In an era of population growth, extreme weather events and innovative but thirsty industries, how can Australia hope to stay within its water budget?
At first glance, the rampant growth of data centres, particularly in and around Sydney and Melbourne, appears to have little to do with the supply and demand of water. But dive a little deeper and the connection becomes clear.
Australia has more than 260 data centres, according to an ABC report. Around a third of these are in Sydney. And more are being developed “at a rapid rate”, said Professor Katherine Daniell MIEAust, Director of ANU’s School of Cybernetics and Deputy Chair of Engineers Australia’s National Committee on Water Engineering.
In fact, the use of AI and cloud computing will necessitate an extra 175 new facilities by 2030, the Australian Financial Review has reported.
“We did research around one of the more efficient data centres in Western Sydney,” Daniell said. “Our approximate calculation showed that on a day over 32°C, that data centre would use the equivalent of 1000-2000 households worth of water.”
Clearly, there is an increasingly complex interrelation between water and other sectors, including:
- Technology: the more we rely on it, the more water it consumes, particularly for cooling
- Energy: the largest extractor of water from the environment
- Agriculture: the largest consumer of water
- Economics: businesses rely on water for their survival
- Society: people rely on water for life, dignity and wellbeing
That interconnectedness means that, on the world’s driest inhabited continent, a careful balance must be engineered as we move forward, particularly as weather events become more powerful and longer lasting.
“In extreme circumstances we effectively turn agricultural water off, except for certain perennial crops,” Daniell said. “We introduce restrictions so people can’t fill their pools or wash their cars. Golf courses and parks don’t get water. But can you turn a data centre off?
“In extreme moments we’re going to see this system cracking, and we have to figure out how many locked-in systems can’t be turned off. We may face a choice between water for firefighting, for drinking and for keeping our data centres on. What do we sacrifice if we haven’t already designed these solutions into the system?”
Competing demands
The production of green hydrogen requires somewhere between 15-110 L of water per kilogram of green hydrogen, depending on various factors, said Dr Kate Holland, a CSIRO Principal Research Scientist and Groundwater Management Group Leader.
Holland led a study into green hydrogen production that found it wasn’t the cost of the water that made certain plants unfeasible. It was the inability to supply the enormous amount of water.
“We had a look at small-scale developments,” Holland said. “They could potentially use existing surface water subject to water planning, or groundwater with some treatment.
“But once you got up to the larger scale hydrogen hubs, you were just looking at desal for supply. The current technology is a really big water user.”
The modelling results depend on such factors as the quality of the water and the cooling technology. Data centres are a similar story.
“In data centres, cooling technology really matters. They’re not in remote regions, so water supply is typically more reliable. What is a problem is the stress they put on water systems.”
That’s where engineering comes in, Holland said. There are choices that can be made around the development of data centres in different geographical locations that can share the load if others must be switched off, for example. And in cooling technologies, greater innovation is required.
For example, the CSIRO-developed groundwater cooling system for the Pawsey Supercomputing Centre in Western Australia involves pumping cool water from a shallow aquifer beneath Kensington in Perth. That water runs through an above-ground heat exchanger to cool the supercomputer, before it is reinjected underground. The geothermal process is estimated to have saved 14 million L of water in its first two years of operation alone.
One other emerging competitor for water, Holland said, is critical minerals.
“Critical minerals will involve an integrated extraction and processing method,” she said. “That potentially requires a lot of water for processing. Each region has its own set of aquifers and its own climate. It really matters where water comes from at a regional scale.”
Intervention strategy
Across Australia’s landmass are countless local weather regions. The simplest way to slice the country is north/wet and south/dry, Daniell said. From there, water researchers can narrow their data down by state, region, city, suburb, coastal, inland, riverine, arid and more.
There is no one-size-fits-all approach to water knowledge, meaning nothing is more powerful than local knowledge.
Data from the Australian Bureau of Statistics (ABS) tells us that Australia extracts around 70,000 GL annually from the environment, much of which is returned. Of this, around 13,000 to 15,000 GL is consumed and not returned.
At a national scale, that understanding is useful. But the most impactful knowledge comes from understanding at a localised resource level how water is being used, and whether regions are at risk of exceeding their budgets.
There are various ways to do that. The Murray Darling Basin has a series of accounting methods, sustainable diversion limits, and other targets and objectives. It has a market mechanism to allocate resources to the most impactful places, overlaid with policy and regulations.
Just as important as policy and regulation is engineering. Water is a space in which engineering innovation and intervention can have a significant and positive effect as weather patterns continue to change.
Here are some top engineering interventions that will shape Australia’s water future.
