Hacker News Reader

Comment by jabl | Hacker News Reader

I'm slightly skeptical of accident tolerant fuels (ATF).

In terms of deaths per kWh generated, nuclear is already the safest form of energy generation we have.

Where nuclear fails, and where improvement is desperately needed, is ability to deliver on budget and within schedule.

What is NOT needed is minor incremental safety improvements at significant cost (e.g. ATF's that the article discusses).

Now, more substantial improvements, e.g. non-LWR designs, passive walk-away safety, modular design, etc., I'm all for those.

View on HN

Hear hear! I think the nuclear industry struggles with finding a balance between sunk costs (e.g. TRISO fuel development under NGNP, the TREAT reactor, ATR, the BISON fuel performance code), and what advances the nuclear industry really needs to survive and thrive. At a NRC meeting a while ago, one of the members of the ACRS asked pointedly what regulatory relief they expect to get from ATF. No one knew. Is the NRC going to relax a bunch of regs when we have ATF and somehow allow nuclear to be cheaper without compromising safety? No way.

We need ways to minimize financing costs during construction, and get the number of staff required at operating plants way down to decrease O&M. My favorite solution? Shipyard-constructed nukes on floating platforms, tugged to locations for power and tugged back to port for maintenance. Solves almost all problems assuming you can get them to operate largely autonomously. In attacks or ship collisions, design it to sink safely and cool itself with seawater. Build in a recovery operation to the design.

Also, the nuclear industry really needs to find ways to consolidate effort. There are literally 50+ SMR designs in work right now, and dozens more larger reactors. The industry is wayyy to small to be fighting over limited investor and government money to develop that many reactors in isolation.

> Shipyard-constructed nukes on floating platforms, tugged to locations for power and tugged back to port for maintenance.

What’s your opinion on Thorcon?

Also I’m wondering if you have some thoughts on this: I’ve been wondering what holds us back from taking a scale out approach rather than a scale up approach we have now (e.g. with EPR design now producing 1.6GW). Instead couldn’t we pump out naval sized nuclear plants in mass production style since we can apparently still build those and my limited understanding is their smaller size and output makes them safer (less decay heat etc). Ship those around as needed for refueling. And just have a couple dozen on each site to produce the equivalent power. It would combine a national strategic need that we have to do anyway (naval nuclear power), with another need that we seem to be failing to do (getting back on the horse with producing reliable clean power).

Thorcon is awesome. I love their style and logic. They're pretty open-door, they have the boldest low-ball cost estimate I've ever seen, and they have a solid set of reasoning for why it's going to be that cheap (e.g. The Tale of Two Ships [1]).

I don't really like that it's fluid fuel though. As I said in another comment, the unknown remote maintenance implications of that are very far from hashed out, and will very likely be terribly expensive for a fairly long shakedown/learning period. Fluid fuel is a good end goal. But don't get me talking about fluid fuel in a near-term first-of-a-kind plant with a low-ball operation/maintenance cost estimate. I also don't like that if they succeed, they have a very large waste stream of super radioactive large-scale components (that get swapped out ever ~5 years and put... where? That's gonna be a problem). Also, working primarily with a country with inexperienced/non-existant nuclear institutions for a very advanced reactor (Indoensia) is probably not going to work. I get the logic though, bootstrap a new nuclear regime without the old regulations. Problem is, fluid fuel reactors have a variety of postulated novel ways to increase source term. That will take work to license. A lot of work. Experimental work, in neutron irradiation.

I just wish they'd try a small light-water reactor with their shipyard construction experience first and take it from there.

I also freaking love that they use LaTeX.

[1] http://thorconpower.com/docs/two_ships.pdf

> e.g. The Tale of Two Ships [1]

Thanks, that's quite a story!

> I don't really like that it's fluid fuel though.

I have some misgivings about this as well. A water-soluble fuel salt in a thin-skinned (compared to a high pressure LWR) vessel on a ship, what could possibly go wrong?

(I'm a fan of MSR's and I'd love to see them deployed, but I'd prefer them on dry land thank-you-very-much)

> As I said in another comment, the unknown remote maintenance implications of that are very far from hashed out, and will very likely be terribly expensive for a fairly long shakedown/learning period. Fluid fuel is a good end goal. But don't get me talking about fluid fuel in a near-term first-of-a-kind plant with a low-ball operation/maintenance cost estimate.

Agreed, but IIUC Thorcon, similarly to Terrestrial, are planning to replace the entire reactor vessel with the primary circuit every few years (5/7/?). So AFAIK the plan is not to do any maintenance of the reactor vessel. Guess you have to hope one of the primary pumps doesn't need a new bearing 1 month after commissioning, then, eh?

