Green hydrogen is produced by splitting water with renewable electricity. The chemistry is fixed: 1 kg of hydrogen requires 9 kg of water. After accounting for the extra volumes used mainly in pretreatment (demineralisation) and small amounts for polishing, cooling and periodic cleaning, most commercial plants consume 20 – 30 L of water per kilogram of hydrogen [1, 2].
Some commentators fear that gigawatt-scale hydrogen production could strain rivers and aquifers; however, the global arithmetic is reassuring. If every kilogram in the Hydrogen Council’s high-demand 2050 scenario (660 million t H₂ yr⁻¹) were made by electrolysis, total withdrawal would equal roughly 0.3 % of today’s global freshwater use. Water, therefore, becomes a design constraint rather than a show-stopper.
How hydrogen’s water use compares with other energy paths
When cooling water and upstream electricity generation are factored in, green hydrogen still appears modest. RMI calculates 20–30 L kg⁻¹ for green hydrogen, 20–40 L kg⁻¹ for grey or blue hydrogen, and significantly larger volumes for coal-based electricity on an energy-equivalent basis.
Coal stations typically evaporate about 1.9 L kWh⁻¹ in their cooling towers. If such coal-heavy grids powered electrolysers, the water footprint would almost double. In contrast, wind- or solar-powered electrolysis adds nearly none [3].
Viewed through a life-cycle lens, the green route is often the most water-efficient option.
Where will the water come from
Seawater via efficient desalination. Modern seawater reverse osmosis (SWRO) plants now consume just 2.5–3.5 kWh m⁻³, down from approximately 6 kWh a decade ago [4].
If electricity costs US $0.13 kWh⁻¹, a cubic metre of desalinated water comes to about US $0.45, or under one US cent per kilogram of hydrogen [5]. Large coastal projects, therefore, choose dedicated SWRO.
One example is NEOM in Saudi Arabia, where ENOWA, Veolia and ITOCHU are building a selective plant that runs on renewable power and sends no brine back to the Red Sea, turning waste into saleable minerals [6].
Treated municipal and industrial effluent. Inland hydrogen hubs often site electrolysers beside wastewater-treatment works. Reverse osmosis followed by electrode-ionisation (or mixed-bed ion-exchange) is the standard pretreatment for both PEM and alkaline electrolysers, polishing secondary effluent to < 0.1 µS cm⁻¹—well below the 1 µS cm⁻¹ conductivity threshold. This adds < 1 % to the electrolyser’s energy demand and < US $0.10 kg⁻¹ H₂ to the production cost.
Atmospheric moisture capture. In deserts, water can be harvested directly from the air. The Desert Bloom scheme in Australia will deploy thousands of off-grid modules that condense roughly nine litres of water for every kilogram of hydrogen, relying wholly on solar energy and avoiding any draw on groundwater [7].
Fresh water only where supplies are ample. Some projects still use fresh water, but only after regulators verify abundance. The ACES Delta project in Utah will withdraw about 730,000 m³ yr⁻¹ to make 100 t day⁻¹ of hydrogen – merely 0.012 % of the state’s annual water use and forty times less than Utah’s golf courses.
Checking the local balance – sustainability lenses
Water impacts are highly local. Namibia, for instance, is adding a 20 million m³ yr⁻¹ desalination plant because its coastal aquifers are already near their limit; the new plant will supply uranium mines and future hydrogen exporters alike [8].
Hyphen Energy’s planned three-gigawatt electrolyser near Lüderitz will share that capacity with towns through a common pipeline network, turning its water demand into a net benefit for the region [9].
Life-cycle analysis also shows that hydrogen can effectively recycle water: a fuel-cell truck or a green-steel furnace emits water vapour, which can be condensed and reused on-site.
Technology levers to cut water further
Several engineering options can further reduce consumption. Closed-loop or dry cooling replaces evaporative towers with air condensers, saving 8 – 12 L kg⁻¹ H₂.
High-recovery SWRO lines, equipped with pressure-exchange turbines, lower seawater intake by roughly one-fifth. Zero-liquid-discharge systems, although still in the early stages of deployment, convert brine into industrial salts, thereby eliminating marine discharge.
