Level: Alchemist
Read Time: 14 minutes
I wanted to give you a big THANK YOU for all the helpful articles you’ve written. Although Dandelion’s “Making Chocolate” got me into making my own bars, it’s been your articles and blog posts that took an overly simplified process (from the book) and given me the tools to understand how I can change the flavors. Thank you!
I’ve read, re-read, and re-read again your articles and blog posts (Ask the Alchemist 200-205) on roasting. As much as I’d like to have all the proper equipment, there’s only so much I can do on a budget in a small apartment with limited storage space. For the time being I’m trying to oven roast. The engineer in me understands momentum, ramp (slope), and I’ve been trying to apply your Drum Roasting profiles to at least have similar concepts with Oven Roasting ... trying to go into the various phases appropriately.
In your drum roasting profile, the temperature of the bean drops after a few minutes of drying. The ramp changes considerably at this point — the beans were ramping fast but after this temperature drop, the ramp tapers off and this gives the beans a good place to go into development phase at a ramp you can control. In your example graphic, it looks like the beans gain ~100°F/min in the first minute, the beans drop in temp, and then raise ~11°F/min for the remaining 9-10 minutes.
Question 1: Does the ramp rate in your profiles change because you’re adjusting the heat source? Or is the ramp rate changing as a function of having the moisture removed?
With my oven roasting I’ve attempted to create a profile (I understand all the caveats that I’m letting heat out every time I open the oven, that the beans will be hotter on the ends, etc). I’m using a deep Pyrex, pre-heated, and giving the beans a quick stir before measuring with an IR thermometer). I’ve roasted a few times now and measuring every minute I’m not seeing the drop in bean temperature in the first few minutes after they first hit ~165°F. Without the drop, I’m ramping too hard and I can start to smell a little bit of a burning flavor come out. It also means the drying stage finishes much sooner than I’d like (~6 min). To avoid over-ramp and bean defect I’ve been turning the oven temp down a bit to lower my ramp rate and accomplish drying in a few extra minutes (yes yes, drying time isn’t so important).
As you’ve pointed out, all of this matters because it impacts the momentum going into development phase. I assume I want the initial ramping in the first few minutes to be quite large to get the moisture out of the beans, however, without the drop in temperature, I have to “unramp” much more heavily to not burn the beans, and this unramping then causes the development stage to be a bit too slow.
So my real 2 questions:
Question 2: Is there something about oven roasting that causes it to not drop in temperature? Is it just a flawed measurement setup due to the nature of using oven + Pyrex?
Question 3: Is the ramp rate for removing moisture in the first few minutes important? Can I slow that down a bit so I don’t have to un-ramp as much going into development phase?
Alright, there is a ton to unpack and answer here. First and foremost, I have to own up to an error I previously posted and state I have no clue what I was thinking. There is no ~100 F/min ramp at the beginning of the roast. In an attempt to make something more clear I made it less clear. Sorry about that. What I’m talking about is this graph.
The intent (and no, I don’t know why I thought this would make it more clear) was to show that the beans start at room temperature but you don’t see that when you load them into a hot roaster. This is what you actually see.
For the first couple minutes there, while you see the temperature dropping, is basically garbage data or what is often called an artifact. What you are seeing at the beginning when the temperature is over 300 F is the temperature of the roaster and consequently the thermocouple probe. When you dump in the room temperature beans, they start to cool the temperature probe but that doesn’t happen instantly. It takes a couple minutes so what you are really seeing is just the cooling probe tip. At 2-3 minutes the probe stops cooling because during that time the beans have been heating up and you finally start getting an accurate measurement of the beans. That turn around point as it is called is a standard artifact of any drum roaster you pre-heat and drop cold beans into.
That basically answers the second question. Yes, there is something in oven roasting that doesn’t cause a temperature drop, namely that you are never measuring the oven temperature so you never see the artifact and more to the point, there is no real temperature drop at all in roasting. My graph was wrong.
So with that out of the way, let’s get down to the other questions and issues.
Question 1: Does the ramp rate in your profiles change because you’re adjusting the heat source? Or is the ramp rate changing as a function of having the moisture removed?