> I just wish they'd try a small light-water reactor with their shipyard construction experience first and take it from there.

Yes, though it seems passively safe LWR's are quite massive. Nuscale reactor vessel is a frickin' 700 tons for a paltry 60 MWe. Thorcon reactor is 250 MWe (or is it 500, their materials are somewhat confusing?). Would a passively safe 250 MWe LWR fit on a ship, along with a pressure dome to handle a LOCA?

"Shipyard-constructed nukes on floating platforms..."

Hmmmm... This reminds me of the nuclear armed b52s discussed in "command and control". Are all of the failure modes actually safe? And is building a movable platform contributing to the risk of failure? Feel like for nuclear, KISS is particularly important.

There is a great deal of experience with putting reactors on ships. On the other hand real world experience for putting reactors on planes is very limited, and what little experience there is suggests that shielding would be minimal (shadow shielding for the pilot/copilot only) or nonexistent.

Because nuclear power such a charged issue (pun intended), transparency and trust is especially important.

The nuclear struggles with genuine transparency, which, imho, would go further than anything else to shifting the public's feelings on nuclear power.

Totally agree. Nuclear technology was born in secrecy but civilian nuclear energy stands to benefit from much more openness now. I've been hearing people talking about finding ways around institutional friction via various open-source inspired collaboration and design practices. Could be hopeful.

> Shipyard-constructed nukes on floating platforms, tugged to locations for power and tugged back to port for maintenance.

Such as Akademik Lomonosov?

Precisely. Hopefully larger though, to really get that economy of scale. You can put up to 2 large GW-class LWRs on something much smaller than the Prelude.

deaths per kWh (or even various adjusted life year metrics) aren’t all that matters. Cost of cleaning up and other economic losses due to accidents also matter.

The Fukushima accident may not have caused much in the way of injury, but the cleanup is projected to cost over $100bn. It’s not so easy for nuclear power to compete when those costs are factored in.

So yes, passively safe reactors are a big deal in my book.

Fossil fuel is causing a whole lot more damage per kWh that we have to clean up. BP estimated 61 billion USD to clean up its oil spill in 2010.

I would say nuclear is already competetive when compared to that.

Also, impending climate disaster that's been building for decades. The cost of that is borderline incalculable.

That sounds like an interview estimation question.

Presume total extinction due to climate emergency.

GWP = annual gross world product.

r = estimate of risk that we kill ourselves through other means per year e.g. 1 in 100 due to nuclear war.

Presume zero growth (e.g. world growth is offset by population growth, environmental damage, discount rate).

So:

Cost = GWP / 2r

”Fossil fuels doesn’t change the climate, humans change the climate, we merely provide the means and are not liable for cleaning up this mess.”

"Economic damage of the Chernobyl accident is estimated at $235 billion for 30 years on after the explosion" (source: UNDP)

Sounds low...

That's like arguing internal combustion cars are better than electric cars, because they're cleaner than horses.

The only thing that matters in reality is the investors making a profit. If nuclear out competed other generation on that basis it would be everywhere now, regardless of whether it was safe or not. cf Hydroelectric. A dam destroys more land for a far longer time than Fukushima has left contaminated today. And as for deaths, dam failures have a truly horrifying track record: https://en.wikipedia.org/wiki/Dam_failure 100 people died just building the Hover dam. Yet we are likely to see more Hydro, not less.

You don't see nuclear for one reason: it's high risk with low returns. It isn't high risk because nuclear reactors regularly go bang. It's high risk because a plant has to operate profitability for 40 years in order to make a return, and 40 years is too far out so safely predict anything. 40 years ago there was no internet and global warming wasn't something we knew about. The plant going bang is just one of many things that could render it unprofitable. There could be fuel supply problems (cf Iran), or the demand could move away (eg, a aluminium smelter could close down), or a cheaper technology could come along, or an earth quake, or a tsunami.

The new designs in the article seem to solve some problems. Being tow-able means if the current market folds you can move it somewhere else for example, and being small means there is less money at risk. But geezzz - 4 years to build. In 4 years it is possible the price of solar or wild could drop by 40%.

What is more important than deaths/twh? Costs? I would disagree. Perhaps, area lost to contamination (e. g. Chernobyl exclusion zone, Fukushima area). But then, how much is lost/flooded due to coal excavation, hydro dam flooding, etc.

I don't think amluto said the clean-up costs were more important than deaths/kwh. The point is that these improvements are necessary and welcome even though nuclear is already safer than the options, for reasons other than safety.

Indeed.

If I were a nuclear regulator, I would not permit construction of a plant that could, by design, be a danger to anything other than itself if it lost power. I would also consider requiring existing plants to retrofit themselves to have the same property if the technology became available.

