Welcome to day 27 or 30. I’m here in rainy, cold Moscow getting over the standard jet lag and thinking about doing a little cafe tour. I’ll try to post some photos and videos to Instagram over the next few days.

For now though, my last couple posts on espresso have me thinking about another aspect in coffee that impacts both filter and espresso brews. This one is all about big impacts in the roasting process, and how those have an effect on bean mass and density. This will be less of the details of how to roast coffee, but rather a look at how coffee behaves differently with some basic roast inputs. 

I want to start out this post by saying that nothing I am presenting here is “science”. I prefer to put forth experiential knowledge of the things I have done and seen personally. You might disagree with something here, or have some numbers based approach to roasting that you feel is better. But the simple fact is that these are the things I have experienced and nobody’s theories or numbers by themselves will convince me otherwise. As soon as I taste results that contradict any philosophies I have on roasting, I will be extremely happy to adjust my discourse on the topic.

Now, some of you are very familiar with roasting and some are less so. For that reason I want to put the roasting process into relatively simple terms of time and temperature. Roasting is a dynamic process between not only time and temperature, but also air flow, machine materials, heat source, heat application, and more. I am also focusing on these two elements because they have a big impact on my focus today, which is the roasted bean density. Also, because there are so many possible outcomes from the roast process this could start turning into a book.

When it comes to time in roast, many of us have different techniques between filter and espresso roasting. Generally, most roasters will say that they roast for espresso longer for more development, caramelization, and perhaps solubility. The goal is often to balance some of the acids with deeper sweetness and body created through ‘maillard reaction’ (essentially an overused term for a quite complex process that includes caramelization). Filter roasts are commonly faster because the acid profile is less intense in filter coffee, and those same maillard flavors can take away from the origin flavors. I will argue that an appropriate level of caramelization is still desirable even in very light roasts though.

Problems come in the form of both too fast of a roast as well as too long. Too fast usually results in sour flavors, potentially because the interior of the bean did not have enough time for the heat to permeate and “cook” the inner cells. Too long of a roast almost always results in “baking” of the coffee. This is when it takes on a flat, muted flavor profile (or for roasters when the rate of rise is too low). The coffee loses its origin character, basically because it is too dried out.

Temperature has huge impacts on coffee flavor and body, especially at extreme levels. Just like with time, many roasters subscribe to a higher finished temperature for espresso, and slightly lower for filter coffee. The reasons are often similar or the same for temperature as timing, and I believe much of the reason for this is that many roasters do not have full control over their machinery. Darker roasting develops more of the “roast” character in the cup, losing more of the origin character. Lighter roasting preserves more of the origin character within the coffee, but lacks some of the potentially balancing roast character. Simply put, temperature by itself adds the energy to perform caramelization up to a point, and then results in pyrolysis or burning of the cells the higher it goes. In specialty coffee the variance of “lighter” and “darker” roasts is usually much more close together than traditional roasting. 

For example, a traditional roaster may have 30°F(@17°C) difference between their light and medium roast, while a specialty roaster may be more like a 5-10°F(@3-6°C) difference. Of course every roaster has a different philosophy on roasting, so these numbers are not a rule by far. These differences make sense though when we look at high quality coffee. There is much more to modulate and play with in high scoring specialty coffee than there is in lower scoring coffee, and a small change can have a big impact on the flavor of a higher scoring coffee.

There is also the element of fear. When someone buys a very expensive coffee (like $100/lb+) they become more and more afraid of over roasting. Unfortunately this results in a lot of extremely priced coffee that is not a great cup. This generally is not an issue with more traditional roasters, as all of their roasts are quite dark.

Problems in temperature, just like with time, come in the form of too high and too low. Too low tends to mean more sour flavors, but also vegetal and “green bean” type characteristics. This is likely because the coffee did not get hot enough to progress into the reactions needed for what we consider “coffee flavor”, especially when in conjunction with too fast of a roast time. Too high of a temperature usually means an increase in burnt flavors like ash, burnt toast, or rubber. When in conjunction with too long of a roast time, high temperature roasts tend to become one dimensional and dry. 

