We're 6 months into roasting. While that's not much time in the grand scheme of things, after nearly 100 batches of coffee on this little Proaster THCR-01, it felt right to reflect and share some lessons learnt while profiling. This blog post is for other newbie roasters or hobby roasters looking for foundations to build their drum profiling system on.
Here's a TLDR for those with less patience or time on their hands.
- Taste your coffee with other professionals.
- When you build heat momentum, keep it.
- Taste drives profiling, not curves. Numbers are there to support not guide you.
- Keep it simple, and don't chase perfection.
Let me start with a short aside before I begin to acknowledge my general newbie-ism. A production roaster might look at this number of batches and scoff. And, I understand why. While I was at my last café, in a short two months, my production roaster buddy roasted somewhere between 300 and 400 batches across probably 8 different coffee profiles. He was roasting 3-4 times a week and hundreds of lbs. I certainly believe that I have plenty of batches to go before I feel any sort of confidence about my skills as a production roaster. There's so much you learn from simple repetition and having to adapt daily to differing roaster and weather circumstances that I have yet to experience. Instead, I want to highlight a feature that our current drop format provides. That feature is flexibility, namely to experiment and profile. In fact, I counted and of the 86 roasts I have to my name in Cropster, 43 are profiling roasts! Including roasts outside of my Cropster account, nearly 50% of my roasting career has been toying with roast parameters.
Having completed near 100 batches on the Proaster, I feel as though I've already learned a ton about washed (mainly Colombian) coffee drum roasting on my roast environment. And, unlike my production buddy, I had no curves or guidelines to follow in planning. However, I admittedly have gotten much advice from several members of my co-roasting space and have largely built my starting settings (charge temp, batch size, and roast times) around three specific roasters' knowledge share... thank you Sarah, Paul and Mike! And yet, the pool of variables to play with still seems endless. So, I plan to share a few general takeaways and an approach to follow that might leave you feeling more confident about the the final profile you decide to offer to your customers... or just friends and family at home.
General Takeaways
Before I begin, I'd like to say that my approach for most coffees is fast and hot. My goal is to roast as light as possible to preserve much of whatever the green possibly has to offer. If there are florals, I certainly will prioritize showcasing those over a strong acid. Inversely, I will take a coffee darker if I taste character I find less appealing. Aside from that, I believe a good roaster has no skill to offer other than paving the way for the consumer to taste what the producer already had on offer.
In making these takeaways, I'm assuming a few things may be true about you. That's okay if not all are. Here are what they are:
a) You're actively roasting on a drum roaster, or you have some understanding of bean temperature and rate of rise on Cropster and are interested in coffee roasting.
b) You enjoy light roast drum roasted coffee (in some capacity).
c) You have not formalized your "style" in roasting and feel a bit overwhelmed in profiling new coffees.
Anyway, here are some of my North Stars that have helped guide me through the vast parameter space of coffee roasting.
1. Taste your coffee with other professionals
This speaks for itself and while it may seem straightforward to share, it doesn't feel as practical in practice. People are busy, you've got deadlines, and you may not be overly confident in what you have to share. But, I urge you, share regardless. I've brought "bad" roasts to cuppings before. In fact, I'd say most of the roasts I am getting feedback on I consider "flat", underdeveloped, cooked unevenly, paper-y or slightly stewed, smoke-y, or murkier than I'd like. Two things happen though.
First, I am consistently surprised with how much people enjoy what I have to share. Unlike your tastebuds, they have not experienced this coffee in an array of formats. The profile is totally fresh to them and as a result, their appreciation of that coffee is totally different. My hot take of the day:
Tasting a flat, over-developed sample roast for the first time is more exciting than discovering your perfect profile of a coffee.
This is because it's fresh flavor and has ton of uneclipsed potential. Now, consider how much better your 70-80% completed profile will taste to someone with no context about your green coffee. Go into that cupping with confidence!
