Revisiting Hamilton’s Tap Water (For Accuracy)

Recently I have been playing with mineralizing distilled water to different GH:KH ratios to see how it affects coffee flavour using the ideal brew water chart from Water for Coffee as a guide and one of Matt Perger’s water recipes (found in the comments here) as a starting point. While doing so, I came across more up-to-date information that makes it necessary to re-examine my two previous posts. The purpose of this post is to explain the changes and then perform this quick re-examination.

My experimentation led me to a lot of math, learning about moles, dilution calculations, and summarizing things in equations. After all of this I couldn’t seem to reconcile Perger’s water recipe with the ideal brew water chart, although this recipe typically produces pleasant coffee. Perger said this recipe should lead to 100 ppm Mg and 50 ppm KH. Whichever way I did the math it added up to 24.5 ppm Mg and 62.1 ppm KH.

Since I was confident in my math at this point, I concluded that maybe Perger wasn’t trying to hit the ideal realm of the brew chart, or that he mistakenly based his math on the chart as though it were presented in ppm with [ion] as [CaCO3], and it tasted good anyway. It turns out that the second conclusion was closer to the truth. But it wasn’t Perger’s mistake.

After a bit of research and reading through coffee’s deep web, I decided to go right to the source and ended up having a short Twitter exchange with Chris Hendon. As it turns out, the ideal brew water chart presented in WFC is presented in ppm with [ion] as [CaCO3] as opposed to [ion] as [ion], despite the book’s advocacy for the use of [ion] as [ion]¹ as the standard for ppm measurement.

This means that my previous two blog posts about Hamilton’s water as it relates to coffee brewing must be re-examined. The data from those posts are correct, however the analysis of the data assumes that the brew chart is presented with ppm as [ion], which is not what we want if we are working with the ideal brew water chart from WFC.

Thus, let’s re-examine.

Testing Hamilton’s Tap Water (For Coffee)
Here is the data presented with [ion] as [CaCO3]:

Mar 17, 2016 – Mg: 65.6 ppm, Ca: 87.5 ppm, KH: 91-93.4 ppm
Mar 18, 2016 – Mg: 32.8-49.2 ppm, Ca: 87.5-92.5 ppm, KH: 93.4 ppm
Mar 19, 2016 – Mg: 65.6 ppm, Ca: 87.5 ppm, KH: 88.5 ppm
Mar 20, 2016 – Mg: 82 ppm, Ca: 93.8 ppm, KH: 91 ppm
Mar 21, 2016 – Mg: 65.6 ppm, Ca: 87.5 ppm, KH: 93.4 ppm
Mar 22, 2016 – Mg: 65.6 ppm, Ca: 81.3 ppm, KH: 93.4 ppm
Mar 23, 2016 – Mg: 82 ppm, Ca: 93.8 ppm, KH: 93.4-98.4 ppm

Average – Mg: 63.6 ppm, Ca: 88.9 ppm, KH: 92.9 ppm

Within this sample set our Mg content fluctuates notably, while Ca and HCO3 (KH) remain fairly constant. If we look at our average reading, we get a GH of 152.5 ppm and KH of 92.9 ppm. In this sample set, our GH varies +/- 30.9 ppm and our KH varies +/- 5 ppm.

In short, our GH will generally be within the ideal range of the ideal brew water chart and the KH will still be pretty far from ideal. Thus, Hamilton tap water may leave our coffee flat and chalky, which is the same conclusion as before. Ideally our water filtration would bring KH levels down to somewhere between 32.5 ppm and 75 ppm while leaving our GH levels (Mg and Ca) relatively the same.

Filtering Hamilton’s Tap Water (For Home Brewing)
The data with [ion] as [CaCO3]:

Tap water – Mg: 49.2 ppm, Ca: 81.3 ppm, KH: 91 ppm
Brita filtered tap water – Mg: 49.2 ppm, Ca: 68.8 ppm, KH: 32 ppm

The tap water in this second test is similar to some of the data from the first test. The tap water had a GH of 130.5 ppm and KH of 91 ppm. After Brita filtration the GH was 118 ppm and the KH was 32 ppm. These results actually put us right on the border of ideal and acceptable water in the ideal brew water chart.

If you brew coffee in Hamilton, the Brita filter should take your tap water from poor for brewing to acceptable or ideal for brewing, depending on the tap water that day.

¹The difference is discussed in the corrigendum to the book posted here on 10-10-15.


Filtering Hamilton’s Tap Water (For Home Brewing)

[Having discovered new information, the data from this post has been revisited here.]

Today I tested my home brewing water with my titration kit. I also tested my tap water again for comparison. I wanted to post the results online in case anyone finds them helpful.

My home brewing water is simply tap water filtered through a Brita pitcher filter. The filter is currently about a month or two old. (Ew. Gross. Yeah. Whatever.) The pitcher has been flashing the “change soon” light for a couple of days, but has not yet flashed the “change” light. (See? It’s fine.)

