Roast Profiling

When we talk about roast profiles , we can be talking about a few different things.

Firstly, we can be talking about the flavor profile of a coffee; how the coffee tastes, the mouthfeel, acidity, balance, etc. Flavor profile characteristics are of course determined by the coffee itself, but are greatly impacted by the roast profile. A roast profile is basically what happened during the roast and what adjustments were made to effect the outcome.

A more direct description: roast profiling is data collection. Your impressions of the cup of coffee itself, whether brewed or on a cupping table, are the most important pieces of data that you can collect.  Those impressions will greatly shape your approach to creating roast profiles.

Precise measurements and data collection hinge on what instruments and equipment you have at your disposal, but you can still collect information about a roast without the use of digital thermoprobes.    Logging your roasts and collecting as much data as possible is key to learning from and reacting to what you find in the cup.  You can log your roasts minute by minute on a chart, adding notes on adjustments that were made, or you can log your roasts on a line graph with the vertical line recording temperature and the horizontal line recording time. This visual way of logging a roast is very helpful in understanding a roast profile because it gives a shape to follow, a physical profile. Replacing the temperature with the color changes of the coffee while roasting on the vertical line can help you chart a similar roast profile on a line graph.

There are also measurements that you can make both before and after the roast that can be very informative. You can even change a roast profile in an air popper by making certain adjustments before roasting, or changing the way the hot air interacts with the charge (the batch of green coffee). It is important to remember that every aspect of the coffee and the roasting process is a variable in the equation.

 

Here's a list of some of those variables:

  • Moisture content of the coffee
  • Batch size (by volume or weight, weight is almost always a better measurement)
  • The start temperature (ambient)
  • Air flow settings and adjustments (where applicable)
  • Gas/Energy settings and adjustments(where applicable)
  • Drum speed adjustments (where applicable. This is fairly uncommon, but noted here because any change in the dynamic between radiant, conductive, and convective heat transfer in a drum style roaster will change the roast profile)
  • Cooling time

 

Here's a list of measurable events during the roast:

  • Turn around time (when did the temp bottom out before starting to rise again. This is usually a bean probe measurement, less distinctive as a ambient temp measurement.)
  • All color changes (the most significant one being when the coffee turns to yellow and browning reactions begin) - beginning of first crack
  • Timing/length of first crack
  • End of first crack
  • Length of time between the end of first crack and the beginning of second crack
  • Finish time, total time of roast
  • End roast color
  • End roast batch size (again, weight is best here but you can do a volumetric measurement
  • Moisture loss (factored by your beginning and ending batch weight)

 

Now, there are even more measurements and readings that we can take before, during, and after the roast, but these are solid and already long lists. Looking at these lists and comparing them to what equipment you have at your disposal will help you to decide what measurements and adjustments you will be able to take and make. If you roast coffee with a method or device that doesn't have air or energy adjustment capabilities, which variables can you still adjust to affect a change in the roast profile? The easiest adjustment that you can make in this case is your batch size, so even with the most limited roasting capabilities, you can still make changes to your roast profile as long as you are keeping track of your data.

Even if you have full control over all of those variables in a roast, it's not wise to make all of them adjustable; in adjusting these variables, it could become very muddy as to which adjustment is creating or affecting which change in the finished roast. Again, this is where data collection truly helps, by comparing your cupping notes (always the most important measurement of all) to your roasting logs and graphs you can begin to see what adjustments are helping to better express the acidity or body or sweetness of a particular coffee.

This also raises some questions about the purpose of your cupping. Cupping as a practice has been most widely used as a way to evaluate the coffee itself;  its quality, how the quality changed over time (shelf life), and whether or not there were noticeable defects. Through the wider use of the practice of cupping, it has come to be used to evaluate the roasting of the coffee as well. This can be really tricky, as certain characteristics added to a coffee through roasting can become quite aggressive through the cupping method. This subject warrants a much larger discussion, but it's important to note here that you are using the cupping in this case to evaluate the changes you made to the roast profile.

In order to do this well, it would be very useful to have already cupped the coffee and evaluated the coffee itself so that you are documenting what your adjustments have done to the original score or impression. You can also use the notes or score of someone you trust and that you've calibrated with, so that you know what a score of, say, 87 means to you.  As far as using cupping to calibrate a new roaster or equipment that you're unfamiliar with, it's again best to have a really solid idea of what the coffee should taste like, this will help you create the roast profile to best express this.

