Hacker News Reader

Comment by nonrandomstring | Hacker News Reader

I'm interested in the "modular" aspect here. It's not so clear what this means, but my must hopeful take is:

- core technology can be switched (pop in a thorium core when it's available, leaving all the cooling, turbines and electrical as is)

- better maintenance, swap out parts with short downtimes

- interoperability, add a cooling system from another manufacturer like TeraPower or even a Chinese or Russian firm (after this war nonsense ends). If we're going to counter climate change with nuclear it must be a global effort.

- easier, safer decommissioning. No need to carefully demolish 5 acre concrete bunker sites, just tow away old parts for disposal at a safe place.

Anyone know what "modular" really means in this context?

View on HN

None of the above. AFAIK, 'modular' here is shorthand for '(mostly) assembled on site from modules made in factories'. The idea is that if you can transform nuclear build-out from a civil engineering problem into a manufacturing problem you can massively lower costs if/when you reach some level of scale.

That was the same song Westinghouse was singing. It ended catastrophically.

No it didn't.

"Westinghouse Electric Company would file for Chapter 11 bankruptcy because of US$9 billion of losses from nuclear reactor construction projects. The projects responsible for this loss are mostly the construction of four AP1000 reactors."

"As of 2019 all four AP1000 reactors in China are operating."

Westinghouse made a large amount of mistakes in their designs, suffered through political climate because of the Fukushima incident and economic hardship because of the ridiculously low gas prices.

How can you reduce that basically to that they tried to mass produce power plants, that was dumb so they failed?

And where's the catastrophe?

China has ditched the AP 1000 technology now. They are building no more of those reactors. So even in China it was a relative failure.

I'll add that there's good reason to think the data from China about nuclear projects being completed on time is invalid. There are cases where at the official start date for construction on some of their plants there was already much visible work that had been completed. Great way to be on schedule, just delay when you say you actually started.

The catastrophe was the financial implosion of Westinghouse and the great damage it did to Toshiba.

I won't dispute that Westinghouse was a failure. They failed at designing a power plant that meets safety criteria for it to be build in the US. Toshiba bet they could do it efficiently, and lost 9 billion on that bet.

None of that has anything to do with Rolls-Royce. There's no reason to assume just because they're trying to solve the same problem that they're going to make the same mistakes and fail as well.

They very well could, and even if they did, your comment would still be useless.

What it has to do with Rolls Royce is that Westinghouse claimed they were going to get cost improvements from modular construction. But they utterly failed at that. The takeaway is that claiming you are going to reduce costs, and actually doing so, are very different things. Talk is cheap, especially without a history to show the words can be relied on.

I think it just means that the whole reactor itself IS a module, can be removed, put on a truck, repaired, upgraded, and replaced as needed. Here is a Wikipedia link that has a picture of one. https://en.wikipedia.org/wiki/NuScale_Power

Since convection depends on absolute height, reactors relying on convection have a certain minimum height. Roughly speaking. So if you scale a reactor down, at some point it becomes very skinny and that might lead to problems. Anyway, 60 megawatts electricity is not too big in my opinion.

I think the biggest contributor is shrinking the size of the engineering effort. The cost scaling curve for large infrastructure projects is often not beneficial - the idea is that a (50%) smaller reactor requires substantially smaller containment, on-site development, etc, which should reduce costs by far more than 50%. More parts can be build in a factory rather than on-site, and higher unit volumes can support a learning curve.

To me modular means being able to buy one or 10 of these off the production line, plug them in and turn them on. A fantastic way of converting coal or gas power plants as they age or otherwise become non-viable, because a big part of the necessary support infrastructure is already there. All it needs is the manufacturing production line to actually exist, and the fuel supply chain and waste management which is unfortunately mostly political and the biggest risk to this sort of project.

Reply on news.ycombinator.com