MORE THAN JUST HOT AIR

Make The Most Of Fluid-Bed Roasting

IN THE WORLD of specialty coffee, we pride ourselves on advancing our knowledge and techniques through innovation and experimentation, from soil amendments and processing techniques at origin, to computerized roast profiling systems, all the way to pulling the “God Shot” of espresso. None of this would be possible without the sharing of the knowledge that we each have gained from our successes and our failures through conversation.

     However, the topic of fluid-bed technology seems to be one conversation that is seldom brought up, and it is sometimes treated more like a curiosity than anything else. Its place in specialty coffee can be debated, just as any other technology or technique can, but no real conversation is being had about how fluid-bed roasters work and how to work with them to produce a consistently good cup of coffee.
     It’s past time: let’s start talking.
     Many companies produce fluid-bed roasters with a wide range of applications, from Neuhaus Neotec’s continuous commercial roasters all the way to the Fresh Roast Plus home roaster. What differentiates a fluid-bed roaster from a drum roaster is that all fluid-bed roasters use the same basic principles to roast coffee: convective heat transfer to roast and fluidization of the coffee while roasting in order to keep the beans in motion. All are based on Michael Sivetz’s original design, and his fluid-bed roasters are found in specialty coffee more than any other.
     The term fluidization refers to the effect that the forced column of heated air has on the charge of coffee within the roast chamber. The column takes individual coffee beans and turns them into a fluidized mass, enabling them to be evenly roasted by convective heat transfer. The transfer of heat in this manner is extremely efficient, not only in terms of fuel consumed but also in the time that it takes to bring the green coffee to the desired degree of roast. Roast levels of 65 Agtron can be achieved in as few as eight to 11 minutes depending on bean density, moisture content and gas pressure.
     At first glance, the most commonly used fluid-bed roaster, Sivetz’s quarter-bag model, looks like two boxes that are connected by a single piece of stovepipe. Aesthetics were of little concern in the design; instead, function was the ultimate goal.
     In this roaster, the smaller of the two boxes houses the burner and burner assembly. A three-horsepower motor draws ambient air over the burner and through the connecting stovepipe. An industrial pressure blower located under the second box forces the now-heated air through a perforated, v-shaped piece of sheet metal that makes up the bottom of the upside-down teardrop-shaped roast chamber. The heated air is then turned into a column that fluidizes and roasts a 37-pound bed of green coffee beans by means of convection. Smoke and chaff exit the roast chamber through ducting that leads to a cyclone, where the chaff is separated from the smoke and collected in a five-gallon pail. The smoke is then burned off by an afterburner in most configurations.
     The controls of this machine, as with most air roasters, are simple to use and easy to understand. Two digital controls are located at the top of the face of the control panel. Each control has two digital displays on it: the upper display reads current temperatures, and the one located directly below it is set by the operator. The control on the left reads bean temperature on the top display, and the bottom display is set to cut off at the desired high-temperature setting. The control on the right reads inlet air temperature at the top; the bottom is set to a cut-off temperature for safety, usually 625° F. These controls couldn’t be easier to read or to set. They typify Sivetz’s function-over-form approach to engineering and do no more, and no less, than they were intended to do.

Benefits and Drawbacks

That, in a nutshell, is the setup of the most commonly operated fluid-bed roaster. But does a fluid bed produce a great cup of coffee with the repeatability that the specialty coffee community demands of its equipment? Norm Killmon, roast master at The Roasterie in Kansas City, Mo., says, “We cup every roast, and the consistency is amazing.” If the proper roasting protocol is set up and followed, I would have to agree that consistency and repeatability are two benefits of fluid-bed roasters.
     But, as Mathew P. Hill, assistant green coffee buyer and roasting conversation specialist at Swiss Water Decaffeinated Coffee Co., says, “There is some give and take with fluid-bed roasters. They produce amazing acidity and sweetness that drum roasters cannot and lose the body that drums are able to develop.” That loss of body is a sacrifice that affects the cup and is seen as an acceptable trade-off for the sweetness and acidity that fluid-bed roasters can bring to the cup.
     One of the main causes of inconsistency, whether roasting in a drum or a fluid bed, is the time that it takes from dropping your charge into the roast chamber to the end of the roast. Most roasters I have spoken with who are using drums are trying for plus or minus 30 seconds of their desired time to temperature. That might be just fine for drum roasters and their use of mainly conductive heat, but in a fluid-bed roaster, 30 seconds one way or the other is a lifetime and may produce a totally different flavor profile. In fact, my cupping results show slight differences even in 20-second variations and are as much as I really find acceptable. The reason for the dramatic difference in such little time is the effectiveness of convective heat transfer.

