Roasting New Blends
Very often, I read articles on roasting, and they are all very practical and scientific. There are lengthy dives into thermal dynamics, airflow, and chemical stages of development taking place inside the coffee bean. These are all valuable things to know; however, they do not help you understand how to create.
The technical aspects of roasting play a role, and it is crucial to understand the factors that can influence flavor in the drum; however, it is more important to perceive those influences in the final product. It is also more important to have a process to create. If you don’t have a target in mind, how can you focus your aim?
For the past ten years, heading F+B Therapy, a coffee consulting company, I am in the unique position of having roasted on a wide variety of roasters. I’ve roasted on almost every type and scale of a roaster, including Brambati, Scolari, Probat, Diedrich, Loring, Ambex, US Roasting Corp. Ikawa, Vittoria (vintage), and a couple at origin didn’t have brands on them.
I create unique blends and roast profiles for clients without the comfort of always using the same roaster. As a result, I have been forced to find alternate paths to my influence and develop methods that are far less driven by the simple mechanics of the machine and much more based on the result in the end cup.
My process typically begins with a flavor target in mind. Let’s say, for brevities sake; I am looking to create a filter coffee that is balanced and approachable for novice drinkers yet sophisticated and quality enough for more experienced and discerning enthusiasts.
For this, we want to hit all the notes from a pleasant yet not overwhelming amount of acidity to a full body and sweet finish.
I start to remember and catalog all my favorite coffees I’ve ever had that fit this description and then try to dissect what about them I loved. Some of this is from memory; however, I also keep coffee journals with flavor parameters, spider graphs, and descriptors for this purpose. This process helps me imagine what I want to accomplish and consider what the key ingredients will be.
Of course, the next consideration is the price point, volume and consistency required. If I need this tasty, balanced coffee to be a high volume and lower cost coffee, this will automatically exclude certain ingredients that may inflate my costs. Also, if the volume is very high, then I may want to have a wider spread of ingredients. For example, instead of a three-bean blend, I may opt for five beans, so I have more options to rotate seasonal ingredients.
Now that I’ve answered these practical questions, my next course is to pull samples, roast, cup, and bring the band together. When you are making music, you select specific notes that work well together to form a song. You blend your high and low notes, your sharps and flats, and wrap it up in a tidy rhythm. This is the same as blending coffee. First, I survey each ingredient on its own merit, knowing what elements I need from each player in the group. Once I’ve heard what each player brings to the table, I can imagine and extrapolate the percentages I need of each element to play the song.
Finally, we are ready to profile roast. I try three different blends at three different profiles. I cup them all and then brew them as they are meant to be brewed as a final product. If I’m making a filter coffee, I will brew it on a machine and a pour-over. Testing the product this way will give a clearer picture of how your end-user will experience the final product. Cupping is excellent for green evaluation, but brewing with the final intention is always best for roast decisions.
Of these nine roasts, you should end up pretty close to where you want to be. After that, you refine and refine. It’s great to understand that while you’re roasting, the beans have chlorogenic and lactic acid inside that is transforming through the heat you apply; however, it is not necessary. The scientific method of trial and error combined with imagination and a well-developed pallet may be all you really need when roasting.