Level: Hell if I know. It is long but I hope approachable.

Read time: 20+ minutes

I know you don’t like chocolate roasted in an oven. There are lots of chocolate out there that have won awards and roast that way. I’m having trouble figuring out how to roast in my oven. The chocolate is only ok or bad. What roasting profile should I use? Can you tell me what temperature I set my oven to and for how long?

It is really interesting to me how the telephone game works. You know about it, right? You start with some story. You whisper it to the person next to you. They in turn tell it to the person next to them and so on until it gets back to you. You then tell the story out loud and compare it to the one that came back to you and everyone is amazed or puzzled how the story went from you skinning your knee when you were 5 and getting soothed by a bowl of ice cream to your mother breaking your legs when you asked for a bowl of ice cream.

I have never once said I don’t like chocolate that was made from beans roasted in an oven. I have said over and over that profile roasting isn’t applicable to oven roasting and that you really don’t roast cocoa by setting a temperature and time like you do for brownies…..but somehow what I have said has changed into the belief I don’t like oven roasted chocolate and that you can roast beans like you roast a chicken.

<minor rant mode> I promise it won’t be too long and there is good stuff coming after it.

What I have said is that I don’t like roasting in an oven and it is fully based on a few reasons.

The biggest reason is that you don’t know what is going on.
If you change how much you are roasting, you can’t use the same settings
There is little to no feedback to know how the roast is progressing.
The settings on my oven most likely are not going to match yours.

You make another comment that you are having trouble figuring out how to roast in your oven. That right there is yet another reason (really just a variation of the last reason) I don’t like oven roasting. It is nearly impossible to teach. It is not set it and forget it.

Finally, I have also said over and over that profile roasting is for drum roasting. The whole point and the backbone of profile roasting is that you can plot the temperature of the beans against time. This is what I mean by not having any feedback. You need a consistent average temperature to plot but because the beans are laying on a tray, the surfaces are all different temperatures. I am not going to get into that again. You can read about it HERE.

Ok, my minor rant is over and really, I have other plans aside from ranting therapy for this article. I have been roasting my own coffee now for over 20 years. When I started way back where, there were a few people (no one professional) that roasted in an oven but they very quickly realized it was a deeply flawed way to roast (see reasons above) and the concept pretty much died the death it deserved. If I achieve nothing else in the chocolate world, my dearest hope is that in 20 years no one will be roasting in an oven because it is generally recognized as the inferior method it is. Before I finish this rant, I am going to go ahead and admit that with very few exceptions, I’ve not been blown away by chocolate made from oven roasted beans. That is NOT the same as saying I don’t like the chocolates….just that in a venn diagram of oven roasted and drum roasted and how those methods really bring out what a bean has to offer, the overlap of success is heavily with drum roasting.

</rant mode>

Seriously, we have too much ranting in the world today and I’ve stopped writing this article twice because I didn’t want it just negative ranting to do things my way. That really came home when I was visiting a chocolate company a couple weeks ago and got chatting with a fledgling chocolate maker there (thanks Cat, this article would not be here if not for you) and she was asking me if I thought she should take a particular class and bemoaning that she was not happy with her chocolate, and doubly so because on the surface, chocolate making is no harder than making brownies. Mix things together, let it refine and bam, you have chocolate. Where in the world was she going wrong? If you have not sussed it out, it was roasting and we quickly zeroed in that she felt like she didn’t know what she was doing, that she was kind of flailing around in the dark trying to roast in her oven….and I agreed, that that is basically what she was doing. She even mentioned she knew I didn’t like oven roasted chocolate (thanks telephone game) and that right there was the birth of this better version of this article. We chatted and I clarified it was the method and not really the result I didn’t like and I ended up giving her some tips that I am now going to give you.

Before that though, and something I didn’t explain to her, I want to take this opportunity to get into the crux of the reason that oven roasting doesn’t work well. To do that I need to introduce some concepts. I think I have come to realize that I think about heat and temperature in a very different way than most people and so to me, it is obvious why oven roasting doesn’t work well and why to many people it isn’t obvious.

The main thing I hope to leave you with is that heat (or energy) and temperature are very different things.

Let me say that again.

Energy and temperature are very different things.

Sure, they are related, but they are not directly related. Once you really understand that, you are then in a position to use the data and tools at your disposal to have a better go at oven roasting, or maybe, just maybe, grasp that it is flawed and that trying to solve a literally unsolvable problem.

We are going to be talking about cute little animals, balloons and how to stack rocks.

Ok. Concept time.

One more time.

Energy and temperature are not the same thing.

