Nuclear Risk Economics

What is the true cost of nuclear power — when you include the risk of accidents?

In 2010 India made private companies liable for these risks — which revealed that the technology is commercially unviable:

The Indian law limits the operator’s liability to Rs 1,500 crore  but in case of an accident, victims could find shelter in the law of Torts that can potentially lead to unlimited damages to be claimed from equipment suppliers.  John Flannery, outgoing President and Chief Executive Officer of GE in India said GE will rather give up business than play within India’s civil nuclear liability rules.

“If the [civil nuclear] liability law stays the way it is, we won’t pursue the business.”

Flannery knows other companies, French and Russian, are nosing ahead in the race but points out that they are backed by the government.

He is very clear about his own: “We are a private enterprise and we just can’t take that kind of risk profile”

GE's position tells the truth about the cost of nuclear risk that all the pro-nuclear industry reports will hide.

I bet the French and Russian Treasuries aren't happy about being exposed to these risks (hopefully GE's position will wake up their governments to the risk).

In 2017, France's President Macron also commented sceptically on the cost of nuclear, saying,

    “Nobody knows the total cost for nuclear energy. I was minister for industry and I could not tell you.”  

The secrecy over costs is partly a consequence of being dependent on a nuclear weapons program.  And although Macron is "generally considered to be pro-nuclear", France then committed to a gradual phase-out of its reliance on uneconomic, subsidised nuclear power (which commenced in 2020 and hopefully will be completed before its ageing, corroding reactors have a major radioactive leak).

I think an insurer would take a hard-nosed look at the industry track record (number of global accidents per power station years of operation), rather than swallow claims about the latest technology being much safer than the likes of Chernobyl (which will have to be managed for centuries), because I recall they said that about the Fukishima technology (before its accident), which, 6 years after the initial 2011 disaster, continues to pollute the entire Pacific Ocean and still poses an extreme risk (with record levels of radiation causing continued reactor damage & even killing clean-up robots).  As the HBO mini-series on Chernobyl tells (notwithstanding some artistic licence), without the fortunately successful emergency action that occurred (including worker sacrifice), it and perhaps also the Fukishima accident (see also here or attached) could have produced dramatically worse explosions, and even in 2019 the situation at Fukishima seemed likely to require deliberate discharge of accumulated radioactive water into the ocean (confirmed by Japan’s environment minister & started in 2023, even though this seems to be the cheap rather than safe option) and is still at risk of further accident, for example from another tsunami.

The initial Fukishima disaster was caused by a 45m high tsunami, but compare that to the following tsunamis listed in this video:

• • 30m, Indian ocean, 2004

  • 85m, Japan, 1971

  • 100m, Japan, 1972

  • 250m, Italy, 1963 (which demolished a dam & 5 villages)

  • 524m, Alaska, 1958 (the most destructive tsunami ever recorded).

At Fukishima it will take at best decades to even locate, let alone remove & make safe the 600 tonnes of highly-radioactive molten nuclear fuel, which depends on inventing new technologies — i.e. there's currently no way of doing it!  Given past disaster sites at Windscale and Chernobyl have still not been made safe after more than 50 years (& surprise, surprise, were much worse than admitted at the time), it seems quite possible that another larger tsunami could obliterate Fukishima before it can be made safe.

Then in case you think Fukishima & Chernobyl  were unique, there's the ongoing radioactive leaks at Sellafield (formerly Windscale) – Europe’s largest nuclear waste treatment facility and most hazardous nuclear site – and the risk of an accident there "worse than at Chernobyl", thanks to cracked & crumbling buildings and failings ranging from nuclear safety to asbestos and fire standards that get “shoved firmly under the rug”, as UK authorities “run something so dangerous on a shoestring budget and without transparency”.

And then there's the risk of failures to cooling power supplies, or an explosion in poorly-managed waste fuel storage pools.  The next nuclear disaster is only a matter of time.

In the absence of any convincing evidence and track-record of superior safety, I also estimate that if the nuclear industry was rapidly expanded in order to achieve the required economies of mass production and to materially combat global warming – with, say, a tripling of the number of reactors over 30 years (adding 30 new reactors p.a., as per the attached xls) – then this major expansion of nuclear power would almost guarantee a very serious accident within the next 10-30 years, and — ignoring even more expensive & less commercially viable 'fast-breeder' reactors, which have been abandoned all over the world (even in France) — would also deplete global uranium resources within the same period, before the end of the planned lifetime of new plants built!  Of course a growing excess of demand over supply of uranium will push up prices, which will hurt the viability of nuclear power even more.

In any case, with the time it takes to build nuclear plants, we have simply run out of time for them to be of any use tackling climate change within the urgent timescale we now face.

And though Thorium reactors may have the potential to be less dangerous and more sustainable as a longer-term power source, they're still far from safe or benign, and after 50 years of failed commercialisation efforts, claims of being possibly better than something that's an unmitigated environmental & economic disaster hardly qualifies the technology for investment-grade status!  So if the private sector won't finance its development – which they assuredly won't – why on earth should taxpayers when there's no prospect of any public benefits compared to renewables and the chances of it being cheaper than solar & wind are now basically zilch? (Whilst the Chinese government may be pursuing Thorium resources and reactors, the one company that was pursuing Thorium reactors – Lightbridge – has now switched to focus on traditional uranium fuel, for which they've also failed to get government subsidies.)

