In the olden days (the 1990s), when I worked on Nuclear Engineering International magazine, I couldn’t see why a European utility would go to the trouble and capital outlay of building a nuclear plant. A gas turbine would go up at a fraction of the price and within about 18 months, at which time investors could start getting their money back.

Come the millennium and things seemed different. The overriding need to cut carbon emissions, and the likelihood that we would move to electricity for heat and transport meant nuclear was needed. The amounts of power needed meant that a fleet of nuclear units would likely be required, along with renewables (at the time at relatively high prices and low volumes).

Now times have changed again. Renewables look like a good bet for bulk supply and, what is more, digitalisation means a flexible partnership with the demand side. That puts nuclear out of the money again.

My point is not whether nuclear is the right or wrong choice. The answer to that has changed over those years.

No, the important point is that the lead nuclear plant under discussion during all three phases – getting on for three decades (yes I do feel old, thanks very much) – is the same one: Hinkley Point C.

It was planned as the second in a series of four reactors in 1990 and still largely waiting for construction decades later.

That’s slightly unfair: we gave up on new nuclear for a while and it’s a mere 15 years since we came back to it. But that makes the comparison with renewables even starker. Offshore wind is true series build. That industry has seen many hundreds of turbines installed, reach maturity, be superseded by new designs, and drive down costs dramatically – all since the Hinkley Point revival started. 

We’ve all been hearing for decades that the way to reduce nuclear costs – like any other technology – is to build a series of standardised units. Yes, more or less true. But that only works if you build them. The original Hinkley Point C was to be a series-build with Sizewell B, not Flamanville and Olkiluoto.

The fact is, most other technologies can go through several generations while nuclear is gearing up for a single unit. (Also, incidentally, one reason why it is hard to get the promised series benefits from the Swansea Bay tidal lagoon. The only slower projects than nuclear are those that involve building large concrete dams in sensitive environments.) 

What is more, non-nuclear technology generations are speeding up, while nuclear evolution gets slower as it becomes more of an outlier in a decentralised industry, and harder to finance and deliver. 

This is why it’s a good thing that the National Infrastructure Commission has turned nuclear conventional wisdom on its head with its proposal to build nuclear one plant at a time, instead of planning a series. Who cares whether series build is cheaper if, in reality, it takes decades to build just one or two units? (I understand that people will say that small modular reactors are different. Maybe so: let’s see how far we have got in licensing designs and sites for them in ten years time.)

We are moving forward step by step into a new energy industry and any nuclear plant is a huge step, not just in investment but in system terms. One at a time is the sensible course. It doesn’t preclude those individual projects from overlapping, but it does mean we will cost those single plants properly. And it will mean we can focus on new industries where development can happen faster.

And if you want to know where the industry is really speeding up