This is oddly a complicated question and it has everything to do with heat transfer, gradients and yes, moisture loss. In my drum roaster, when I have everything set up just perfect (perfect being the propane flow and thus heat input) I can do an entire roast and never turn down the fuel/heat. The reason for this is that as the roaster and the beans get hotter, you also tend to loose more heat to the environment and this is because of a basic thermodynamic law - the larger the difference in temperatures (called a gradient) the faster heat/energy moves. If that doesn’t make immediate sense, just think of driving up a hill. The steeper the hill, the quicker you slow down. You also have moisture loss that takes both a lot of energy out of the system so it naturally slows the roast. But there is nuance here. If you don’t have your heat input balanced and are putting too much energy into the roast, your roast can instead of slowing down, speed up as the mass of the beans drop again due to that moisture loss. So it is a little of this and a little of that. One part is pre-heating the roaster so it has enough energy to give to the beans but not so much to make it run away. That is balanced by the heat loss which is further complicated by how cold the area is you are roasting. I certainly have to turn my fuel down in summer compared to winter because the gradient just isn’t as steep. In short it is a multi-variable system that doesn’t allow one either or answer.
You may get some of the issues with oven roasting but unfortunately you are a really missing a big key.
It is that you basically can’t apply drum roaster profiles to oven roasting. You can try all you like but at the end of the day they don’t behave the same. You are caught between the proverbial rock and a hard place. In order to get a drum roasting ramp rate you have to have the oven really hot….but a really hot oven will scorch the surface of the bean because they are laying on a tray and not moving…..and there is no way around this. Turn it down enough to not scorch and you won’t get the ramp you need and there is just no Goldilocks zone to get both. This is basically the reason the pretty common “long and low” roasting method is used so often with people roasting in ovens (convection or not). Since you absolutely don’t want scorched beans, the only direction to take for consistency is to keep the temperature low enough that it doesn’t happen. Sadly, it is fraught with issues, one of the biggest is that it is nearly impossible to share those times and temperatures with anyone and have them be transferable. The other is it isn’t scalable. One pound will require different settings than 2 lb let along 5 or 10 or 15 lb. But I’ve gone over that before.
On a minor note, you mention understanding that there is heat loss every time you open the oven. While technically correct, the density of air and consequently the amount of actual energy it contains is virtually negligible compared to the mass of the oven and beans. As long as you are not keeping the oven open minutes at a time or stirring every couple minutes, you can not worry about this.
Question 3: Is the ramp rate for removing moisture in the first few minutes important? Can I slow that down a bit so I don’t have to un-ramp as much going into development phase?
First off, due to my error in that graph, this really isn’t the same question any more as there is no ‘un-ramp’ now. But to still answer it, again, the answer is a little complicated as it is both yes and no. Is it important for removing moisture? No, not really. How long it takes you to drive off the moisture taken in isolation will not affect your flavor profile assuming it isn’t so fast as to scorch the beans. I’ve done roasts with 8 minute drying phases (the time for the beans to get to 212 f) and 28 minute drying phases and as long as the rest of the profile is the same (that is absolutely critical!!), the beans will taste the same. That said, it IS important because it directly relates to how fast you are coming into your development phase. If your goal is to have a development phase at 8 F/min then a drying phase at 10 F/min works nicely as it is pretty easy to slow it down that little bit (see above). But if you are coming in at either 5 F/min or heavens forbid, 20 F/min you are going to have one hell of a time hitting your goal of 8 F/min, so in that sense it is quite important. In short, your drying phase should be either the same or just a little faster (because it will probably naturally slow down) than your desired development phase for the sake of being able to hit it and consistency.
All that said, I’m going to reiterate that pretty much all of my information about profile roasting is applicable to drum roasting and except for understanding the underlying thermodynamic principles, isn’t going to be useful to you for oven roasting. I wish it weren’t so, but even Scotty couldn’t break the laws of physics (mostly). Yes it IS possible to make good chocolate from roasting in an oven but it will take you a LOT of personal trial and error and aside from those thermodynamic principles, which can help, I’m not going to be able to help you much and you will find you can’t do much to pull out and develop specific flavors like you can predictably do in a drum roaster.