[imtringued]

With a new projected cost of 25 billion USD the cost overruns of the Vogtle Plant are worse than I thought. There is a reason why nuclear is losing to natural gas and coal. It's too damn expensive and takes too long. If this money was instead spent on building more storage capacity we'd have solved our energy problems a long time ago and even if we started today we'd be done long before the nuclear plant is online.

The nuclear plant consists of two units each with 1200MW production capacity. Let's be pessimistic and assume that we need 24 hours of storage or roughly 57600MWh. A modern redox flow or lithium ion battery can cost as little as $100 per kWh but Tesla's grid battery with 129MWh cost 66 million which is around $500 per kWh so we will use that.

Well it turns out 57600MWh * 500$/kWH is exactly 28.8 billion USD. No power grid on earth needs a 24 hour battery but even with this crazy assumption grid storage isn't significantly more expensive than nuclear power.

[thatcat]

Your cost analysis neglects replacement, disposal costs for batteries, long term costs of carbon emission, and coal flyash disposal. The only reason nuclear is losing to gas and coal is because the externalities of the mess they make are dumped on the public, whereas nuclear pollution and waste processing is heavily regulated.

[skybrian]

Doesn't nuclear power have subsized costs as well? I'm thinking of the Price-Anderson Act.

[crdoconnor]

There's also the nuclear liability cap. Taxpayers cover disaster cleanup costs that exceed some fairly low level ($200 mil I think).

It's very expensive but also free insurance, basically.

[cma]

That's the Price-Anderson "Nuclear Industries Indemnity" Act he was talking about.

[toomuchtodo]

Batteries can be recycled, nuclear waste is stockpiled with no permanent disposal or recycling plan in sight.

Solar and wind, unsubsidized, with storage, are already cheaper than nuclear. Those costs will continue to plummet, especially as EV battery demand ramps (rapidly expanding battery manufacturing capacity).

Nuclear lost because it’s too expensive, and has its own externalities to ignore (decommissioning, waste disposal, liability insurance).

[petre]

Most of this nuclear "waste" could be recycled as well by processing and further burning it in breeder reactors. Of course the processng equipment gets contaminated but that's life.

[beaconstudios]

this might sound silly (and it probably is), but I don't understand why we don't build a hulking great cannon and shoot our nuclear waste into space.

[ben-schaaf]

I know others have addressed the dangers of this, but I had a little fun looking at the logistics. The UK alone produces 2Mt of radioactive waste[0] per year, ~9% is of "intermediate" level. Just sending those 18Mt into LEO would take 129,000 Saturn V launches per year. That's about one launch every 4 minutes. And LEO is not where we want to send anything. Lets assume a Trans-Lunar Injection is enough, bringing that number up to 370,000, almost one launch per minute.

Just the RP-1 & LH2 fuel cost would be around $600,000 per launch[1], so about 222 billion USD per year (The fuel is the cheapest part of the launch). Each launch releases around 440,000 kg of CO2[2] into the atmosphere. That's 163Mt, an increase of 44% of the UK's current CO2 pollution from the launches alone.

Just for fun, if we wanted to send all nuclear waste all the way to pluto using only Falcon Heavies we'd have to launch around 1.8 per second. Going off the 80 million USD price tag for these launches, that's 4.5 quadrillion USD. That's about 60 times the Gross World Product to cover the UK alone.

[0] https://nda.blog.gov.uk/2017/04/03/how-much-radioactive-wast... [1] https://www.quora.com/How-much-fuel-does-it-take-to-travel-t... [2] https://space.stackexchange.com/questions/13082/calculate-fa...

[chongli]

Launching waste into space with a rocket is a silly idea. We don't really care if the waste gets damaged as long as it doesn't break up and make a mess. Why not build a giant railgun to launch it?

[ben-schaaf]

I'd consider them equally as silly, considering we're actually capable of launching rockets into space yet the closest we've gotten with a space gun is a measly 180km apogee @ 3.6km/s[0]. The original question was why we don't send our nuclear waste into space. We can't send our nuclear waste into space using a rail gun because we can't currently send anything into space with a rail gun.

Otherwise here's a good discussion on the viability of space guns: https://space.stackexchange.com/questions/2370/what-technolo....

[0] https://en.wikipedia.org/wiki/Project_HARP

[acidburnNSA]

It's cheaper and safer to store it in deep crystalline bedrock. I guess if we geologists can't convince we the people of that maybe the railgun idea would work, especially if you could hit the sun where the fast fusion neutrons would consume the hell out of it.

