Water for coffee: hardness, alkalinity and mineral content explained
Short answer: water is nearly all of the drink, so its composition governs how coffee extracts and how it tastes. Two properties matter most and are constantly confused: hardness, which is dissolved calcium and magnesium, and alkalinity, which is buffering capacity from bicarbonate. Hardness affects extraction and forms scale. Alkalinity neutralises perceived acidity. They are measured separately and treated separately.
Almost every piece of water advice in coffee fails for the same reason: it treats hardness and alkalinity as one thing, usually calling the whole business "hard water". Once the two are separated, most of the confusing advice resolves — including why softened water can still taste flat, and why two supplies with identical hardness readings produce different cups.
Why water composition is not a detail
Brewed filter coffee is roughly 98–99% water by mass. Espresso, being far more concentrated, is still around nine parts water to one part dissolved and suspended solids. Whatever else you change, the majority of what a customer swallows arrived through the mains.
Water is also not a passive carrier. It is the solvent doing the extraction, and its dissolved content changes both how much it pulls out of the coffee and which compounds it pulls out preferentially. A grinder adjustment moves the surface area available to that solvent; it cannot change what the solvent is capable of.
This is why water sits alongside grind, dose, temperature and pressure in the complete guide to espresso extraction variables rather than below them. It is also the variable most bars never check, which makes it the one most likely to be silently limiting the cup.
Hardness and alkalinity: the distinction that everything depends on
Definition. Hardness is the concentration of dissolved divalent metal ions in the water — in practice, calcium and magnesium. Alkalinity is the water's capacity to neutralise acid, contributed mainly by bicarbonate ions. They are separate measurements of separate things.
They are related in origin, which is why they get conflated. Water moving through limestone or chalk dissolves calcium carbonate, and that single process raises both the calcium content and the bicarbonate content together. So in many natural supplies, hardness and alkalinity do rise and fall together, and treating them as one number appears to work.
It stops working the moment anything intervenes. A conventional ion-exchange softener removes calcium and magnesium and replaces them with sodium. Hardness falls to almost nothing. Alkalinity is essentially untouched — the bicarbonate is still there. The water is now soft and still heavily buffered, which is exactly the combination that produces a flat, dull, slightly saline cup. Operators who soften their water and find the coffee got worse rather than better have met this directly.
The reverse happens too. Some treatment reduces alkalinity while leaving hardness largely in place. That water may taste noticeably brighter and still form scale.
Two practical consequences:
- Ask for both numbers. A water report or test giving you only "hardness" has told you half the story.
- Do not reason from one to the other. Knowing hardness does not let you predict alkalinity, or the reverse, in any water that has been treated or blended.
What hardness does
Two distinct jobs, one desirable and one not.
It participates in extraction. Calcium and magnesium ions are not inert passengers. They interact with several of the organic acids and other flavour compounds in coffee, forming associations that help those compounds move into solution and stay there. Water with some hardness therefore extracts more effectively than water without it — you get more dissolved coffee from the same dose, same grind and same time. This is the mechanism behind the observation that very soft water produces weak, hollow, thin coffee no matter how you adjust everything else.
Calcium and magnesium do not do this identically, which is the subject of the companion piece on how water's mineral content changes espresso flavour.
It forms scale. Calcium and magnesium are precisely the ions that come out of solution as hard mineral deposits when water is heated. This is not a coincidence or an unrelated nuisance — the ions that help you extract are the ions that fur up a boiler. Scale forms preferentially on the hottest surfaces, which are the heat-transfer surfaces the machine depends on, and the consequences run from quiet cup-quality drift to component failure. That progression is the subject of why hard water damages your coffee and your machine.
So hardness is the property with a genuine conflict inside it. Every water decision you make is negotiating between those two facts.
What alkalinity does
Alkalinity is a buffer. Bicarbonate reacts with acids and neutralises them, and the more bicarbonate the water carries, the more acid it can absorb before its pH moves.