The crux of the issue is that to roast well, we have to measure energy but the only thing we can control in an oven is temperature.

Say we have a bunny named Cocoa and we want to lift her to a certain height. To do that we attach a few balloon to Cocoa and she flies into the air some amount.

How high the bunny is lifted is its temperature but how many balloons we use is how much energy is required to lift it. You can see they are related but two different things.

If you instead want to lift a tortoise, you are going to need a lot more balloons to lift it to the same height (temperature)

Say we want to lift the bunny a foot off the ground and it takes 3 balloons to do that, then we can define that and say it takes 3 balloons/foot to lift the bunny. The tortoise on the other hand is heavier so requires 14 balloons to lift it the same distance.

What we are basically talking about is a property of materials called its specific heat and is related to another property everyone is familiar with, the calorie. But what is a calorie in this context? It is basically a balloon of a certain size that does a certain job…like lifting a hedgehog named Calvin (who we’ll call Cal for short).

In our cute little analogy here it is the size of the balloon needed to lift one hedgehog one foot off the ground.

In the real world, a calorie is the amount of energy (Cal the hedgehog’s balloon of course) required to raise the temperature of 1 gram of water 1 degree Celsius. Now that energy can be in the form of electricity, propane or even I guess hedgehog balloons.

The whole point here is that there IS a relationship between height (temperature), energy (balloons) and specific heat (what you are lifting/heating) but the are very much not the same thing.

And this is where we can use what I’ve laid out to talk about why ovens are less than optimal to roast in.

When you roast in the oven, the oven itself must heat up. It is the tortoise and you can control how high the tortoise gets with the thermostat (the tortoisestat?). Let’s get tortoise 35 feet into the air (350 Fahrenheit as it were) . You turn it on and it works by opening up a valve and letting balloons flow in. Since we need it to be 35 feet high and it requires 13 balloons per foot, you need (35 x 13) 455 balloons. When there are 455 balloons and the tortoise is 35 feet high, the valve closes and no more balloons are let in.

Sounds good. So ask yourself what do you do if you want to lift up Cocoa the bunny to 25 feet? Well, you would set your tortoisestat to 25. But there is a problem here. In roasting it isn’t just about temperature (height). It is about how long it takes too. Through out the years I have learned that if I can get Cocoa Bunny 25 feet high in about 25 minutes then everyone is happy. To do that, the math isn’t that hard. Cocoa requires three hedgehog balloons per foot. That means I need 75 balloons in 25 minutes.

And this is where we run into a major issue. I have no real clue how fast the balloons are being let into the oven when our tortoisestat opens the valve. If the balloons are coming in at 3 balloons a minute, well are good, but alas, most ovens don’t run that fast. They maybe are 1 balloon a minute. So what do you do? Well, you could collect up some balloons by attaching them to the tortoise and then they can share the balloons. It is sort of like some weird Rube Goldberg pulley system with bunnies, tortoises and hedgehogs. You raise your tortoise 35 feet by giving him 455 balloons (oven pre-heat if that isn’t obvious). Then you toss in Cocoa Bunny and Tortoise, being the great guy he is starts to share his balloons. As he does that, since he no longer has as many balloons he starts to sink, and bunny starts to float and because they are such great friends (laws of thermodynamics and all) they will eventually settle at the same height.

In this case, they will be at 28.4 feet high. Why? Remember, there are 455 balloons total and bunny needs 3 per foot and tortoise needs 13 so they have to divide them up. If you felt like it you could do the maths like I did in the side bar to the left. Otherwise trust me you that Bunny and Tortoise are going to settle a little over 28 feet high.

But there are two problems again.

Maths Side bar

13 balloons/foot x 35 feet (for tortoise) = 455 balloons total
**3x + 13 x = 455**
**16x = 455**
**x = 28.4 feet**
3 balloons/foot x 28.4 feet = 85.2 balloons for Bunny
13 balloons/foot x 28.4 feet = 369.2 balloons for Tortoise
and to check our work
85.2 + 369.2 = 454 hedgehog balloons
Or the amount we preheated.

One issue is that Cocoa doesn’t like going 28 feet (280 F). Bunny feels really bad. And sure, we could keep an eye on Cocoa bunny and pull him away from Tortoise when he is only 25 feet high but that still isn’t the real issue. Remember we were trying to compensate for only 1 balloon coming in per minute and we want to get Bunny 25 feet high in 25 minutes. This time it took 30 minutes. Our only real solution is to add more balloons (pre-heat hotter) and hope Tortoise doesn’t share too fast.

