Water
What the water out of my tap does to green tea
Tea is somewhere between ninety-nine and ninety-nine and a half percent water by weight, so it always seemed obvious that the water mattered. What was not obvious was how much, in which direction, and whether I could actually taste it rather than simply expecting to.
The water that comes out of this tap
The supply here is hard. The published figure for my part of the city sits around thirty French degrees, which works out at roughly three hundred milligrams of calcium carbonate per litre, and the kettle agrees: left to itself it grows a crust on the element in about five weeks. On top of a cup of green tea that hardness shows as a dull metallic film that catches the light, which people often assume is oil from the leaf. Some of it is, but most of it is calcium carbonate coming out of solution as the water cools, and you can prove that to yourself by making the same cup with distilled water and watching the film not appear.
The other half of the story is the treatment. There is chlorine in the supply, not much, and I can smell it if I run the cold tap into a glass and put my nose straight in. That smell is gone if the water stands uncovered for an hour, which is the cheapest single improvement I have found and the one I most often forget to do.
For the first two years I brewed with whatever came out of the tap at whatever moment I wanted tea, and I do not think that was wrong, exactly. It just meant that a meaningful part of the variation between my cups had nothing to do with anything I was recording.
Six weeks of three jugs
The comparison I ran between March and May 2023 used one green tea, a flat-leaf spring picking that I had about a hundred and eighty grams of, which is enough for a long test without the leaf changing under me. Three waters: straight from the tap, the same water through a jug filter with a cartridge four weeks into its life, and a bottled still water with a stated dry residue of thirty-eight milligrams per litre. Everything else held constant at three grams to two hundred millilitres, eighty degrees, ninety seconds, in the same two glass cups washed between rounds.
I tasted in pairs, never three at once, because by the third cup my palate flattens out and I start reporting differences I want to find. Twelve sessions, two pairs each, with the left and right positions swapped halfway through each session so that a preference for the cup on my right would show up as noise rather than as a result.
The honest summary is that two of the three comparisons produced a difference I could name reliably and one did not. Tap against low mineral: obvious, every session, no hesitation. Filtered against low mineral: I got it right about seven times in twelve, which is barely better than guessing. Tap against filtered: clear in nine of twelve, and clearer in the second half of the test than the first, which probably says more about my attention than about the water.
Kept on the card
- One tea, one weight, one temperature; only the water changes
- Taste in pairs, and swap which side each cup sits on
- Count how often you get it right, do not rely on remembering that you did
- A four-week-old filter cartridge is not the same as a fresh one, so write down its age
What the low mineral water actually changed
With the soft bottled water the liquor came out visibly paler and tasted distinctly more sour, in the pleasant sense of the word, with the sweetness sitting further forward and a clean finish that stopped rather than lingering. The tap water cup was fuller in the mouth, rounder, slightly heavier, and noticeably flatter at the end, as if somebody had turned down the top of the sound. Both are recognisably the same tea. If I had drunk them a week apart I would have said the leaf had aged.
The reason is mostly the alkalinity rather than the calcium as such. Hard water here carries enough bicarbonate to buffer the small amount of acid that the leaf contributes, so the brewed pH lands higher, and at a higher pH the catechins and the tea's own acids read differently on the tongue. The same effect is why a squeeze of lemon lightens a cup of black tea visibly within seconds.
It does not follow that soft water is better. For the green tea in the test I preferred it clearly. For a strong black tea drunk with milk in the morning I tried the same comparison over a fortnight in the autumn and came down mildly on the side of the hard tap water, which gave a fuller and less thin cup. So the working conclusion here is that the water suits some leaf and not others, which is unsatisfying but seems to be true.
What I do now, on an ordinary morning
Filtered water for everything green or lightly oxidised, drawn the evening before into an open jug so the chlorine has gone by morning. Tap water, boiled fresh, for anything strong and black. The filter cartridge gets changed on the first Sunday of alternate months whether or not it looks tired, because the one time I let one run to four months the difference had crept up on me and I could not tell when it had started.
I descale the kettle on the same schedule with a weak acid and a full rinse, and I have stopped topping up a half full kettle and reboiling. Reboiled water is a thing people argue about and I cannot taste the difference at all in a blind pair, but a kettle that has been half emptied and refilled five times has a mineral concentration nothing like what I wrote in the notebook, and that alone is enough reason to start fresh.
None of this is expensive and none of it is difficult. It took six weeks of slightly tedious paired tasting to find out which parts of it mattered, and the part I would keep if I could only keep one is the jug standing out overnight.