Exploring Water in Filter Coffee

Filter coffee provides us some additional flexibility that isn’t available in an enclosed boiler due to the relative low cost of replacing kettles, ease of descaling, and the benefit of less concentration of minerals in the boiling apparatus. This opens the water space to higher hardness, different hardness sources, and the use of some potentially damaging water additives like chlorides.

I’ve spent some time walking through a few different waters for filter coffee. The resulting waters are *not* universal and simply fit the grinder and coffees that I am working with; what I’m actually interested in sharing is the framework for water experimentation.

It’s useful to have a very basic understanding of how water in coffee is discussed, especially an understanding of the terms GH, KG, ppm as CaCO3. It’s also useful to know how to prepare a concentrate of mineralized water in advance. A good resource is https://espressoaf.com/guides/water.html

A Narrative of Dialing a Recipe

I started these experiments with the classic Holy Water (62gh/23kh) from strivefortone. This water is Epsom salts and potassium bicarbonate mixed into distilled. There is a convention in coffee water of reporting GH and KH as CaCO3, which I’ve used here. I thought the water excellent but that it might be a little overbuffered for the coffee and ratio I wanted to use. I verified this by cutting my brewing ratio and tasting for the perceptual acid balance. This improved, so the next step was to cut buffer down (62gh/11kh). In this case I cut it roughly in half and tasted again, which seemed right. In order to make sure I was on the right track I decided to briefly try higher hardness and the original buffer to verify (75gh/23kh) but preferred the lower buffer option. You could instead make a hardness only water and titrate buffer to taste, but this may only to get you into the ballpark if the minerals do have an effect during extraction.

Once I got buffer into the right ballpark I kept tasting for quality. I noticed two things I didn’t like. One was a kind of metallic sweetness, like what you’d get with Splenda or Aspartame. The other was a lack of what I’ll call sparkle, or the feeling that you’re drinking something quite exciting.

I started by trying to fix the sparkle first. I iterated through a few brewing parameters to make sure that the brew method wasn’t an issue and eventually settled on adding a small amount of NaCl to my brew water in the form of sea salt. .01g/l isn’t a lot, but it was enough to create that sparkle I was missing. However, the metallic sweetness remained. At this point I was using (62gh/11kh/.01g/l NaCL)

There were three potential causes of the metallic sweetness, so I tried to make some relatively simple changes to figure out what was causing it.

  • Potassium
  • Magnesium
  • Sulfates

I started out by replacing my buffer with baking soda to eliminate potassium; at this concentration this didn’t seem to be the issue. This was the easiest change to make since I had it on hand. Next, I replaced my Epsom salts with calcium chloride. This resolved the perceptual issue for me. Eventually I intend to test this again with magnesium chloride to determine if I just don’t like the sulfates, but that hasn’t happened yet. One small thing you might notice is that I’m using baking soda and calcium chloride, so I was able to drop the sea salt from the recipe. At this point the hardness and buffer totals were unchanged though the sources obviously were. (63gh/11kh)

Creating a Process for Developing Water Recipes

The narrative above was an interesting experiment for a few coffees from a single roaster, but many others have not responded quite as well to this water.

The most likely outcome is that I determine the coffee I’m drinking has a large variance of ‘best’ water. I say this because some wildly different waters already show much promise, but the presentation is all quite different. Whether I’m willing to direct mix single doses of water for a given coffee or not is up in the air at this point, but with a good concentrate system shouldn’t be too hard for filter brewing.

Using the narrative above we can derive a simple approach for iterating water design. We know several things from the above.

  • Chlorides may be interesting in filter applications
  • Different hardness sources may have different presentation of sweetness
  • Different hardness sources likely have different presentation of clarity
  • Buffer levels are easy to dial using multiple methods.

First, a note on making your life easier. I recommend using concentrates to build small water batches instead of mixing gallons or many liters at a time. Once you feel like you’re close go ahead and mix bigger batches to verify you like something. You can pour DI water into a glass, add hardness and buffer concentrates and single dose your kettle.

  • Create concentrates of at least magnesium hardness and buffer.
    • Consider having concentrated solutions of several common sources
      • Calcium Citrate
      • Calcium Chloride
      • Magnesium Chloride
      • Magnesium Sulphate
      • Potassium Bicarbonate
      • Sodium Bicarbonate
      • Sodium Chloride
    • Consider how you want to compare
      • This guide is built for filter experiments and uses water that is potentially corrosive; it’s likely a bad idea to use this framework for espresso without a lot of foresight.
      • Pourover
        • Possibly how you’re planning on making coffee, so if you are I’d eventually verify that the water suits this method.
        • Harder to be certain your brewing variables are well controlled.
        • Pourover has been great for illustrating water differences for me.
      • Cupping
        • If you’re primarily an immersion brewer this should be sufficient for dialing your water.
        • Pretty trivial to control brewing variables
        • Easier to side by side water recipes, especially if you have more than one way to heat water. Less reliant on taste memory which is a good thing.
          • Thomas at EAF mentioned that you could add concentrate drops to cups and then boil distilled, making this even easier.
    • Start with a moderate Magnesium based hardness, 0 buffer water
      • Titrate split samples of a single brew with buffer solution and taste
      • Compare with several different coffees
      • Compare with multiple grind sizes
      • Compare with multiple ratios
      • Try the other buffer source if available
      • Brew with this buffer included in the water to verify that the level is correct when mixed.
    • Adjust hardness for taste
      • Single dose your kettle with varying hardness levels using your concentrates and distilled water. Keep buffer fixed.
      • If you are getting muddy brews and using calcium hardness try replacing half with magnesium hardness.
        • If it’s persistent at this point try using only magnesium hardness, and/or add chlorides
      • If you are getting metallic sweetness try either magnesium chloride or calcium chloride.
        • If the metallic note persists with the magnesium chloride let me know and use calcium chloride instead!
      • If your brews lack interest try using a chloride source like NaCl or using a chloride based hardness source.

Leave a comment

Your email address will not be published. Required fields are marked *

Go to top