Secondly, if they're experienced roasters, they will likely provide you with tangible feedback that will guide you with next steps to take in profiling. While tasting with friends in the industry, I discovered that: I could take my coffee lighter than I had been, I needed to slow down my momentum before entering first crack, faster did not always mean better, weight loss is far from a North Star, and I still have a way to go to properly penetrate the inside of the coffee bean. Tips shared with me in the cupping room have directly influenced my profiling afterwards. Sometimes you're far too in the weeds to see the bigger picture.
2. When you build heat momentum, keep it
Probably the best advice I have received to date was that my coffee was weak, low intensity, and not very acidic. One roaster finally said it to me, "it's unevenly cooked." From a visual inspection, you would never be able to tell either. My weight loss numbers were far into the normal range (10-12%), batch-to-batch and inter-batch color was extremely consistent. Color readers would not reveal what the experienced palate could though. What actually gave it away was the fact that some of my ultralights had better body than my light profiles. This was because I, somewhat misguidedly, kept my heat momentum better in my ultralight roasts than I did in my light profiles. See the example below.
In the two curves above, red represents my Light Profile and blue represents Ultralight profile from Drop 3. The red curve had a weight loss of 12.7% and an agtron score near 105. The blue curve had a weight loss of 10.2% and an agtron near 120. So, how could the blue curve be anymore evenly cooked than the red curve, especially considering how similar the rate of rise was throughout most of the roast (until about minute 5).
The proof was in my gas adjustments. See below.
On this roaster, 200 is full gas, or 100%. Notice that in my ultralight profile (blue), I carry full gas all the way to about 5:20 or 355°F while in my light profile (red) I begin to taper off as early as 4:30 or 330°F. On weight loss and color reading alone, I've created a darker coffee in red than I have in blue. Yet, blue had a bolder body. This is because, although it's easier to cook a coffee at the later stages in a roast, coffee takes on heat "better" (less aggressively) while its endothermic (earlier in the roast); cell structure is most intact and the coffee has more moisture to counteract heat. Applying heat evenly throughout a single seed gets harder and harder the longer you are in a roast. This is because heat application in a coffee seed gets easier and easier to apply to the external walls of the coffee seed, but not necessarily the interior. What heat you give up early will be very hard to replace later.
This is also a good foundation to roast from. Find a steady place to decrease consistent momentum from, and then, play with how late or early in the roast you drop your gas application. For example, your goal in profiling could be:
With the same momentum, how early can I ease off on gas while still cooking evenly (i.e. not tasting "green" in the cupping).
This would correlate with trying to maximize agtron score as well. If you don't taste green in your current batch, next batch ease off earlier and see if your coffee is more expressive. If you do taste green in your current batch, hold gas longer before applying your constant step down of gas. From here, you have a standardized way of profiling without introducing a pool of complicated variables.
In the future, you can still introduce moderate complexity by playing with the rate at which you step down on gas. If you started with 20 units of gas pressure per 5 degrees of bean mass temperature, try 20 units in less or more degree increments, or try larger unit adjustments. This will all depend on how reactive your adjustments are to your current batch size or burner you have at your disposal.
3. Taste drives profiling, not curves. Numbers are there to support, not guide, you.
Yes, you can develop rules of thumb as starting points. I've heard roasters have a plethora of North Star parameters to aim for when profiling. This might include: having a delta in temperature change post first crack that equals your moisture content, or measuring a weight loss that equals moisture content. These are great foundations to begin profiling. Find yours and use it for your first batches. But, good roasters often break their own profiling rules to find where a specific coffee tastes best to them. You should too.
Here's some ideas for how numbers can guide you when first building your profiles. Note that these are all based on the notion that we want to roast hot and fast.
- Use what you know about your roast system. I know that I've typically liked my extra light roasts in the 6:45 to 7:15 minute roast times. So, when I profiling a new coffee, I'm hoping to fall within that time window.
- I have typically liked my extra light coffees in this roaster that had bean temps of around 388 to 392°F at drop. Even if my coffee is cracking later, I'm going to aim for that range in post crack development and then adjust based on taste. If I'm cracking super late, I might think about how to build better heat initially.
