Extraction is an Average
Refractometers can tell you the total amount of coffee solids extracted from the grinds, but they can’t tell you what that extraction tastes like or what it’s made of. This is a limitation but not a fatal flaw of the method. There’s still a lot of actionable info to be gained from refractometry.
Within every brew, there is a mix of extractions. The millions of coffee grinds all experience a slightly different combination of temperature, time, and flow, and each has their own unique ratio of surface area to volume. It’s helpful to think of a coffee not as one extraction, but as a mix of millions of extractions of every single grind particle.
This giant mix of extractions becomes quite easy to understand once we measure every single grind from a typical sample. It turns out there are some incredibly counter-intuitive and surprising conclusions. The following videos explain how we measure our grinds, and the implications of this data on how we think about extractions.
In this video Matt shows you how coffee particles are analysed using a lazer diffraction particle size analyser (LDPSA) machine.
This video explains how to interpret the data obtained from a LDPSA machine.
Note: for every particle above 100um, there are 100,000,000 below that.
(See this Grinder Research Paper and the BH summary of the original research here)
A graph from the Grinder Research Paper depicting the high frequency of particles with a very small diameter
It’s also helpful to assume all coffee extractions aren’t perfectly even — they contain a gradient of extractions. All of these ‘micro’ extractions are mixed together and present the refractometer with an averaged strength. The refractometer only tells you the average because the coffee is now a homogenous beverage.
If you could pause time and extract the liquid from different sections of the coffee bed during the making of a brew, you would see varying strengths. As this is difficult and tedious to do, an average reading of coffee strength is taken instead. These different kinds of gradients as an average can be misleading. Taste and flavour imbalance can still occur even when your TDS number matches up with your targets.
Perhaps some of the coffee is extremely over-extracted and some of it is under-extracted, creating an average strength close to what was hoped to be achieved. If you rely solely on the refractometer to “taste” for you, you’ll be blind to the problem. Deliciousness must always be the aim — the refractometer is merely a tool to diagnose problems and help guide towards an ideal flavour profile.