Green hydrogen withdraws and consumes around a third less water than blue hydrogen — but the large-scale rollout of both to meet climate targets could present a burden on water-stressed regions, according to a new report by the International Renewable Energy Agency (Irena).

Stay ahead on hydrogen with our free newsletter
Keep up with the latest developments in the international hydrogen industry with the free Accelerate Hydrogen newsletter. Sign up now for an unbiased, clear-sighted view of the fast-growing hydrogen sector.

Water is not only needed for production processes, whether electrolysis, steam methane reforming or gasification, but also for cooling and H2 purification.

Irena cautions that many of the studies to date have only looked at water used in small-scale production, without accounting for the extra volumes needed for these other steps.

Additionally, more water tends to be withdrawn from a source than actually consumed within a system, as some volumes are rejected or intentionally drained to prevent mineral build-up.

For green hydrogen, this means that although, on paper, nine litres of water is needed to produce one kilogram of H2, the actual amount withdrawn from a municipal water source is closer to 15.2 litres, of which 5.2 litres are rejected, with one litre drained to prevent mineral build-up. An extra 19.5 litres is needed for cooling, of which 14.6 litres is evaporated, and 4.9 litres drained off.

Chart from the Irena report showing the average amounts of water withdrawn and consumed per kg of hydrogen for the different H2 production methods (including cooling). The straight black lines show the ranges of withdrawal and consumption. Photo: Irena

Irena’s baseline water intensity for each production method is based on the use of tap water, with even greater volumes needed if using river water, groundwater, or seawater, due to additional pre-treatment — although volumes used in cooling do not need to be treated or desalinated.

This means for green hydrogen, water withdrawal before cooling increases to 17.2 litres if sourced from a river or groundwater, and 28.6 litres if using seawater.

However, while grey hydrogen production via unabated steam methane reforming (SMR) has a relatively low water intensity — drawing on 20 litres and consuming 17.5 litres per kilogram of H2 — adding CCUS is likely to drastically increase how much water is needed for cooling.

Irena estimates that SMR with CCS would need to draw on 36.7 litres and use 32.2 litres of water per kilogram of hydrogen, with this falling to 30.8 litres withdrawn and 24.2 litres consumed by autothermal reforming with CCS.

Coal gasification is already the most water-intensive way to produce hydrogen, drawing 50 litres and consuming 31 litres to produce one kilo of H2. Adding CCS would increase water withdrawal for each kilogram of hydrogen to 80.2 litres and consumption to 49.4 litres, Irena calculates.

This would mean coal-based blue hydrogen would withdraw even more water per year than a 1GW coal-fired power plant.

Alkaline electrolysis meanwhile withdraws 32.2 litres and uses 22.3 litres, while proton exchange membrane electrolysers draw on 25.7 litres and consume 17.5 litres — comparable to grey’s relatively low water consumption.

This is because PEM electrolysers are more energy-efficient than their alkaline equivalents, which means less water is needed to cool down waste heat.

Water stress

Irena warns that even though green hydrogen can have a water footprint equivalent to or lower than existing H2 production, expanding hydrogen production will require a growth in water demand.

The world produced around 86 million tonnes of hydrogen (mainly from coal gasification and SMR) in 2021, withdrawing around 2.2 billion cubic metres of freshwater.

However, in its 1.5°C scenario, Irena projects that the world will have to produce 166 million tonnes of green hydrogen and 81 million tonnes of blue by 2040, while 2050 will see 493 million tonnes of green and 30 million tonnes of blue.

Such amounts would require water withdrawal for hydrogen production to more than triple to 7.3 billion cubic metres by 2040, and more than quintuple to 12.1 billion cubic metres by 2050.

Other organisations, such as the Rocky Mountain Institute (RMI), have suggested that this overall increase in the global hydrogen sector’s water use will likely be offset by a decrease in water consumption from incumbent production, refining and use of fossil fuels.

However, water scarcity is often a localised problem — which could hinder hydrogen project development in some of the best sites for renewable resource or carbon storage.

Irena notes that around 12.3% of the 1.7 million tonnes a year of operational green and blue hydrogen production capacity is located in highly water-stressed areas, ie, where the ratio of withdrawal to available renewable surface and groundwater supplies is higher than 40%.

This measure generally indicates increased competition for water, which could make it more difficult for water-intensive clean hydrogen projects to secure planning permission.

Meanwhile, 35.7% of the 56.3 million tonnes of planned capacity is also set to be built in water-stressed regions.

However, some countries expected to produce vast quantities of green hydrogen production are likely to have more projects built in water-stressed regions than others.