> "I would not permit construction of a plant that could, by design, be a danger to anything other than itself if it lost power."

What about a plant that's a danger to others when it's running optimally within design spec, as every single coal fired plant on this planet is? Do you apply this standard to all power plants, or only nuclear?

I think that, even in an appropriately regulated and taxed environment, coal has a place. In particular, fly ash can replace a respectable fraction of the cement used in concrete. If a coal-fired power plant can scrub its emissions to remove particulates and CO2, it may be cost effective as a source of fly ash compared to the (taxed) cost of cement production.

I'm skeptical that "clean coal" will ever be cost-effective when considered purely as a source of power in most locations.

As far as I'm aware coal (particularly coke) plays a vital role in the production of steel, so I don't expect we'll do away with it anytime soon. But it's very frustrating to see coal being given a free pass by critics of nuclear, particularly when those critics call themselves environmentalists.

One can make steel without using coal or coke. Some steel is made this way even today ("Direct Reduced Iron", particularly the mode where iron ore is reduced with syngas made with natural gas). Hydrogen could be used instead of syngas to do this reduction, and the amount of carbon that would be needed to make the final alloy would be small, and could be sourced from non-fossil sources.

> coal has a place [snip] in particular, fly ash can replace a respectable fraction of the cement used in concrete.

AFAIK fly ash is mostly used as a replacement for other filler (sand etc)... Using fly ash doesn't change the amount of CO2 produced per m3 of concrete.

I am struggling to understand what your point is here.

Also fly ash is generally a single digit percentage by coal weight burnt. So say 90% of coal burns with oxygen, most of which ends up as CO2 (by weight). Using the fly ash waste (from electricity generation) in concrete makes sense. Burning coal to indirectly justify making cement does not.

As I understand it, fly ash can replace a fair amount of the cement, and cement production is CO2-intensive.

Your understanding is correct, fly ash can be used as a substitute for portland cement. It's apparently chemically similar to how the Romans used to use volcanic ash in their concrete mixes. A lot of cement sold these days is a mixture of fly ash and portland cement.

My impression is that fly ash involves the production of more CO2 than burning coal to produce an equivalent amount of portland cement, but it's worth it when you're burning coal anyway and fly ash is basically free as a byproduct. So in practice it essentially offsets CO2 emissions from burning coal, but only partially.

That's just the impression I've gotten though, I don't have any numbers for any of it.

I would say any plant that was built before maybe the 2000s, which is admittedly a lot, should just be decommissioned, with more modern plants replacing them. Our main risks with nuclear plants is that most of the plants are at or around the 50 year point, which not only is about their original planned lifespan, but also absolutely ancient in terms of design and technology. Yeah they have been improved, but only so much as an already built plant can be modified for cheap. When people talk about nuclear safety, they are thinking about plants designed only 15 years after we invented nuclear technology. Talking about the safety of Fukushima, or Chernobyl, or many other plants is the equivalent of discussing the crash safety of a Model-T instead of a 2010+ model car, and has people wondering if cars should even be allowed based on the Model-T's design.

It's important in terms of adoption. Fukushima destroyed the economics of Japanese nuclear power. One accident like that per 100 years and nuclear suddenly becomes very expensive.

global warming has sever long term consequences for economies as well. People need energy, problem is that the costs of dirty energy are spread out and “average Joe” doesn’t think of them as energy costs (pollution, worse health, wars for oil, etc.)

That’s not accurate. “Rooftop” solar is minimally dangerous to install, but large scale solar fields on empty land is extremely safe per kWh.

PS: Nuclear also has many deaths not generally included in these statistics. As building and operating a power plant is not the only risky things you need to do to have nuclear power. Ex: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4164879/

It depends if you're including the carbon footprint of the entire solar panel build/shipment process (from raw materials to shipping/etc). I'm not sure of the numbers, but installation is a negligible part of the total solar impact to the environment.

If you use that level of accounting nuclear plants also have a sizeable footprint just from their concrete useage in construction.

I am a fan of nuclear but current economics seem against it even if they are driven by irrational reactions.

But doesn't that carbon footprint assume that the energy to manufacture and transport the panels comes from fossil fuels? Then it seems we would eventually reach a point where a sufficient fraction of the energy to produce and transport would itself come from solar, leading to a reduction and eventual elimination of that carbon footprint.

The ability to deliver on budget and within schedule is dependent more on political than technological development. Arbitrary delays caused by political meddling (including with pretext of safety concerns) are a significant part of the schedule problems.

However, minor incremental safety improvements may be a tool in addressing this political risk - though it is true that cost still needs to be controlled.

From a tech point of view, you're probably right. But one of the biggest things holding back nuclear energy is public perception and politics. Maybe the biggest thing.