[petre]

What if it fails and the waste falls back to Earth spreading over inhabited areas or agriculture/fishing grounds? It's a bad idea to begin with.

[richardburden]

2Mt is two metric tons, right? It surely is not 2 million metric tons. See the IAEA's estimates of total radwaste in storage or disposal worldwide: http://www.world-nuclear.org/information-library/nuclear-fue...

Because of the extremely high energy density of nuclear fuel, the amount of waste produced per Gigawatt-year of electricity generated is very small.

A Saturn V has a payload of 140 Mt to low earth orbit. Total high-level waste is 22,000 cubic meters, possibly 220,000 Mt, assuming a density 10 times as great as water. 1,572 Saturn V launches would carry away all of that. All the high-level waste ever produced that has not been recycled. Most of it can be recycled in breeder reactors, so there is no need to carry it away!

[ben-schaaf]

2Mt is 2 mega tonnes. If you look at the source only 0.03% of radioactive waste in the UK is "High Level". The expended fuel itself is a tiny portion of the nuclear waste. At lot of things other than nuclear reactors also produce radioactive waste that requires proper disposing, as your source confirms.

[joe_the_user]

Oh, this reminds me of high school, when I asked a physicist my mother dating about the same idea. He worked for the Rand corporation and was overall pro-nuclear. Still, He considered for a second and did his best to explain...

What I remember of his explanation was the point that if you launch a payload of material into space, you have essentially placed that payload in an elliptical orbital path, one that intersects earth's elliptical path. And since orbits are periodic, that payload is going to be intersecting earth's orbit regularly, meaning that in this instance, the chance of the material coming right back into the atmosphere is quite high.

[BurningFrog]

I like to remind people that we already live on a radioactive planet. We can build nuclear reactors because radioactive stuff naturally is everywhere. There probably wouldn't be life here without it.

So to me the "let's get it off the planet" idea is an overreaction. Radioactive material just need to be handled with common sense.

[jazzyjackson]

Hmm... what's the over-under on dropping barrels into a volcano?

You get a little radioactive steam surely...but eventually turns into radioactive dust...

As an aside, ignorant physics question, does the half-life of something decrease with heat? Like does more beta-decay mean it gets to safe levels faster?

Maybe a few years in between tectonic plates would do a planet good eh?

[archgoon]

> As an aside, ignorant physics question, does the half-life of something decrease with heat?

Short answer: No. Definitely not in a Volcano.

Long Answer:

Decay rates are ultimately determined by the subatomic structure of the nucleus. This is why different chemicals, and different isotopes of said chemicals, will decay differently. A lone neutron will decay after about 15 minutes (which is a long time in subatomic scales), however, the two neutrons in (4/2)He will happily stick around forever.

So melting a material won't have any impact on the structure of the nucleus, you're just disassociating the intramolecular bonds; same goes for turning said material into a gas.

If you were to raise the temperature high enough, and we're talking sun-like (not the measly 5000k that you get on the surface, I mean like 27-million degrees you find in the core), then you'll actually start stripping and fusing nuclei, which will at that point change (but not necessarily decrease) the radioactive rates of whatever material you started with.

[AngryData]

In most cases a faster decay rate does mean the material becomes safer at a faster rate, but consequently produces more radiation in that same time. If it was reprocessed we could reuse the high-output waste over and over until all that was left was the low energy but long half life material, but old regulations based on poorly understood science in the past, public ill-will, and potential international political conflicts could appear as the reprocessing technology is basically the same technology needed to make nuclear or radiation bomb source material.

[BurningFrog]

The classic answer is to bury extremely long lived waste in a subduction zone at the bottom of the ocean, such as the Mariana Trench.

A "subduction zone" is where two tectonic plates collide and fold into the depth of the planet. So anything you bury there would naturally travel deeper into the planet over the eons.

I haven't heard any serious arguments against this, but I suspect it's impossible because people think it would "contaminate the oceans".

[eseehausen]

Because we already know how to store it for thousands of years with minimal leakage without creating a contraption that could lead to nuclear waste raining down on our heads.

[pvaldes]

To be fair, we already know how to store it with minimal leakage for thousands of years, but the tentation to just dump it in the Mediterranean (1980's), Somalia (1980-90's), seed it over Irak (Gulf's war) or leak it into the Japanese Sea (2011-17) is irresistible.

[ethbro]

Uranium (elemental) has a density of 19.1 g/cm^3.

For comparison, lead has a density of 11.34 g/cm^3.

It's heavy.