But let’s go ahead and say we do this a few more times and finally can get Cocoa bunny to 25 feet in 25 minutes, and we are watching really closely and pull Cocoa away from Tortoise just as Cocoa gets to 25 feet. Great. We have scienced it. We did something, changed something, recorded it, and kept going until we could lift Cocoa 25 feet in 25 minutes.

Now that we have dialed that all in, it is time to get serious. Now because we are in production we need to raise Cocoa and his brother Cacao. We now have twice the rabbit to lift. How high do we raise Tortoise to account for twice the weight? I hope you see it starts to get really complicated. Clearly you aren’t going to lift Tortoise twice as high but how much?

Now I know some of you could do the maths and work it out but the reality is that the maths up there are just examples to prove a point. The situation is frankly way way more complicated and really isn’t readily solvable because of the way Tortoise shares his balloons. It gets even more of a mess because we are not really talking about one bunny or one cocoa bean. We are talking about hundreds or thousands. Remember from Alchemist Ask the 302 I said there were about 100 cocoa beans per 100 grams of beans. That means you have about 1000 Cocoa bunnies per kilogram that need balloons and Tortoise, try as he might just can’t hand them out evenly. Ovens are uneven. The bunnies closer to Tortoise are going to get more than those in the center, and sure, they can pass them along and eventually all the bunnies would have the same number of balloons but that is going to take a lot of time and if you recall, time is what we don’t have. We have to pull out Cocoa when he is 25 feet high…..but what do you do when you have 1000 bunnies and some at the edge are at 26 feet while some in the middle are at 22 feet.

When do you pull them out? It seems basically hopeless doesn’t it?

The thing is, it is even more complicated than I’ve made it out to be. This is the simplified version.

Consider these things.

In reality we don’t turn off the oven. That means that in addition to the 455 balloons you started with, every minute there are another 1-6 balloons being added depending on your oven..
Some of them are popping (heat loss).
The more balloons you have, the more that escape (the hotter the oven, the greater and faster the heat loss).
The pan you put your beans on require hedgehog balloons.
If you are stirring to try and help out generous Tortious distribute his balloons (stirring), then balloons are escaping and not even a constant amount. The more balloons in the oven, the more than will pour out.

That last item is probably the one I run into day in and day out. Ovens are different. As you can see by now, the oven temperature is only a way to add hedgehog balloons to your roast and it is intimately connected to how many balloons your oven can supply in a set amount of time AND how big your Tortoise is. The one in the example needed 13 balloons/foot. Yours might be 9, or 17 or 27. There is just no way to know. Ratings like 1800 W, 15 amps and 25,000 BTU/hour (different ways of saying hedgehog balloons/minute) can be wildly off and are token estimates at best. To connect the dots, that means if I tell you to preheat to 350 F but your oven is twice as powerful as mine, that is going to be way too many balloons.

That right there is why I (nor anyone) can’t give you a time and temperature.

And yes, I know, I know, I know, you read it somewhere….but not a week goes by that someone like yourself (or Cat the aspiring chocolate maker) writes in bemoaning they are ‘following the directions’ but it just isn’t working. Do you see now why it isn’t working? I certainly hope so or I have not done my job today.

I want to mention one more thing people ask when I explain the above. “But Alchemist, if all that was true, that should also mean that cookbooks shouldn’t tell me to bake a cake for 35 minutes at 350 F but they do”. My response to that is that they also tell you to check if the cake is done…you know, that stick in a toothpick or knife and see if it comes out clean? This is said precisely because every oven is different. They also tell you to not double the recipe, watch out for hot spots and a bunch of other warnings because of all the reasons I outlined above. I bet if you give it some thought, a light bulb just might click on as to why it is pretty easy to roast a chicken well but damn near impossible to roast a turkey well (think balloon distribution).

Ok, I promised I would not just be ranting about what you can’t do. I just needed to lay out why it just isn’t as easy as setting the oven to a temperature and putting your beans in for a set amount of time. I promise you’ll leave this with some useful tips and tricks.

Clearly as was pointed out, people do make good chocolate with oven roasted beans. So how do you do it? Let’s get into that. The short of it is that you will have to most likely ruin some beans or at least roast them to less than their full potential. Milk chocolate is a great use for some of those since in essence dilution is the solution to the problem.

I hinted at what you are going to do. The key is that you have grasp that you are calibrating YOUR oven with ONE set of beans. Some of the maths might make you think you can extend it to more or less beans but I’m here to tell you I have tried and all those variables just make it fail. I was very tempted to lay out a page of complicated thermodynamic calculations but at the end of the day it is a lot of work for really just the shock value since your kitchen isn’t a controlled laboratory and you most likely aren’t a scientist with a background in properly control for a bunch of variables.

