On nuclear, Coalition prefers the optimism of misleading, decade-old, unverified claims
The Coalition is a fan of quoting the Australian Nuclear Science and Technology Organisation's optimism on nuclear timelines compared to the CSIRO. But do the numbers add up?
To the extent that most Australians have heard of the Australian Nuclear Science and Technology Organisation (ANSTO), we know it as a supplier of radio-isotopes for use in medicine and as the operator of a small research reactor at Lucas Heights in the suburbs of Sydney.
So, it may have come as a surprise to hear shadow energy minister Ted O’Brien cite ANSTO as the source for an estimate that a small modular reactor (SMR) could be constructed in three to five years, and a large reactor in eight to 12 years.
Appearing on the ABC’s 7:30 report in mid-March, O’Brien stated “that is the advice from ANSTO. That is the advice of the Albanese government’s nuclear agency”. In view of the fact that widely publicised advice from an extensive study undertaken by CSIRO yields much less optimistic conclusions, that seems like a surprising claim.
But O’Brien is correct. ANSTO is indeed the government agency officially advising on nuclear technology, including nuclear power.
Section 5 of the ANSTO Act mandates the organisation provide advice on aspects of and the application and use of nuclear science and nuclear technology. ANSTO provides such advice to government, parliaments, ministers, departments and agencies, inquiries and investigations, members of the public, and international, multilateral and bilateral partners — in pursuit of the national interest.
In a submission to the Senate standing committee on environment and the communications inquiry into the environment and other legislation amendment (Removing Nuclear Energy Prohibitions) Bill 2022, ANSTO stated that SMRs “have the potential to reduce build costs using a variety of strategies, including reducing plant build times from six to eight years for large reactors to two and a half to four years for SMRs via the use of series-production methods“.
These numbers are even more optimistic than those cited by Ted O’Brien. But terms like “potential” can do a lot of work in claims of this kind. Nuclear fusion, for example, has the potential to meet all the energy needs of the planet, but it won’t do so any time soon.
A natural response from an interested member of the public would be to visit the ANSTO website to get more detailed information on the assessment of nuclear technology. This leads us to a webpage titled “What are small modular reactors and what makes them different?”, which leads with the claim “the USA is expected to have its first SMR operating by 2026” and includes the timeframe of three to five years for construction.
A note hastily added in the last week states: “Please note that this content was current at the time of publishing (July 2020), and the projected construction time of SMRs (three to five years) is referenced from a University of Leeds research paper. In November 2023, NuScale [the subject of the 2026 claim] announced it was discontinuing its SMR project in Idaho.”
Even in 2020, this research was out of date. The NuScale project, originally projected to be delivering power in 2023, had already pushed its target past 2026 by then. But given that the project has been abandoned, there’s no need to look too closely at this.
The University of Leeds paper is more interesting. It turns out to be a literature survey covering the period 2004-19. The three- to five-year estimate for the construction time for SMRs is taken from a non-peer-reviewed 2016 report by consulting firm Ernst and Young (which worked with one of the authors on the University of Leeds study). The information used to compile the report is even older, going back to 2014 or earlier. To put it bluntly, this is worthless.
Rather than complying with its legal obligation to keep abreast of nuclear power technology and inform the public of its findings, ANSTO has relied on decade-old, unverified claims, made by a consulting company. This sloppy treatment of an issue that should be a central focus of ANSTO analysis contrasts sharply with the careful assessment undertaken by CSIRO.
I went to ANSTO for a response but didn’t hear back.
Ted O’Brien can scarcely be blamed for taking ANSTO’s word on these matters, particularly when its claims are so convenient to his case. But ANSTO needs to retract its misleading claims as soon as possible. That would give the LNP an opportunity, if it wants it, to drop its nuclear policy and put the blame on an Albanese government agency for misleading it.
One final irony. The ban on nuclear power, which is now the subject of so much controversy, was introduced by the Howard government to secure the passage of legislation that allowed ANSTO to build a new research reactor at Lucas Heights. In light of this history, maybe ANSTO’s remit should be revised to steer the organisation clear of nuclear power once and for all.
Read my newsletter
Nuclear proponents, Including Dr Alan Finkle (https://www.theguardian.com/commentisfree/2024/mar/22/heres-why-there-is-no-nuclear-option-for-australia-to-reach-net-zero) often state the safety record is comparable to that of wind and solar in terms of deaths per unit of energy produced. What they don’t mention is the cost of making nuclear safe. Arguments that fewer safety measures are needed founder on the fact you can't test to failure and therefore we can't know the precise level of safety margins needed. Add to this the high consequences of getting risk analysis wrong and no-one in the free world are prepared to fully accept those risks.
I also had a look at the ANSTO website recently and was disappointed to find cut-and-paste content in the nuclear power section which could have come from the industry lobby. There was no detailed description informing the public on the state of the art in Nuclear power technology, no detailed lists of various reactor technologies the number of producing reactors and prototypes etc.
The notion that we MUST be able to get usable energy out a substance tens of thousands times the energy density of coal is highly compelling, so it’s no wonder people keep trying. The problem is nearly all technological development depends on trial and error, but with reactor technology error can kill you. You can’t “fail upwards” Elon Musk style.
Before you even build a prototype, you have to have your have an orders of magnitude higher degree of certainty of the probability for failure than for any other technology, and for every aspect of the system because you can’t know how a design will function until it’s functioning. New, previously unknown potential points of failure can then crop up when you are building the prototype, running the prototype, building the working reactor and running the working reactor.
Faults or malfunctions in an active reactor core can’t be directly repaired so you have to make sure nothing goes wrong before you load up the fuel, and if something does go wrong, hope that it is the kind of problem remote repair systems are designed to cope with, that it will resolve itself, that, if all else fails, it won’t result in a leak or explosion and the reactor can be shutdown and decommissioned safely. All of this means a precautionary principle must be applied to design and construction simply due to the fact you can’t test a reactor to failure, at least, not on purpose.
Wow. You did a great job tracking the Coalition's claim back to its underwhelming source. I'll climb on board the nuclear bandwagon if the numbers stack up but for now they very clearly don't.