Hacker News Reader

    building a fission reactor takes one or two decades 
    and its kWh is costlier than alternatives

It didn't and doesn't need to be that way. Obviously we need strong regulatory oversight over nuclear power, but a big part of the cost is the need to satisfy incredibly hostile regulations imposed by politicians who are (a) pandering to public fear (b) heavily influence by the fossil fuel industry.

Also, talking about kWh cost in the short term is... missing a large portion of the point. Burning fossil fuels is only "cheap" if all the long term damage is ignored. Let's talk about how cheap it is once we start truly paying the price for climate change to the tune of billions of lives and many quadrillions of dollars.

> > building a fission reactor takes one or two decades and its kWh is costlier than alternatives

> It didn't and doesn't need to be that way [..] a big part of the cost is the need to satisfy incredibly hostile regulations [..]

(Alleged over-)regulation is only part of the story.

"analysis, done by a team of researchers at MIT, is remarkably comprehensive. For many nuclear plants, they have detailed construction records, broken out by which building different materials and labor went to, and how much each of them cost. There's also a detailed record of safety regulations and when they were instituted relative to construction. Finally, they've also brought in the patent applications filed by the companies who designed the reactors. The documents describe the motivations for design changes and the problems those changes were intended to solve."[0]

"while safety regulations added to the costs, they were far from the primary factor. And deciding whether they were worthwhile costs would require a detailed analysis of every regulatory change in light of accidents like Three Mile Island and Fukushima"

[0] https://arstechnica.com/science/2020/11/why-are-nuclear-plan...

So how are other countries e.g. South Korea building them faster than the US, yet not suffering nuclear accidents?

I presume the problem is not a lack of engineering talent in America.

I think you underestimate how much lobbying the nuclear industry does.

Here's the former head of the US Nuclear regulator talking about how he worked to reduce safety regulations to make it easier to build nuclear plants. Now he thinks no new nuclear power should ever be built: https://www.washingtonpost.com/outlook/i-oversaw-the-us-nucl...

It stands to reason that the nuclear lobby is out for its bottom line, but let’s not lose perspective on just how devastatingly deadly fossil fuel combustion is by comparison, killing an estimated 8.7 million people per year [1]. That’s 1 in 5 of all deaths globally.

Our perceptions of risk are massively skewed by the (literally) explosive nature of nuclear disasters compared to this silent holocaust to which we’re shockingly normalized.

From Our World in Data [2]:

> Nuclear energy, for example, results in 99.9% fewer deaths than brown coal; 99.8% fewer than coal; 99.7% fewer than oil; and 97.6% fewer than gas. Wind and solar are just as safe.

That’s per unit of energy generated.

Curiously, while most can likely name Chernobyl and Fukushima (perhaps fewer Windscale and Three Mile Island), what of the Banqiao Dam disaster, which killed an estimated 171,000 people the 1970s?

All that said, extrapolating the lethality of nuclear generation to a world with many more nuclear plants is fraught, precisely because there are so few data points.

There’s no escaping the fact that these are incredibly complex and expensive machines, which can fail in unexpected ways, no matter how scrupulously they’re designed to be passively safe — especially when compared to a solar PV park.

[1]: https://www.sciencedirect.com/science/article/abs/pii/S00139...

[2]: https://ourworldindata.org/safest-sources-of-energy

> building a fission reactor takes one or two decades

> and its kWh is costlier than alternatives

> It didn't and doesn't need to be that way. Obviously we need strong regulatory oversight over nuclear power, but a big part of the cost is the need to satisfy incredibly hostile regulations imposed by politicians who are (a) pandering to public fear (b) heavily influence by the fossil fuel industry.

That is not true, there was an HN submission which broke down the cost of nuclear construction (am on mobile and can't easily find it right now) and cost is largely dominated by construction cost, which to a large degree (>50%) are the same as a regular thermal power plant. Regarding regulations, the nuclear lobby is actually very strong, they even managed to reduce regulations for the steam-generating cycle compared to other power plants (I think this was in the US).

> Also, talking about kWh cost in the short term is... missing a large portion of the point. Burning fossil fuels is only "cheap" if all the long term damage is ignored. Let's talk about how cheap it is once we start truly paying the price for climate change to the tune of billions of lives and many quadrillions of dollars.

But the comparison is not to fossil fuels, the comparison is to renewables. If renewables are cheaper and faster (which is the case) they will enable us to move of fossil sources faster than nuclear, so the overall emitted CO2 is less.

Fuel diversification is important too. That's something Germany seems to have forgotten when they went all in on natural gas.

Not only are the costs comparable to a carbon thermal plant, the generating equipment is identical, and the cooling needs are also identical.

Solution: build nuclear cores as close to existing coal plants as possible, shut the coal plant down, and move everything from the steam boiler to the wires to the nuclear plant.

Repeat until no coal.

The cooling needs on a LWR are actually a bit higher, per unit of electrical power output, because the steam temperature is lower than that of a coal plant. The discrepancy is much larger for combined cycle plants.