In addition, use what you know about the coffee when you buy it to inform your first batches.
- I'm roasting a Colombia natural coffee with a high water content. Cell structure will be easier to break down in a natural vs. a washed. This implies a lower charge temperature. Higher water content though means that I can apply greater maximum heat early in the roast since it's release of moisture will counteract the heat. I'll charge lower, but apply heat rapidly out of my soak.
- I'm roasting a washed Kenya coffee that has a uniform screen size and was grown at high elevation. I'm expecting a super dense coffee that will require higher RoR later in the roast. I also know that this coffee will probably crack at similar times to one another due to the screen size sorting. This might cause my RoR to spike after first crack. I'm going to roast fast and hot, but be ready to kill the gas right at first crack if I see a spike.
Lastly, use the resultant data to guide your roast goals. For example, set ground rules. For this new Chiroso, I'm going to follow gas adjustments and profile I loved of the Chiroso I had a few months back. I'm also going to shoot for an agtron of 100 and a weight loss of +1 of whatever my moisture reading was. Here are some examples of how to think about adjusting profiles from sensory response based on your foundation.
- I cup my coffee and I'm tasting green still. My agtron was 100 but my weight loss was much lower. This tells me I should either carry RoR hotter longer and/or I probably need to lower my charge.
-
I cup my coffee and it tastes cooked but flat. In my next roast, I'm looking for more weight loss but the same, if not higher color score. This means a higher RoR later into the roast but a quicker ramp down of gas.
- I cup my coffee and I taste roast character. I'm now wondering if my RoR was too hot approaching first crack. I may also consider taking my development time down or shooting for a lower final bean temp (probably this!). And finally, could I ease up earlier on gas going into first crack?
4. Keep it simple and don't chase perfection.
This is a lesson I am still learning. Most of coffee flavor has already been imparted on the green you are roasting either a) at the farm or b) at the processing mill. My opinion (and the opinion of many other specialty coffee roasters) is that our goal as the roaster is to do our best to stay clear of imparting our own character. Instead, we want to emphasize the work done before us.
Here are two recent times I have had to use context about a lot to inform myself how light I can take a coffee.
- I can taste what I know is beer yeast in my yeast-inoculated coffee. While not a terrible profile, it does ruin the magic of the aroma the esters provide. I decided to take development longer, as well as slow down momentum to "stew" flavors.
- My coffee has intense acid that really shines but hides the more delicate flavors that were on offer. I decided to take development down by 10 seconds, but take maillard longer (slow down a little) to ensure my color score and weight loss was similar.
Besides that, I follow Rob Hoos's opinion about profiling. If my coffee tastes red: cherry, strawberry, apple, then I cannot possibly make it taste blue: blueberries, huckleberry, or plum. Lean into what you taste instead. Do these flavors taste best in their full structure (i.e. spread out on a charcuterie board) or do they taste better stewed together (i.e. jammed into a sandwich)? This informs me of roast time, momentum, and modulation stage times I might hope to achieve within the same development time, drop temperature, and (for the most part) color reading.
While we cannot change the color we perceive, we can change the intensity. Acid, body, sweetness, and clarity will all play a part in what flavors we do get to perceive in most brews. It's easy to get lost in this world, but my advice is high clarity should always be the goal. If I could possibly see four "red" flavors, I want the roast where those come through in a cupping and might let a good brewer accentuate for themselves. That might mean lowering acidity (the malic of cherry or apple) to make way for florals or berry sweetness. As a brewer, I can probably alter my bloom to adhere to the level of acidity I think balances my cup profile anyway.
My parting disclosure to this blog is that I'd like to note that this is my way and where I may get other roasters to disagree. More importantly, find YOUR way and what you prioritize in those four intensities I mentioned earlier. For example, your priority may be building body so that you can pair with milk, or acid to accentuate the fruit-forwardness of your coffee, or even just a "balance" of all the elements. Let that guide your judgement. There are many ways to roast great tasting coffee!