If you can say "prevents the type of accident that happened at Fukushima", those could be powerful words. Fear is often not rational, and it's human nature to be acutely fearful of things you have seen happen (even if they aren't actually larger risks). Even if the actual safety impact is small, something that hits directly at the thing people are worried about could have a large impact on the political reality.

Yes, it's inefficient to have to placate people who don't know a lot about it, but if nuclear is to succeed, we have to realize that the majority of people are never going to understand the risks in depth. It seems worthwhile to spend some resources just to make it so "nuclear" is less of a scary word.

In the HBO series Chernobyl they claim to have been hours away from causing a much larger incident killing about 60 million people and making large parts of Europe and Asia uninhabitable.

Is there truth to this claim?

If so, and if such failure modes are possible, I’d say that current deaths per kWh is not be the best way to estimate risk.

> Is there truth to this claim?

No, there isn't. They invented quite some nonsense to add drama and the reactors blowing up like thermonuclear bombs is among that.

https://theconversation.com/ten-times-the-chernobyl-televisi... point 5

It's entertainment, not a documentary.

Everytime i read this claim i feel bored, because there is no substance to it, justified by uncounted people saying so because someone said so, therefore it has to be "the truth".

What do you make of people who have been there and worked there afterwards, which came to a different conclusion? Conspiracy theo- or terrorists?

https://www.neimagazine.com/features/featurea-different-view...

https://www.neimagazine.com/features/featurereinterpreting-c...

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1867971/

https://www.tandfonline.com/doi/full/10.1080/00295450.2017.1...

The whole point of the "docudrama" was pointing this out, but failing to point out the next level of this error.

And most of you are regurgitating the same shit over and over again, in blissfull ignorance.

In this emotional response full of links, you forgot to articulate your point.

Sigh. Someone asked if what has been shown in the tv-series could really happen. Someone answered naaa, never ever, absolute BS. That would be the feared steam explosion if the molten corium reached ground water and would produce another large "dirty bomb". That was the feared scenario afterwards. The links i posted support a wholly different chain of events, meaning that the explosions did not occur for the reasons generally assumed, and have been indeed in parts been of a nuclear nature, but not like a "fizzling nuke", rather something different, not seen up until then. Therefore not understood. All based on the detected isotopes at different places. Which makes no sense if the story is told as it is told. Which leads to the question why building a big dome over that thing, if 90% of the reactor mass have been vaporized, as the reinterpretation suggests. This is also supported by people who have been in the wreckage with floor plans several times over the years and did not find enough material in it to account for the 90% reactor mass assumed in the wreckage, and about 10% blown away.

So my point should have been:

The Tv show had good production values, but told the story wrong, ignoring the new interpretations which meanwhile became available. Then someone asks about something which couldn't even happen according to that new interpretation, because not enough molten corium in there anymore to make a good and large dirty bomb. Which could have happened IF the corium had been there in large amounts as has been assumed up until now. But it is mostly gone. So not even wrong!

Wasn't the whole point of the show to show how wrong assumptions and mis/disinformation leads to catatsrophe? Especially in the end of the last episode? During the trial?

I think his point was that, regardless of the chances that these doomsday scenarios could occur, experts on site at the time warned that they would, which is precisely what the show depicted.

One of the comments I heard was that while the physicists knew that wasn't possible, at that point, they were 1) battling with politicians who weren't overly concerned with anything except perception, and their own asses, and 2) wouldn't know enough to call them on the lie, and felt that claiming such a thing was possible might "wake them up" and hopefully provide some forward progress.

Did it actually happen that way though? It’s my understanding that we don’t know (the people at those meetings are dead) and HBO/Sky made it up in a way to add drama.

Likewise with the 3 plant workers on a suicide mission to open the sluice gates, who were supposedly sent to their deaths. They’re still alive and healthy today (except one who died of non-nuclear causes).

Well, at least with those workers? That actually happened, with those three men, and they were expected not to survive. (Water does a better job than many solids at absorbing radiation, I believe...).

Noted. Thanks for clearing that up.

Since they did state that some of the characters were made up I had some hope they’d be a bit careful about other things, such as this claim that didn’t add that much to the actual drama.

They probably should have added "Based on a true story" so you know it is partially fake.

Probably overblown. The show was good, but some parts were technically scientifically inaccurate. The claims about megaton explosion certainly were.

What precisely was said there? Perhaps they meant the fallout would be similar to that of a megaton range explosion? The long-lived fission products in a large NPP near the end of a fuel load are about the same as those from a 100MT nuclear bomb.

I don't have the HBO scene at hand, but I found this documentary on youtube which the scene was probably based on. Physicist Nesterenko says the experts concluded there was a possibility that the corium explosion due to contact with water tank could have magnitude of 3-5 megaton and that Minsk would be erased [1].

