Coffee Data Science

Laser Particle Measurement for Coffee: Wet vs Dry

The limits of measurement

Robert McKeon Aloe


A few months ago, I had some coffee samples run Mastersizer 3000 Particle Distribution Analyzer. I was particularly interested in looking at how it performed for coffee under a few conditions. I was particularly interested in wet vs dry measurement of coffee.

I looked at 3 samples:

  1. Fresh grounds at 100 um setting on the Zerno Grinder
  2. Fresh grounds at 200 um setting on the Zerno Grinder
  3. Spent (dry) coffee from some experiments of the same coffee used for fresh grounds.

Then I looked at dry vs wet.

Starting with dry measurement, the particles are fed in a certain rate with some vibration, and some adjustments were needed to make sure there was enough particle separation. You could almost say that one has to dial in the feedrate.

I looked at these on a log scale. The 100 um had more fines as expected.

Then I compared dry and wet for the 200 um grinder gap setting because that is my typical setting for espresso. I looked at the cumulative distribution to better understand the percent of the total for something like fines. Depending on how you define fines, that point in the graph varies. I think of fines as < 100 um. Some believe fines are less or more than 100 um.

I noticed a gap. It seems the dry setting dropped to 0% around 10 um, but the wet setting went further. Let’s zoom in:

Then let’s make it more log:

This suggests even the microfines (<10 um) stay attached to larger particles.

Spent Coffee

When I looked at spent coffee, the distributions mostly aligned, but not quite.

Zooming-in, the dry sample goes down to a slightly smaller number, but not by much.

Making the graphs more log helps to show there are not many micro fines left. So they were either washed away during extraction or they dissolved completely.

I then plotted the difference of the cumulative distributions for both.

These simple tests clarified the complexity of the tool but also show how much information can be understood by using the two different sensor modalities. I look forward to designing more experiments for laser particle analyzers given the accuracy of the measurement, and I suspect they have been under-utilized in studying coffee.

If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.

Further readings of mine:

My Second Book: Advanced Espresso

My First Book: Engineering Better Espresso

My Links

Collection of Espresso Articles

A Collection of Work and School Stories



Robert McKeon Aloe

I’m in love with my Wife, my Kids, Espresso, Data Science, tomatoes, cooking, engineering, talking, family, Paris, and Italy, not necessarily in that order.