[ryacko]

Yes, but how much coal versus uranium would a household use, if used to generate electricity in the most efficient way possible (cogeneration presumably)?

[skywhopper]

One answer is that it probably isn’t possible to do what you say, but if it were the failure states are wildly unacceptable. Any “cannon” (or rail gun) capable of accelerating an object to escape velocity from the surface would likely burn the object up during its flight through the atmosphere, spreading its radioactive contents into the air to float around as low level global fallout for decades.

But if instead you just launched it into space on a rocket... well what happens when a launch fails and the rocket blows up on the launch pad? If you’re lucky, you mark a few square miles as no-go zones and then move on to the next launch pad? Doesn’t strike me as a sustainable solution.

One of my favorite ideas is to encase the waste into giant glass blocks, and build pyramids in the desert... but I doubt that’s very practical either.

[Eric_WVGG]

because a misfire would result in a fine mist of nuclear fallout spread over the entire planet

[pvaldes]

We'll have to pack it very carefully in any case. Perhaps we should assume that this will happen and work with that scenery in mind. The concept of a fire resistant box is not new at a smaller scale; We can vitrify the less dangerous waste and store the crystal in steel containers. Then we could theoretically design explosion-proof containers for it with some outer sacrificial disposable layers made of a plastic clay-like matherial or so.

[erikpukinskis]

There’s some more weight.

[beaconstudios]

hmm, I guess that's not ideal.

[WalterBright]

If it's spread out over the planet, the resulting radiation would be insignificant.

[ThomPete]

Uranium can be reused so that pretty much solves that.

Its not too expensive its very cheap, its the regulation around that makes it expensive which can be solved too.

[ben_w]

> Uranium can be reused so that pretty much solves that.

No it can’t. Any given atom stops being uranium after it fissions. Some of the waste can be reprocessed because not all is fissioned, but even then there is a lot of literally untouchable waste left over — well, literally untouchable if you want to live: https://what-if.xkcd.com/29/

[cm2187]

But I don’t get this untouchable concept. Lava is untouchable. Deep oceans are unreachable. Why is it a requirement that we have to be able to go and touch every single spot on earth? What is the problem with having one minuscule location on earth where you are not supposed to go and stock pile this material there? It is contained, we have full control over it, unlike the immense volumes of chemicals currently released in rivers or the atmosphere. And there are already countless locations where you are not supposed to go because it is private or military property, polluted with chemicals, or too cold or too warm for humans, or a protected natural reserve, etc.

I don’t think waste is a legitimate concern of nuclear energy, as long as they are stored and protected responsively. The risk of an uncontained explosion of a reactor is a more legitimate concern. But I understand many modern designs like molten salt reactors reduce this risk to pretty much zero.

[ben_w]

I am somewhat sympathetic to your points, but was responding to a comment that seemed to claim it was a non-issue.

I’m only “somewhat” in agreement because humans are terrible at reading warning signs, especially from long-dead people. “Oh,” they say, “that was ages ago. It’s just a primitive superstition. Anyway, we’re special.”

[PavlovsCat]

> What is the problem with having one minuscule location on earth where you are not supposed to go and stock pile this material there?

How much time do you have? This article isn't just long, it has many, many interesting links.

https://en.wikipedia.org/wiki/High-level_radioactive_waste_m...

> There is a debate over what should constitute an acceptable scientific and engineering foundation for proceeding with radioactive waste disposal strategies. There are those who have argued, on the basis of complex geochemical simulation models, that relinquishing control over radioactive materials to geohydrologic processes at repository closure is an acceptable risk. They maintain that so-called "natural analogues" inhibit subterranean movement of radionuclides, making disposal of radioactive wastes in stable geologic formations unnecessary. However, existing models of these processes are empirically underdetermined: due to the subterranean nature of such processes in solid geologic formations, the accuracy of computer simulation models has not been verified by empirical observation, certainly not over periods of time equivalent to the lethal half-lives of high-level radioactive waste. On the other hand, some insist deep geologic repositories in stable geologic formations are necessary. National management plans of various countries display a variety of approaches to resolving this debate.

So, various countries display a variety of approaches to resolving this debate.

That's all we have for now.

> I don’t think waste is a legitimate concern of nuclear energy, as long as they are stored and protected responsively.

Yeah, but so far we have no way of doing that, so it's a legitimate concern. I might as well say "we can just turn off all nuclear reactors now, we just need other means oto generate the energy and do the shutdown responsibly". Leave it allll up to the reader, or in this case, future generations, why not.

https://en.wikipedia.org/wiki/Radioactive_waste

https://grist.org/article/stang/

> At Hanford, a rough rule of thumb for planners is to look ahead 1,000 years. That’s like a Viking trying to conceive of an astronaut, then trying to pass a note to him.

