Drum Roasters vs Fluid Bed

Despite my best efforts all were failures and I never achieved the professional roast quality I was interested in.

In hindsight, I’d estimate that exactly half of my failure was directly attributable to my ignorance of coffee. I’d wrongly assumed, exactly the same as every other engineer and shade tree mechanic that has tried their hand at coffee, that roasting would be easy to figure out.

TLDR: it’s not simple.

Over the last 10 years, I’ve focused exclusively on drum roasting and I’ve learned an enormous amount about how coffee roasts and the pros and cons of drum vs. fluid bed coffee roasters.

The guy that popularized fluid bed roasting, Michael Sievetz was bright, but he did not invent fluid bed technology. Fluidized bed technology was used for eons in sugar plants and grain dryers and even foundries to create lost wax investment casting molds. He merely applied it to coffee and got the USPTO to cough up a patent.

Michael seized on fluid bed roasting for two reasons. First, fluidized bed convection heating is a very efficient way to heat coffee seeds. Highly convective roasts tend to progress faster allowing more throughput (aka greater production) per hour.

I’m convinced that actual roast quality is more of an afterthought, and I think the original patent’s reference to continuous process roasting was/is a code word for industrial coffee.

Second, fluid bed roasters are simple and cheap to build. Literally exactly the same methodology as popcorn poppers. The fan is typically the single moving part.

If your goal is the simplest and least expensive production of small batch fresh roasted coffee, fluid bed tech may be your jam.

Specialty coffee roast quality is typically a more ambitious proposition and drum roasters remain more popular with professional roasteries for several reasons.

Drum roasters mechanically fluidize the seed mass with the rotation of the drum and provide much more precise temperature control as the seed physically changes during the roast.

The fluid bed system lofts the seed mass with heated air. The loft required to evenly roast coffee is dependent on the geometry of the seed and batch size. The machine must then precisely heat the volume of air required to evenly agitate the seed mass. One big problem is that coffee expands as it roasts and airflow induced loft tends to be very different from raw seeds to roasted seeds. This is one reason data logged thermocouple readings tend to be less reliable and show much more noise relative to drum roasters. Poorly size sorted coffee and blends with differing seed sizes often roast unevenly or actually char on aggressive roast profiles.

Air loft machines also require such a great volume of heated air that they tend to desicate the interior of the seed early in the roast cycle. This condition greatly affects inner seed development. The inner portion of the seed is depleted of sweetness and acidity and the coffee frequently presents as thinner with less complexity and less aftertaste.

There’s just a little less coffee flavor in the coffee.

This doesn’t matter when expediency takes priority over quality. It may especially matter less when you haven’t yet gained the experience to develop the professional skills to understand the difference in cup quality across a range of brewing methods.

I am not judging. Coffee is a huge industry. There is room for everyone to participate as a roaster no matter what you choose to roast on. There are no wrong answers, just the answer that is most right for you.

Drum roasters remain popular because drum roasters do have some real advantages. Roast control is a function of stability and drum roasters are inherently more stable. Drum roasters can be equipped with much more sophisticated and effective sensors. Drum roasters scale enormously better than fluid bed roasters. As a matter of fact, where coffee quality is important, no fluid bed roaster of greater than about 6kg per batch can achieve the roast quality of even the humblest of competently operated drum roasters.

Lastly, drum roasters can handle a range of batch sizes. My roasters are specifically designed to provide optimal results from 20% to 120% capacity. Fluid bed roasters are typically configured for one batch size only.

I suspect that fluid bed devotees claiming “the best coffee I’ve ever had” are unknowingly comparing the fluid bed roaster to coffee birthed out of a poorly designed drum roaster run by an amateur operator.

It has been my experience that a better operator with a better green coffee inside a better roaster will seriously raise the “best coffee I’ve ever had” bar.

Steve