Pour a good mineral water next to your tap water and taste them side by side. The bottle usually wins, and it wins for reasons you can name. It is rounder. It carries no chlorine edge. It leaves none of that faint metallic scrape at the back of the tongue. It finishes clean and quickly. So you buy a case of it, brew with it, and the coffee comes out duller than the tap water version you were trying to escape. Nothing went wrong. You got exactly what you paid for, and what you paid for is the problem.
The smoothness you like in a good drinking water is very often its bicarbonate, and bicarbonate neutralises the acids that make coffee taste bright. The same property that rounds off a water's edges on the tongue rounds off the coffee's. Water can be excellent to drink and poor to brew with, and it is usually the identical component doing both jobs.
That is the whole counterintuitive part, and once you have it, most confusing water advice sorts itself out.
Alkalinity is not hardness, and the difference is the article
Two numbers get muddled constantly. Hardness is the dissolved calcium and magnesium in your water. Alkalinity is the water's capacity to neutralise acid, and it comes mostly from bicarbonate. They tend to move together in nature, which is why people treat them as one thing, but they do opposite jobs in your brewer.
Calcium and magnesium are participants. They are divalent cations, and they take part in pulling flavour compounds out of the coffee bed rather than merely dissolving what the coffee offers up. Worth keeping the proportions in view here. Brewed coffee is about 98 to 99 percent water by mass, so the water is not a supporting ingredient with a chemistry footnote attached. It is nearly the whole drink, and the beans are the flavouring.
Bicarbonate is a buffer. Its function is to absorb acid and hold pH steady, which is a useful property in a drinking water and a destructive one in a brewer. Coffee's acids are not a defect. They are most of what separates a bright Kenyan from brown liquid. Send high-alkalinity water into the coffee bed and a share of those acids is neutralised on contact, before you ever lift the cup.
Now go back to the taste test at the top. What does a buffered water feel like on the tongue? Soft, and rounded off at the edges, without the small sharpness that low-alkalinity water can carry. Bottlers know this, drinkers reward it, and a water marketed on its smoothness is frequently a water marketed on the exact number you want low.
Why brightness disappears rather than shifting
There is a nice piece of sensory research here that explains why the failure feels like flatness rather than like weakness.
Jessica Easto, in her book on tasting coffee, sets out the distinction the trade draws between acidity and sourness. Both words point at the same basic taste. Acidity is the term people reach for when that taste is proportionate to everything else going on, and it reads as brightness or tang. Sourness is the term they reach for when it takes over. The chemistry is identical and the judgement is about proportion, which is worth knowing before you decide your water ruined the coffee.
Easto also summarises research that undercuts an assumption most of us carry. Perceived sourness did not track pH. It tracked titratable acidity instead, and the gap between those two measurements is the mechanism you need. Think about how each number is produced. A pH meter reads only the acid that has already released its hydrogen ions into solution at that moment, which is a fraction of what the cup contains. Titration works differently: you keep adding base until every acid in the sample has been answered for, loose or not, so the figure you end up with is the total stock rather than the active slice of it. A buffer does not merely nudge the meter reading around. It consumes acid outright, which means bicarbonate reduces the quantity your palate was reading in the first place.
So the cup does not go sour, and it does not go weak. It goes flat. The fruit character thins out, the finish shortens, and what is left leans bitter, because the bitter compounds were never the ones being neutralised. If you have been calling that result "weak" and reaching for more coffee, strength and extraction are two different variables and you are adjusting the wrong one.
Reading a bottle label like a brewer
Most European bottled waters print their mineral analysis, usually on the back, usually in mg per litre. Two lines matter. Bicarbonate, written as bicarbonate or as HCO3. And dry residue at 180 degrees, which is close enough to total dissolved solids for this purpose.
