“Post hoc ergo propter hoc”
This is an opinion post on my thoughts on density and roasting.
Currently, there is a trend in which roasters are measuring the water content (moisture) of their coffee. This is fine. This is good. It is certainly never a bad thing to know more about the coffee we are roasting. It provides some insight into how that green coffee will keep; how shelf stable it is. Although, I feel that this correlation is still in need of a lot of testing.
What measuring moisture in coffee does not give a roaster, in my opinion, (and I would wager… my controversial opinion) is how to better roast coffee. Really, it tells us nothing concerning how a coffee will roast… at all.
Now, I know that at first this sounds CRAZY. Bear with me.
The moisture that is in the coffee when it is green is free moisture, meaning it is not a molecule of potential H2O that is currently part of a larger complex molecule. It is only water in the seed. H2O that we see during first crack is created during reactions (Maillard Reactions, Caramelization.)
Moisture is a GREAT conductor, and while present in the coffee, can work to wick heat into the core of the seed. Thus, its existence in the coffee is meaningful to how we heat the seed. But, the information a moisture meter gives us is being misused as a directive to how we should approach roasting a coffee.
The density of an object is simply Mass over Volume: D=M/V. We can measure this very easily for coffee by using a graduated cylinder. I recommend at least 250ml, but the bigger it is the more the more precise it will be. Fill it, and set it on a scale. There will be data noise with displacement (space between seeds) when measuring different screen sizes. The larger the tools, the less variable displacement exists, giving a more accurate reading. Some moisture meters come with a density tool, which is just awesome. Use it! That would be my preference, as it is much more accurate and easy.
Density is usually defined by: Grams/Milliliters, Grams/Cubic Centimeters, Pounds/Bushels, etc… This is important to know, because there is no standard within our coffee industry currently as to what definition we assess coffee to. Maybe we should have that. I would love to see a standardization in our language here.
OK! Back to moisture.
Let’s do a hypothetical exercise: pretend with me for a moment.
- There are two 100ml cylinders.
- Each cylinder is filled to the 100ml line.
- There are two different coffees.
- Each coffee measures 10% moisture.
(So far everything looks identical… right? The inference many people would make right now would be to roast these two coffees similarly. BUT WAIT!! THERE IS MORE!!)
Let’s, for the sake of this hypothetical, say that each cylinder is filled with exactly 100 coffee seeds. (I know… nonsense, but we are pretending.)
- When weighing the cylinders:
- #1 weighs 100g
- #2 weighs 90g
Now, we can see through this hypothetical, that even though our moisture meter is showing we have 10% moisture for the coffee that we are checking, the moisture meter is not telling us what that is 10% of. By measuring for density, we now see that in coffee #1, 10% shows that in 100ml of coffee there are 10g of water. We can see that in 100ml of coffee #2 there is only 9g of water. So, even though they have the same percentage of water, one coffee has more water than the other.
Furthermore, I would argue that this difference in moisture does not matter to how heat is transferred to the coffee. The density of the coffee DOES matter to how heat is transferred to the coffee.
The physics of heat exchange are such that: heat transfers more quickly (conducts) through more dense objects, and more slowly, (is insulated against) through less dense objects.*
Therefore, density is key.
Having water present adds to the density and conductivity of a seed (ability to absorb heat.) This is, of course, very important. However, if a roaster looks at the moisture percentage only, they are being led to believe something about the coffee that is incomplete. In order to qualify and quantify the percentage of moisture in coffee, a roaster must know what that percentage correlates to. In other words, a low density coffee may have the same percentage of moisture as a high density coffee.
Think about two water balloons. Balloon #1 is 5 inches in diameter and balloon #2 is 10 inches in diameter. If each is filled with 10% moisture, the 10 inch balloon has about 8 times more volume of water. (V=4/3πr3)
Imagine weighing out batches of coffee.
- Batch #1 stats: 100g, 10% moisture
- Batch #2 weighs 100g, 12% moisture
How should a roaster approach roasting these? Well, according to the current philosophies, they would treat them differently based on that moisture reading. The school of thought is that for Batch #1 the roaster should start with a lower heat level, as the coffee seed will take on heat more quickly because it has less water.
Why is this wrong?
Let’s hypothetically say that batch #1 is 100 seeds and batch #2 is 120 seeds.
Batch #2 is a lot more coffee that is a lot less dense, but will be roasted as if the inverse is true, based on how we currently look at moisture for roast preparation.
When thinking in terms of differences in our moisture levels for specialty coffee, there is very little difference from one coffee to another. However, in terms of density, that difference is much more varied. A low grown robusta from India could potentially have the same moisture level as a very high grown coffee from Colombia. Yet, would we suggest that we roast these in the same way? No! Yet, that is what we as an industry are currently teaching.
While it is true that there are many other factors beyond density that determine how a coffee seed should be roasted, measuring density can lay the best groundwork for how to initially approach a coffee before that coffee has been fully profiled. It will give you a head start that is meaningful and helpful, far more important that measuring for moisture can.
This will also lead to thinking about batches not only in terms of weight, but also volume. Volume is very important to how air will flow through the drum, and how much time each seed will be spending conductive time, bean to drum and bean to bean.
In my opinion, measuring moisture and using that measurement has created a correlation of information to results that is not truly a causation. I look forward to spelling out some of these correlations that are actually kind of true, but not fully true, so that we can be ready for the outliers, and basically be better at our jobs.
I am working on some research that (so far) supports what I am saying (or may completely trash it.) Stay tuned! I want a very large data set, so it may take some time. Discussion can start now, however-let me know you think! If you want to join me in the research, and you have have good tools for measuring and collecting data, I would be happy to assign tasks and crowd source. Thank you for taking the time to read this!