hankua wrote:Here's a graph illustrating Steve's point, between 7-8 min. there's a ROR spike.

I have experienced the same ROR spike on my electric quest with an increase in airflow. I believe the ROR spike you're seeing here may not be due to more efficient combustion. Rather, I believe it's due to more efficient airflow - in other words, the increase in airflow (and change in the complex way the air migrates through the roaster) means that hotter air is now hitting your probes.

I'm not sure about all of this, but my question about the "more efficient combustion theory" this this: At low airflow settings, isn't there still enough air available to the burner for just about max efficiency? My diedrich does 290 CFM out the exhaust. At a low damper, that's probably about 50 CFM through the roaster. Does a roaster doing 48,000 BTU/HR need more than 50 CFM of air for efficient combustion?

Ok, rather than ask the above questions, I looked them up (to the best of my abilities). Efficient combustion of propane requires 5 O2 molecules for every propane molecule. Accounting for air vs O2 and we're at roughly 1:25 for propane to air. I don't want to convert molecules to volumes, but they should be pretty close - so does your roaster get at least 25x as much air as propane?

Now i've gotten to the bottom of this post and i'm not sure.

I think that it is clear that changes in airflow can affect ROR, and that users might choose to vary airflow to achieve a particular profile. It is pretty easy to learn how changes in airflow influence the overall profile by trial and error on one's own roaster. A harder, and more interesting, question is what would happen if we execute two nearly identical profiles, but with higher airflow in one and lower airflow in the other. (Obviously heat input would need to be adjusted to keep the profile curves similar). Likewise, with respect to any given profile we could attempt to locate the airflow "sweet spot" on a particular roaster, which would be the setting that maximizes taste for that profile. I'm not sure this would be the same setting for all coffees or all profile curves.

The space of possible profiles and airflow settings is very large, which makes systematic investigations of these questions quite hard. I doubt we will do more than scratch the surface in Nov/Dec, but it will be interesting to spend some time trying to learn more about this.

The space of possible profiles and airflow settings is very large....

Add variable drum speed and we have 3 variables at play. Holy smokes, Batman, the possibilities are virtually endless!

popeye wrote:I have experienced the same ROR spike on my electric quest with an increase in airflow. I believe the ROR spike you're seeing here may not be due to more efficient combustion. Rather, I believe it's due to more efficient airflow - in other words, the increase in airflow (and change in the complex way the air migrates through the roaster) means that hotter air is now hitting your probes.

I'm not sure about all of this, but my question about the "more efficient combustion theory" this this: At low airflow settings, isn't there still enough air available to the burner for just about max efficiency? My diedrich does 290 CFM out the exhaust. At a low damper, that's probably about 50 CFM through the roaster. Does a roaster doing 48,000 BTU/HR need more than 50 CFM of air for efficient combustion?

Ok, rather than ask the above questions, I looked them up (to the best of my abilities). Efficient combustion of propane requires 5 O2 molecules for every propane molecule. Accounting for air vs O2 and we're at roughly 1:25 for propane to air. I don't want to convert molecules to volumes, but they should be pretty close - so does your roaster get at least 25x as much air as propane?

Now i've gotten to the bottom of this post and i'm not sure.

It's not just volume of air, but also the fuel/air mixture. I'm only guessing that to a certain point and through a variety of mechanisms, greater air flow probably allows for a more homogenous molecular 5:1 O2 to fuel ratio. I worked on combustion simulations years ago (when dinosaurs still roamed the earth) and, although it was pretty neat, just thinking about it still makes my head hurt. This is a crazy stupid complicated thing to model which, when we were doing it anyway, too often yielded different theoretical results from empirical observation.

I'm certain advances in software and computing power provide better answers now, but all I'm trying to do is roast some coffee.

I think I answered this the same last time this question was asked, but airflow affects the coffee exactly how you think it would. It increases the heat penetration to the inside of the seed. Just like EVERYTHING in else, its going to depend on the coffee if that's beneficial or not. Ive experimented with repeating exact profiles with only changing the airflow. 3 different roasts: High air flow, mid airflow, and low airflow. Obviously things get a bit tricky when you have very low airflow to control a roast. So I made sure to do the low airflow roast first, and then copied that profile with the mid and high airflow roasts.