But even excluding these risks and long term radioactive waste issues (on top of the ongoing cancer-causing impacts of radiation for people – children especially – living near power stations during routine operation), the cost of just constructing nuclear power plants repeatedly turns out to be many times what proponents claim, and vastly more than renewables + storage (several nuclear projects are over 7 years late, with costs multiplying by up to 10 times original estimates and averaging twice initial budgets — far worse than for other energy technologies), so just about the only advocates for nuclear these days are vested interests like the Minerals Council of Australia (whose members just want to dig up & flog uranium) and right-wing lunatics who seem to back it just to "generate another fight with the green left", even though the average economic loss for a 1GW nuclear plant has been about Є5bn — which is why in 2017 Germany, the US, France & others were all moving away from it (with Germany's nuclear plants all shut down in 2023, fully offset by renewables while fossil fuels continued to decline).

Undeterred, the latest self-serving, fantasy promise from the nuclear industry is to achieve low costs through mass production of "Small Modular Reactors" (SMRs).  But even if you believe their BS about low costs (see also here, here & here), SMRs seem likely to produce much more radioactive waste (two to thirty times more, including new types of more radioactive waste), and safety remains a serious issue, no matter what reactor type you choose (whether LWR, HTGR, liquid metal/"PRISM", Thorium or supposed "Waste-Annihilating Molten-Salt Reactors" — which actually don't reduce pollution or nuclear weapons proliferation risks as claimed).  Writing in July 2023, the former chairman of the US Nuclear Regulatory Commission notes that most SMRs are just designs, very few have been demonstrated and none are commercially available (& likely never will be), let alone licensed by a nuclear regulator.  He partly blames recent "hype around new nuclear power technologies that so far, largely, don’t exist and will likely be very costly", on, ”tech bro” or “technofideism” culture that valorizes new technology, whilst ignoring the reality that new nuclear power is actually many decades away from having any measurable impact on climate change – if at all (which, given their history of failure, is extremely unlikely).

Much of the madness driving the recent enthusiasm for SMRs stems from illogical desperation of failing industries, backed by parochial, short-termist & corrupt politics seeking government subsidies (& to avoid real action on climate change), and the self-interest of egotistical "billionaires surrounded by a chorus of sycophants who reinforce their opinions rather than challenge them as data changes" (data demonstrating the now vastly superior economics of renewables, even including network integration & storage costs).  And now in November 2023, it looks like the nonsense fad of nuclear SMRs is over before it began, as the first planned small nuclear reactor in the US has been cancelled due to a combination of the ongoing dramatic reductions in the price of renewables, plus the usual nuclear cost blowouts — tripling in $/MW over 3 years, before they even started.  So having blown US$600m of taxpayers' money, the proponents, NuScale, now want to pursue SMRs in countries with less safety regulations & more subsidies!

Meanwhile the Financial Times writes that the first new US nuclear reactor in three decades may also be its last, as Georgia Power's "Vogtle" units started entering service in 2023, seven years late and costing $30bn compared to an original budget of $14bn.  The FT notes that, "The only reason there’s a nuclear renaissance is because the federal government is throwing tens of billions of dollars at nuclear”, whilst “Investors aren’t interested”.  Despite all the enthusiasm, in 2023 net global nuclear power capacity went backwards, and this seems likely to continue.  From 1990 to 2025, the mean age of the global nuclear power fleet increased from 11 to 32 years, and 88% of this ageing capacity is projected to close by 2050, which means the industry faces a daunting challenge just to avoid going backwards.

In Australia, nuclear power is a completely unrealistic non-starter that would likely cost at least 50% more than "firmed renewables" (including batteries or other storage, even with up to 90% renewables), even it were feasible in any remotely useful timeframe (and more likely 2-3 times more costly, and still more for the first reactors installed or if solar reaches its ultra-low cost targets).

As for the ever-distant pipe dream of fusion power (which isn't the unlimited source of free power it claims to be), perhaps the artificial intelligence that researchers now hope will help them will more quickly conclude and recommend that we stop wasting so much money on this misrepresented & overhyped concept that is so technically challenging (if not completely flawed) as to be economically delusional! (and totally irrelevant to solving climate change problems in the required timeframe)

What total environmental and economic madness the nuclear power industry is!  No rational investor would continue to chase the illusory nuclear dream, and if we don’t put it to bed soon, it will start to generate real nightmares.

Much as I may dislike the modern fad of blaming everything bad in society on “toxic masculinity”, I can't help think that irrational macho attitudes are at least partly to blame for the billions of $ wasted on nuclear power — along with other big science projects like particle accelerators (which can't answer questions about quantum or modified gravity, and so instead aim to prove the non-existence of made-up particles), or fantasies of Mars colonisation and most of the military.   

However, whilst politicians in Australia's COALition parties profess support for nuclear power as an excuse for not tackling the fossil-fuel industries they're beholden to (or as a purely political diversion), elsewhere the only real reason for building nuclear power stations has been to develop nuclear weapons — as is clear from the UK's 2024 nuclear power plan and its absurd pursuit of "Hinkley Point C", which, like every other new nuclear-power station construction, has vastly exceeded its already-massive budget & schedule and has contributed to the French nuclear industry's “slow descent to hell”.  Yet nuclear weapons are no justification either, since no-one with any morality would use such weapons anyway, so how can they seriously be considered a deterrent?  (We need to establish a "Ministry for National Security" to develop a more sensible and strategic approach to national security issues that challenges current warmongering and military waste.)

So it's good to see that increasingly — except when nuclear power is propped up by subsidies from states with political/military motives — renewables are leaving nuclear behind, where it belongs, in the shade of solar.