Australia's Datacentre Boom: Boon or Burden?

On Mamre Road, in Sydney’s outer western suburbs, there are plans to build a “hyperscale” datacentre that will be one of the biggest in the world.

If approved, the 52-hectare site will include six four-storey buildings that stretch 40 metres high, alongside 936 cooling units and 852 diesel backup power generators.

The Mamre Road project is part of an estimated $155bn investment pipeline over the coming decade, amid a worldwide rush to build the infrastructure enabling the artificial intelligence revolution.

There are about 160 datacentres operating in the country now, with another 90 proposed, according to the Climate Council.

Australia is in the middle of a sudden and enormous datacentre boom and not everyone is convinced it’s good for the country.

Alex Hooper, the head of climate and energy economics at Oxford Economics Australia, says the global appetite to invest in datacentres is so large that “in theory we could have any size of datacentre industry we want”.

“There are huge opportunities but we need to be smart. There is some level of datacentre investment that is good for Australia. The question is: how much?” Hooper says.

“Once we have that level of investment here, the next thing we need to ask is what role datacentres can play in our economic growth and prosperity. The productivity argument is there to be made but it’s not a given.”

These are questions that are now preoccupying experts, policymakers and governments at all levels.

Most immediately, the answer revolves around how much power and water Australia can affordably and sustainably provide, and what the nation receives in return.

Datacentres are a key part of the digital economy. They are essentially large, climate-controlled warehouses full of racks of servers and other IT equipment that store and process the vast amounts of data that keep websites, cloud services and artificial intelligence running smoothly.

The Mamre Road datacentre, once operational, would become the country’s single biggest energy user, consuming more power than the Tomago aluminium smelter.

That’s a big reason why Penrith council says it does not want the project to go ahead.

Hooper calculates that datacentres now account for 2.8% of the electricity consumption on Australia’s east coast.

“By 2030 we expect that to be 7%, rising to above 10% by the mid-2030s. There is also significant upside risk to that outlook given the strength of the project pipeline,” she says.

The Climate Council estimates wholesale electricity prices on the east coast could be 20% higher by 2035 if the potential extra datacentre demand is not offset by additional renewable energy sources.

Datacentres also consume a large amount of water for the evaporative cooling that prevents overheating servers, and require big tracts of land – with most located in cities and sometimes too near to homes.

And for all that, they are not big job creators.

Studies in the US show it typically takes thousands of workers in the construction phase, but only hundreds once operational.

Which raises the question: why do we want them at all?

Beth Webster, an economics professor at Melbourne University who specialises in advanced manufacturing, is optimistic that Australia can avoid the pitfalls that have given datacentres a bad name overseas.

“America has revealed all the problems,” Webster says, including not requiring developers to deliver additional energy and water to feed the centres, and placing them too close to where people live.

“I think it’s going to be a win. Foreign direct investment is a very important source of knowledge exchange, and I can’t see too many downsides as long as they have rules around it.”

Pat Bustamante, a senior economist at Westpac, says there is “absolutely no question” that the datacentre boom is good for the economy.

“It’s laying the foundation for the next wave of productivity growth. We saw this during the PC and IT revolution during the late 90s, and this is going to be bigger than that,” Bustamante says.

“It’s during that production and usage phase of the datacentres that you get that productivity boost, and that is going to come next. Once this high-speed computing power is in place, that’s where we expect to see that productivity boost.”

To a large degree, scepticism around whether datacentres will be a boon or a burden for Australia is rooted in broader and deeper community angst about AI.

As Sally Auld, the chief economist at NAB, says: “People and society have big question marks around where AI will take us

“So are we just sowing the seeds of our own destruction by building these things?”

Auld doesn’t believe so. Her research suggests the new technology will augment rather than replace most jobs.

“But it’s hard to know what the net effect of that will be.”

Hooper says policymakers need to develop a proper overarching industrial policy to make sure datacentres are something that works for us, rather than just happens to us.

“We have a lot of interest here to make sure the public sees the benefit, and for that we will need to have a vision of the future,” she says.

“I think there are huge opportunities but we have to be smart.”