On the particular coffee I used(a Kenya) the highest airflow roast was the best. Which makes sense. Kenyan coffees are generally big seeds and they also take heat really well. The high airflow roast was sweeter and more intense in flavor.

On coffees that are less dense (like a natural Ethiopian or Brazil), I almost always have a lower airflow setting. With a higher airflow, these coffee will pick up roast notes faster (because of the deeper heat penetration).

+1
I'd like to add a comment/question: In addition to choosong certain coffees as best candidates for higher heat penetration facilitated via airflow, there seems to also be a place in time within any given roast that higher airflow/ heat penetration is most advantageous, as well as potentially damaging to flavor. In my experience, highish airflow early on when the beans are still green seems to produce a harsh, hollow cup, and to little air later on after the coffee is browning results in a lingering acrid note. What's your experience?

My standard approach these days is similar to what Dustin describes, and for similar reasons.

dustin360 wrote:... Ive experimented with repeating exact profiles with only changing the airflow. 3 different roasts: High air flow, mid airflow, and low airflow.

Those are all "set it and forget it" approaches. In the video, Marocco discusses reasons he favors adjusting according to time and stages of roasting. I have found his approach produces better results than choosing *one* airflow setting and not changing it.

On coffees that are less dense (like a natural Ethiopian or Brazil)...

While natural Ethiopians like less heat than a WP from Ethiopia, I hesitate to call them "less dense." I have begun to measure volume to corroborate elevation as an indicator of density. Take a lined, graduated beaker. $10 on Amazon. Measure a half kilo (or any amount), place into the measured volume container and record. For me, a natural Ethiopian would not fall into the "less dense" camp with a Brazil.

Replace "less dense" with "likes less heat"then. The point is they more fragile. I also started placing a set weight in a beaker and messurejnf volume. Then abandoned it as I didn't find it useful.

And obviously if you are trying to figure out what are flow does "setting it and forgetting it" is much better than changing it all over the place.

Just like EVERYTHING in coffee, sometime varrying airflow is better and sometimes it's not. I'll do either, depending on the coffee.

One other point about airflow, is I don't buy the people who say keep it high to remove smoke so you don't end up with smoky flavor a in the coffee. I feel like that would only be true if you had the airflow set so low that the air temp INSIDE the drum gets super hot as well as the drum itself. But I suppose if the roaster had weak airflow it could be a concern.

The fan control dials on my roaster allows step less adjustment from 0.0 to 8.5, and the manufacturer suggests a minimal of 3.0 for efficient ventilation. I did the lighter flame on the bean tryer test. At 0.0, the cyclone generated sufficient suction to bend the flame. At 5.7, the air flow was strong enough to blow off the flame from the lighter. Therefore, my useful fan range is from 30 to 55 if readers here try to correlate my chart to your own roaster.

While I was testing the low fan speed on my roaster, the flame of my roaster was blue (not orange), indicating sufficient propane-air mixing. However, somehow I could smell the "gas" (actually it's ethanethiol) at the 0.0 setting. I would put safety first, and stick to my manufacturer's recommended minimal fan speed (3.0 in my case) in the future. Although the airflow/fan design varies in each roaster, I think a properly vented burner at working condition should always have virtually 100% gas burnt. Airflow may cause faster heating (increased convection) or drag the roast (more cold air, removal of hot air), but I would not think it's due to less efficient burning.

Anyhow, my $0.02.

I played with the air flow last weekend. The Ethiopia (for drip) was my typical approach, with more aggressive air applied in the development phase.



I then did two batches of Tanzania (for SO espresso), one with slower overall airflow than the other. I tried to break down the power control into several smaller steps and apply these steps at different time points to fine tune the curve. I also tried to keep the roast time around 11 mins (4-4-3).





In the Dry phase, the effect was not significant, and I could still combine various charge temperatures and bean masses for my desire Dry phase duration (4:00 to 4:30). The effect of airflow was more pronounced at the development phase, as my high airflow batch reached 400F, while the low airflow batch got 409F. Higher airflow removes heat, and drags the roast in my settings.

As can be seen, the above two profiles are still not quite the same. I'll need to compare the [high fan/high heat] vs. [low fan/low heat] which produces identical time segment durations and drop temperatures to tell the effect of airflow.