Desalination and Water Reuse in 2026: Why Fresh Water Got Cheaper to Make, What Happens to the Brine, and When Recycling Beats a New Plant
- Internet Pros Team
- September 11, 2026
- AI & Technology
Taking the salt out of seawater is one of the oldest ideas in engineering and, for most of its history, one of the worst. It was the option of last resort for places with money and no rivers. That reputation has not caught up with the numbers. Over four decades the energy needed to produce a cubic metre of fresh water from the sea has fallen by roughly an order of magnitude, and the technology that did it is not a reactor or a turbine. It is a sheet of plastic film.
The Membrane That Changed the Economics
Early desalination worked by boiling. Multi stage flash and multi effect distillation plants evaporate seawater and condense the vapour, which eats enormous amounts of heat. They still run in the Gulf, bolted to power stations so the waste heat is free, but almost nothing new is built that way.
Reverse osmosis does something different. Push salt water against a thin film composite membrane hard enough and water molecules pass while dissolved salts do not. The pressure required is set by physics, not engineering skill, and for seawater it sits near sixty bar. That pressure is the entire energy bill, which is why the most important invention in modern desalination is the device that gets most of it back.
The pressurised brine leaving the membrane still carries the energy that was pushed into it. A pressure exchanger transfers that energy directly into the incoming feed water with efficiency above ninety five percent. Combined with better membranes and larger pumps, this dropped seawater reverse osmosis from roughly eight kilowatt hours per cubic metre to around three, and at large scale the best plants now operate below that.
Where the Energy Actually Goes
Comparing water sources by cost alone hides what is really being paid for. Energy intensity is the more honest measure, and it separates the options clearly.
| Source | Rough energy use | Where it fits | Main constraint |
|---|---|---|---|
| Thermal distillation | Very high, heat dominated | Legacy Gulf plants beside power stations | Only viable with cheap waste heat |
| Seawater reverse osmosis | About three kilowatt hours per cubic metre | Coastal cities with no freshwater option | Brine disposal and intake ecology |
| Brackish reverse osmosis | Roughly a third of seawater | Inland aquifers and industrial supply | Nowhere to put the concentrate |
| Treated wastewater reuse | Lower still, often under one | Any city that already has sewers | Public acceptance and regulation |
The cheapest cubic metre of water is almost always the one already inside the pipe network. Desalination is what a region buys when it has run out of easier answers, not the first thing it should reach for.
The Brine Problem
A seawater plant recovers between forty and fifty percent of what it takes in. The rest leaves as brine at roughly twice the salinity of the sea, warmer, and carrying antiscalants and cleaning chemicals. Discharged carelessly through a single pipe it sinks and pools, and the seabed community underneath does not survive.
The engineering fix is unglamorous and works: diffuser arrays that break the discharge into dozens of high velocity jets, sited where natural currents help, so dilution happens within tens of metres. Modern permits increasingly specify a salinity limit at a defined distance from the outfall rather than trusting a design assumption.
Inland plants have no such option, which is where zero liquid discharge enters. Evaporating concentrate to a solid is expensive and energy hungry, so the interesting work is in mineral recovery, pulling magnesium, lithium and industrial salts out of the concentrate so a disposal cost becomes a smaller one. Nobody should pretend this is solved economically. It is the honest limit of the technology today.
Why Reuse Usually Wins First
- The water is already collected. Wastewater arrives at a treatment plant under gravity. No intake, no marine permit, no coastline required.
- It is far less salty. Membranes work against osmotic pressure, and treated effluent has a fraction of the dissolved solids seawater carries. Less pressure means less energy.
- The concentrate is smaller. A reuse train produces a much lower volume of reject, and it can usually go back into the existing sewer.
- It is drought proof. Supply tracks population rather than rainfall, which is exactly the property a water utility is trying to buy.
- The barrier is trust, not chemistry. Multi barrier treatment with online monitoring produces water cleaner than most sources. Getting the public and the regulator comfortable takes longer than building the plant.
How This Became a Technology Industry Problem
Water used to be someone else to worry about. Two industries changed that. Semiconductor fabrication needs ultrapure water in enormous volumes, tens of thousands of cubic metres a day for a leading edge fab, and every litre of it has been through reverse osmosis and more besides. Fabs have been recycling aggressively for years because the alternative is not being allowed to build.
Data centres arrived at the same conclusion more recently and more publicly. Evaporative cooling trades electricity for water, and it is a good trade right up to the moment a facility is sited where the aquifer is already stressed. Water usage effectiveness now appears alongside power usage effectiveness in sustainability reporting, and several operators have pledged to replenish more water than they consume. That pledge is met through closed loop and air cooled designs, reclaimed non potable supply, and funding restoration projects elsewhere. The last of those deserves careful reading, because a replenishment credit in one watershed does nothing for a well in another.
The practical consequence for anyone planning infrastructure is that water availability has joined power availability as a site selection constraint. In several regions it is now the binding one.
Questions Worth Asking
- What is the actual water draw? For a facility with evaporative cooling, ask for litres per kilowatt hour, not a percentage improvement against an unnamed baseline.
- Where does it come from? Potable municipal supply, reclaimed non potable supply and on site groundwater are three very different answers in a stressed basin.
- Is the offset in the same watershed? Replenishment that happens two thousand kilometres away is an accounting statement, not a local benefit.
- What happens to the reject stream? Any on site treatment produces concentrate. Ask where it goes and under what permit.
- What is the drought plan? Municipal restrictions apply to large commercial users too, and a cooling design with no fallback is a business continuity risk wearing a sustainability costume.
The Direction of Travel
Desalination is not going to become cheap the way solar panels became cheap. Osmotic pressure sets a hard floor, the best plants already run within a factor of two of it, and the remaining gains are in pumps, pretreatment fouling and uptime rather than in a breakthrough membrane. What is changing is where it sits in the hierarchy: a firm, weather independent supply that pairs well with cheap renewable electricity, ramping production when power is abundant.
The more consequential shift is that water has become a design input rather than a utility bill. For decades an engineer could assume it would be there. In a growing number of places that assumption now has to be justified, and the organisations that treat it as a constraint early will be the ones still building where others are told to wait.