Now obviously the nuclear explosion of the corium is not entirely impossible, there are known mechanisms which could set it up. But even if it happened, it would not erase Minsk, that is not believable, because Minsk is too far away. So that puts the main claim into question. How probable such corium explosion is is not clear, some things about nuclear bombs which those experts knew are probably still classified, but internet pundits say not very likely.

[1] https://youtu.be/ZKV1piLjoa0?t=58

> Where nuclear fails, and where improvement is desperately needed, is ability to deliver on budget and within schedule.

Indeed, this is what is glanced over way too often with the urge of politicians and industry bureaucrats to rush builds of the bleeding edge and experimental designs.

The only economic nuclear power stations today were big, multi-gigawatt PWR and BRW designs with more than 4 reactors. In comparison to power output/fixed costs, all other factors pale in comparison.

Very good point. Incremental safety measures with an eye to swaying public opinion is pretty pointless as public opinion changes very slowly I imagine and we must be hitting the law of diminishing returns w.r.t this particular set of tech.

On the other hand a pure win would be figuring out how to enlarge the market share of nuclear by persuading those who are already willing but balk at the large up-front costs, the insurance, the regulatory environment, the long planning processes (etc. etc. – basically everything but the tech).

If the USA and the EU both made a commitment to doubling the share of nuclear tech in their jurisdictions to combat climate change (and we're going to need loads of cheap juice to power all those EVs) by streamlining the processes and creating an Atoms for Peace mark II umbrella that would make this HNer a happy camper.

Don't see it happening unfortunately. Every time this topic comes up I get so disheartened.

Public opinion won't matter at all until construction itself becomes a reliable process. There are plenty of sites willing to host nuclear, and as our current fleet ages out many sites would welcome replacement reactors to keep the community's economic base going.

However it's abundantly clear that the current tech's time has come and gone. The current designs can't be constructed in the US, as it becomes a Mexican standoff between utility, designer, and contractor, each just waiting to sue the others for what they assume will be inevitable and costly delays and construction/design errors.

The Rankine cycle for converting heat to electricity is just not super exciting any more, and wind/solar/storage may outcompete thermal electricity generation as an entire class of technologies.

From what I hear the US can barely maintain its roads, let alone build nuclear power plants, but why the gloom? We did it before, this time it really matters, so let’s put some money into saving the planet and get that x1e6 energy density going.

Maintaining roads is pretty trivial compared to building more nuclear, and we are very good at that as long as there's money allocated. When money is allocated for nuclear, we are not good at building, just as we are bad at nearly all massive multi-billion dollar builds.

We no longer have the project management skills and construction skills we had 50 years ago.

If you look back at the pre-Three Mike Island track record for construction, it isn't much better than our current, and fairly disastrous construction record from the past decade. There is a survivor bias when we look around at the currently running and constructed nuclear reactors.

Meanwhile, there's a route that requires far more scalable expansion, our current path of wind, water, solar and storage.

The US seems to be OK at building wind farms and solar farms.

Screw up 1% of the modules in a solar farm and output is down 1%.

Screw up 1% of the concrete pours for a nuclear power plant and it's a disaster.

The distributed nature of wind and solar makes building them, and training the workforce to build them, easier.

Exciting? Who cares if it is exciting? It's the most dense means of energy production that we have, in terms of cost, square footage, and like, every other positive metric I can think of.

In terms of cost, we are on the exponential improvement phase of the logistic technology curve for wind/solar/storage. For thermal energy, we are long past that. That's why "exciting" matters.

Already wind/solar/storage have beat nuclear on cost, are beating coal, and in many areas solar+storage is beating natural gas for new installs.

We don't know how much wind/solar/storage will improve on cost, but we know it will be quite a bit. We know that thermal generation of electricity is optimized to its max.

As far as dense energy production, what benefit does that have? If you want to stick a steam turbine on wheels, or a nuclear reactor in a submarine, that would make sense, but I'm not sure how it helps for the grid.

At some point it will dawn on us that the resources needed to build solar/wind is defined by the energy density available in these resources. Nuclear is basically invoking a deeper level of reality, one where energy is measured in MeV rather than paltry eV of chemical binding energy. Solar and wind are even less dense. It may be an old Physicist’s ramblings, but we have backed away from Prometheus’ fire way too soon to save ourselves.

With wind and solar we are capturing leftover energy from the sun's fusion; the internal amount of storage is not relevant.

With fossil fuels, we are simply using the stored energy from past fusion.

It's likely that we will start creating synthetic fuels from the electricity that comes from wind and solar, if you want a dense storage mechanism for keeping fuel close by for mobile applications or isolated sites.

