Short Musings on Extraction Theory

Posted on February 19, 2024 in Opinion

I'm far from being a coffee expert, but I started getting into coffee a bit and thought I could use a summary of what I came to learn.

If you buy an espresso machine and want to get the most out of it, you will start looking for how-to’s on the web. You will soon find out you also need a grinder and a proper scale. And then most likely you quickly come across the following recipe:

18 grams, 1/2 ratio, 20s-30s time

Although it’s pretty straightforward to understand, all sorts of follow-up questions come up. First of all, it is not at all clear how to achieve this. Why grams and not dl? Why is time important if you know the ratio and the input weight? Why 18 grams if this is actually a recipe for a double, and a single basket can only take up to 10 grams? The interesting concept that connects all of these questions is extraction theory.


I would call extraction a measure specifying the intensity of getting active substances from the ground beans into water per a unit of time. The higher the extraction, the more substances are dissolved per a unit of time. One can influence extraction in many ways, with relations given in the following table.

Here ^ means, e.g., that the higher the temperature, the higher the extraction. One can see that roast level also affects extraction, where darker beans need less time to extract. Darker beans are also drier, so water flows much quicker through them compared to ligher beans ground at the same grind size - maybe this is why some people tend to spray beans before grinding them. In the case of the amount of water, it is a bit controversial because more water can extract more from the beans, but also dillutes the already extracted coffee. As for >, for example, the higher the dose, the lower the extraction. In other words, it takes more “work” to extract a higher dose of coffee, and more “work” to extract from finely ground beans. The rightmost column points to our original “ideal” recipe that tells us how to get a proper extraction.

Extraction is only useful insofar as it is consistent across the entire basket. It is a measure per particle in the basket. This is why one needs to ensure an even extraction, where each particle is in contact with roughly the same amount of water. There are many ways to achieve this in normal ways, like

  • proper tamping,
  • using a WDT tool,
  • having a dry basket before using.

Working with the recipe

We usually try to fix all variables except for one when trying to improve the taste, in our case, the recipe allows for changing the brew time. In theory, it is easy what to do when we do not like the taste and the beans are ok. When the taste is sour, we want to increase extraction so we increase the brew time. When the taste is bitter, we want to decrease extraction so we decrease the brew time.

The difficult thing is that changing any one of these factors in our table can influence the others. For example, if we fix 18 grams and a 1/2 ratio, how do we change just the brew time? Observe that we cannot simply stop the machine sooner, because then the 1/2 ratio is broken. What do we do?

It depends on what tools we have at our disposal. Based on my machine, I will assume that we cannot change the temperature and cannot change the pressure. If the machine is already “dialed in” (or reaching our recipe within reasonable bounds), everything is ideal and stays the same, we can increase (decrease) the brew time simply by grinding finer (coarser). The next time we are making coffee, we are looking at the output grams and reaching a ratio of 1/2 should take a bigger (smaller) amount of time.

However, we usually try to grind as fine as possible. And if we start grinding too fine when increasing the brew time, and our machine chokes, then we have to figure out a different strategy. Either we have to keep a good enough range for both directions on the grinder that our machine can handle, which might not be possible if we can’t change the pressure, or change some of the other variables. On that maybe next time.

Steaming milk

People like to use more technical terms like microfoam and macrofoam. When you are actually steaming milk, most likely you will not stare at it with a microscope. The milk will be in a stainless steel jug, and you will not be able to peek in. For an amateur like me, it is easier to simply differentiate between milk and foam, and this brought me better results.

The technique depends on our expectations, i.e., what we want to get from the milk. Cappuccinos in this “Kapuziner” style are easier to make because there is no latte art involved. You start with a jug filled up to a quarter, and then you try to push roughly the same amount of air into the milk while keeping the milk spinning. Once you reach one half of the jug, just spin the milk until you reach the ideal temperature. If successful, at the end you will get steamed milk with a blob of foam on top that you simply pour to your espresso shot.

Where I live, cappuccinos are basically lattes served with less milk and just a tiny bit of foam on top. This is (I presume) to be able to keep latte art. Instead of pushing too much air, we have to keep it at a lower level, I would say roughly in a 1/4 ratio (foam to milk). Then simply keep the milk spinning until reaching the ideal temperature. After it is done, first you will pour in the steamed milk, and at the end there will remain a tiny layer of foam in the jug that you can do your latte art with. The last part, of course, is less technical and more depends on one’s ability to draw with liquids. I regretfully admit that I am usually unsuccessful, as I cannot draw very well even with a pencil :-)


I will try to keep expanding this list.

  1. Puck is too mushy… Increase dose in the basket.

  2. I can’t make good steamed milk… Buy a thermometer and try to maximize the steaming time, e.g., by keeping the jug (and the milk) in the refrigerator, and warming up the steaming wand beforehand.

  3. The consistency of my foam is not good for latte art… We pushed too little/too much air into the milk or we didn’t give it enough of a spin.

Go up