I used the Red Sea Pro Test titration kit. I followed the same procedure as I did here, except this time I rinsed my sample vials and sample syringe with tap water before measuring the tap water, and with Brita filtered water before measuring the Brita filtered water.

Yellow rubber gloves were worn, and titration ensued.

Here are my results:

Tap water – Mg: 8-12 ppm, Ca: 32.5 ppm, KH: 111 ppm
Brita filtered tap water – Mg: 12 ppm, Ca: 27.5 ppm, KH: 39 ppm

There are a few things I would like to point out.

Firstly, the tap water was quite consistent with the samples from my first experiment. This is good to see. Consistency = more frequently good coffee.

Secondly, the GH. GH stands for general hardness, the “G” coming from “general”, and the “H” coming from “hardness”. Brilliant. General hardness is our calcium and magnesium content combined, in ppm. According to today’s measurements, our GH is in the 39.5 ppm to 44.5 ppm range for both waters. It’s a bit low. We know from Water For Coffee that we want our GH to be 50ppm or higher for ideal extraction capabilities.

It also looks like the Brita filter may have lowered the calcium content just a little bit. I don’t want to make any major conclusions about this. However, the Brita website says that it does reduce the concentration of calcium and magnesium. And it may have put us just a bit further off from the 50 ppm GH minimum.

Thirdly, the KH – carbonate or temporary hardness. This one I am willing to give a little more weight to. This is more in line with other results I’ve had measuring Brita filtered tap water. It is also a pretty drastic shift. And it is a step in the right direction.

As I mentioned in my last post (I’m just going to keep linking you to that), we also know from Water For Coffee that too much carbonate hardness can leave coffee tasting flat and chalky. 111 ppm is probably going to be too much. At least according to Colonna-Dashwood and Hendon, the authors of the book. And they know their stuff.

Luckily, the Brita filter brings Hamilton water down to 39 ppm, a much more reasonable KH level. Reasonable, but not ideal. As consistent as it is, the water composition will vary slightly day-to-day. Today the KH happened to be at a point where it would be ideal – if the water had a higher GH, in the 50 ppm to ~65 ppm range. It’s about balance. Even if we had 50 ppm in GH, our KH might fluctuate a little bit tomorrow and put us off of ideal water composition. At these low levels of GH and KH there isn’t a lot of leeway.

However, the Brita filter does take Hamilton water from less-than-acceptable to acceptable. It’s progress. And my coffee is tasting pretty good at home. Actually, quite good, more often than not. I guess it can only get better.

I will keep experimenting with ways to have better brewing water at home. For now, the Brita filter is doing pretty well and is a definite step in the right direction for Hamilton home brewers looking to take their brew game to the next level.

I’d love to hear what people are using for their brewing water. Let me know in the comments below.

[Edit: The KH measurements in this post were originally posted in [ion] as [CaCO3]. However, since we are more interested in bicarbonate, I have converted those measurements to display ppm as HCO3.]

Testing Hamilton’s Tap Water (For Coffee)

[Having discovered new information, the data from this post has been revisited here.]

I recently finished reading Water For Coffee by Maxwell Colonna-Dashwood and Christopher H. Hendon, an incredibly insightful book on the importance of water composition as it relates to coffee flavour and extraction. As the book was partly intended to spur discussion about water for coffee, I thought it would be fun to get in on the conversation.

The book focuses heavily on the roles of calcium and magnesium (general hardness, or GH), and bicarbonate (temporary/carbonate hardness, or KH) in determining the flavour in a cup of coffee. Naturally, after reading the book I decided I must buy a titration drop test kit to test my brew water composition. The book recommends titration as an affordable and accessible means of testing one’s water composition relatively accurately. And so I bought the Red Sea Reef Foundation Pro Test Kit, as seen on the Red Sea website and the feeds of every science-enjoying coffee enthusiast on Instagram.

After testing my own brew water I performed an experiment. I took and measured samples of water from my tap at home every day for a week to get a general idea of how the tap water in Hamilton is for coffee brewing so that I could post my results online for whoever may be interested, e.g. local coffee nerds and reefers (the latter apparently as nerdy as us coffee folk).

For those of you who do not know what a titration session consists of: in this case it is when you take vials of water, add specified amounts of specific chemicals, then add more drops of a different chemical until the water in the vial changes colour. The number of drops it takes to reach the end colour correlates to a certain ppm (parts per million) measurement of the measured ion. Very science.

There are three separate tests in the Pro Test kit: Mg, Ca, and KH. The Mg and Ca tests were pretty straightforward, however I had some trouble identifying whether I had reached my end colour with the KH kit, reaching a sort of pale salmon colour rather than the expected murky purple colour. Perhaps had I added more titrant I would have reached the final purple colour; although I did use quite a bit and the colour didn’t seem to change any further. For the sake of conservative estimation, we can interpret my KH measurements as meaning “at least” x ppm. The Mg and Ca tests were much more obvious.