To compare a bicarbonate figure against coffee guidance, you need one conversion, because the two are reported in different units. Coffee targets are given as alkalinity expressed as calcium carbonate. Divide the bicarbonate number on the label by roughly 1.22 and you have the equivalent figure. A water listing 300 mg/L of bicarbonate is carrying about 245 mg/L of alkalinity as calcium carbonate. The long-standing coffee guidance aims at something nearer 40.
That is not a rounding error. That is six times the target, and it is why a bottle can flatten a cup so completely that people assume they got a bad bag of beans.
One caution on the guidance itself, because it is misquoted almost everywhere. The familiar numbers, around 150 mg/L total dissolved solids and alkalinity near 40, descend from the legacy SCAA Water Quality Handbook. When we checked the SCA's published standards register on 27 July 2026, no water standard appeared in it. The targets are a reasonable place to aim and they were written by serious people. They are not a numbered standard, and calling them one is the most common error in the category.
The honest counterweight
Water chemistry is oversold, and this article would be dishonest without saying so.
Christopher Hendon and Maxwell Colonna-Dashwood published work in the Journal of Agricultural and Food Chemistry arguing that counting hardness is the wrong move. What the water is hard with decides the outcome. Bicarbonate suppresses extraction. Magnesium lifts it, at the same total hardness. That is peer-reviewed and we have no quarrel with it.
Then there is the other half of the picture. A blind panel result circulates alongside that chemistry and gets far less airtime: fourteen tasters were served water tuned to the higher magnesium proportion, the version the chemistry predicts should extract better, and they could not pick it out. Fourteen people is a small panel and a null result from it proves very little in either direction. What it does tell you is where fine mineral tuning belongs in your spending order, which is well down the list.
The gap between those two findings is worth holding on to. Alkalinity control and chlorine removal are the parts that reliably show up in the cup. Optimising the magnesium-to-calcium ratio is a hobby, and a legitimate one, but nobody should sell it to you as a fix.
What changing your water will not do
It will not rescue stale coffee, and it will not compensate for a grinder producing dust and boulders in the same dose. It will not correct a ratio you are guessing at rather than weighing. If the cup is already wrong in an obvious direction, work out whether you are looking at a sour cup or a bitter one before touching the water, because that single question routes you to grind, dose and time, which between them account for far more of the outcome.
It will also not work in reverse. Going to zero minerals is not the answer to bicarbonate. Distilled water extracts poorly and produces a hollow, thin cup, because the calcium and magnesium doing the extracting have been removed along with the buffer you were targeting. Anette Moldvaer's published target names sodium at 5 to 10 mg/L and zero chlorine, iron, manganese and copper, and it names a calcium range rather than an absence of calcium. Zero hardness is not a goal anybody in the literature recommends.
And a domestic water softener does not solve this. Ion-exchange softeners swap calcium and magnesium for sodium. They leave alkalinity untouched and raise sodium, which can push the cup toward tasting salty. Softened water is usually worse for coffee than the hard water it replaced, and people who install one for their pipes are frequently surprised by what happens to their coffee.
The version of this you can run tonight
Buy two bottles. One with the lowest dry residue on the shelf, and one premium mineral water with a bicarbonate figure in the hundreds. Brew the same coffee twice, same grind, same dose, same time. Taste them in pairs rather than one after the other with an hour in between, because a single cup on its own tells you almost nothing.
The difference is usually not subtle. The high-bicarbonate cup will read as rounder, quieter and shorter in the finish, and about half the people who run this test say they prefer it. That is a legitimate preference and it is worth knowing about yourself. If you like your coffee soft and low in acidity, buffered water is a tool rather than a fault, and it is cheaper than buying a coffee roasted to taste that way.
For everyone else, the practical answer is not a bottle at all. Find out what comes out of your tap first, since your utility publishes an annual report for free, then decide whether you have an alkalinity problem worth spending money on. If you do, the 2014 answer of buying bottled water has aged badly and there are better routes now.
The bottled water on the shelf was optimised by people solving a different problem than yours. They were making something pleasant to drink cold, in a glass, by itself. They succeeded. It is just not the same brief.