Best Protocols

So, what kind of roasting protocol can be set up to defeat the ravages of such small increments of time?
     The first protocol to set is when you are starting your roast. Not having a starting point that is set and adhered to creates a variable that should not exist. The lack of a set start procedure will lengthen or shorten your roasts, and this problem should be eliminated right away. Using the inlet air temperature reading is the best method for determining the start of your roast. Starting the roast when the display is reading temperatures of 185° F or higher will most likely end up tripping the safety setting and result in the roaster shutting itself down, possibly losing an entire roast; thus it should be avoided. Roasters in hotter climates should watch for this because of higher ambient air temperatures and lower humidity. In my opinion, the best place to start is 175° F. That will prevent the inlet air temperature from going above 580° F, well within the high limit, even when roasting your darkest roasts (Agtron 30 or below). It will also give you a great reference point at the cupping table and allow you to find out where a particular coffee develops its best flavor and aroma profile according to the high bean temperature setting and time that it took to reach that temperature.
     The second protocol involves gas pressure. The time that it takes to reach a set temperature is also directly affected by how high the gas pressure is set. Gas pressure is read in pounds per square inch of gas (PSIG) by a Dwyer Magnahelix gauge, which is mounted to the bottom of the control panel. With a gas setting of four inches of water column, you will achieve a roast of around Agtron 65 in about eight minutes and 30 seconds (under optimum roasting conditions). Adjusting your gas pressure will allow you to lengthen or shorten your roast time according to your personal interpretation of desirable flavors and aromas. The pressure may also need to be adjusted seasonally due to changes in formulation made by your natural gas company.
     Now, the third and most important part of eliminating the time variable is warming the roaster. With a drum roaster, it’s pretty straightforward: light it and wait for the drum and environmental air to reach a workable temperature. Unlike with a drum roaster, lighting a fluid bed and waiting for it to warm doesn’t work because there is no direct source of heat around the chamber itself. All you are lighting is a pilot light, not the actual burner, which doesn’t light until the roast cycle starts. So how do you get the metal that makes up the roast chamber to a workable temperature?
     One way to do this is to run the roaster without beans in it. However, it takes a long time to achieve a workable temperature due to the fact that there is nothing keeping the heated air within the chamber, and this method burns a serious amount of fuel. All roasters know that time and fuel are money, and no one wants to waste either.
     How can you circumvent the roast chamber’s need to preheat to achieve your desired time to temperature without re-engineering the entire roaster? You have to increase the inlet air temperature in order to compensate for the absorption of heated air by the cold metal of the roasting chamber and the unheated coffee. Turning up the gas pressure will do the trick, and the operations manual will show you just how to do it.
     The proper amount of gas depends upon many variables, such as the formulation of the natural gas and the temperature of the roastery, so you’ll want to experiment with this technique using less expensive coffees and increase the gas pressure in a column increment of one-quarter inch of water. Increasing gas pressure too much will result in exceeding the inlet air temperature safety setting, possibly shutting down the entire roaster. Finding the proper starting pressure will take a few tries and, since it’s only applicable to the first roast of the day, a few tries will take a few days. This is most easily accomplished after the roaster’s pilot light has been lit long enough for the inlet air temperature reading to be 150–160° F. Due to the placement of the thermocouple by the manufacturer, this is not the reading of the environmental air temperature, just the temperature of air directly below the perforated V at the bottom of the roast chamber.
     Be patient, keep notes, and once you find the right pressure, mark the metal surrounding the gas pressure adjustment screw for reference and repeatability. Once you find this sweet spot, roast two batches of the same coffee consecutively, the first with the increased gas pressure setting, the second reduced back to your normal gas pressure. If the difference in the cup is out of acceptable range, try reducing the set temperature of the first roast—the one with the gas increase—by one or two degrees. That reduction will offset the higher inlet air temperature caused by the increased pressure. Cup those results the following day, and you should find variations that are well within an acceptable range. I have found the best results with pre-blended Agtron 65 roasts containing three or more types of beans and dark-roasted (Agtron 35) origin coffees.
     After breaks from roasting, other gas pressure increases need to be made in accordance with the amount of heat lost in the roast chamber. Since the thermocouple located within the roast chamber is meant to read bean temperature only, it is useless in determining what pressure you should start with after the individual breaks. Instead, observing the inlet air temperatures reached during the start of the day should be the best guide. That temperature should never get above 610° F and will drop dramatically—around 10° or 20° F—after the bean temperature reading is between 380° and 390° F, or first crack. Again, keep notes on the gas settings of the failures, mark the successes the same way you did before, and cup all the results that land within the plus- or minus-20-seconds range.
     Those of you out there roasting on drums are probably shaking your heads right about now, but as Colin O’Callaghan, president and founder of Heis Roasters, a new generation air roaster manufacturer, says, “The simplicity of the original design made [fluid-bed] roasters very accessible, but the basic design hasn’t changed in 30 years, and the lack of control of the roast has kept them from playing a major role in the roasting community.”
     On the other hand, drum manufacturers have been innovating since the design was first manufactured. The protocols that I have laid out here are merely the first step in starting the conversation on how to hone your skills as a fluid-bed roaster and bring forth ideas for improving an existing design by adding more control. If implemented, these steps can save a roaster using a fluid bed about 20 minutes a day in labor and fuel costs. If you do the math and extrapolate that cost over a year, you’ll come up with a grand total of almost 11 eight-hour days in savings.
     Fluid-bed roasting may not be the most popular girl at the party, but she’s not just a lot of hot air either. You just may discover there’s more to her than meets the eye.

JUSTIN JOHNSON is the roast master and green coffee buyer at World Cup Roasters in Portland, Ore.