There's nothing magical about nuclear, it's just a heat source. Unfortunately it's difficult to transmit heat long distances.

> As far as dense energy production, what benefit does that have?

So that we don't blanket the countryside with wind and solar farms? I would have thought that'd be obvious enough.

Also, very few people are really factoring in the environmental costs of repairing and recycling huge turbines and masses of solar pv panels – it's externalities like these (and other externalities besides) that are masking the true cost of renewable tech in the same way that environmental pollution masks the true cost of so-called fossil fuels.

Just you wait and see. The Chinese and Russians are forward thinking, increasing the allocation of atoms for peace in their energy mix.

The land requirements for renewable energy are a tiny fraction of that used for farming, density is not a concern.

Why would you think recycling turbines or panels or inverters is going to be onerous? The quantity of materials, and the nature of the materials do not seem in the least bit onerous.

> density is not a concern

I don't think this is true. We would need about 20,000 sq mi of solar to cover just our electricity needs today. Problem is, electricity is only ~38% of our energy usage. To replace all energy used (transportation, heating, industrial, etc) we need something more like 53,000 sq mi of solar. This would be like covering all of Iowa or New York with solar. Since panels only last around 30 years, we also need the infrastructure to produce, install, remove and recycle ~1,750 sq mi of solar panels per year.

Listing these numbers without context, one might think they are big. We have 1,400,000 sq mi of farmland (according to the 2007 stay I saw), so even if we use this prime space thats already been taken from nature, thats 1%-3% of farm land. We also have many other areas that are currently used by humans that we can cover. We cover something like 100,000 square miles with asphalt in the US, and that doesn't even directly produce revenue.

The numbers, with or without context, are big. Let's just say that it is an area the size of Iowa – that's an enormous amount of land to cover for electricity generation. To argue against that is ridiculous.

And I bet it would actually be a fair bit larger because of all the extra infrastructure you'd need to produce, process and recycle the solar panels. For some reason nuclear waste is a very big deal but wind and solar never appear to have any externalities. Funny that.

Imagine covering all that land with forest, or growing vegetables?

France is well known for nuclear being a high percentage of its electricity production. 75% – how many reactors? 58. Only 58!

I can't find your stat for asphalt. The one I found was: https://www.quora.com/How-much-land-in-the-United-States-is-... which says, “The estimate of roads in America is about 65000 square miles but Maybe 8% of that isn't asphalt. Plus This number would not include parking lots, driveways, private roads and other uses.”

I think an energy production source that takes up as much land as roads _is_ a huge amount of land. Imagine taking each road and covering some land somewhere in the equivalent amount in space with solar panels. That's mind-boggling.

> doesn't even directly produce revenue.

So what? Infrastructure doesn't directly produce revenue. Yet it enables all sorts of revenue generation. What sort of argument is that? You know what else doesn't even directly produce revenue? Electricity production.

You're using lots of subjective comparisons here.

Why is it an enormous amount of land to cover for electricity? Iowa is already 100% covered with farms, which use the sun for growing things. And we have 30x more land covered just for farming, much of it make-work to ensure that we have a massive oversupply of field corn and soybeans. Solar panels are a far less intensive use of the land than that sort of farming.

Why is only 58 sites for France's nuclear reactors a good thing? Can you connect to some sort of value system where such a count becomes a good thing?

My point about "doesn't even produce revenue" is that we covered up all that space without any sort of direct financial incentive to whoever lost that land; with solar, farmers can, and are, devoting their land to electricity production because it's a positive revenue stream for them.

I'm also having trouble contemplating the problem with solar panel waste. What problem do you think is created by that disposal? Logistics? Environmental?

"Mind boggling" could describe any of the industrial processes that happen every single day in our modern society. One could look at the vast amount of corn produced during harvest, and think "that's unbelievable," but it's not an argument against doing it.

The land issue is a red herring. This can be seen by looking at the contribution of the cost of land to the cost of renewable energy. It's small, especially in places where the land isn't even good enough for farming.

Now, maybe in the densest parts of Europe things could be different. This just means that in a renewable future, energy intensive industry leaves Europe. Europe going nuclear would not change this, as they still couldn't compete with the sun-drenched countries closer to the equator.

Problems with nuclear:

  - waste mgmt
  - safety
  - proliferation risk
    (for uranium-/plutonium-based reactors,
     not so much for thorium-based reactors)
  - outrageous construction costs
  - heat pollution (heat discharge into local waters)
  - baseload-only
  - slow to blackstart

Thorium reactor designs solve the first three issues only (unless they can be blackstarted soon after a blackout -- do thorium reactors suffer from poisoning?), in which case they'd solve four of the above problems. Or perhaps thorium reactors can be cheap to build? (I'll believe that when I see it, but only then will nuclear [fission] begin go be appealing.)