> a big part of the cost is the need to satisfy incredibly hostile regulations imposed by politicians who are (a) pandering to public fear (b) heavily influence by the fossil fuel industry

People keep on parroting this, but could you list what these "incredibly hostile regulations" entail?

The only "new" thing I know of is the requirement in certain places (like in Sweden) to have ICSS, Independent Core Cooling System, to prevent a Fukushima situation, plus to prevent a meltdown caused by what happened at Forsmark Nuclear Plant in Sweden 2006[1]

ICSS isn't that expensive BTW. Vattenfall cited the cost of adding ICSS to their 5 reactors to about 3 billion SEK in 2020. That's about 300 million US dollars with today's exchange rate.[2]

[1]: https://analys.se/wp-content/uploads/2015/05/forsmark-incide...

[2]: https://www.world-nuclear-news.org/Articles/Swedish-reactors...

>ICSS isn't that expensive BTW. Vattenfall cited the cost of adding ICSS to their 5 reactors to about 3 billion SEK in 2020. That's about 300 million US dollars with today's exchange rate.

$300m seems pretty expensive to me as a cost to add to what is already the safest energy source in the world by terawatt-hour produced[0]. One-fifth the death rate of rooftop solar. 0.025% as dangerous as oil. Every terawatt-hour of energy a coal power plant produces results in as many deaths as one hundred Fukushima "situations."

[0]: https://www.statista.com/statistics/494425/death-rate-worldw...

Rubbish.

Considering a single reactor costs €11-19 billion[1][2] to build in Western Europe currently (Olkiluoto 3, Flamanville (we haven't seen the final bill for that one yet)), an additional 300 million dollars is a drop in the bucket and not the thing that will make the project go from viable to nonviable.

[1]: https://en.wikipedia.org/wiki/Olkiluoto_Nuclear_Power_Plant#... (final)

[2]: https://en.wikipedia.org/wiki/Flamanville_Nuclear_Power_Plan... (projected, not final)

Step 1: Nuclear reactors are so expensive already, additional costs to increase safety barely matter in their overall price! This additional regulation isn't enough to make a viable plant nonviable.

Step 2: Nuclear reactors are so expensive it makes no sense to provision new ones when renewables are just around the corner! Just keep the current coal power plants running while we take another decade to increase solar grid capacity by a few terawatts.

Step 3: Go to step 1.

See also; heap fallacy. Seriously, coal power generation is so bad that if we had to reduce safety regulations to the point that we were having a Chernobyl-level meltdown every month to replace all coal with nuclear plants, we would be significantly better off for it. It's not even close. We could literally completely deregulate safety of nuclear power plants and be safer overall.

(Per the stats I shared earlier, coal power kills 100,000 people per thousand terawatt-hours produced. The world produces roughly 44,000 tWh of coal energy, resulting in 4.4 million deaths per year. Casualty estimates of Chernobyl vary wildly, but even the most pessimistic estimate produced by Greenpeace, avowed anti-nuclear activists that they are, only totals 200,000. Coal power is almost twice as bad as having a Chernobyl every month)

The problem is that the sun is not always shining (and half the year not at the right angle), and wind is not always blowing. Unless a country has access to always-on sources (like sea currents), energy must be stored or things must be burned. Storage is hard and energy intensive to build, and many countries would need a few months worth of storage to never bother about burning stuff again. And that’s _a lot_, you wouldn’t believe the kind of power needed to support some branches of industry, and they won’t shut down in winter of course.

Storage is certainly not harder to build than nuclear power plants. Power-to-Gas is technologically quite straightforward. The research that's currently happening is just to make it cheaper until we have enough renewables for grid-scale storage to make sense. The really hard part of becoming carbon neutral is the sectors other than electricity, e.g. heating and transportation, but nuclear power won't help you there.

Why do you think fission can't help with heating? Heat is one of the by-products of running a reactor.

E.g.: https://www.wsj.com/articles/nuclear-power-could-heat-your-h...

https://www.powermag.com/district-heating-supply-from-nuclea...

Because the problematic part is switching millions and millions of buildings from gas and oil to either heat pumps or district heating, not generating heat or electricity.

Power to gas (incl reverse) is terribly lossy right now.

Surplus renewable electricity is incredibly cheap. Losses matter a lot less than the cost of the infrastructure you need for storage. Batteries have great efficiency, but they're not cheap. We already have a bunch of infrastructure that can handle gas.

If your grid is large enough (and that is already happening in Europe for economic reasons), you have enough geographic distribution to average out variations. There have been studies that showed you could run the US on something like 500% overcapacity with a fully integrated grid using only renewables and no storage.

Moreover nuclear are slow moving, they typically don't load follow, so even with a combined nuclear/renewables you still either need significant overcapacity or some sort of peaker. So you haven't actually solved the variation problem.