1. Desalination innovation
Desalination generates one unavoidable byproduct: brine. But engineers such as Jerome Douziech, Vice President of Operations and Performance at SUEZ, see this not as waste but as an opportunity.
“There is a research trend gathering momentum around recovering minerals from the brine,” he said. “For instance, could we recover the minerals that we need to rebalance the water, or minerals that have a value on the market, like magnesium?
“If we could do that, we would not only reduce our carbon footprint, but also reduce our dependencies on other parties.”
2. Managed aquifer recharge and water banking
Just as the energy sector has made advances in storage with big batteries, water has also made moves into storage by recharging aquifers, which eliminates evaporation.
“Water banking can increase water security for regional communities and industries during drought,” Holland said. “However, a key challenge is navigating the rules for how we manage this water stored in each aquifer in each state and territory across Australia.”
3. Reuse of industrial water
Whether for production of green hydrogen, data centre cooling, water use around gardens and parks of new suburbs or even recycled drinking water, industrial and other wastewater could be applied more effectively to help solve challenges, Holland said.
“Our reuse, looking through the ABS numbers, is only around 2 per cent – higher in South Australia and WA, probably because they’re the driest. We can do much better.”
4. Circular wastewater treatment and resource recovery
Sewage treatment plants are beginning to evolve into resource recovery centres, said Adam Lovell, Executive Director of the Water Services Association of Australia. The further engineers can take this idea, the better.
“Wastewater is not waste. It’s a resource,” Lovell said. “That mindset shift underpins the infrastructure we’re building now.”
Integrated systems
The overwhelming message from experts is that everything is connected to water, and water is connected to everything.
“My advice to engineers is to broaden their perspective, because no engineering today is isolated,” Douziech said. “You need to have a larger skill set, be curious and open in your perspective, and keep learning.”
We’re at the point where engineers can’t just think about “pipes and pumps”, Lovell said. They’ve got to consider cities, geographies, data, customers, carbon and resilience – because everything connects. Engineering in water has become systems work.
“It’s not enough to optimise a single asset or a single system anymore,” Daniell said. “You need to understand how the whole system interacts, including the social, ecological and technological. The future of engineering is not just about materials or hydraulics. It’s about governance, behaviour and adaptation.”
READ: Installing Western Australia’s next major water source
Changing the flow
An interconnected view is beginning to be mirrored in government frameworks, where decisions increasingly link water security to energy transition, housing growth and climate adaptation.
The National Water Initiative (NWI), currently under review by the Federal Government, is seeking to address some of those systematic integration needs in a modern way.
What we know of the NWI so far makes it clear that water policy can no longer sit in isolation, Lovell said. It connects to housing, energy, climate, digital and more.
“How do we weave AI into a water utility’s data?” Lovell said. “How do we use urban stormwater or recycled water to keep the urban areas green and healthy? How do we break down bureaucracy and manage challenges from an engineering and technical perspective?
“We believe the upcoming national policy reform is a major step forward. It enables everybody to get on the same page and, in the face of climate change, start acting quickly on water supply security, on flood risk management and on a range of weather extremes, which the current NWI is really not designed to do.”
Supply and demand
How much water do we have in Australia? How much water do we use annually, and how? And is it possible to blow the budget?
For water usage data, most experts refer to the Water Account Australia, published by the Australian Bureau of Statistics (ABS).
The latest report, from 2021-22, said the total volume of self-extracted water – water extracted directly from the environment, including rivers, lakes, groundwater and desalinated seawater – was 66,205 GL, an increase of 3 per cent on the year before.
However, it’s important to note that “self-extracted” does not mean consumed. Much of the self-extracted water, for hydroelectric plants, for example, is returned to the same place it was taken from.
Households consumed 1773 GL, the rest extracted by industry. The lion’s share (49,853 GL) goes to electricity and gas supply, with water supply, sewerage and drainage services the next most thirsty sector, at 11,899 GL.
Agriculture, forestry and fishing use just 2555 GL, but these sectors represent the largest consumers.
How does this compare to the total available?
It depends on the weather in various regions, but during the 2021-22 period, the ABS said, we extracted 93 per cent of our total surface water supply, 4 per cent of groundwater and 2 per cent of potential seawater for desalination.
Could we over-spend on our water budget?
That’s absolutely possible in local areas, particularly in drought years and without fallback plans and local management strategies. Those in the Adelaide Hills are no stranger to such challenges, as water carters had to be brought in during early 2025, with some households reporting a wait of several weeks.
This story was originally published in the February 2026 edition of create with the headline “Thirsty work”.
Desalination plants are subjected to complex hydrodynamic forces, making accurate load prediction essential. Discover an alternative approach to this process.
