Nuclear Power

Mainstream: It accounts for why most net-zero pathways keep nuclear, and for why grids with a lot of wind and solar still value round-the-clock supply.

Alternative: It accounts for the overruns at Vogtle, Hinkley and Flamanville, and for the fast growth of solar and batteries.

Mainstream: Repeat builds of the same design should get cheaper and faster. Some have (Korea, China, and later units at Vogtle and Barakah). Western repeat builds and SMRs have not yet shown it at scale.

Alternative: The claim fails if nuclear programmes deliver on time and at costs comparable to renewables plus firming. Korea, China and 1970s France did so, which counts against a universal claim.

Mainstream: The promised cost declines from standardisation and SMRs are still mostly unrealised in the West. Meanwhile the stated reason for building (firmness) is rarely priced against cheaper storage.

Alternative: It underweights the cost of firming very high renewable shares, and it treats Western overruns as inherent even though other countries built faster and cheaper.

Mainstream: It accounts for the low death rate per unit of energy and for the absence of any harm from stored waste.

Alternative: It accounts for the displacement caused by Chernobyl and Fukushima and for public distrust.

Mainstream: Repository leaks, or severe accidents in modern passive-safety plants, would count against it. None has been seen in a modern design, and Onkalo has only just begun operating.

Alternative: A working repository with no problems over decades would count against it. Finland's is only just starting.

Mainstream: The long-term repository claim rests on very little operating experience, and 'political reasons' is used to explain away slow progress.

Alternative: The measured harm from waste and from the Fukushima displacement is small compared with the harm from fossil fuels. Evacuation harms are partly policy choices, and the 'unsolved' label does not fit Finland's repository.

Mainstream: It accounts for overruns in projects that were poorly managed or redesigned during construction, and for cheaper builds where teams repeat a design.

Alternative: It accounts for the rise in US costs over time and for the extra cost of safety margins below very low radiation doses.

Mainstream: Costs should stay high even after regulation is relaxed. A regulatory reform that left costs unchanged would support it. Recent US licensing reforms are too new to show this.

Alternative: Costs should fall markedly once rules are relaxed or fast-tracked. Reforms are only now being tried, so this has not yet been observed.

Mainstream: It does not explain why nuclear costs in the US rose so much over time while comparable safety performance stayed flat. Regulatory process creep looks like part of the cause.

Alternative: Korea and China built cheaply under strict regulators, which suggests that learning-by-repetition is a bigger part of the answer than rule changes. Meanwhile the claim does not say which rules to remove.

Mainstream: It accounts for the need for firm capacity and for the energy-security motives behind the expansion in China, Russia and South Korea.

Alternative: It accounts for the poor fit between inflexible output and grids with a lot of solar and wind, and for the lock-in of large projects.

Mainstream: The policy fails if new plants keep arriving years late and over budget. That could be tested by the next round of builds, and a failure is hard to undo once the money is spent.

Alternative: The policy fails if grids with high shares of renewables cannot get reliable firm supply without large new low-carbon capacity, and if data-centre demand outruns efficiency gains. This is a live test now, and the policy is easy to adjust.

Mainstream: It assumes cost declines that have not yet happened in the West, and the plan is hard to reverse once capital is committed.

Alternative: Demand growth from data centres and electrification is much larger than efficiency alone covers, and the plan has no firm low-carbon replacement for long-lasting needs.