Roast profiling: from sample to production roasting

How the world’s leading coffee roasters translate their roast profiles using ROEST and simple steps for how you can too

The first time you roast a new coffee can be nerve-wracking. All the hard work of the producers is now resting on your shoulders— not to mention the ever-increasing cost of a full batch of beans. Even a small mistake can ruin the delicate flavors of the coffee. Wouldn’t it be great if you could already know exactly how to roast that coffee?

That is the promise of profiling: the idea that with a good sample roaster, you can try different roast curves, 100 grams at a time, until you find the perfect profile. Then, when you load up your first batch in your production roaster, you already know exactly what you need to do to get the best flavor from that coffee.

The problem is that no two roasters are exactly alike. Even if you have two of the same model roasters standing side by side, the chances are that you need to make small adjustments to the profile if you want to get perfectly identical results. That makes it all but impossible to directly transfer a profile from a sample roaster to your production roaster.

Armed with a good sample roaster, however, you can learn a lot about how a coffee behaves during roasting. With its five temperature sensors, high-precision controls, and unique approach to monitoring first crack, the ROEST sample roaster gives you more information about your beans than any other sample roaster. This gives you a huge head start on developing your final roast profile, saving you both time and money.

 
A roaster checking the bean color using a trier on a ROEST sample roaster

The ROEST’s precision sensors help you understand how beans will respond to your production roast profile.

 

Roast profiling is a huge topic, so we have split this article into two parts. In this post, we talk to some of the world’s leading coffee roasters to find out how they use the ROEST to get to know the full potential of a coffee and help them develop their profiles for production roasting. We also outline a simple strategy for the ROEST that anyone can use to learn how the roasting variables affect the flavor of the coffee.

In part two (coming soon!), we will go in-depth with one specialty roaster who has developed a sophisticated system for creating and refining roast profiles in a 120-kg production roaster, based solely on what he learns from his ROEST.

But first, let’s briefly look at why a profile can never be exactly transferred from one roaster to another.

The limitations of profile transfer

The first challenge is that a temperature reading on one machine doesn’t match to a temperature reading on the next. The bean temperature probe in your roaster, for example, doesn’t only measure the temperature of the beans, but also of the air around them. Even small changes in the probe's position can affect how much the air influences the temperature reading. So even if you manage to replicate a bean temperature curve perfectly on two roasters, you are not guaranteed to get identical results!

Furthermore, if you have two roasters at the exact same temperature, but the airflow is slightly different, then the heat transfer will also be different. More air means more heat can get into the beans, even if the temperature is the same. Just changing the length of the chimney can be enough to affect your roast profile, because it changes the airflow through the roaster.

So when it comes to roasting with two completely different roasters, it’s no surprise that you can’t use exactly the same profile on both. But what you can do is learn how the coffee responds to roasting in your sample roaster, and use that information to guide the profile you create on your production machine.

We asked Tim Wendelboe, a world-champion barista and multi-award-winning roaster, about how he uses his ROEST when he is developing a profile for a new coffee. “I don’t really think it is possible to convert a small batch profile directly to a bigger batch — but I do use the ROEST to get to know a coffee,” he explains.

 
Tim Wendelboe pouring green beans into a ROEST sample roaster, getting ready to roast
 

“I roast samples of the coffee before it is bought, before it is shipped, and after arrival. That means I taste it many times and understand what the coffee has to offer. If I then feel that the coffee is not properly developed on the big roaster, or if it is too dark, I will have a reference to go on afterward, from my memory of tasting the coffee over time.”

Learning about your coffee

Roasting with the ROEST gives you a lot of information about how the beans respond to roasting. Events like First Crack, for example, change the way the beans absorb heat. In the ROEST you can measure the temperature of the air before, during, and after it contacts the beans, helping you to understand exactly how that heat is being absorbed.

The ROEST also shows you what’s happening during first crack: not just the moment it starts, but also the rate of cracks, and the total number that the roaster ‘hears’. The rate of cracks relates to the amount of energy in the roast as first crack begins.

Being able to visualize every single crack also gives you some extra information about how the coffee behaves during first crack — for example if a few beans tend to crack earlier than the rest, it’s helpful to know this before starting your production roast. No other roaster can give you this level of information about what happens to the beans during roasting.

 
Graph from the ROEST web portal showing the different temperature sensors and rate of rise

A handful of early cracks on this roast curve (in orange) show you what to expect from this coffee when you roast a full-sized batch.

