Wind and solar aren't guaranteed sources of energy. Could nuclear power fill the reliability gap?
Professor Ken Baldwin
ANU Research School of Physics
Nuclear power has popped in and out of Australian public discourse for over 50 years, and was banned in 1998 under a Howard government deal with the Greens. Public opinion has also fluctuated over time, with a recent 2022 poll in favour of removing the ban on nuclear power, just as the Federal Opposition this year put nuclear power back on the political agenda.
In 2024, Australian public opinion is firmly for combating climate change by replacing fossil fuels with clean energy. The big question is: is there a role for nuclear in Australia’s energy transition?
Considering costs
Australia is the world leader in renewables , installing solar and wind at the fastest rate per capita of any nation.
Renewables are cheaper than any form of new-build energy, even accounting for the storage and transmission needed to ensure reliable supply. The CSIRO Gencost project that annually reports the levelised cost of electricity (the lifetime costs of a power plant divided by energy production) is based on the best Australian and global data applied to Australian conditions - and shows that nuclear is the most expensive option.
The key determinant of any future role for nuclear in Australia is timescales. Even given the social licence to build a regulatory framework which might take five years, it might take another 10 years to construct a nuclear power station.
By the late 2030s, the Australian Energy Market Operator’s Integrated System Plan (ISP) projects that almost all’Australia’s aging coal-fired power stations will have retired, replaced by renewables with perhaps a small amount of flexible gas power in the system to guarantee 24/7 operation given the intermittency of solar and wind.
This leaves no room for large-scale nuclear power stations, whose lunch will literally be eaten by abundant cheap solar during the middle of the day, and by wind at night. The economics of solar and wind will become increasingly affordable, while most international experience shows that nuclear costs are increasing.
The bottom line is that in Australia, large-scale nuclear will be outcompeted by renewables by the time it is built, so will become a stranded asset on 2040 timeframes.
The renewable reliability gap
But that’s not quite the end of the story.
As shown in the , the last few percent of decarbonising the Australian electricity system by 2050 will be very expensive to guarantee 24/7 supply when a week of cloudy, windless weather might sideline the great majority of renewables.
A limited number of options present themselves:
- Vastly overbuild solar, wind, storage and transmission so that somewhere in the country, renewables/storage are always operating
- Use gas turbines with (as yet unproven at scale) carbon capture and storage
- Produce hydrogen using water electrolysis and renewable electricity, then store and burn the hydrogen in turbines to create electricity when needed (an inefficient cycle)
- Nuclear energy
All of these are very expensive, and it isn’t clear which option will be the cheapest in the 2040s when the last few percent of the electricity sector is being decarbonised.
The nuclear option
If we pursue the nuclear energy option, it’s not just a case of building one reactor to help cover those last few megawatts of power that would be needed when the sun doesn’t shine and the wind stops blowing.
Several reactors would be required to guarantee operation during both scheduled refuelling and maintenance, and unscheduled downtime. Typical large-scale reactors produce around 1,000 megawatts (MW) of electricity, several of which are likely to be more than what is needed for this purpose. Further, large-scale reactors are not easy to fully ramp up and down quickly to fill gaps in variable renewable generation.
So large-scale nuclear power has no realistic prospect of playing any role in Australia - either for coal replacement by 2040, or as a backstop for renewables beyond.
This raises one final prospect: building nuclear reactors at smaller scale - so-called Small Modular Reactors (SMRs).
These reactors can be designed to be fail-safe, often require less downtime, and generate up to 300 MW each. While the levelised cost of electricity might be higher than for a large-scale reactor of similar capacity, there are potential advantages.
A bank of SMRs could be sized to provide cheaper backup for renewables than a several large-scale reactors. SMR facilities could be built gradually, with other modules added as needed that can be switched out to respond to renewable variability. This incremental approach would reduce the ’project risk’ compared to a large-scale reactor, thereby reducing the cost of finance.
Component manufacturers would benefit from economies of scale and the learning rate from manufacturing hundreds of SMRs globally would drive down the cost more rapidly than for large-scale nuclear. However, while there are many designs for SMRs there are very few in operation : one in Russia based on an old marine design, and a prototype in China - both produced by governments with little available cost information. An SMR project approved for construction in the US was recently terminated because of cost overruns following COVID supply chain price increases, providing the sole data point for the CSIRO Gencost study. So there’s some uncertainty as to whether SMRs are cost competitive with other options.
Nevertheless, if there’s even a small chance that nuclear power could fill the reliability gap in a 100 per cent clean energy system, a technology-neutral approach requires that all’options compete on a level playing field.
This is a strong argument ( currently favoured by public opinion ) for removing Australia’s legislated ban on nuclear power, so the nation can evaluate the best option without one hand tied behind its back.
This is an edited version of an article that appeared at ANU Policy Brief .