Coffee is acidic, and a good deal of what makes a coffee taste lively — the citrus, the stone fruit, the sensation of brightness — is the perception of those acids. Water with high alkalinity neutralises a share of them on the way into the cup. The compounds were extracted; you simply taste less of their acidity.
The result is predictable in direction:
- High alkalinity produces a rounder, flatter, duller cup. A washed Ethiopian arrives tasting like a competent but unremarkable blend. Its distinguishing character was removed in the brewing water.
- Low alkalinity leaves acidity unbuffered, so the cup reads sharper and brighter. Taken far enough this is not a virtue: with almost no buffering, a coffee that was intended to taste bright tastes sour and harsh instead, and small variations in extraction produce large swings in the cup because nothing is damping them.
- Very low alkalinity also matters to the machine. Water with minimal buffering capacity is more chemically aggressive toward metals, and can attack the copper, brass and steel it passes through. Removing buffering entirely trades a scale problem for a corrosion problem.
There is a workable middle, and where exactly it sits is a matter of published standards and taste preference rather than a single correct value.
Alkalinity is not pH. These get swapped constantly. pH tells you how acidic or alkaline the water is at that moment. Alkalinity tells you how strongly it will resist being made more acidic. Two waters can share a pH and have completely different buffering capacity, and it is the buffering capacity that determines what happens to your coffee's acidity. For brewing purposes, alkalinity is the useful number and pH is close to irrelevant on its own.
Why TDS on its own tells you almost nothing
TDS — total dissolved solids — is what a cheap handheld meter reads. It is usually not even a direct measurement: most meters read electrical conductivity and convert it to an estimated TDS figure using an assumed relationship.
The problem is that it is a single total. It reports roughly how much dissolved material is present and says nothing about what that material is. Consider three waters with a similar TDS reading:
| Water | What it contains | How it brews |
|---|---|---|
| A | Mostly calcium and magnesium, moderate bicarbonate | Extracts well, forms scale |
| B | Mostly sodium and bicarbonate, after softening | Extracts poorly, tastes flat, forms little scale |
| C | Mostly sodium chloride | Tastes saline, extracts unremarkably |
Same reading, three results in the cup and three different consequences for the equipment. A TDS meter is useful for one narrow job — confirming that a treatment stage is still doing roughly what it did last month — and close to useless for deciding what your water needs. Hardness and alkalinity are measured separately for exactly this reason.
Why "pure" water is the wrong target
The intuition is strong: coffee is the flavour, water should get out of the way, therefore the purer the water the better. It is wrong on both counts.
Distilled and fully deionised water extract badly. With no calcium or magnesium to assist the process, extraction yield drops and the cup comes out thin, hollow and lacking in body. This is a repeatable result, not a matter of taste — the solvent has been made worse at its job.
They are also aggressive. Water with nothing dissolved in it and no buffering capacity is chemically hungry. It will take metal into solution from the surfaces it contacts. Running an espresso machine on deionised water is a maintenance decision, and not a good one.
This is why serious treatment for coffee rarely stops at removal. Reverse osmosis strips nearly everything, and where it is used for coffee it is normally followed by remineralisation or by blending a proportion of untreated water back in, precisely to rebuild the mineral content that extraction needs. Specifying that correctly is a job for a water treatment specialist and is covered in choosing a water treatment system.
The two objectives, and why treatment is always a compromise
Set the two goals side by side and the tension is obvious.
| What it wants | |
|---|---|
| Good extraction and flavour | Enough calcium and magnesium to extract properly; enough alkalinity to keep the cup from turning harsh; not so much alkalinity that brightness disappears |
| Protecting the machine | As little calcium and magnesium as practical, to limit scale; enough alkalinity and dissolved content to avoid corrosive water; stability over time |
The overlap is real but narrow, and where you sit inside it depends on your situation. A specialty bar serving single origins on a machine it services attentively will accept more scale risk for a better cup. A site with a fleet of machines, unsupervised hours and an expensive callout will weight the other way, and should. Neither is wrong; they are different optimisations.
