TDS vs Hardness: What the Numbers Really Mean

Consider two samples. Sample A reads 850 mg/L TDS and 15 mg/L hardness: it is sodium chloride water, soft but salty. Sample B reads 210 mg/L TDS and 160 mg/L hardness: it is calcium bicarbonate water, low in total solids but hard enough to scale a 70 °C heater. A TDS meter shows the first as the worse water and the second as the better water, which is exactly backwards for equipment protection.

Coastal and island sources often show the first pattern, where seawater intrusion raises sodium and chloride. Limestone and gypsum geology often shows the second, where calcium, magnesium and bicarbonate dominate a modest total. The treatment response is opposite: sample A may need RO for salinity, while sample B needs a softener and possibly no RO at all.

Hardness as CaCO₃ equals calcium in mg/L × 2.5 plus magnesium in mg/L × 4.1. The multipliers come from equivalent weights: the CaCO₃ equivalent weight of 50 g/eq divided by the calcium equivalent weight of 20 g/eq gives 2.5, and divided by the magnesium equivalent weight of 12.15 g/eq gives 4.1. A water with 40 mg/L calcium and 12 mg/L magnesium therefore carries 100 mg/L as CaCO₃ from calcium plus 49 mg/L from magnesium, or 149 mg/L total. That is about 8.7 grains per US gallon and 1.5 mmol/L of divalent ions.

Laboratory reports use several units, and mixing them is a routine source of sizing errors. Convert everything to mg/L as CaCO₃ before comparing with a threshold, then convert to the unit used by the softener datasheet only at the end.

Hardness drives the softener; TDS drives the RO. They cross only when both are elevated. The matrix below is the practical version of that rule: hardness above 120 mg/L points to softening, TDS above 600–1,000 mg/L points to reverse osmosis, and the two decisions are independent. Installing a softener to fix a salty taste, or an RO to fix scale, solves neither problem.

A softener removes calcium and magnesium and releases sodium. Hardness at the outlet falls to 0–17 mg/L as CaCO₃, typically below 1 gpg, while TDS stays within a few mg/L of the inlet because one divalent ion is replaced by two monovalent ions. Sodium rises by about 46 mg/L per 100 mg/L of hardness removed. Alkalinity, sulphate, chloride and nitrate pass through unchanged.

The consequences are specific. A softener does not make water low in TDS, does not remove nitrate, and does not protect an RO from silica or boron. It does protect the RO from calcium carbonate and calcium sulphate scaling, which is why softening raises achievable recovery. On the sodium side, water softened from 300 mg/L hardness carries roughly 138 mg/L more sodium, which matters for low-sodium diets and for irrigation.

A TDS meter misses everything uncharged or poorly conductive and everything present at low concentration: bacteria, viruses, chlorine, pesticides, and the identity of the ions. A hardness test misses sodium, potassium, nitrate, fluoride, arsenic and lead. Neither test detects a problem that has no electrical or titrimetric signature.

Both also hide the difference between safe water and unsafe water. Coliform bacteria at 1 CFU/100 mL, nitrate at 50 mg/L and arsenic at 0.01 mg/L change the design far more than a TDS reading of 400 or 450 mg/L, yet none of them moves a TDS pen. Screening is not analysis, and a treatment train designed from two handheld readings is a guess.

For TDS, use a meter with temperature compensation to 25 °C and record the conversion factor; two meters set to 0.5 and 0.7 differ by 40% on the same water. For hardness, use a drop-count titration kit or a photometer; titration reads total hardness as CaCO₃ directly and is accurate to about 5 mg/L. Test at the raw tap, after the softener and at the point of use, because hardness can rise between them if resin is exhausted or a bypass is open.

No, and this is the most common field mistake. Because softening swaps calcium for sodium, the conductivity change is small: removing 200 mg/L of hardness as CaCO₃ removes about 80 mg/L of calcium and adds about 92 mg/L of sodium, so TDS moves by roughly 12 mg/L, well inside meter error and temperature drift. A softener that has completely exhausted its resin can still show a normal TDS.

Verify with a hardness titration instead. Outlet hardness below 17 mg/L as CaCO₃ confirms exchange; a rising outlet value over consecutive days predicts exhaustion and tells you when to regenerate. TDS is the right tool for sizing an RO and the wrong tool for proving a softener works.