Perspective: A Festival of Fallacies

Stanford Engineering adjunct professor Amory Lovins disputes recent claims by Tim Yeo, chairman of the New Nuclear Watch Institute (NNWI), that nuclear expansion is vital for national and global security.

A recent report by the NNWI extols export-led nuclear growth, justified by “security concerns” and “system level contributions” but “no longer ... by unit cost estimates” -- despite claiming nuclear's levelized cost of energy (LCOE) below wind's or solar power’s. NNWI cites a 2019 OECD Nuclear Energy Agency (NEA) and International Energy Agency (IEA) report that copies their 2015 numbers. Yet their 2020 edition, reported on NNWI’s website, says only Japan, Korea and Russia still claimed nuclear beats renewables; 10 other countries -- now 11 since Japan recanted -- found renewables cheaper. The 2020 NEA-IEA report found nuclear the cheapest dispatchable resource, but locally optimized Clean Energy Portfolios obviate classical dispatchability and “baseload” capacity. IEA confirmed utility-scale photovoltaic (PV) is “now the most cost-effective new source of electricity in many countries around the world.” BNEF finds wind and PV the cheapest bulk source for 91% of global generation and below fossil operating expenses for 48%.

NEA’s low LCOEs assume low capital expenditure, construction time three years below and capacity factor around 10 percentage points above global averages, and 2020 US operating expenses 19% below 2019 actual. No standard data set validates new-nuclear costs below unsubsidized renewables: Lazard finds US nuclear LCOE three to eight times higher, BNEF 5-13 times higher.

NEA assumes all technologies pay the same for capital, but private capital typically costs several to many percentage points more for nuclear than renewables. Market experience illustrates why: of 259 US power reactors ordered, just 93 remain operable; by mid-2017, only 28 (11% of original orders), some slated for closure, remained competitive in their regional markets and hadn’t suffered at least one year-plus outage -- an 89% dry-hole risk.

Yet NNWI still finds nuclear “the most cost-effective option” for decarbonizing electricity -- although curiously needing and meriting vast new subsidies -- by conjuring four kinds of high renewable “system costs” from the 2019 NEA-IEA report (much softened in 2020). How? As NEA’s 2015 report warned, “Any assessment of system effects is a complex undertaking, subject to ad hoc assumptions.” Quite.

Such models’ high variable-renewable “system costs” are caused mainly by using the two costliest grid integration tools --giant batteries and interconnectors -- but not eight cheaper carbon-free ones, notably end-use efficiency (now severalfold larger), strong demand response (ditto), accurate renewable forecasting, thermal storage, electric-vehicle, CHP, and dispatchable-renewable integration.

NNWI's Tim Yeo claims no country except France and Sweden “has yet gotten close to providing its citizens with continuously available electricity without significant consumption of fossil fuels.” Actually, 2019 renewables alone surpassed France’s 90.5% non-fossil generation share in 20 countries, exceeded Sweden’s 97.9% in 12, and achieved 100% in six. Denmark, with 0.06% hydro, was 76%-78% renewably powered, Scotland 79% without or 97% with hydro. Renewables also often outpace nuclear growth.

NNWI decries geopolitical risks from energy-transition materials -- an innovation bonanza masquerading as an exaggerated threat; curtails renewable reserves rather than profitably decarbonizing heavy transport and industrial heat; never competes or compares efficient use with new supply; ignores opportunity cost and path dependence; and falsely blames the February 2021 Texas blackout on its energy-only market. (Capacity was adequate but often unwinterized. Renewables underperformed for 10 hours -- at most by 1.4 gigawatts for one hour -- while over 30 GW of gas plants failed. A 1.35 GW nuclear unit was off line for 63 hours. Texas was minutes away from losing another 2.42 GW nuclear plant whose restart might have taken one to two weeks.)

NEA’s and NNWI’s charmingly antique view of grid resilience echoes other central-plant advocates’ discredited claims. Central plants require, but distributed resources bypass, the grid where ~98%-99% of US outages begin. Reactors’ vulnerabilities to climate change are substantial, rising and potential triggers of early retirement, yet NNWI calls only renewables “weather-dependent.” Unmentioned are major accidents’ knock-on shutdowns, like Japan’s post-Fukushima nuclear convulsions; capital-market rejection; nuclear proliferation and vulnerability to attack.

US nuclear advocates warn of Russian and Chinese export dominance; NNWI is unworried. Corruption and debt-trap concerns manifest in South Africa are cheerily dismissed -- another case of “ever more sales-starved vendors, seeking ever less solvent customers, [offering] a risky project the seller can’t finance to a customer who can’t pay.” Prospective buyers like Vietnam are fleeing. So why should the US join its main geopolitical rivals in competitive folly? Leaving their dodgy nuclear adventures to die of market forces should enhance US security.

Most fundamentally, NNWI ignores convincing evidence that nuclear power makes climate change worse than if cheaper and hence more climate-effective efficiency and renewables were bought instead. Its basic thesis of nuclear climate protection is false. But that hardly matters. In 2020, nuclear capacity grew by 0.4 GW net of closures while renewables added 278.3 GW -- 781 times more, or about 220 times more added annual output. Game over.