What is wrong is not making the choice — running whatever the mains delivers, unmeasured, and then attributing the results to the coffee or the grinder.
Where the target numbers come from, and why they are not printed here
You will find specific target figures for brewing water in several places: industry bodies publish recommended ranges, machine manufacturers specify acceptable water conditions in their documentation, and competition and certification schemes have their own criteria.
Those sources do not all agree, and they get revised. Ranges published a decade ago have been superseded, some standards specify hardness and alkalinity independently while others give a single band, and the units differ between regions. A number quoted here would look authoritative while being detached from both your supply and your machine, which is worse than no number at all.
The reliable procedure:
- Get a current published standard for brewing water, in the units your local reports use.
- Get an analysis of your own supply, covering hardness and alkalinity separately.
- Get the water conditions your machine's documentation specifies, which are about protecting the equipment rather than optimising the cup and may be stricter.
- Treat toward the overlap between them, and re-check when anything changes.
How to find out what your water actually is
Three routes, in increasing order of usefulness.
Your water utility's report. Public supplies publish periodic quality reports covering hardness, alkalinity and major ions by supply zone. Free, reasonably thorough, and the right starting point. Two limitations: it describes the zone rather than your building, and it is an average over a period. Supplies blended from more than one source vary seasonally, sometimes substantially.
A test kit. Titration kits for hardness and alkalinity are inexpensive, take a few minutes and give you the numbers at your own tap. Less precise than a laboratory analysis, and quite precise enough to tell you which region you are in and whether anything has changed. For a bar this is the practical tool.
A laboratory or supplier analysis. A water treatment supplier will normally analyse a sample as part of specifying a system, covering ions a kit does not — chloride, sulphate, sodium, iron, and anything relevant to corrosion. This is the analysis a specification gets built on, so get it before committing to treatment.
Whichever route, sample from the tap the machine will actually use, after running it, and record the result with a date. Water is not a fixed property of a site.
Where this stops paying. Some supplies are already close to a sensible brewing range and need nothing but a filter for taste and particulates and an eye on the readings. If your water tests well, your coffee tastes the way you intend and your machine is not scaling, the correct action is no action. Water chemistry is a genuine lever, but confirm that grind consistency, dose accuracy, coffee freshness and cleaning are in order first. A treatment system fixes water. It does not fix a worn burr set or a stale hopper — see grinder maintenance and calibration schedules.
Frequently asked questions
Is soft water good for coffee? Not by itself. Softening removes the calcium and magnesium that help extraction while leaving bicarbonate buffering in place, so softened water commonly produces a flatter, weaker cup than the untreated supply. It does reduce scale, which is why softeners exist. Treat it as protecting the machine at a cost to the coffee, and specify accordingly.
What is the difference between hardness and alkalinity in one sentence? Hardness is how much calcium and magnesium the water carries; alkalinity is how strongly it resists a change in acidity. The first affects extraction and scale, the second affects how bright the coffee tastes.
Can I just use bottled water? For a domestic machine, sometimes — but bottled waters vary enormously in mineral content, most are not formulated for brewing, and some are harder than the mains. Read the analysis on the label rather than assuming bottled means suitable. At commercial volumes it is neither practical nor sensible.
Does a filter jug or a carbon filter fix my water? Carbon filtration removes chlorine, some organics and taste and odour compounds, which is worth doing and often noticeably improves the cup. It does not meaningfully change hardness or alkalinity. If scale or flatness is your problem, carbon alone will not address it.
Why does the same coffee taste different at two sites of the same business? Water is one of the most likely explanations, particularly across supply zones or where one site has treatment and the other does not. It is also one of the easiest to check: two test kits and ten minutes will tell you whether the water differs before you start adjusting recipes.
This article is part of the Vea Group knowledge base. Vea Group S.p.A. is an Italian designer and manufacturer of professional and premium coffee machines, with heritage dating to 1919 and production in Chignolo d'Isola, Italy, and Suzhou, China.