Moving on. A couple disclaimers. What I’m going to try to do here is teach you how to stack rocks. And really, that analogy is pretty good. There IS a technique but I have no idea what your rocks are like……..

If you want to stack rocks like these, there isn’t much too it.

But rock stacking can get really complicated. It is all about balance, and surface, and grip and counter balance, and I can talk until I’m blue in the face but until you try it, you aren’t really going to learn.

So let’s get in there, find where those flat surfaces are, which surface likes to grip to another, which ones just are not going to work, and make you into a zen rock stacker.

If you have not caught on yet, this is not an exact science and I’ve tried to make it as approachable as possible. That means there are tradeoffs. I could make this procedure pages long with lots of complicated and sexy (if that is your kink) equations but even those would only get you kind of close. I’ve opted for the minimum steps with the greatest chance of success with just a few tests.

Here is what I want you to do.

You will need:

Baking tray
Oven, gas or electric
Cocoa beans (2 lb)
Infrared thermometer.
Notebook

Pre-heat your oven to 350 F. Give it a good 15 minutes of soak time after it says it is at temperature. Check out the roasting profiles I give and note the EOR (end of roast) temperature. Record that in your notebook. You are NOT going to care about any of the other numbers of the profile.

Put 2 lb of raw cocoa beans on a room temperature tray, spread out as much as possible and put them in the hot oven.

I don’t care if the tray is thick or thin, perforated or not. Those are just hacks people have tried to deal with all that stuff I wrote above.

After 5 minutes, get your infrared thermometer and take a bunch of readings around the beans to see how hot your beans are.

Remember, the bunnies ARE going to be different heights. You are checking the average.

Give the beans a pretty quick stir and take another set of temperature readings.

I would expect there are going to bounce around 140-170 F.

After another 5 minutes do it again. A number or readings, stir, a few more.

At this point you might be getting close to 200-210 F.

The goal here is to continue at your start temperature (350 F in this case), checking and stirring every five minutes, until you are in the 205-215 F window. At that point turn your oven down to 15 F ABOVE the EOR temperature you recorded. If it is 255 F, you’ll turn the oven down to 270 F. I know many ovens have increments of 5 F so I would round down. If the EOR is 262 F and you want to go to 277, just go to 275 F and see how it works.

And here is the beauty of this method. Let the beans mostly be. Check them every 10 minutes, taking temperatures, stirring and a temperature again. Because you are only 15 F above your end point, the roast is going to slow down and the beans are going to be in very little to no danger of going too far. Basically, you are matching balloon passing to balloon loss to balloon input.

It might take another 10-40 minutes to get there. THAT IS OK. This is a calibration run. You will be roasting again, taking what you learned and methodically dialing in your roast based on YOUR oven.

Before we get to that, I want to talk about why this works. Over the years I have roasted 10s of thousands of pounds of beans a couple dozen pounds at a time. I’ve taught countless people how to drum roast. I have a methodology fully laid out. And it can look rather daunting. But I’ve also found that over time, I’ve simplified it or maybe noted, the really critical part that determines whether a roast is successful or not and that piece is the Development phase, or the time it takes to go from 212 F to 232 F. You nail that, and everything else falls into place. And how do you nail it? By having the drying phase at a certain speed, which is about 10-12 minutes to 212 F. After that, really it is just a matter of roasting long enough to have full heat penetration without going too hot….and this method above does that. Let’s look at why.

What are the goals?

Get to 212 F in 10-12 minutes
Roast at least 6 minutes more (2.5 minutes in the Development phase, 3.5 minutes in the Finishing phase)\
Don’t go past the EOR temperature

How are we doing this?

We are pre-heating so there are lots of balloons. I’ve done the work for you and am telling you 350 F is a fine STARTING point for most residential ovens (300-325 F is safer for commercial ovens).
Monitoring. We are watching pretty closely that we are not exceeding 212 F. Due to how much temperature spread there is in an oven where the beans are not moving constantly like in a drum roaster, I’ve worked out 205-215 is fine, and really 195-225 is even ok.
Once the roast has hit 212 F, we turn the oven WAY down trusting the physics of hedgehog balloons to carry the roast on through the Development phase.
We then just sit back knowing that the nature of the Tortoise is to distribute those balloons if we give them enough time with the assurance that with the oven so low, we just can’t over roast.

Before I get into the next step, I want to address one more really sweet piece of thermodynamics or as we are talking about it, the science of lifting cute things with balloons. We have a massive safety net. You know how you literally can’t burn something if you are boiling it? That is why simmering is so great for cooking things. You can’t get above 212 F because that is as hot as water can get before it turns into steam. But there is something even more interesting.

