Roughly ninety-seven percent of the Earth's water is salt water, and a further two percent is locked in glaciers and ice caps. The remaining one percent is the resource on which all agriculture, industry, and human consumption depends, and it is not distributed evenly. Some of the world's fastest-growing cities sit beside vast oceans whose water they cannot drink. Against this background, the removal of salt from seawater - desalination - has steadily moved from a fringe technology into a significant element of the global water supply. What began as small experimental plants serving island communities has grown into an industry supplying more than three hundred million people in over 150 countries.

The oldest method is distillation. Seawater is heated, the pure water vapour rises, and the salts remain behind. Sailors have used versions of the method on board ship for centuries, and large multi-stage distillation plants were the first major industrial application of the idea, installed along the Persian Gulf in the 1950s and 1960s. Such plants are energy-intensive, which limits their economic range to locations with access to cheap fuel or to the waste heat of an adjacent power station. They are still widely used in the Gulf region, where natural gas is abundant, but they are rarely chosen for new projects elsewhere.

The dominant method today is reverse osmosis. A high-pressure pump forces seawater through a membrane whose pores are large enough to pass water molecules but too small to pass dissolved salts. The rejected brine is discharged back to the sea, while the filtered water is collected for further treatment. Modern polymer membranes can produce fresh water at pressures of sixty bar or higher, and the energy required to produce a cubic metre of drinking water has fallen from about eight kilowatt-hours in 2000 to around three in 2024. The improvement reflects incremental advances in membrane materials, energy-recovery devices that recapture pressure from the discharge stream, and the steady refinement of large-scale plant design.

The largest reverse-osmosis plants are impressive pieces of infrastructure. The Sorek B plant in Israel, completed in 2022, produces two hundred thousand cubic metres of fresh water per day, enough to supply approximately one and a half million people. The Victorian desalination plant south of Melbourne, a somewhat smaller facility, sits alongside a large solar farm that provides much of its electricity, reducing the carbon emissions associated with the production to well below those of other Australian water sources. Similar projects under construction in Chile, Morocco, and the western United States are explicitly paired with renewable generation, signalling a broader shift in which desalination and clean energy are treated as joint investments.

The industry's great environmental concern is brine. Desalination leaves a stream of water that is saltier than seawater and often warmer, and if that stream is simply discharged close to the shore it can smother bottom-dwelling communities. Modern designs reduce the impact in two ways: by diluting the brine with additional seawater before discharge, so that the plume mixes quickly with the surrounding body of water, and by placing the outfall at a depth and distance where the dispersion is strong. Research by the Spanish oceanographer Luis Ortega has shown that a well-designed Mediterranean plant can reduce local salinity increases to less than a tenth of what older designs produced.

Costs remain uneven. On a per-unit basis, desalinated water is still more expensive than water from conventional surface sources, though the gap has narrowed. In Singapore, where imported water has been a strategic concern for decades, desalinated water accounts for roughly a quarter of supply, and the authorities consider the marginal cost acceptable for the security it provides. In coastal Spain, where agriculture competes with urban users during summer droughts, desalination has been adopted more reluctantly and is still sometimes described by farmers as the water of last resort. Ortega has argued that the economic case for desalination is ultimately a political question about acceptable levels of water stress, not a purely technical one.

Innovation continues on several fronts. Forward osmosis, a method that uses a concentrated "draw solution" to pull water through a membrane without applied pressure, has been proposed for several niche applications including the treatment of particularly contaminated brines. Graphene and related two-dimensional materials offer, in principle, thinner and more permeable membranes than the current polymer products, though no graphene-based membrane has yet been produced at commercial scale. Solar-thermal desalination, using a field of mirrors to heat seawater without any intermediate electricity generation, has been built on a small scale in Saudi Arabia and could compete in locations with reliable sunshine and no ready grid connection.

For all these developments, desalination is not a universal solution. It remains most useful in coastal regions with significant water demand, and its economics are poor in inland areas where the fresh water would have to be pumped a long way uphill. The world still relies, and will continue to rely, on rainfall, river management, and careful use of groundwater for the bulk of its supply. What desalination offers is a backstop: a source of water that does not depend on rainfall and that can be scaled up when needed. In regions where that assurance is worth the cost, the technology has become a central element of long-term planning, and the trajectory of the last quarter-century suggests that its role will continue to grow.