Moisture Loss

One of the most important measurements that you can take in regards to roast profiles is your moisture loss after roasting. This is also one of the easiest measurements to take; simply subtract your end weight from your start weight and then divide that number by the start weight.

Moisture loss can tell you a lot about a coffee that roast color, or time, or temp cannot. This is because you can produce a fairly light roast, yet still have a higher moisture loss which will affect the brightness, body, sweetness, and overall impression of the cup.

Generally, if you are getting moisture loss readings over 18%, you are either producing darker roasts beyond the Full City level, or you may have too much airflow which is excessively stripping the moisture from the coffee (this could also come from excessive heat input towards the end of the roast, which could also be affected by airflow depending on the roaster type). No matter what the actual reading is, it can also tell you how consistently you are achieving your desired profile from batch to batch. Consistency from batch to batch is at the end of the day the goal of creating a roast profile.

Once you have found a cup profile that you're happy with you want to be able to achieve the same results each time you roast the coffee, which you can do by following the roast profile, right? Well, this is where it gets a little more tricky. As many of us know, green coffee changes over time. These changes can be noticed in the cup, but can also be noticed in how the coffee behaves in the roaster and how well you're able to get the same results. You can keep track of your coffees' changes by taking regular moisture readings of the green coffee, or you can actually take volumetric readings in a graduated cylinder or even a measuring cup. If you see that it takes a greater volume of coffee to achieve the same weight as compared to readings from when you first got the coffee, then you know that you have lost some of the moisture content of the green coffee and that it will behave differently in the roaster.

The big lesson from this is once again collecting as much data as possible, but also that you should never become a slave to your roast profiles. It is true that not only will you have to change the roast profile to achieve similar or acceptable results, but also it is true that the same roast profile is not necessarily going to be the optimum one for every coffee!

An Exercise:

This is one of my favorite exercises, not just for learning about roast profiles, but for learning about roasting in general. The gist is that you'll roast 2 batches of coffee and based on the outcomes of those profiles (which you'll document) you'll try to create a new profile that gives you the cup profile (i.e. taste) that you desire. Here's a step by step:

  1. Select a coffee
  2. Document any data that you have about that particular coffee with an emphasis on physical characteristics (bulk volume density, varietal, SHG or lower grown, moisture content if you can)
  3. Weigh out 2 batches to the exact same size 
  4. Roast batches to 2 different roast profiles (this can be based on roast levels, or total roast time, etc., etc.), documenting all adjustments and important events. (refer to lists above)
  5. Chart both roasts on a line graph in order to create 2 visual profiles on the same graph (use different color pens for each roast)
  6. Cup the roasts the next day, taking careful notes
  7. Based on the cupping, decide what you'd like the coffee to be more or less of. This could be somewhere in between the 2 profiles, or darker or lighter, or longer or shorter, etc.
  8. Referring to your cupping notes as well as the roast graph, create/write out a new roast profile that you think will result in the desired cup profile
  9. Perform roast based on written profile. How well were you able to stay along the predetermined profile?
  10. Cup the new roast the next day. Did you achieve the result that you hoped for? If not, try again.
  11. Post your results on the Sweet Maria’s Forum!

 

Great start to the discussion

Thanks Chris for a cool set of thoughts on profiling. This is a topic I've always wished for more discussion on: sure, we talk about degree of roast all the time, but I feel like the process of selecting profiles is something that people keep to themselves. I know for myself, through trial and error I found a couple of profiles that seem to work well, and I use those, but there's no reason to think I couldn't be doing things much, much better.

I'd be particularly curious to hear what people think the effect of varying various parameters is. That is, if I hold all else constant, what happens if I charge at a higher temp or use more airflow? What's the effect of drawing out the drying phase, or the phase between first crack and the end of the roast?

Looking at it from the perspective of taste: what variables would you change if, say, you wanted to bring out sweetness or acidity?

Certainly each person should experiment and try things out on their own, but I'd be really curious to hear what general rules of thumb people are using.

 +1 to Josh's comment. 

 +1 to Josh's comment.