Hardness and salinity describe different problems
Hardness mainly reflects calcium and magnesium, often reported as equivalent mg/L CaCO₃. Conductivity reflects electrical transport by the mixture of ions; total dissolved solids is another measurement. Low hardness does not demonstrate low chloride, sodium or silica concentrations.
Start with a water specification at the point of use. It may include hardness, conductivity, silica, alkalinity and microbiological requirements. Each parameter needs a technical response. Choosing a treatment name does not establish whether the complete specification will be met.
What sodium-cycle softening does
A cation-exchange resin in the sodium form retains hardness cations and releases sodium. The bed is regenerated with brine and then rinsed. This is targeted treatment: replacing calcium with sodium is different from removing all dissolved salts.
An equivalent balance makes the distinction clear. Removing 250 mg/L of hardness as CaCO₃ adds approximately 115 mg/L of sodium on top of the sodium already present: 250 × 23 / 50. This stoichiometric example assumes no hardness leakage or brine carryover and does not describe the complete water composition.
Conductivity may change during exchange, but it is insufficient evidence of residual hardness. Measure hardness at the outlet and monitor progress towards resin exhaustion. Set regeneration around a demonstrated operating capacity rather than assuming every resin cycle will behave identically.
What reverse osmosis adds
RO applies pressure to produce permeate with lower concentrations of many salts and a separate concentrate stream. Rejection depends on membrane selection, the individual constituent and operating conditions. A laboratory-tested element specification does not guarantee the performance of the assembled plant.
For process water, RO is worth assessing when sodium, chloride or conductivity prevents the use of softened water. It also needs a concentrate outlet, defined net production and protection against fouling and scaling.
| Requirement | Sodium-cycle softening | Reverse osmosis |
|---|---|---|
| Reduce calcium and magnesium | Primary purpose | Reduced within system rejection performance |
| Reduce overall salinity | Not its purpose | Can reduce it; requires a projection |
| Manage residual streams | Regeneration brine and rinse water | Concentrate, flushing and cleaning streams |
| Verify quality | Residual hardness and service cycle | Permeate analysis and normalised production |
When combining softening and RO makes sense
A softener before RO can reduce certain scaling risks associated with divalent cations. Selection depends on composition, recovery, flow and the regeneration cycle. It is not compulsory for every plant: other scaling-control approaches should be compared for the actual feedwater.
Removing hardness does not automatically deal with silica, organics, oxidants or particles. Design the remaining barriers separately. If the process cannot stop during regeneration, assess storage, alternating duty units and hardness alarms within the appropriate supply scope. Confirm what happens if regeneration is incomplete or a valve allows untreated water to reach the membranes.
Compare useful output and residuals
For resin treatment, include salt, regeneration and rinse water, replacement and maintenance. For RO, include feedwater, electricity, concentrate, chemicals, cleaning and membranes. A continuous service-flow rating does not describe whole-cycle water use.
Compare the same annual volume at the same delivered quality. If both options comply, cost and operating requirements may decide. If one fails conductivity or another critical constituent, its lower purchase price is not a valid substitute for compliance. Include space, operator attention and the available outlet for saline streams.
The minimum data needed for selection
Share feedwater analysis, average and peak demand, operating schedule and the process water specification. Include hardness, sodium, chloride, alkalinity, silica, conductivity and temperature where relevant. A photograph of a conductivity meter cannot size a resin bed or establish an RO design.
For existing equipment, add salt consumption, regeneration records and episodes of high outlet hardness. This helps distinguish an incorrectly adjusted plant from a need for a different treatment objective. It also gives a realistic basis for comparing improvements with complete replacement.
Common questions
Is softened water demineralised water?
No. Sodium-cycle exchange replaces hardness cations with sodium while other dissolved salts remain. Demineralisation requires a different treatment objective.
Does RO always need an upstream softener?
No. The decision depends on scaling constituents, recovery and alternative pretreatment strategies. It should be justified in the design.