Now I have heard preferences for both high and low temperature roasts (very light and very dark in common terms), and I have actually tasted versions of both that were surprisingly palatable. In the US there is a common consumer expectation of ashy and darkly roasted coffee. It is even requested of roasting companies specifically because that seems to be the preference. I believe this preference is more a byproduct of a lack of comparative options, a desire for “intensity” in a large cup with a large amount of milk and sugar added, as well as some misconceptions(?) about the level of caffeine in darker coffee. The desire by some companies to source as low of a green price as possible has also led to darker roasting simply to cover defects. 

When most consumers taste a reasonably roasted (ie. not too much acidity) high quality coffee next to a darkly roasted (ie. burnt) low quality coffee, they can taste the quality differences easily. Their preference then needs to be influenced by their perceived value. Is the price not only fair, but could they afford it on a regular basis? If not, is the quality so great that they will adjust their budget to keep getting that experience? Many consumers will adjust their spending in order to justify those great experiences, they just don’t always have them.

There are numerous issues that could be addressed in discussion about roasting styles and philosophies, but to stay on track I want to talk about the physical nature of the beans being roasted when our time and temperature extremes come into play. This is important as it will be a big element of tomorrow’s post.

Let’s say for the sake of our examples today that we have the exact same green coffee for all of the roast scenarios. We want to think about what is physically happening in the process. Please don’t ignore that every green coffee needs a different roast technique for its best flavor outcome while we are talking about this.

Physically, a green coffee bean goes through loss of moisture, color changes, a wrinkling of the exterior skin, and depending on how high the temperature got, an expansion of the bean size, which results in a smoothing of the exterior skin and an ever darker and darker color as temperature rises. At the same time there are chemical reactions that begin internally, such as the maillard reaction I mentioned earlier. I have heard many roasters who subscribe to a requirement that the bean exterior must be smooth before it is finished roasting. This of course is a bit of an old philosophy, and largely based on green bean density or hardness (denser coffee will usually wrinkle more and take longer to smooth back out).

Otherwise, I want to look at bean density. I currently measure bean density with a graduated cylinder, and it is expressed as g/cl (or g/cc). As roast time and temperature increases, a bean gets lighter in weight from the loss of moisture. Most commonly (but not always) the higher the roast temperature the longer the time used, which means that darker roasts consistently have less density and lighter roasts have more. 

At the same time, temperature has the majority role in the bean size after roast. As coffee gets hotter and hotter, the bean begins to stretch and expand. Some of you may have never even seen the extremes of this. This expansion is not just the bean, but the cellular structure, which means that more physical space is taken by the same amount of beans. It is even more space when measured by weight, since each bean is slightly larger and considerably lighter. 

Have you ever seen a bag of Starbucks dark roast next to a specialty light roast of the same weight? It is almost double the size!

So in conclusion we have an increase in volume per gram weight (lower density), a loss of weight per bean, and increased dryness in the beans the longer the time and higher the temperature. Coffee beans are bigger, more brittle (you can see this by trying to snap it open with your fingers), and of course darker in color.

The inverse is true of shorter roasting times and lower temperatures. Lower volume per gram weight (higher density), a comparatively heavier weight per bean, and more moisture in the beans. Beans are smaller, sometimes more wrinkly (depending on starting density/hardness), less brittle, and lighter in color. 

None of this is talking about the flavor of differing roasts, and it is interesting again to note that the variance for specialty coffee tends to be more subtle and closer in time and temperature to each other than traditional roasting. 

Some additional notes here. Coffee that was lightly roasted, but took a long time, may be baked, and though the color is light and the bean is smaller in size it has still lost more weight and is likely a lower density overall. A coffee that was darkly roasted but in a very short time, might actually be quite dense. This is because while the exterior is quite dark, the interior did not absorb the heat as fast, and may have lost less moisture than it seems. These are conditions of coffee that have commonly been tested with an Agtron color test of both the exterior and interior of the bean in the past. A density test here is also quite useful though.

Now that we have run down all of these basic concepts in roasting, we are ready to talk about the post tomorrow. If you are wondering, I will be talking about how these variances impact extraction in both filter and espresso coffee. While the basic impacts on flavor are pretty obvious, there is a lot more to consider than you might think.

See you tomorrow,