The only good things about nuclear are that the fuel is cheap (maybe) and it has no air pollution (barring accidents). That's not enough to justify the issues.

There's one more problem: cost per kwh. It might be competitive in the current market (though arguably it actually isn't). But to stay competitive in the future it will need to keep up with mass produced solar/wind continuing to drop in price. It seems solar + storage bids are continuing to break records and are already underbidding nuclear, even before considering subsidies. IMHO, prices will continue to drop over the next years/decades. Bids are currently going as low as 2 cents per kwh in some areas and prices dropping to below a cent might actually happen at some point.

Energy storage is much less of an issue if you can simply produce more than you need to offset e.g. cloudy days and use the excess energy to synthesize any of a wide range of fuels that can be stored, top up batteries, pump water to some reservoir, or power any of the many ways in which we can store energy. People are getting creative when it comes to energy storage. Energy storage cost is dropping rapidly as well.

Nuclear doesn't just need to get safer, it also needs to get vastly cheaper to keep up with this. IMHO if you are not designing for half a cent or less per kwh, you might as well not bother. My understanding is that current designs are off by factor 10 at least. At those prices, even the security needed to protect the facilities will be prohibitively expensive.

Nuclear isn’t intrinsically expensive. Most of the cost is due to safety regulations, which in principle safer designs would need less of.

Safer designs are the result of safety regulations, if you lift those regulations you will get less safe reactors. They might be cheaper but also less safe.

I think you misinterpret what I was saying. An actively cooled core needs lots of safety measures in place in order to make sure that the water pumps never fail, or if they do there are backup systems in place, and backup systems for those backup systems, etc. There's a physical cost to making all those mechanisms. Furthermore there is a regulatory and administrative cost to ensuring that those mechanisms would work across the entire industry.

On the other hand, something like the NuScale design takes an all-passive approach: it places a smaller fission core directly in a massive swimming pool that has enough water to passively cool the design all the way down. There are no moving parts to switch on in the case of failure. The way you handle a catastrophic failure is: you do nothing. It solves itself.

No moving-part safety mechanisms to install, just a big tank of water. No fallback mechanisms for those safety mechanisms, etc. Inspection is pretty easy: did the water level remain in range? Yes/no.

Cutting costs of safety inspections due to an inherently safer design doesn't mean a less-safe outcome.

And I simply disagreed with this assumption.

For reactors built here a significant part of cost is plain mismanagement, corruption and/or profiteering. This results in costly and less-safe reactors.

Don't know about others.

Partially this is certainly true and it is easy, since everybody assumes that the costs will increase, nobody is surprised that the reactor which should have costed 3 billion euros will cost over 10 billion euros in the end.

For a technology where the prices are decreasing, this is harder to do.

s/would/should/

"Where nuclear fails, and where improvement is desperately needed, is ability to deliver on budget and within schedule."

Something that see nobody discussing and that I gave almost no thought to myself is the embedded carbon cost of the mining and enrichment of Uranium itself.

Never mind the building of the plant, etc., but the actual carbon expended to achieve a workable fissionable fuel.

This process was brought to my attention in the long, detailed chapter on Uranium enrichment in _No Immediate Danger: Volume One of Carbon Ideologies_ by William Vollmann.[1]

It's really quite amazing how many steps are involved and how expensive those steps are in terms of embedded carbon cost. In the absence of fully non-fossil-fuel power sources to do that mining and enrichment it is difficult to see how that process makes any sense absent the massive economic and military subsidies we have in place for gasoline/diesel.

[1] https://en.wikipedia.org/wiki/William_T._Vollmann

The cost of enriching uranium used to be a problem, then gas centrifuges replaced diffusion and reduced the energy cost by a factor of 50.

I'd say one of the biggest roadblocks to nuclear power right now is public opinion. Nobody wants one in their backyard because there's a lot of false information (and shows like Chernobyl don't help). Even though we're well into the safe area, showing that we're making them even safer might help sway public opinion to build one where it otherwise wouldn't have been at all, regardless of the multi-billion dollar cost and decade+ timetable.

But that's not the big roadblock to nuclear power. The main obstacle is that it's grossly uneconomic. Even in locations where the locals are nearly 100% for reactor, the reactors are not being built. The people controlling the money don't see them as good investments.

I personally would not mind having a nuclear plant in my backyard, but please, may I have something in return for the increased risk?

Something like lower property taxes would do. Or free electricity and heating for all residents.

I agree with you that it’s important to make nuclear reactors more cost efficient but also what you are leaving out is twofold. One, The costs of a cleanup after an accident is massive which affects total cost. Two, the public perception of nuclear being scary and an invisible killer inhibits adoption. Just look at Germany’s knee jerk reaction to Fukushima to see that in action.