Direct seawater electrolysis remains laboratory-scale because chloride ions corrode catalysts and yield unwanted chlorine gas. Finally, large-scale atmospheric water capture, although energy-intensive, enables entirely off-grid hydrogen production in deserts – a trade-off that is sensible in regions rich in solar radiation but poor in surface water.
Researchers note that conventional SWRO already consumes < 0.2 % of the stack’s electricity, so the economic case for skipping RO is limited unless transport or footprint dominates [10].
Policy and permitting – turning risk into requirement
Regulators now embed water stewardship into project approvals. The EU Sustainable Finance Taxonomy requires hydrogen plants to demonstrate they “do no significant harm” to water resources, providing evidence of sustainable withdrawal volumes and safe discharge in environmental-impact assessments [11].
In the United States, projects seeking federal support under the Hydrogen Hubs programme must report baseline water availability, planned consumption and community consultations; robust water plans gain extra evaluation points [12].
Australia’s 2024 National Hydrogen Strategy update directs projects in arid states to rely on desalination or recycled water and to collaborate with state water agencies.
Namibia’s licensing process goes further, obliging developers to prove they will not unduly affect existing users; Hyphen’s pledge to pipe surplus potable water to Lüderitz became a concession condition [9].
In short, policy is pushing water-efficient design from the outset.
Project snapshots illustrate diverse strategies.
NEOM (Saudi Arabia) plans to produce 650 t day⁻¹ of hydrogen using renewable electricity and a selective, zero-liquid-discharge SWRO plant that recovers minerals from brine [6].
Desert Bloom (Australia) will rely entirely on atmospheric moisture, demonstrating that hyper-arid landscapes can support hydrogen farms without drawing on scarce aquifers [7].
Hyphen (Namibia) couples a renewable-powered desalination facility to its three-gigawatt electrolyser and shares the pipeline with local communities, demonstrating hydrogen as a catalyst for broader water infrastructure [9].
ACES Delta (Utah) relies on modest volumes of freshwater because supplies are plentiful, and its demand is dwarfed by ornamental uses, such as golf-course irrigation [2]. In Europe, Air Liquide’s 200 MW project at the Port of Rotterdam will utilise harbour wastewater in an industrial symbiosis model [13].
At the same time, BP’s planned two-gigawatt cluster in Valencia will recycle effluent from its adjacent refinery [14].
Across these cases, a common thread is the alignment with local hydrology and the delivery of community benefits.
Economics – why water rarely tops the cost stack
At the upper end of recent operating-cost estimates – roughly US $1.50 m⁻³ – seawater desalination adds only about US $0.014 per kilogram of hydrogen, or 0.7 % of a US $2 kg⁻¹ cost target [10].
Polishing the water to ultrapure quality raises the figure by fractions of a cent. Electricity, by contrast, typically accounts for 60–80% of the levelized cost of hydrogen.
For most developers, reducing power prices and improving electrolyser efficiency dominate the economics, while achieving water stewardship can be done at a minimal marginal cost when considered early.
Community engagement – water as a social-licence issue
Even where the numbers are small, perceptions matter. Communities in drought-prone regions may oppose any project they believe will “use a lot of water”.
Transparent data and clear co-benefits help. NEOM’s planned Internet of Water will allow residents to view real-time withdrawals [6].
Hyphen’s commitment to supply surplus potable water to Lüderitz reframed the project from a potential competitor for resources into a provider of new infrastructure [9].
The US Department of Energy also stresses the importance of early engagement with indigenous and agricultural stakeholders [12].
Conclusions
Green hydrogen does require water, but in modest quantities by energy industry standards. Including auxiliaries, a modern plant consumes around 20 – 30 litres per kilogram – about a household bucket.
Three proven sourcing routes – renewable-powered desalination, treated effluent, and atmospheric capture – enable projects to match local conditions, while closed-loop cooling, high-recovery SWRO, and zero-liquid-discharge designs further reduce the footprint.
Regulations now require developers to show exactly how they will secure water sustainably and discharge responsibly, and leading projects go beyond compliance by funding new water infrastructure or recycling schemes that leave communities better off.
Water is therefore neither a fatal flaw nor a free pass; it is a design brief that, when met, strengthens the social licence of the green-hydrogen economy.