 

We were lucky enough to attend coffee consultant Scott Rao’s roasting course in Chicago last month. He explained that he sometimes uses the number of cracks that his ROEST counts as a guide to when to terminate a roast if he has been trying out an experimental profile and can’t rely on his usual methods.

Like Tim, Scott says it’s impossible to transfer settings directly between roasters — but he does use the ROEST to explore the potential of a coffee. “I’m not profiling, I’m using the batches to inform me what my target end results are— what color, what weight loss,” he says. He then uses that information to adapt the roast profile he uses on his production roaster to suit the coffee. “If I prefer a certain development time on the ROEST, I know how to translate that to my IMF [production roaster],” he says.

For Scott, ROEST’s automation also reduces the amount of time and effort he needs to put in when trying out different roast profiles since he can start the roast following an automated profile, and then switch over to manual controls to adjust the profile towards the end of the roast. “I do my emails while I wait for the beep that announces first crack is coming,” he says.

The flavor map

You can even eliminate the roast profile as a variable, by simply using a flat inlet temperature to roast the coffee. With this type of profile, you can be sure that any change in the rate of rise curve is due only to the coffee itself. You can then adjust your production roast profile accordingly.

 
Graph from ROEST web portal showing flat inlet temperature

A flat temperature inlet profile is a very effective way to explore the variables in roasting one at a time.

 

A flat inlet profile is also a great way to explore the flavor of a new coffee. By roasting coffee beans at different temperatures and with different development times, you can build a map of the potential flavors of the coffee. The ROEST’s automation makes this process very quick and easy to do.

Here’s how it works: create a profile with a fixed inlet temperature, for example at 280°C. Now roast the coffee to three (or more) different development times: for example, 00:50, 1:00, and 1:10. This gives you an idea of how the roast degree or end color affects the flavor of this coffee.

Now repeat the process at different temperatures— say, 300°C and 320°C. This shows you how the flavor varies with the speed of the roast. Hotter, faster roasts tend to develop more acidity.

 
Chart map showing temperatures compared to time and how dark or fast the roast is
 

You now have 9 coffees, representing different roast times and different degrees of roast. By cupping these coffees all together, you can start to build a mental map of what kind of roast profile gives you the best flavor in this coffee. Perhaps you like the flavor of the lighter roasts, but got the best balance with the slower roast time to moderate the acidity, for example. Armed with this information, you can now adapt your standard production roast profile to suit this coffee.

This approach might be simple, but it has the potential to capture nearly all of the potential flavor that a coffee can have. In a recent Filter Stories podcast, Morten Münchow, founder of CoffeeMind, claims that as much as 95% of the flavor difference between roasts of the same coffee can be attributed to the roast color and the development time. The ROEST’s automation makes it easy to explore these parameters and find the sweet spot for a coffee.

Three simple steps to roasting on your own

1.⁠ ⁠Experiment with sample roasting

  • Start with a baseline profile on the ROEST sample roaster.

  • Roast small batches with different curves to understand the coffee's behavior.

  • Monitor key variables like bean temperature and airflow.

2.⁠ ⁠Create a flavor map

  • Use ROEST's precision to roast at different temperatures and development times.

  • Build a flavor map to identify optimal roast profiles for your coffee.

  • Document results and flavor preferences for each profile.

3.⁠ ⁠Transition to production roasting

  • Replicate successful sample roast profiles on your production roaster.

  • Fine-tune profiles based on differences in heat transfer and airflow.

  • Continuously evaluate and adjust profiles to maintain consistency and quality.

The next step

This method is a simple way to explore the potential of a coffee, and to understand how to adjust your standard profile to bring out its best flavor. Based on the results, you might try a few more combinations of time and temperature to zero in on the best possible roast for this coffee.

By mapping out the flavor like this, you not only have an idea of what your profile needs to look like, but you also have a reference of what the coffee can taste like at its best. For many roasters, this is enough to create their profile based on their knowledge of how their production roasts relate to what they see in the ROEST.

But Jan Komarek, Head of Coffee at Bailies Coffee Roasters in Belfast, takes this approach a few steps further and has developed a sophisticated system for perfecting his roast profile with the ROEST. Look out for part two of this post, where he shares his method with us in detail.

Previous
Previous

A systematic approach to roast profiling

Next
Next

Updates on P3000 - the Automated Production Roaster