Recently, a coffee roaster reached out looking for some help with dialing in their new sample roaster. They had purchased a 1-kilo roaster with plenty of control options like variable drum and fan speeds, airflow control and a few thermocouples for real-time roast data logging. Being new to roasting, the number of variables were almost overwhelming to them. They had questions for me like,

– “When should I adjust airflow?”
– “How long should my turn around be?”
– “When do I apply gas changes?”
– “Do I need to change my drum speed while roasting?”
– “What should my rate of rise be?”
I was a little taken aback by all these questions, because they weren’t really options I had considered for sample roasting. While still sample roasting for purchasing decisions, the machine I used was a vintage 4 barrel Jabez Burns. If you are unfamiliar with this type of sample roaster, just know that you essentially have two options: flame on, flame off. On top of this, each barrel will more than likely have different gas pressure. Yet, these sample roasters are still scattered across the world and are being used even a century after their construction. The reason why is because they were designed with the intention to use as consistently few variables as possible while roasting.

From my perspective, my roaster friend was trying to profile his coffee, not sample roast it. While sample roasting, your objective should be to determine if a coffee fits the quality needs you desire for your company. The best way to do this is by barely altering a coffee’s inherent flavor characteristics. When you are able to sample roast consistently like this, you are also speaking the international language of coffee evaluation.

Developing a consistent sample roast workflow with only a small number of variables can be a tricky thing to do though. Caravela has taken this approach into consideration when constructing techniques and workflows for sample roasting with newer built machines. I have always admired Caravela’s Quality Assurance from the years of sourcing coffee through their supply chain. Now that I work for Caravela, I’ve been able to see first-hand how our sample roasting is done the same way across the world on 2 barrel Probat sample roasters, and below I will walk you through this profile.

Metal Drum Sample Roasting

Step 1: Warm up sample roaster for at least 30 minutes between 5-10% of total gas pressure. This is done to ensure heat stability and full thermal capacity of the metal drum and all other proximity parts.

Step 2: Perform a “warm up” roast on each drum being used. These roasts should charge around 390 degrees Fahrenheit. A “warm up” roast helps make sure that your coffee will roast in the time frame I will speak to below. It also will build additional thermal energy in the drum, helping your roasts become more consistent. It’s very common for the first roast on a machine to slow down too much around the first crack due to the lack of thermal energy. This can cause a coffee to “bake”, tasting like bread, or paper with very little sweetness. There are a lot of other roast defects you can cause by adding too much heat, or not enough heat, here are some more if you are curious.

Step 3: When your “warm up” roasts have been completed, let the drums heat back up between 5-10% of total gas pressure. Charge your first roast when your temperature readout is hovering around 370-390 degrees Fahrenheit while keeping your gas between 5-10%.

Step 4: To create a uniform delta or turnaround, Caravela applies between 50-75% gas pressure at 1:40. It will take some trial and error to find out what is the correct amount of gas pressure you should be applying at this point. As an example, if the coffee starts yellowing before 4 minutes, you’ve applied too much gas. If the coffee hasn’t yellowed by 5 minutes, then you have not applied enough gas. Once you figure this out for each drum, it is important that you stay consistent. You might notice that as you get to sample roast number 10 for the day, you are starting to hit yellowing closer and closer to 4 minutes. If you see this, you will want to slowly start dialing back the gas percentage being introduced at 1:40.

Step 5: Around 6:00-6:20 the roast operator should notice more smoke and chaff beginning to fall away from the coffee. Without making any adjustments, your roast should reach first crack around 6:20-6:30. At this point, it is best to trier, and smell the roast in an attempt to identify the aroma of vinegar or a tingling sensation in the sinus cavity. The audible first crack is usually around the same point, however not all coffees make that loud “popping” noise.

Step 6: The roast operator should continue to use the trier to smell the roast as it goes through first crack. When the aroma of vinegar or that tingle sensation is muted to the point of being less over-powering but still present, your coffee is ready to be dropped. This should be around 8:00, adhering to the 80% / 20% (Roast time / Development time) roast philosophy.

This is a basic, but easy to follow and consistent profile. When making adjustments to this profile to suit your company’s needs or adhere to it better, it’s important to only change one variable at a time. Like their operators, roasters are fickle machines. If you change too much at once, your rhythm and consistency will be off, and it will be very hard to figure out what you’ve done. Some additional things to consider for this profile include;

1. Find an Agtron reading between 55-62 degrees that fits your company’s taste. Once you’ve agreed on a specific Agtron reading for your sample roasts, you can also use this as a tool to determine when your coffee is done roasting.

2. If you are running fast on all of your times try charging at a lower temperature or making a smaller gas adjustment for your artificial turn around.

3. If all of your times through first crack are working but you are stalling at first crack, you can try adding airflow around 6:00 to introduce more thermal energy into the drum. If your machine does not have airflow controls, you can experiment with the addition of a small percentage of gas.

Ultimately, whether it’s green offering, physical/sensory evaluation, sample roasting, or any other protocol; consistency is the virtue to live by. With so many variables that are uncontrollable in our industry, my philosophy has always been to control the variables you can to the utmost degree. I hope this glance into Caravela’s sample roasting workflow helps your business become more consistent in sample roasting for your purchasing decisions.