[Edit: After watching Christopher Hendon’s recent video on titration, I have discovered that the pale salmon colour that I described above is, in fact, the correct end colour for the KH test.]

I used 10 ml water samples for all three tests despite the recommended 2 ml, 5 ml, and 10 ml for Mg, Ca, and KH respectively, which, provided I understood the book correctly, will increase the precision of the Mg and Ca measurements (Edition One, page 95). The Mg test measures in 20 ppm increments per 0.01 ml of titrant added to a 2 ml sample, so in a 10 ml sample each 0.01 ml of titrant should correlate to 4 ppm. The Ca test measures in 5 ppm increments per 0.01 ml of titrant added to a 5 ml sample, so in a 10 ml sample each 0.01 ml of titrant should correlate to 2.5 ppm. I believe this is allowable.

The KH test measures in meq/L, which is a unit of measurement that eludes me, so to obtain the measurement in ppm I used this neat little converter that I will just blindly assume is accurate until somebody with more understanding corrects me.

[Edit: The converter was very close to the math from Hendon’s titration video.]

I think it is also important that I mention that, according to the corrigendum posted on the Water For Coffee website on 10-10-15, the Red Sea kit “measures and reports individual ions rather than in [ion] as [CaCO3]” (Chapter 1, page 10). This essentially means that our measurements report ppm for each ion as though each ion were itself, and not CaCO3, which is, surprisingly, a distinction that must be made. As such, no conversions should be necessary for final ppm measurements, as would be necessary for ppm reported in [ion] as [CaCO3].

[Edit: According to Hendon’s titration video, the Red Sea kit reports in individual ions for the Ca and Mg measurements, but in [ion] as [CaCO3] for KH measurements. Since learning this, I have converted the following measurements to display ppm as HCO3 (bicarbonate), as they were originally posted in ppm as CaCO3.]

Anyway, here is the data:

Mar 17, 2016 – Mg: 16 ppm, Ca: 35 ppm, KH: 111-114 ppm as HCO3
Mar 18, 2016 – Mg: 8-12 ppm, Ca: 35-37 ppm, KH: 114 ppm as HCO3
Mar 19, 2016 – Mg: 16 ppm, Ca: 35 ppm, KH: 108 ppm as HCO3
Mar 20, 2016 – Mg: 20 ppm, Ca: 37.5 ppm, KH: 111 ppm as HCO3
Mar 21, 2016 – Mg: 16 ppm, Ca: 35 ppm, KH: 114 ppm as HCO3
Mar 22, 2016 – Mg: 16 ppm, Ca: 32.5 ppm, KH: 114 ppm as HCO3
Mar 23, 2016 – Mg: 20 ppm, Ca: 37.5 ppm, KH: 114-120 ppm as HCO3

Average – Mg: 16.29 ppm, Ca: 35.5 ppm, KH: 113.33 ppm as HCO3

Basically this means that my tap water (our generic Hamilton water sample) is not ideal for brewing coffee, according to the standard set by Colonna-Dashwood and Hendon, as it is too high in KH. The GH (combined Mg and Ca) is ideal for a very narrow set of circumstances, and at least acceptable for a much wider range of circumstances (i.e. if the KH were lower). To give you a good idea how far off our KH is, at this GH we would need our KH right around 37.5 ppm to find ourselves within ideal parameters (although with a higher GH, our KH could be as high as 75 ppm, depending on the GH).

Why does this matter? Well, KH functions as a buffer for the acidity in coffee – it balances out the high amounts of acidity. High enough amounts of KH can make the acidic compounds in a brew function as bases, leaving the coffee tasting flat and chalky.

Luckily my Brita filter seems to do the trick for home brewing, according to the quick, preliminary test I performed while learning to use my titration kit. That is, if the trick is to have acceptable brewing water. It’s still not ideal, but I will be experimenting further.

A few other details for those of you who might be interested:
-Where I accidentally pressed more than one drop of titrant into the test water vial, a range of ppm is given, correlating to the range of drops in which the colour change took place.
-It should be noted that tap water composition will vary from city to city, and will presumably even fluctuate between different parts of the city. This experiment is more to provide an example of the water at some “random” point in Hamilton to give a general idea of what the water may be like in any given part of the city.
-I followed the instructions provided with the kit almost exactly. Apart from my previously mentioned divergences, I did not flush the vials I used with distilled water, but with tap water before and after each test. For our purposes, I think this should be fine.
-I wore yellow rubber gloves for each titration session because some of the chemicals are corrosive and I was scared.
-Hamilton, as in Hamilton, Ontario, Canada.