Aillio Bullet R1 Roaster: Testing Airflow with Rwanda Kageyo

Details on the Airflow pattern in the Bullet Roaster and how to improve your roasts using airflow.

While this blog post is about the Bullet’s airflow, I don’t touch on the practical applications too much. This exercise is more about looking at two extreme examples of airflow settings on the Bullet to see what effect they have on the roast progression, and ultimately, flavor.

Airflow is used to pull chaff from the roasting chamber, and in conjunction with factors such as batch size, heat, and drum speed, it can change the roast dynamics from steady progression to stalling out altogether. I plan on diving deeper into the intersection of these variables in future roast profile blog posts, how to vary the airflow setting to control roast, extend or shorten a phase of the roast, moderate the heat, and also how air flow and drum speed settings relate to each other. But this article is just about a basic understanding of airflow in the Bullet roaster. 

I used Rwanda Kivu Kageyo Station for this roast assessment, a coffee that has proven to be versatile in terms of cup flavors presented at different roast levels. You can pull out plenty of chocolate roast tone with darker roasting, as well as round off the acidic edge, and so I figured that any significant changes in roast development would be more easily tasted in the cup than a more neutral coffee (plus, I just happen to really like this coffee and so it doesn’t hurt to have a couple lbs. on hand!). Read more about the coffee on Sweet Marias or Coffee Shrub.

A view at the rear of the Bullet with the chaff collector removed

One of the most compelling features of the Aillio Bullet R1 is no doubt the amount of control you have over roast dynamics. Central to this control is the induction heat source. Heat is transferred to the steel drum via electromagnetic field, allowing you to make changes in thermodynamics almost instantaneously.

But airflow also plays an integral role in shaping the roast profile, and as with many of our other manual roasters, is effective at doing a lot more than just pulling chaff and smoke from the drum. For this blog, I take a look at the Bullet’s airflow circuit and run a couple of test roasts to see what affect sweeping changes in airflow has on roasting, if any.

In order to better understand how the Bullet’s airflow circuit works, I think it helps to start at the rear of the machine where both the air intake and exhaust points are located. Ambient air enters the roaster through the inlet behind the drum pulley and continues its path across the drum (and coffee, of course), up and through the exhaust pipe along the top of the roaster, and finally exits through the exhaust point in back. Airflow velocity is regulated by the “squirrel cage” fan you see in the photo above, which is what you’re controlling.

Airflow path on the Bullet R1 – in the back, across the drum, and out the top

I ran a pretty simple test of two 440 grams roasts (just shy of 1 LB), one with airflow at 1 and the other with airflow at 9. Power/heat settings were set to 9 all the way up until the bean temp reached 320F at which point I dropped the power to P5 in order to keep from a violent 1st crack (1C). I kept drum speed on the slow side since I was only roasting at half capacity.

I’m going to out myself as a fairly new Bullet user, and had I reviewed the instructions a little more thoroughly, I would’ve realized the airflow settings go to 12! (heat only goes to 9) Nevertheless, this difference of 1 to 9 was enough to see a visible difference in progression on the roast chart and impart discernible differences in taste.

Roast #1 of Rwanda Cyato with airflow setting at 1

After the drying phase (where the temperature bottoms out at about 1:00) the rate of rise, or “ROR”, holds strong at roughly 30F (ROR is the rate at which the roast progresses over the course of a minute). But then after the yellowing point it drops all the way down to 20F around 6:00 and continues to decline until the end of the roast, bottoming out at 10F.

You can see that 1st Crack (“FC”) occurred at  7:14/343F, and I pulled the roast at 9:02/377F. I believe the steep climb of 28F in 1:48 to be a false reading due to the original Bullet sensor that is still installed in my machine. That steep increase in heat is something I see in all my roasts after hitting 1st C and if it were true, the roasts would be much darker than what I wind up with. Other than that, this is a fairly normal roast trajectory, at least given these heat, drum speed and airflow parameters.

On my second pass, the overall times aren’t that much longer. There’s about a 30 second lag on all roast stages – yellowing, first crack, and pull. I also pulled my roast at a slightly lower temperature (-6F). I made the decision to pull batch #2 after the same amount of time post-1st C had passed, rather than wait for it to reach the same bean temp of 377F.

No surprise, the 2nd batch had a much lower weight loss than the 1st. If you follow the ROR line on the graph, the 2nd roast never got above 22F, an overall slower progression than roast #1. The coffee was visibly lighter coming out of the roaster, and the weight loss came in at 11.5% compared to the almost 13% loss of roast #1.

Round 2 on the Bullet with airflow setting at 9
Comparitive chart of the 2 coffee roast curves
A crude overlay of Roast 2 (the dark green data) over Roast 1.

We can see just by looking at the graphs that the introduction and increase of outside ambient air into the roast chamber cools down the roast. At least with these particular parameters. Knowing this, it stands to reason that a sharp increase in airflow will slow down a roast – the higher the airflow, the greater the affect on cooling the drum temperature.

It would also stand to reason that there are thresholds here that will be dependent on heat, and to some degree, drum speed setting. For example, if I were roasting the same batch size at a lower power setting, I would probably introduce less airflow, or perhaps wait until after 1st C was underway to introduce the same airflow in order to keep from extending the roast too long, or worse, stalling. Another way to slow down the roast on a small batch size like this would be to increase drum speed in order to suspend the beans in the airflow path longer which also has a cooling affect.

The cupping results are where the roast differences really showed. The high airflow coffee was a much brighter cup, the first difference I noted when cupping these two side by side. Sometimes longer roasting can sort of bake that vibrance right out. But roast time really wasn’t that much longer, and the 6F difference in temp along with a slower overall development really came through as a brighter cup of coffee.

I found both coffees to have fairly developed sweetness for City roast level, though roast #1 with low airflow tasted a bit more developed. It had acidity too – kind of a lemon water flavor and vibrance – but not quite the level of the higher airflow roast. The second roast was just generally a more complex coffee, and for my personal taste, more in line with what I look for in Rwandan coffees (black tea, spiced, bright, sweet, etc).

You can see from the thickness of the crust that the coffee on the right is more developed – Roast #2 at left / Roast #1 at right

Practically speaking, I am most likely to ramp up airflow before the beginnings of first cracks to keep the roast from taking off. I’m pretty conservative in the adjustments I make when roasting and only make minimal shifts until the tail end of the roast. That’s when I’ll increase the amount of air pulling through the drum and dial back my heat settings (around 20F before 1st C), the degree to which is determined by how dark a roast I’m shooting for.

While I preferred the roast with airflow at 9, the takeaway isn’t that roasting coffee with high airflow is somehow better. In this particular test with the Bullet the final cup was brighter and more complex than the coffee with lower airflow. But more than anything else, the difference in cup flavors and roast development are an illustration of how airflow can be used to affect roast dynamics. When it comes to roasting coffee there is no “one size fits all” approach and different coffees will show “better” or suit individual taste when roasted different ways.


Other Articles of Interest:

Using Sight to Determine Degree of Roast

Need a Visual Guide to Determine Coffee Roast Color?

Coffee Science – Green Coffee Science and Cup Quality