The prospect of Australia turning to nuclear power has been little more than a politically radioactive thought bubble – until the Coalition this week confirmed it wants to put reactors at the sites of soon-to-be-closed coal plants.
Energy experts have previously derided the idea, saying some of the technologies being touted did not exist, and that nuclear would be too slow, too expensive and unnecessary in a country with so much free solar and wind available to harness.
But as the Coalition promises to take a pro-nuclear policy to the next election, the prospect of Australia trying to build nuclear reactors raises many questions.
Could it even work, or would it be an economic and logistical disaster?
What’s being proposed?
The Coalition is understood to be looking at both conventional large-scale nuclear reactors and small modular reactors (SMR) which are not expected to be available commercially until the early 2030s, or potentially later.
The Coalition has said more details will come before its reply to the federal budget, expected in May.
What about timing?
The Australian Energy Market Operator’s latest draft plan to develop the national electricity market (that’s everywhere except Western Australia and the Northern Territory) says under the most likely future scenario, 90% of coal plants will have retired by 2035, with the rest gone by 2038.
That means governments and electricity generators are already having to work at breakneck speed to add enough renewable energy and storage, such as batteries and pumped hydro, to make sure Australia can meet its obligations to bring down greenhouse gas emissions while delivering the cheapest and most reliable electricity system possible.
Those decisions are being made based on the technology that is available now, and based on economics. Renewables backed up by storage are the cheapest option, according to the vast majority of experts.
The Coalition argues building reactors on the sites of old coal plants would avoid some of the costs of building new transmission lines and the storage needed to allow for the times when wind and solar production is low.
Tony Wood, the director of the Grattan Institute’s energy program, says by the time a hypothetical nuclear plant could be built and operating – which he says could be 20 years away – “the coal-fired power has been closed down and you’re in deep trouble”.
“I just think the economics and the timescales don’t fit.”
But Wood says Australia should remove its ban on nuclear.
Future technologies such as SMRs “if they’re available” could be worth exploring to deliver the final 10% or so of electricity generation after renewables and storage have filled most of the electricity system, he says.
“It’s purely about economics and timing that make it unrealistic as something for today, but we should have a conversation about it,” he says.
When could reactors be built in Australia?
A Coalition government, if elected, would have to agree regulations to govern a brand new industry and create new government institutions such as the US Nuclear Regulatory Commission to run the industry.
Dr Dylan McConnell, an energy systems analyst at the University of New South Wales, says: “Recent nuclear power projects in comparable countries have experienced considerable delays and cost overruns. There is little prospect of a conventional nuclear power plant delivering electricity before the 2040s.”
Construction on two nuclear units in the US started in 2009 and were seven years late and reportedly US$17bn (A$25bn) over budget when they started generating power last year.
In the UK, French company EDF’s 3,200MW Hinkley Point C plant began construction in 2017 and, after several revisions, the company says it may not be delivering electricity until 2031. Initial cost estimates of £18bn (A$34bn) have been revised as high as £46bn (A$89bn).
A deal to buy the electricity using a surcharge on UK customer bills has put “an albatross around the neck of UK consumers for a very long time”, says Woods.
And who would build them?
Glenne Drover, the secretary of the Victorian branch of the Australian Institute of Energy and a broad supporter of nuclear power, thinks the biggest potential block for nuclear in Australia would be finding overseas companies to actually build reactors.
At last year’s Cop28 climate talks, more than 20 countries pledged to triple nuclear power capacity by 2050. The books of many nuclear builders are filling up.
“Australia could do what the United Arab Emirates did and go out with a tender process, but who would take that up?” he asks.
“Would we be happy if one of those companies was Chinese or Russian? Kepco [a South Korean company] maybe, but could you get them?”
Drover argues Australia should continue on its path building renewables but “keep all options open” and continue to review nuclear’s appropriateness for Australia.
What about water use?
Drover says nuclear power stations use a comparable amount of water to existing coal-fired power stations for cooling.
“Water is a valuable resource,” he says, but argues that for a small number of reactors enough non-drinkable water would be available, as it has been for coal plants.
Could nuclear work in Australia’s grid?
McConnell says conventional nuclear plants need to be almost constantly running to be economically viable. But as we head towards 2040, Australia’s energy system will be dominated by renewable energy and storage, such as batteries and pumped hydro.
This would be a “very challenging power system for nuclear power to economically operate in”, says McConnell.
“Some of the challenges facing the economic viability of coal generation are directly transferable to nuclear power generation.
“Low-cost renewable generation is eroding both the utilisation and market value of coal generation and would have similar effects on a nuclear generator.”
But large nuclear reactor units could also create a headache for the electricity grid. The biggest single generating source on Australia’s grid is a 750MW coal unit in Queensland.
“The typical size of conventional nuclear units today is larger than any existing coal units in the national Eeectricity market,” says McConnell. “The loss of such a single large unit represents a risk to the operation of the electricity system. This risk would need to be managed and would come at some additional cost.”
McConnell questions if the sites of old coal plants would be available to buy. “The grid connection points are valuable assets,” he says. “Old power station sites are already being developed to make use of these valuable connection points.”
What about costs?
Without knowing what kind of reactors might be built, and how they would be paid for, and the price of establishing an industry, it’s difficult to know the future costs of nuclear. But it’s unlikely to be cheap.
For SMRs, the CSIRO gives a theoretical range of $382 to $636 per MWh in the year 2030, compared with $91 to $130 for wind and solar. CSIRO has so far not provided estimates for conventional nuclear but, using US data, nuclear is among the more expensive power options.
John Quiggin, a professor of economics at the University of Queensland, says without a carbon price, nuclear plants would need very large taxpayer subsidies in Australia.
He says one of the only countries to have recently started a nuclear industry is the United Arab Emirates that drew up its first nuclear policy in 2008, commissioning South Korean company Kepco to build four 1,400MW units.
Quiggin says these four reactors will likely have cost the UAE as much as $100bn – enough money to put a large solar system on the roof of every Australian house, he says.
The first of the country’s four nuclear power units came online in 2020 and a fourth is expected to start producing electricity in the coming months – all between three and four years later than expected.
That’s a 16-year process in a country that, without a democratic system, can make arbitrary decisions to get plants built. It is not a good comparison for what might happen in Australia.
“When the economics of this comes up in Australia, it will look very bad,” he says.
