The Coalition’s nuclear plan: does it add up?

Nuclear reactor at sunset
Nuclear reactor at sunset

The Coalition has announced a plan to replace Australia’s outgoing coal-fired power stations with nuclear reactors. But does it make sense?

Professor Ken Baldwin

ANU Research School of Physics

The new Coalition plan to build seven nuclear power plants to replace Australia’s ageing coal-fired power stations could be viewed as a way of leveling the energy transition playing field, and of opening up discussion on options for the nation’s decarbonisation strategy.

Indeed, a truly level playing field would require a significant change: removing the present prohibition on nuclear power legislated by the Howard Coalition government in 1998 - a reform often called for by technology-neutral commentators.

However, even were this social license granted to a future government, there remain three burning questions about the new policy. Will this be in time to contribute to the energy transition? Will this cost the taxpayer more than the renewable alternative? Do these nuclear power station locations make sense?

The time

The Australian Energy Market Operator (AEMO) projects that all coal-fired power stations will likely have retired by 2037 - 90 per cent of them in the next decade.  Under this scenario, solar and wind will have replaced all coal-fired power stations by 2037.

That could be a conservative estimate.

The AEMO 2024 draft Integrated System Plan states: "Coal retirements may occur even faster than these forecasts. Ownership has become less attractive, with higher operating costs, reduced fuel security, high maintenance costs and greater competition from renewable energy in the wholesale market... Replacement capacity must be put in place well in advance."

Given this 2037 timeframe, we should examine what the global experience tells us about construction times for nuclear power stations.

In western countries with appropriate regulatory frameworks, recent construction times have far exceeded a decade. In countries like the United Arab Emirates with different regulatory and governance standards, it’s under nine years. It’s not just the construction times that Australia needs to consider. Before any nuclear power plant can be built here, we would first need to establish a regulatory system. That could take up to five years.

The Coalition plan states that nuclear plants "will start producing electricity by 2035 (with small modular reactors) or 2037 (if modern larger plants are found to be the best option)." 

Even if this unlikely proposition were possible, this effectively means that nuclear won’t replace coal-fired power stations. Most, if not all, of those coal-fired stations would be gone by the time that nuclear comes online, and coal would have been replaced by renewables. 

This negates the argument that ’baseload’ nuclear generation will replace ’baseload’ coal. It will have been replaced by renewables ’firmed’ (supported) by storage to deliver flexible supply, which large thermal power stations struggle to provide because they are difficult to fully ’ramp’ up and down quickly.

The cost

According to the Coalition’s plan, after 2037 nuclear will battle with renewables for the remaining slice of the electricity market. Will it be able to compete economically? 

The answer is no.

The CSIRO/AEMO GenCost study shows that, even including the cost of additional transmission and storage required for solar and wind farms, renewables are almost half the price of large-scale nuclear - and even cheaper still compared to more expensive small-scale modular reactors (SMRs).

There’s good reason why Australia hasn’t followed the lead of the 32 other countries in adopting nuclear power. With our more abundant solar and wind resources and greater land area, we don’t need to.


Finally, does it make sense to place nuclear power stations in the Coalition’s chosen locations?  Certainly, plans for offshore wind make the same argument that they can be located close to retiring coal-fired power stations such as in the Latrobe Valley, to utilise existing transmission systems.  Nuclear makes a similar argument about re-using the cooling water opportunities left by coal.

Here it is helpful to examine the proposed nuclear power locations, their power capacity in gigawatts (GW) and their current projected end-of-life date:

  • Liddell, NSW - already closed
  • Mt Piper, NSW, 1.4GW (2040)
  • Loy Yang VIC, 2.2GW (2035); 1.1GW (2047)
  • Tarong, QLD, 1.8GW (2037)
  • Callide, QLD, 0.7GW (2028); 0.8GW (2050)
  • Northern, SA (SMR only) - already closed
  • Muja, WA (SMR only), 0.5GW (2029)

On a 2037 timescale, 5.2GW of coal power stations will be retired, with a further 1.4GW (Mt Piper) before 2040, 1.1GW (Loy Yang B) by 2047 and 0.8GW (Callide C) by 2050. That’s a total of 8.5GW.  However, as predicted by AEMO, these closures may happen many years earlier under the unrelenting competition from renewables, meaning that this coal capacity will have already been replaced by renewables. 

The Coalition then plans to progressively introduce five full-scale nuclear reactors (at around 1GW each) plus two SMRs (at around 0.5GW each) starting in 2037 and 2035 respectively, by which time all the present 21GW of coal capacity will have already retired and been replaced by renewables.

Meanwhile, AEMO projects around 120GW of extra capacity will be required by 2037 to cope with the electrification of transport, households and industry, meaning that the projected 6GW of nuclear will be swamped by much cheaper renewables being installed on an industrial scale to cope with this extra demand.

This 6GW of nuclear will make very little difference to the total energy landscape - and at the expense of introducing a much more costly technology.

The Coalition plan is silent on what remains the greatest challenge for Australia’s energy transition - how do we decarbonise the last few per cent of the electricity sector with fully flexible generation post-2040, to ensure electricity reliability during a week of windless, cloudy weather?

Options include: vastly overbuilding renewables and storage; using gas with carbon capture and storage or with carbon offsets; or using small modular reactors.  Which will be the cheapest option is an open question. But it’s unlikely to be addressed by introducing expensive, less flexible large-scale nuclear before then, on a scale that won’t be needed.

This article was co-published with ANU Policy Brief .