Coffee Data Science

Stumbling in Water: Espresso Water Experiment

A first set of experiments in brew water

Published in

Towards Data Science

5 min readJan 5, 2021

The idea of brew water recipes is certainly not new, nor has there not been at least some documentation on which recipes people like. There has yet to be a proper comparison of recipes, and I wasn’t looking to fill that void anytime soon. However, with James Hoffmann’s world wide coffee tasting, I had a chance to look at Third Wave Water.

Third Wave Water produces different mineral recipes to mix with distilled water, and James sent out one type to help everyone have the same water recipe during the coffee tasting. This was very important considering people were using various water sources around the world.

Afterwards, I wanted to start looking at brew water, so I started some paired shots with the gallon of water I had left over. In summary, I found a negative effect in six paired shots. I know this is a small sample, and I’m writing this up not as some conclusive statement but rather, initial results. I stopped collecting data because I’m more interested in what could improve my shots, and I just didn’t see a path forward for a great improvement. I

It is quite possible that the Classic Profile of Third Wave Water is not great for espresso or that my water is already pretty good. I’m still open to trying other brew water, but I moved onto other experiments in the meantime with a desire to circle back.

Baseline Water Setup

For comparison, I used filtered water from the tap, so the water is from Santa Clara County, California. I then use a Brita filter before pouring the water into my Kim Express. I haven’t had the scale problems on the Kim Express at home that I have had with the water at work, and it’s probably the Brita filter.

Shot Performance Metrics

I’m putting these definitions here because just after this section, I evaluate different amounts of paper filters using these metrics.

I used two metrics for evaluating the differences between shots: Final Score and Coffee Extraction.

Final score is the average of a scorecard of 7 metrics (Sharp, Rich, Syrup, Sweet, Sour, Bitter, and Aftertaste). These scores were subjective, of course, but they were calibrated to my tastes and helped me improve my shots. There is some variation in the scores. My aim was to be consistent for each metric, but some times the granularity was difficult and affected the final score.

Total Dissolved Solids (TDS) is measured using a refractometer, and this number is used to determine the percentage of coffee extracted into the cup in combined with the output weight of the shot and the input weight of the coffee, called Extraction Yield (EY).

Data

I collected six pairs of shots across two roasts on my Kim Express. Typically, my shots had a 1.5 output to input ratio, and most of them were on the higher end of extraction. These shots were all staccato tamped shots with a paper filter in the middle.

For the taste score (Final Score), Third Wave was only better once in a large way, while it was much worse 3 times. For Extraction Yield (EY), Third Wave was only as good as or worse than the filtered water.

In looking at TDS vs EY, I didn’t see any telling pattern for Third Wave expect being out performed by Filtered water.

I looked at time metrics as well, TCF and Total Time. TCF is the Time to Cover the Filter which I found is a good indicator of when to end pre-infusion (3*TCF seems to be the best time to end pre-infusion). Total Time includes pre-infusion and infusion time, and pre-infusion time was constant for each pair.

Third Wave shots ran slower, and I don’t have an explanation as to why. It is curious that they extracted less even though they took longer.

For the purposes of data, it is good to keep in mind that 6 paired samples is just an exploratory study. It doesn’t hold statistical significance. If anything, this small study suggests I should try a few brew recipes at the same time to narrow down what recipes to go more in-depth on.

I’m sure there is an optimal brew water recipe for espresso, but this was not it. I’m also sure there is a better recipe to brew water than my filtered water, and I will continue to search that out.

If you like, follow me on Twitter and YouTube 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.