For example, 99% of India’s operational and planned clean hydrogen projects are likely to be built in areas under extreme water stress, compared to 56% and 19% of the project pipeline in China and the EU, respectively.

However, Irena warns that by 2040, rising global temperatures and an increase in water demand across sectors could create water stress in areas that do not currently experience it.

Transitioning from coal vs gas

Whether a region is moving away from coal or gas as a feedstock for hydrogen could impact how much extra water is needed for clean H2.

In its examination of existing coal-based chemicals plants in the Yellow River Basin in northern China, where more than half of all hydrogen produced today is made primarily from coal gasification, Irena suggests that adopting CCS for all current production capacity would increase water withdrawal and consumption by around 77%.

“If coal-based hydrogen production were to be replaced with SMR+CCUS, alkaline electrolysis or a mixture of both, the Yellow River Basin would be able to produce more hydrogen with less water withdrawal in 2030 than in 2020,” the report adds.

However, while SMR with CCS would decrease water withdrawal from the Yellow River Basin by around 160 million cubic metres per year compared to current levels, it would also slightly increase consumption from 520 million cubic metres to 600 million.

Replacing coal gasification with alkaline electrolysis, meanwhile, would increase hydrogen production by 11%, decrease water withdrawal by 28%, and cut water consumption by 20%.

However, Europe — which currently produces most of its hydrogen from SMR — is likely to see much more water used for clean H2 production amid increasing water stress.

The report highlights that Europe saw its driest year in 500 years in 2022, which massively disrupted energy throughout France (as temperatures for cooling water in nuclear reactors grew too high), Italy (due to lower hydropower output) and Germany (as low river levels hindered barge-based coal transport to power plants).

And while most blue hydrogen projects are set to be built in the UK, the Netherlands and Norway — which see relatively low water stress levels — many of Europe’s green H2 facilities are planned in highly water-stressed countries, such as Spain and Portugal, due to their sunnier (and drier) climates, which are good for solar power.

Irena adds that the continent’s current 7.5 million tonnes of annual hydrogen production withdraws 150 million cubic metres and consumes 132 million cubic metres of fresh water.

“In the years until 2040, while hydrogen production rises by about 243% from 7.5 [million tonnes] to 25.7 [million tonnes], the sector’s total water withdrawal and consumption could increase by 419% and 334%, respectively,” the report warns, although this is mainly due to a shift from grey hydrogen to blue H2 and alkaline electrolysis.

The desalination question

Some regions proposed as potential hydrogen export hubs, particularly in the Middle East and North Africa, may have limited access to fresh or groundwater but plenty of access to seawater, which can be desalinated for use in H2 production.

Desalination is only expected to add $0.02-0.05 to the overall cost of producing a kilogram of green hydrogen, according to the report. And although Irena describes the process as extremely energy-intensive, the RMI’s analysis suggests that it only uses an additional 1% of the energy input for electrolysis.

Oversizing new desalination capacity built for a project could also supply extra water to local communities.

However, the report warns that brine by-product from desalination and thermal pollution (ie, seawater used in cooling systems being discharged at a higher temperature than the surrounding water) from hydrogen production could disrupt aquatic ecosystems.

While a number of blue hydrogen projects have been proposed in the Middle East, with producers planning to draw on cheap gas, Irena argues that alkaline and PEM electrolysis will need less seawater for cooling and fewer volumes of desalinated water.


The Irena report makes a list of recommendations, as follows:

  • Green hydrogen projects should be prioritised for future hydrogen development.
  • Water-related impacts and potential risks need to be carefully evaluated in hydrogen production development plans, particularly in water-stressed regions where stringent water use regulations must be established for the sector, and enforced.
  • Retiring fossil-fuel-based hydrogen plants and replacing them with green hydrogen should be prioritised in hydrogen development plans, particularly in areas where water is already scarce.
  • Water withdrawal and consumption should be considered as performance indicators of hydrogen production projects for pre-operational evaluation purposes and be metered and monitored during operation.
  • Regulations and financial incentives should favour projects demonstrating higher efficiency in energy conversion and water consumption.
  • More investment and research are required to improve the efficiency of commercial-scale electrolysers and reduce the consumption of fresh water for cooling.
  • Hydrogen production projects in regions where water is already scarce should be incentivised to use water-efficient cooling technologies such as air cooling.
  • In present and future freshwater-stressed coastal areas, utilising seawater for hydrogen production and cooling processes should be incentivised, even as regulations for thermal pollution and brine management are enforced.