> Experts inside and outside of DOE have pondered this communication conundrum. The agency has assembled panels of scientists, historians, artists, and others to tackle from all angles the question of how a 21st century sign should look to a 31st century person. From symbols to colors to materials to size, everything’s up for grabs — and nothing’s been decided.

We don't have solutions. Pretending we have solutions will not help us get solutions, either.

[ThomPete]

The ability to reuse more and more will improve over time so yes it's perfectly feasible.

Furthermore, the price of uranium is so cheap that it's financially feasible right now. It will be over time but that should also show you just how abundant and available it is as a resource.

If we actually got politicians to sit down and go through the requirements with engineers and scientists they would realize how cheap and still safe it can be.

[PavlovsCat]

> The ability to reuse more and more will improve over time so yes it's perfectly feasible.

Well, if that "counts", then solar power will become totally free and have a net positive environmental impact, we'll just plant some nanobots that grow and repair solar panels inside a marked area, and can be scooped up should we need them elsewhere, hence "100% free once we perfected them". They'll smell like vanilla, and change color according to the moods of people using the energy.

Latter on we make nanobots and send them to the sun to hang out there for a few thousand years and grow a scaffolding from, uhh, space dust or something, and then we shoot nanobots at it that grow solar panels on it, and we have our Dyson sphere. Add some nanostuff that creates material from energy and flings that at Earth, then convert material back to energy.

This didn't even take me 5 minutes, so I really don't get what the problem is :P

[toomuchtodo]

If it was easy to solve, it’d be solved already. It’s not easy and it’s not cheap, and for those reasons it’s likely it won’t be done.

It’s already cheaper to build new solar plants and wind farms than continue to run an existing coal plant (which itself is cheaper to run than nuclear).

[ThomPete]

No it wouldnt be solved already since the political environment is completely against nuclear, so simply false. Solar is a fraction of energysupply and will not even be close to being signifficant whare it really matters which is in transportation.

[toomuchtodo]

How is nuclear being expensive political? It is fundementally untenable as a commercial enterprise when compared to natural gas, solar, and wind.

90-95% of all new generation coming online each year are renewables in the US. Renewables cost continues to decline year over year, speeding its uptake. I cannot fathom how one would think nuclear can compete at all in such an environment.

https://www.eia.gov/todayinenergy/detail.php?id=31192

“Nearly all nuclear plants now in use began operation between 1970 and 1990. These plants would require a subsequent license renewal before 2050 to operate beyond the 60-year period covered by their original 40-year operating license and the 20-year license extension that nearly 90% of plants currently operating have either already received or have applied for. The AEO2017 Reference case projections do not envision a large amount of new nuclear capacity additions. By 2050, only four reactors currently under construction and some uprates at existing plants are projected to come online.”

[rblion]

I love Hacker News. Reddit is elementary compared to the quality of comments like these.

[TrainedMonkey]

The missing link is to build up solar capacity to refill your battery.

To refill the battery you need to generate 57600MWh each day. Assuming that we get 12 hours of sunlight per day at 50% efficiency, you that comes down to 2 (efficiency factor) * 57600MWh / (12 hours / day) = 9600 MWh solar installation.

At $1 / Watt, 9600 MWh solar installation comes down to $9.6 billion.

Assuming my math is correct, this seems pretty competitive. In particular, I am assuming that $1 / Watt means $1 per Watt of electricity at max luminosity. If $1 / Watt means $1 per 1 Watt of absorbed solar energy, the efficiency should be 10% - 15%.

Large scale installation cost per watt: https://www.greentechmedia.com/articles/read/doe-officially-...

[wbl]

The winter vs. summer amount can be far more, and then you have heating costs to provide power for.

[sokoloff]

And not infrequent spells of winter weather that are a week or more long without ever seeing the sun.

[strainer]

> assume that we need 24 hours of storage or roughly 57600MWh. A modern redox flow or lithium ion battery can cost as little as $100 per kWh

I think the flow battery can get even cheaper than this for 24hr storage. Recent price claims[1] were $150/kWh at 4hr, and $100/kWh at 8hr - this puts the cost of charge/discharge rate at $400/kW and the cost of the energy storage at $50/kWh.