Finally, because cost for nuclear is largely dominated by capex (construction cost, both in dollars and CO2 foodprint), not running the nuclear plant as close to capacity as possible will even more increase the price and also reduce the CO2 lifetime emission. In usual comparisons which puts nuclear on par with renewables, nuclear is assumed to run essentially 24/7 while solar/wind are based on some statistical uptime. If we operate a nuclear not close to capacity its lifetime carbon footprint becomes significantly worse.

> Moreover nuclear are slow moving, they typically don't load follow

Which is not entirely true: https://www.oecd-nea.org/upload/docs/application/pdf/2021-12...

Even older designs could do load following.

The problem isn't really to what extent nuclear is load-following, but that the economics of nuclear look even worse than they already do if we want to follow the cheapest generator.

To be blunt: if the sun is shining and it's windy, no-one really wants to buy a nuclear plant's output. Not at an agreed fixed price, or possibly at any price.

The idea of nuclear getting paid the same price - or worse, an index-linked price - for the lifetime of the plant, regardless of what the future holds, and even on those sunny and windy days, just seems horrendously anticompetitive.

If nuclear is as necessary, competitive and flexible as some make out, then go right ahead and build your plant(s). Just don't expect taxpayers to underwrite anything.

> if we want to follow the cheapest generator.

Energy isn't just about generation

> If nuclear is as necessary, competitive and flexible as some make out, then go right ahead and build your plant(s). Just don't expect taxpayers to underwrite anything.

By the same logic taxpayers shouldn't underwrite any renewables: they are significantly slower than nuclear, and have literally zero base load capacity.

> By the same logic taxpayers shouldn't underwrite any renewables: they are significantly slower than nuclear, and have literally zero base load capacity.

The cost for offshore wind projects has fallen so fast in the UK that the many of the latest projects don't need subsidies, see this report from Imperial College (London)[0], in fact they'll be paying the government, see this article from Bloomberg.[1]

I'm pretty sceptical about the phrase 'base load', when it comes up, such as in a HN discussion[2] from a last week, it seems to be used to describe wanting to choose slow and/or expensive power plants.

EDIT: See also this[3] recent HN discussion, in which it was pointed out "California has put emphasis on renewables and if the nuclear power station isn't guaranteed to provide base load then it's too expensive to operate"

[0] https://www.imperial.ac.uk/news/200353/offshore-wind-power-c... [1] https://www.bloomberg.com/news/articles/2022-01-13/high-powe... [2] https://news.ycombinator.com/item?id=32152588 [3] https://news.ycombinator.com/item?id=31610996

> The cost for offshore wind projects has fallen so fast in the UK that the many of the latest projects don't need subsidies

I'll have to check that. Sensational news usually omit quite a few important details.

> I'm pretty sceptical about the phrase 'base load', when it comes up, such as in a HN discussion[2] from a last week, it seems to be used to describe wanting to choose slow and/or expensive power plants.

No. It means that wind and solar literally produce zero output when there's no wind or there's no sun.

However, life goes on: trains run, homes are being heated, businesses operate, factories produce goods. This is base load.

So, to cover a wind farm or a solar farm that produces inadequate power you need to get that power from somewhere.

Where from?

And this is the question that renewables enthusiasts just brush off as not important.

"We can store energy". No. We can't. Not in the amounts required.

"We just build more". How much more? Does this "more" remain cheap then? Yea, you can transmit it from afar, but it's no mean task in itself. And not cheap either.

But sure. "It's all about keeping slow power plants". What the hell does "slow" even mean in this context, when we're talking about wind/solar which can't even be properly used in load following precisely because they are extremely slow.

Building wind and solar to 500% overcapacity sounds like a great waste of resources and space though.

And makes nuclear look a lot more cost-competitive by comparison.

But currently solar/wind are ~3x cheaper than nuclear and falling rapidly. So the 500% overcapacity would likely in the end cost the same as 100% capacity of nuclear. However, with nuclear you need at least 200% capacity as well (maintenance, hot days, not being able to load follow fast enough). So which one is the waste of resources?

Long-distance transmission is at least as slow to build as nuclear, at least in the US. There are projects that have foundered for decades.

> building a fission reactor takes one or two decades

This is not universally true. Looking at South Korea's construction times for example[1], you'll see that it's averaging between 5-7 years per reactor, all the way into the 2000s and 2010s. Japan shows similar numbers[2], and they're currently in the process of restarting their nuclear investments following the accident at Fukushima Daiichi. Same story for China[3].

[1] https://en.wikipedia.org/wiki/Nuclear_power_in_South_Korea#B...

[2] https://en.wikipedia.org/wiki/List_of_commercial_nuclear_rea...

[3] https://en.wikipedia.org/wiki/List_of_commercial_nuclear_rea...

I feel you are cherry-picking those South Korean numbers. If you look at the latest ones that started construction in 2009-2012, it's taken 10 years (and some are still not started). That's "one decade". Extrapolating the construction time inflation, you could imagine any future developments will take longer still.

> What's the point?

Energy density.

Oh look, a metric no one cares about.