Desalination converts the 97% of Earth’s water that is salty into water that can be used. It is not a new idea — sailors were distilling seawater centuries ago — but it only became economically viable at scale when membranes replaced boiling.

Thermal versus membrane desalination

Two ways to separate salt from water: boil the water off, or push it through a membrane. Membranes won on energy, which is why almost all new capacity is reverse osmosis.

Thermal versus membrane desalinationTwo parallel routes from seawater to fresh water. The thermal route heats seawater so it flashes to vapour in a series of chambers at falling pressure, condensing to distillate, and uses roughly 10 to 25 kilowatt-hours per cubic metre equivalent. The membrane route pressurises seawater through reverse osmosis membranes with energy recovery, using roughly 3 to 4 kilowatt-hours per cubic metre. Both produce a concentrated brine that must be disposed of.Seawater~35 g/lHeatingthermal routeFlash chambersfalling pressureDistillate10–25 kWh/m³ equiv.Pressurisationmembrane routeRO membranesPermeate3–4 kWh/m³Brine disposalthe shared problem
Why thermal persists
Thermal plants tolerate poor feedwater that would foul membranes, and make sense where waste heat from power generation is free.
The theoretical floor
Separating salt from seawater cannot take less than about 1 kWh/m³ thermodynamically. Modern RO at 3 kWh/m³ is already within a factor of three of that limit.
Brine
Roughly twice the salinity of the feed, and often warmer. Discharge design determines whether it disperses or forms a dense layer on the seabed.

The energy floor

Separating salt from seawater has an irreducible thermodynamic cost. The minimum work required to produce fresh water from seawater at typical recovery is around 1 kWh per cubic metre. No technology can go below it.

Modern seawater reverse osmosis plants use roughly 3 to 4 kWh per cubic metre — within a factor of three of the theoretical limit, which is a remarkable engineering achievement and also means the scope for further reduction is limited. Efficiency gains from here are incremental, not transformational.

Energy intensity by technology
TechnologySpecific energy (kWh/m³)Note
Thermodynamic minimum, seawater~1.0Physics; unattainable in practice
Seawater RO, modern with energy recovery3–4The dominant technology for new capacity
Seawater RO, 1980s8–10Before pressure exchangers
Brackish water RO0.5–1.5Far cheaper — lower salinity means lower osmotic pressure
Multi-stage flash (thermal)10–25 equivalentViable where waste heat is available
Multi-effect distillation6–15 equivalentMore efficient than MSF, still well above RO
Conventional surface water treatment, for comparison0.2–0.4An order of magnitude below desalination

Thermal figures are converted to electrical equivalent and depend heavily on the assumed conversion. Ranges reflect plant design, feedwater salinity and recovery ratio.

Reverse osmosis

Osmosis moves water toward higher salt concentration. Applying pressure greater than the osmotic pressure reverses it, pushing pure water out and leaving salts behind.

Reverse osmosisTwo chambers separated by a semi-permeable membrane. In natural osmosis, water moves from the fresh side to the salty side until pressure balances. In reverse osmosis, a pump applies pressure to the salty side that exceeds the osmotic pressure, forcing water molecules through the membrane to the fresh side while dissolved ions are rejected. The output is permeate on one side and concentrated brine on the other.residual pressureFeed waterseawater or brackishHigh pressure pump55–70 bar for seawaterRO membranerejects ionsPermeatefresh waterConcentrateroughly 2× feed salinityEnergy recoverypressure exchangerEnergy recovery: Recovers pressure from the concentrate stream. It is what took seawater RO from roughly 8 kWh/m³ to around 3.Remineralisationrestores hardness
Osmotic pressure
Seawater has an osmotic pressure of roughly 27 bar. Plants operate well above it, typically 55 to 70 bar, to achieve useful flow.
Recovery ratio
The share of feed that becomes product. Seawater plants typically recover 40–50%; the rest leaves as concentrate.
Why remineralise
Permeate is almost pure water, which is slightly acidic and aggressive to pipework, and tastes flat. Calcium and alkalinity are added back.

The brine problem

Every desalination plant produces two streams: fresh water and concentrate. A seawater plant recovering 45% of its feed produces roughly 1.2 cubic metres of concentrate for every cubic metre of product, at roughly twice the feed salinity, often a few degrees warmer, and containing the antiscalants and cleaning chemicals used in the process.

Discharged badly, it sinks and forms a dense hypersaline layer on the seabed that suppresses benthic life. Discharged well — through a diffuser, into an area with strong currents, or blended with power station cooling water — it disperses to near-background salinity within a short distance. Brine management is a design problem with known solutions rather than an unsolved obstacle, but it is a real cost and a real constraint on siting.

When desalination makes sense

  • Where there is no adequate freshwater alternative — the Gulf states, Malta, many small islands
  • Where drought resilience justifies the cost of standby capacity, as in Australian coastal cities
  • Where energy is cheap relative to water, which is the underlying economics of the Gulf
  • For brackish groundwater, where the energy cost is a fraction of seawater and the case is much easier

Sources