1kW @ 4hrs = (400+ 50* 4) /4 = 150$ (50+400/4)

1kW @ 8hrs = (400+ 50* 8) /8 = 100$ (50+400/8)

1kW @24hrs = (400+ 50* 24) /24 = 66$ (50+400/24)

So it really was a bit crazy to estimate 500$/kWh for that storage, 200$kWh would be absolute maximum present price for flow battery, and under $75kWh seems likely.

[1] https://news.ycombinator.com/item?id=18751639

[erikpukinskis]

Is it really sensible to believe any press release about future batteries?

[strainer]

The press release contained current and next years prices - 200$/kWh @ 8hr storage is the current cost of flow storage.

Its not really sensible to price 24hrs of li-on storage considering li-ons charge/discharge rate capability is fixed. There is no use for the capability to charge and discharge 24hours of power within a couple of hours, but you cant avoid buying it with li-on. With flow batteries you price the discharge rate and the capacity separately. This is why flow batteries should work out a fraction of the cost even at todays price.

[epistasis]

I would rate the reliability of a battery press release slightly higher than a signed-off budget and timeline for nuclear construction.

And when it's press from a battery maker, rather than a battery researcher's university press department, the battery press release is far far more reliable. I'd say their cost overruns will be at most 20%-30%, compared to nuclear's usual 200%-300%.

[selimthegrim]

The other one in South Carolina was mothballed because they couldn’t find any lenders to finance the rest of it at <30% interest

[8bitsrule]

How many wind turbines could already be spinning for the $25B thrown at this endless project? The answer, at $10M each, is 2500.

[maxerickson]

Where's the over-capacity to make that thing work going to come from?

There's ~12 hour batteries on the grid (pumped storage), but they mostly recharge from coal and nuclear at night.

[CodeWriter23]

The 57,600 MWh figure is the peak usage in California on the hottest of summer days. We only need that for roughly 2 hours of one of those days, and we can typically meet that without engaging rotating blackouts. The generation capacity is there. And it’s a lot more efficient to operate plants as needed to recharge storage than to spool them up in response to demand, the Tesla battery farm in Australia has already proven this. The key reason is there is lest waste if you have storage as a buffer.

And think of all the CA plants sitting idle on a day like today, where peak load was 28,000 MW.

http://www.caiso.com/Pages/default.aspx

[cm2187]

Not an expert but I think solar power would also be uneconomical if it wasn’t done at scale. The problem is that most recent new power plant constructions are unique designs, with ever changing regulations. I am sure nuclear energy would be economical if you were renewing a large number of plants at once (which will need to happen as a lot of plants are becoming really old).

[shanghaiaway]

Interesting armchair analysis. Nuclear is both the cheapest and cleanest source of energy.

[pfdietz]

Nuclear is by no means the cheapest source of energy. That's the big problem it's facing: it has priced itself out of the market.

[bovine3dom]

It hasn't priced itself out of the market - Governments insisting that it is many orders of magnitude safer [1] than any other form of power generation have unsurprisingly made it very expensive and over-engineered.

Nuclear quite simply needs to be given its fair share of blood to be competitive - to get wind power down to the same level of safety we'd have to build robots to climb the turbines to maintain them.

[1] https://en.wikipedia.org/wiki/Energy_accidents

[pfdietz]

The argument that governments aren't letting reactors melt down enough is a nonstarter. If nuclear proponents have to resort to that, you might as well give up.

I suggest instead pushing for designs that can achieve acceptable levels of safety at lower cost. In particular, molten salt designs, such as the one from Moltex, that allow the size of the containment building to be shrunk by a factor of five from that of an LWR (for a given power output). Moltex's design does not dissolve the fuel in the primary coolant and has other attractive features vs. other MSR designs.

Such designs are not on the market today, and will take time and money to bring to market. I think R&D on them is a good idea, and supporting that is a proper role for governments, just as it was for renewables, but they are still something of a longshot, given where the competition is likely to be in 20 years.

[bovine3dom]

It isn't just the reactor that is over-specced - anything to do with nuclear is held to a much higher standard than anything else, relative to the actual risks.

Thorium molten salt reactors have been bandied around for a while, especially in India due to necessity, but I fully expect governments to still make unrealistic demands of them, and therefore drive them up to unrealistic prices.

Stuff like this depresses me - I don't see how politics can handle grown-up conversations about trade-offs like this when people will just vote in the cake-and-eat-it parties.

[pfdietz]

When it comes to MSRs, thorium is a distraction. The benefits of MSRs have very little to do with the presence of thorium.

It's important for MSR proponents to focus on the actual benefits, rather than fake benefits, so they don't make bad engineering tradeoffs.