The debate about nuclear power in Australia flickered to life with the suggestion by the New South Wales deputy premier, John Barilaro, that small modular reactors could be constructed in regional centres. Prominent backbencher Barnaby Joyce followed up with a call for a parliamentary inquiry.
It would perhaps be churlish to observe that the small reactors advocated by Barilaro exist only as designs and may never be built. There is a much bigger obstacle which is essentially impossible to overcome.
To make the central point as bluntly as possible: even with a crash program there is no chance of deploying nuclear power in Australia in the required timeframe. I looked at this question in a submission to the South Australian royal commission into the nuclear fuel cycle and concluded that “there is no serious prospect of Australia producing nuclear energy before 2040”.
That was in 2015, and the news for nuclear power since then has all been bad. All of the nuclear power plant construction projects under way in the developed world have experienced substantial delays (the VC Summer plant in the US has been cancelled with a loss of billions of dollars).
Most of these projects (Flamanville in France, Olkiluoto in Finland and Vogtle in Georgia) received their initial approval around 2005, and none is now likely to start before 2020. So, to be sure of getting nuclear power going by 2040, we’d need to have projects in their initial stages before 2025, in the term of the next parliament.
To see how absurd this is, consider some of the steps that will be needed before a project could begin.
First, both major parties need to be convinced of the case for nuclear power. That’s highly unlikely but let’s suppose it can somehow be done by 2020. Next, the current ban on nuclear power needs to be repealed. This ban looms large in the minds of nuclear advocates but actually it’s such a minor problem we can ignore it.
The first big problem is the need to set up, from scratch, a legislative and regulatory framework for nuclear power. That would require adapting an overseas model such as that of the US, where the nuclear industry is regulated by at least eight different acts, covering more than 500 pages. Back in the 1970s the French government could do this kind of thing by fiat, without parliamentary debate, but that’s not a feasible option for Australia.
Having passed the necessary legislation, the next task would be to establish and staff a regulator similar to the US Nuclear Regulatory Commission or the UK Office for Nuclear Regulation. The only Australian body with any relevant expertise is the Australian Nuclear Science and Technology Organisation which operates a 20 megawatt (research reactor at Lucas Heights. Ansto has little or no capacity to deal with the problem of licensing and regulating commercial reactors of 1,000MW.
Even with a massive effort, and assuming no political obstacles, it’s hard to see these tasks being accomplished within five years, which would already take us to 2025. But there are many more remaining difficulties.
Most obviously, as the Industry Super report states, we would need a carbon price or a market mechanism with similar effects, such as an emissions trading scheme. On any realistic political analysis, that’s impossible – the overlap between supporters of nuclear power and advocates of carbon pricing in Australia is virtually zero. At a minimum, the adoption of a carbon price would require a change of government at the next election, which may happen though it doesn’t seem likely at the moment. Even if it would occur (and assuming, improbably, that a Labor government relying on Green support could be persuaded to back nuclear power), there would be further delays before the carbon price could be put in place.
But that’s just the beginning. Before any project could be considered, it would be necessary to license designs that could be built and operated here. The processes of the NRC in the US, which were expedited in the hope of spurring a “nuclear renaissance” typically take three to four years.
We could simply accept the judgement of overseas regulators, but even then we would have a problem – there may be no designs available.
In my submission to the SA royal commission, I argued that the only serious contender for Australia was the Westinghouse AP1000. Since then, however, cost overruns and cancellation have sent Westinghouse bankrupt, almost taking its owner, Toshiba, with it. There is no serious prospect of any more plants of this design being built. Areva, which is building its EPR model in Europe, is in similar difficulty. There’s a serious risk that the only contenders would be Chinese or Russian designs, which would pose some obvious problems.
The most difficult step would be the need to identify greenfield sites for multiple nuclear power plants, almost certainly on the east coast, and go through the relevant environmental processes. Reliance on overseas models won’t be of much use here. All the plants under construction in western countries are “brownfield”, that is, situated next to existing plants, built last century, and approved as far back as the 1970s.
In summary, it would be a heroic endeavour to get construction started on a nuclear plant even by 2030. Getting it finished and generating electricity by 2040 is virtually impossible.
Fortunately, there are alternatives, though the Industry Super report dismisses them. The combination of solar photovoltaics and battery storage is already cheaper than new coal-fired power. As a backup, Australia has huge potential for storage using pumped hydroelectricity. We don’t need to call on the phantom of nuclear power to secure a reliable, carbon-free electricity supply for the future.