>Where nuclear fails…

Nuclear fails because the public is uncomfortable with these plants, not for any technical or cost reason. Though to your point you could avoid spending the R&D on safety technologies and instead spend it on marketing which would likely yield better results.

Come on, the classical nuclear power plant is highly expensive to build and decommission. If the problem was purely public perception, we could have hundreds of plants far away in deserted places and transport the power on power lines. In reality, many nuclear constructions stalled or were stopped when cheaper alternatives became viable.

I think the Yucca mountain debacle is enough evidence to show that putting nuclear facilities, granted not generation, “in the middle of nowhere” is not a panacea.

It’s a third rail after Chernobyl and 3 Mile Island, the public perceives it as far too risky to allow anywhere near them. No one wants to touch the idea and yes, it doesn’t help matters that nuclear is insanely expensive.

You have that totally backwards. Nuclear has failed because the people building the plants have consistently blown through their cost estimates. Like Lucy pulling the football away from Charlie Brown, at some point they have to be held responsible for this.

Solar panels can't kill. You could have a very large field of panels, 0 deaths.

People falling off roofs while installing or maintaining panels is the biggest contributor to deaths from solar electricity. Per TwH, current solar is far from the safest: https://www.nextbigfuture.com/2008/03/deaths-per-twh-for-all...

That chart strangely leaves off grid scale solar? I wonder what explains that mysterious oversight?

And of course if you fit the panels at the same time as building the house you don't get any additional deaths and even rooftop solar sails to the top of the list.

I'm sure we'll find another obscure stat to focus on though (even if we have to leave off anything that beats nuclear). Dont let the facts get in the way of a good story.

Rooftop solar is a small fraction of solar installed worldwide, those stats are incomplete without also listing solar farms.

Yeah, agreed entirely. It's not the best document. But "solar panels never killed anyone" just turns out to be not true is all. It turns out everything has risks. I find post-construction rooftop solar to be a perfectly acceptable risk, personally. Grid-scale solar not on rooftops is even better.

What about the carbon footprint from the manufacturing and materials for the solar panels?

It's weird you think that use of fossil fuels is inherent in the production of solar panels, and cannot be substituted for.

I like nuclear but the death numbers are not as relevant when exposure time to radiation is the relevant factor and we've abandoned many cities to prevent those types of deaths.

> Where nuclear fails, and where improvement is desperately needed, is ability to deliver on budget and within schedule

The onerous safety regulations are mostly the reason nuclear plants run over budget and schedule. Safety improvements that can relieve some or all of that burde arr probably worth it to scale nuclear.

That's a big reason for why it got so expensive though :)

You could probably remove quite a lot of requirements and still have the safest energy generator and that would bring down the cost.

It is misleading to claim current generation plants are safe, for all relevant meanings of safe. If they were actually safer than other power sources, the plants would be insurable. You could call them safe if they did not pose a bomb proliferation risk. You could call them safe if cleanup after accidents was a well-understood and feasible process. You could call them safe if the risks did not reach to somewhere between hundreds of miles and the entire planet.

Safer nuclear power is not just about reducing the possibility of immediate death, which, of course can't be entirely eliminated. You still have risk to employees from non-radiological parts of a power plant.

What you won't have from safer nuclear power plants is the risk of spreading stuff like Strontium 90 over a wide area. That by itself is a very valuable goal that will vastly reduce the maximum downside risk and the cost of mitigating that risk.

Cost is part of risk, in the form of financial risk. Smaller, less expensive plants are less likely to become white elephants on rate payers' bills. They won't lead to To Big To Fail entities that may need bailing out because they are the custodians of a big nuclear mess.

Safety is multi-dimensional. People opposed to building more nuclear plants with current technology understand that. They want to keep their wallets safe AND their milk safe. They want to keep the value of their property safe. That's perfectly rational.

Safe is measuring how many deaths per user various power generations sources cause. Coal etc is less safe then nuclear because it causes more deaths. Yes when a nuclear accident happens, it can be bad, but that doesn't happen that often at all. Compared to the numerous accidents and deaths that happen every day with say coal. Also coal is outputting pollutants into the atmosphere which causes more deaths (a slower death but dead is dead). Cleaning up that air is also just as difficult as cleaning up nuclear waste.

> Safe is measuring how many deaths per user various power generations sources cause.

No. Safety depends not only on deaths but on all kinds of damage. It also does not depend on how many people nuclear kills per year. That number is predictable so it isn't even a risk (to society), it's a cost.

Nuclear is unsafe because of the damage it causes in the worst case. What happens when NPPs are operating correctly isn't even relevant to the discussion.

Reply on news.ycombinator.com