting and more aromatic coffee bean.

Original Sivetz coffee roaster, circa 1978

Cleaner Product & Machinery

The Sivetz process continually passes new hot air through the suspended bean bed. The significance of this procedure is that any and all loose chaff is immediately swept away from the beans into the cyclone collector. The chaff is unsigned. Whereas, in a drum roaster most of the chaff rides on the tumbling beans and burns and smokes. This smoke, fumigates all other beans, giving them a harsh bitey taste. The smoke also deposits on the surface of all other beans, resulting in darker bean surfaces than interiors, and oil surfaces. This carbonization of the drum roasted chaff creates volatile products similar to those found in cigarette smoke and charcoal preparation from wood. These are classified as carcinogenic chemicals. The metal drums build up in oily carbonized lacquers, causing poorer heat transfer rates from the metal drum to the beans (conduction), and the roaster operator needs to increase the temperature of the drum, in order to keep roasting at minimal times. Since most of the heat passes through the steel drum wall, usually the drum deposits need to be removed about once per week, down to the metal surfaces. Such frequent cleaning takes many hours of labor, and is never totally completed. People, who own and operate drum roasters, especially when preparing dark roasted beans, need to realize that they are doing work that is not done on the Sivetz maintenance free fluid bed roaster, with only one moving part. The reasons for this difference, are not only due to higher steel drum temperatures, but also the slower bean and air movement, with metal tipping hot spots on the beans releasing oils, and the smokey chaff tar deposits both on the beans and metal parts. The fluid bed roaster, on the other hand, because of its lesser inlet air temperatures, and vigorous bean movement keeps the roast chamber wall scoured and clean to the metal surfaces. There is no need to spend time and labor to clean the roast chamber weekly. The roasted beans from a Sivetz fluid bed roaster have a clean smooth taste, that does not leave an after taste “coating” in one’s mouth, that requires a glass of water.


The biggest benefit offered by the Sivetz fluid bed roaster, is its ability to sense the digital bean temperature accurately with a thermocouple. Hence with this information, the degree of roast is known at all times, as well as allowing easy settings for different degrees of roast and offering accurate reproducibility of roasts.

This is not easily done on a drum roaster, when intermittent probed bean samples and operator eye judgments are used.

In today’s market, many large (say 4 bag) drum roasters have bean catching troughs where bean temperatures are measured. However, these methods have not been used on the smaller specialty sized drum roasters. Such bean color samplings, if not verified, can often give inaccurate bean temperatures, hence inaccurate degrees of roast.

Many production people in the specialty trade are not even aware of this bean temperature relationship to degree of roast taste, weight loss, and light reflectance. Systems that measure the reflectance of ground coffee, can only tell what has been produced, and is not a way to directly influence the termination of the roast, while roasting.

One must keep in mind that most large roasting firms all over the world have purchased Neuhaus-neotic fluid bed roasters based on the Sivetz principles (previously licensed). Many of these machines are doing 8 minute roasts. So one can appreciate that not only have the “majors” quickly accepted this way to roast, but that in using a more productive Sivetz fluid bed methods, there is a rapidly growing portion of the world’s coffees being so roasted and displacing the older less efficient drum roasters.


There have been a few so-called fluid bed roasters built and operated in the past, namely Lurgi 1 Kg unit, Smitherm continuous pressure roaster and Wolverine hot air down blast design on to a vibrating conveyor. Each had some deficiency which limited their commercial use. The basic principle of a fluid bed roaster or drier is that the granular bed is totally uniform in temperature and other properties regardless of where you stick your probe. This is true of the Sivetz fluid bed roaster. The Wolverine method did not give uniform roasts as evidenced by bean colors alone. So it could be used for ground coffee sales but not bean sales. Also none of the listed prior methods measured bean temperature accurately, if at all. A body of beans needs to be uniform in color or chemistry from bean to bean as well as from the surface to the center of the bean.

As already noted, drum roasters deposit smokey tars on the bean surfaces, especially in darker roasts. Beans roasted to the same weight loss are darker on the surface than within from a drum roaster, whereas, the bean surface is lighter than the interior of the bean in a fluid bed roaster. One reason for this is that the Sivetz fluid bed processed beans have no dark tars on their surfaces.


In so far as I’m concerned, the Sivetz fluid bed system (also used by Neuhaus Neotec), from a technical as well as taste aspect is far superior to any drum roaster operation, just based on resulting tastes, uniformity of roast, ease and accuracy (digital) in operations. And it appears that most major roasting firms have agreed to this, because they are buying fluid bed roasters and that is the product we are seeing more and more.

It needs to be stated however, that large scale commercial roasting and blends and pricing are not the same as specialty coffee operations and products. However, the Sivetz fluid bed roaster offers to both parties what they seek.

The example at the right is from NEUHAUS NEOTEC RFB Roasting Chamber

Sivetz Patented Fluidized Air Bed

The Sivetz fluid bed roasters have only begun to be used in the past 20 years, and so it remains to be seen what the ultimate commercial results will be. In any case we can say two things: 1) There is a taste difference between drum and Sivetz fluid bed roaster products, and 2) Traditional tastes tend to persist in the food industry over time. One can only expect a newer method and product to take some years to replace a traditional one.

It is evident that the roasting process is only one of several influential factors in successful coffee sales. Tradition, marketing, flavoring, advertising, etc. as well as management will influence sales. There are many traditional processing firms that have been switching to more efficient roasting methods, and that alone will not much change their bottom line.

Some of our early users of Sivetz fluid bed roasters, like market-oriented Green Mt, have reverted to drum roasters, mollifying customer complaints about taste change, and have gone on to successfully grow in the specialty coffee market.

The reasons for switching types of roasters are varied and often lack technical logic. The changer wants to show a European machine of traditional drum construction. The changer usually has learned to roast on a Sivetz fluid bed roaster, and now has made enough profit to buy a more costly traditional European drum machine. Comparing costs and technical features have not been the basis of such decisions.

The buyer of the roasting machine will be influenced by many factors, and the ultimate success of his business will only be partly influenced by whether it is a drum or fluid bed roaster.

One overriding factor for the Sivetz fluid bed roaster vs. drum is that it is a clean machine requiring little to no maintenance, delivering a clean-tasting bean, in a controlled manner so that any number of operators can be doing the work, and still, the same degree of roast is obtained from different people at any time of the day, and at the least investment in machinery, maintenance and training.

I can only present the technical facts. It is up to the buyer to decide whether a bright red Italian drum roaster is purchased.