Coffee Data Science

Fines Migrate in Espresso, but Not Far: Part 3

A Spent Coffee Experiment

Robert McKeon Aloe
Towards Data Science

--

One of the more difficult variables to observe is how particles shift in a shot from the original grind distribution to after extraction as extraction changes the size of the particles involved. To get around that, I used a puck with just spent coffee (well mixed), so I can measure before and after the shot. This experiment still lacks how having solubles can impact flow as well as a lack of CO2, but it stands to represent the best case scenario for shifting particle distributions.

This is a continuation of previous experiments in examining fines migration.

The Puck

After I pulled the shot, I had to pop the puck out and prepare it for slicing.

After wetting the puck, I was able to slice it into the bottom, middle, and top.

Measuring Particle Distributions

I measured the particle distributions before the shot and then for the three layers. For each layer, I took three samples, and I averaged the samples for this first plot.

There was a big shift after the shot, and the shot itself still had a 0.9% extraction yield. So there were still some solubles. Here is the cumulative distribution and a zoomed in version for 200um and smaller.

We can focus just on the top and bottom, and this shows a larger shift of the bottom samples from the top samples.

Another way to look at these samples is a cumulative bar chart. It shows the bottom has a shift, but it isn’t that large of a shift similar to my previous experiments on this topic.

To understand the impact better, we can take the difference between the top and bottom. This is a relatively small percent, and I suspect it represents the upper limit of migrating particles in a coffee shot. It seems the cumulative difference around 200um is around 1% of all the coffee.

This experiment helped isolate solubles, CO2 gas, and coffee grounds distribution. It supports the theory that finer particles do migrate, but the total amount of migration is around 1% of the total grounds in the puck.

--

--

No responses yet