You know how it takes 1 balloon to raise 1 hedgehog 1 foot? That is only true until the hedgehog vaporizes. Alright, that is a little more gruesome that I meant it to be…but now that I think about it, it is not a bad analogy. Vaporizing a hedgehog takes a LOT of energy. Let’s stop the analogies that would be more appropriate in The Boys and just talk about water.

It takes 1 calorie to raise 1 gram of water 1 degree Celsius. You are following right? Well, that is only until the water is at 100 C (or 212 F). At that point, if you add one more calorie, nothing happens to the temperature of that gram of water. It just absorbs it. And it keeps absorbing calorie after calories until there is enough energy in there to rather literally explode or vaporize into steam. But it isn’t just 10 more calories. Or even 100 more. It is 540 more calories. In hedgehog balloons that looks like this.

The implication of this is that as long as there is water in the cocoa, there is this HUGE energy sink around 212 F as it starts absorbing energy like crazy and the temperature barely goes up. Mind you, it does go up and there is only 6-8% water, and much of that is deep inside, but there is still a buffer. This is why I am not concerned about you putting the beans into a really hot oven for a short time. The water in the beans helps protect them. And this also is why I am having you turn the tortoisestat/thermostat WAY down after that, as now much of that buffer is going away as the water vaporizes.

Alright folks, home stretch. I really appreciate you sticking around.

You have done your first calibration roast. There are basically two things that have to be adjusted for.

The time to 212 F
The final temperature not going (much) over your target EOR.

At this point, it is mostly intuition or maybe I would say common sense as to what you have to do next.

If your time to 212 (205-215 F) is under 10 minutes, drop your pre-heat temperature and roast again. I’ve found you should start kind of big. 25 F is a good place. If you were at 350 F, try 325 F. If you are over 12 minutes, increase your pre-heat temperature by 25 F.

The same goes for your EOR. If you find you are just coasting over the EOR in a short amount of time, then turn the oven down a little more. Instead of 15 F over your target, try 10 F. If the roast is stalling, and just not getting to your EOR target, or maybe just taking way too long (like over 30 minutes) then don’t turn it down as much.

And then do it again and adjust again. What you will find is that each time you are going to get closer and closer to your targets…..BUT…the best part is that unlike flailing around in the dark for temperature and times, here you are methodically dialing in two variables in a very controlled fashion.

AND

With possibly the exception of the first one, they should all be good roasts that make good if not great chocolate.

Ok, a few other things to keep in mind.

Trust the method. If your oven is so powerful and holds so much energy that you need to pre-heat to 280 F and turn it down to exactly the EOR temperature, then do that.
Do NOT try to extrapolate (i.e. estimate or correlate) one weight of beans for another. You are just going to make your life miserable. Instead, DO USE the energy specifications of the oven and roughly match them via my Drum Roasting Ask the Alchemist 300 to work out how many beans you should be roasting. Oh, and this is really important. If the DRUM calculations tell you 20 lb of beans, you are going to want to cut that in about half to 10 lb due to lack of REAL convection. This holds for 95% of convection ovens out there. Trust me here please.
Keep notes….but minimum notes. Do not try to record the 12 temperature measurements and average them when you are seeing if you are close to 212 F. It can get overwhelming and will not improve your ability to dial in. Again, please trust me. I tried. But do keep notes. There is a quote from Adam Savage that I’m going to paraphrase, that the difference in science and fucking around is recording the data.
Don’t just fuck around. Record your data.
Finally, please reach out if you have questions, comments, suggestions or problems. That is why I’m here. I just said a few times to trust me….but well, verify what I’m saying too and if you find your experience is different from mine, maybe I didn’t communicate what I wanted you to know and understand. At the end of the day, trust yourself and your experience.

P.S. One of the jumping off points for my brain were these videos that talk about convection and even though the equipment used was called convection, the reality was much different. it is a good watch for theory vs reality

I Turned My Dumb Food Dehydrator Into a Smart Pasta Dryer (PC fans, Sensors and all)

This is Why Drying Pasta At Home IS ALMOST IMPOSSIBLE

P.P.S The cute little characters above are not mine. I did my level best to contact the places I found them and did not get any response. All of them I found spread all over a variety of mediums to the point they very much appear fair use. To that end I would love to credit the artists. If one of these are yours, please reach out as I would love to license them if possible and of course give credit.

Ask the Alchemist #304

Energy and temperature are very different things.

Energy and temperature are not the same thing.

Important pages: