Hacker News new | past | comments | ask | show | jobs | submit login
Energy Department teams up with Bill Gates to move mini-nuclear plants to market (washingtonexaminer.com)
207 points by rmason on Aug 25, 2018 | hide | past | favorite | 114 comments



Terrapower's reactor uses molten chloride salts, with fuel dissolved in the salt.

An issue with this sort of reactor is the entire primary loop becomes loaded with fission products, and becomes intensely radioactive.

I prefer the MSR (molten salt reactor) design by Moltex. http://www.moltexenergy.com/

Their design also uses molten salt, but contains the fuel-bearing chloride salt inside tubes. A separate, non-fueled salt circulates around the tubes.

For reasons of chemistry, this enables them to use a fluoride salt outside the tubes, and unlike homogenous fluoride-based MSRs the reactor structure can be made of stainless steel rather than more exotic (and expensive) nickel-based alloys. It also reduces the volume of chloride needed, which is an important cost savings if the chlorine has to be isotopically separated to reduce production of 36Cl (which has a halflife of 3x10^5 years.) The high cross section of chlorine for thermal neutrons is also why chloride MSRs are necessarily fast reactors.

Moltex's design retains the major advantage of MSRs, which is the absence of volatile materials in the containment building. It is the vaporization and pressurization of these materials in an accident which forces the containment to be large and strong.


Have you looked at the press release that yazr linked to? It sounds like this is a new design for Terrapower that is similar to the concept that you prefer. It would be great for someone that is familiar with the design issues to confim if this is the case.


Fair points. The challenge with fluid in tubes is fuel density. Very hard to get it high enough to be economical.

Good old solid nuclear fuel can be upwards of 15 g/cm3.


With the traditional flibe salt, you have to enrich the Li to reduce tritium production. And Be is a pretty nasty material to handle as well.

But in the end the decisive factor is the neutronics. Turns out that in the fast spectrum chloride salt are better, in the thermal spectrum flourides.


> in the fast spectrum chloride salt are better, in the thermal spectrum flourides

Would you mind expanding on this? Why?


He's talking about FLiBe. Both 7Li and 9Be are fairly light isotopes, so they moderate neutrons, better than carbon in graphite does. Unfortunately both are expensive.

The fluoride cooling salt used by Moltex does not use lithium or beryllium. Instead, it's sodium-zirconium fluoride or sodium-potassium-zirconium fluoride, with no isotope separation required. The atomic mass of these isotopes is high enough that the neutron spectrum remains hard. In fact, Moltex's design adds a bit of hafnium (a strong thermal neutron absorber that is chemically very similar to zirconium) to the coolant salt to shield the walls of the reactor from thermal neutrons.

One issue with fluoride salts in homogenous reactors is that the salt must be kept a bit oxidized, so uranium does not disproportionate into UF3 and U metal. The metal would plate out on surfaces, which is a no-no. Unfortunately, this means the salt must be oxidizing enough that it will dissolve chromium, so ordinary stainless steels cannot be used.

By using a barren coolant salt, Moltex can make the salt reducing enough that chromium is stable (they do this by using a sacrificial anode of zirconium metal.) Their design can use stainless steels that have already been tested and proven for use with the neutron exposures of fast reactors.

The chloride fuel salt in Moltex (and Terrapower) don't have that chemistry issue; uranium is more chemically stable there.


In addition to what the sibling poster said, https://info.ornl.gov/sites/publications/files/Pub29596.pdf and http://moltensalt.org/references/static/downloads/pdf/MCFR_B... contain some discussion of the salt options, particularly for fast reactors.

For thermal reactors, particularly if you want to breed, then you have to be very careful wrt losing neutrons (since you're barely with your nose above water to begin with), so there parasitic neutron capture becomes a very important parameter. And yes, in that regime FLiBe turns out to be a pretty good choice.

For fast reactors, you have more excess neutrons so you have a bit more freedom. Cl, being heavier, allows a slightly harder spectrum, which is good, though not that critical. That being said, there are many fast spectrum paper reactors using flouride salts as well (not FLiBe, but FLiNaK, or some ZrF salts etc.).


Isn't stainless steel not so resistant to chloride ?


The challenge in a traditional MSR design is that you have half the periodic table swimming around in your primary loop at high temperature. From a corrosion perspective that's pretty challenging.


Correct. But there's some fancy high-nickel and other stuff that can likely handle it if you control the chemistry carefully.


Salty water is a problem, but pure chloride salt is less corrosive to stainless steel than pure water, according to a presentation I saw by someone from another chloride reactor company.


Here is the DoE press release

https://www.energy.gov/ne/articles/southern-company-and-terr...

This is a new sodium liquid fuel design.

It is NOT the travelling wave reactor, which has been championed by TerraPower for the past 10 years.

These are 2 very very different designs which is confusing to me.


From their website, it seems work on the traveling wave concept wound down several years ago (not spelled out explicitly, but lack of updates).

I'd guess they figured out the traveling wave concept wasn't workable, and switched to another design.


As of last fall, they've partnered with someone in China to build a traveling wave reactor there.

https://www.smh.com.au/business/bill-gates-and-china-partner...


The link clearly states that this is the MCFR - Molten Chloride Fast Reactor.

I think the confusion may lie in that people equate fast reactor with sodium. It's a fast MSR.


The submitted article seems to be mostly cribbed from the DoE's tweet; this link is much better.

dang or sctb, could you update the link to the energy.gov one?


Mini-nuclear plants. 1100 megawatts. These are completely in disagreement. Terrapower focuses mostly on large plants to power an increasingly urban world. The media has gotten that Terrapower does small reactors wrong for over a decade.


Does anyone know how big this thing is, physically?

If you can put the reactor vessel, and associated 'hot' parts into a 40ft Iso container, and ship it cross country for installation/recycling, that might improve on the economics of existing reactors. If you can't do that, I'm not sure how it's going to be cheaper than other options.


It's pretty compact, the whole reactor is about the size of a large house IIRC. Putting the whole core in a container and shipping it off is pointless, though- it's designed to operate for many decades and uses a continuous refueling process. They're also designed to be able to get really high utilization of fuel and not produce much waste. This design is very different from current commercial plant designs.

Operating costs are also not the problem for nuclear plants, they are dirt cheap to run relative to their energy production. It's the upfront capital cost of building a large plant that makes them uneconomical currently, which is why smaller-scale designs like this one are appealing.

Edit: http://www.world-nuclear.org/information-library/economic-as... is informative regarding the economics.


The commercial versions of these are bigger than that. Economies of scale are still very much at play in nuclear due to the unique regulatory and supply chain issues, as well as how much not-highly-enriched fuel you need to go critical. But some folks working on nuclear farms (fields of small nukes with one security force) are dreaming of interesting new options.

Regarding economics, there are lots of reactors in the USA that have their capital cost completely paid off but they're still expensive because of high Operation and Maintenance costs. A lot of this is expensive retrofits and added security costs post 9/11. More of it is that cheap natural gas in deregulated markets has brought electricity prices way down and therefore electricity sales revenue way down. This is why some nuclear plants in the US are shutting down early before their lifetime is up. It's a great tragedy because these plants are massive carbon-free baseload electricity sources. Nukes in the US make 60%+ of the carbon-free energy of the country.


Size wise I think even the commercial units they are planning have the core and associated support machinery at about that size, not including things like turbines etc that would be needed for a proper plant. It's been a while since I've seen any technical presentations about TerraPower though so maybe that's out of date.

The cost of retrofits is a good point, especially given that our nuclear fleet is getting pretty old. I am also worried about the glut of cheap natural gas and its tendency to stifle other technologies. It's almost unbeatable right now price wise, but I'm a bit concerned about what will happen when that changes. Renewables are great, but base-load power is still a problem.


> 1100 megawatts

Isn’t that smaller than the average power plant?


there are 99 nuclear reactors operating in the US, producing about 100,000mw. So about 1000mw/ea.

Many of the plants in the US have more than one reactor.. but you can see on this page, plant capacity/units is about 1gw/ea [edit: corrected, thanks], and many of the single reactor plants are about that size also: https://en.wikipedia.org/wiki/List_of_the_largest_nuclear_po...

Note, that page has the largest plants.. about ~45. There are only 61 in the US. So there are a few smaller ones not in the list. If you built a 1100mw reactor, it would be ~#42 largest plant.


Isn’t this not about size of energy produced but rather the physical size of the reactor?


*1 GW each.

A small reactor is under a few hundred megawatts. A micro reactor is on the order of tens of megawatts. It's quite hard to make economic reactors below that.


Some Russian satellites went into space with two to a whopping six kilowatt reactors (https://en.wikipedia.org/wiki/US-A). The Americans in contrast only got a 500 watt reactor into the sky, and it failed after 43 days (https://en.wikipedia.org/wiki/SNAP-10A).


Yes but when you aren't allowed to use highly-enriched (weapons-grade) uranium it's a different story.

And space is a much different story for economics. Out there, fully enriching your fuel probably makes sense because weight is so expensive. My comment was limited to terrestrial reactors.


1100 MW is big enough that you would need to connect to the high voltage transmission network. This will require you to build the station close to an existing line with capacity, or spend millions on your own cable.


“Southern Company and TerraPower plan to develop and license a test reactor before developing a 1,100-megawatt prototype by 2030.“

Seems entirely too little too late. We need enough of these to power US infrastructure and electric transportation ASAP.


I don't think we're facing any critical electricity shortage, at least in the continental US. This is just one of many new power generation systems being pursued. It is neither too little, because any amount helps, nor too late, because the need for more electricity is going to continue forever. I'm not sure how this is a negative thing.


We, the world, need to move off fossils fuels and begin cleaning up our greenhouse gas contributions. We have something like 150 years of emissions to clean up. And we all still want to drive cars and do all the other modern, energy intensive things we do. Preferably, we would have started this before irreversible changes occurred, but we’re probably beyond that point already. So yes, we need more nuclear plans and the sooner the better.

As for the waste, that is a problem, and it’s entirely solvable. Much easier to solve than raging forest fires and the growing inhospitablity of climate change.


Isn't it unrealistic to assume that 150 years of emissions can be reverted? At this point it's more about slowing down the global warming rather than reverting it.


I doubt it will happen but there’s no technical reason we couldn’t reduce atmospheric CO2 to pre-industrial levels if it was a high enough priority. Terra preta sequesters carbon for centuries and we know how to manufacture it, it’s just soil enriched with powdered charcoal. Turn cellulose into charcoal, spread it on soil, repeat.

https://en.m.wikipedia.org/wiki/Terra_preta


Negative? Reliable and ultra-long term fuel disposal (i.e., the lack there of) is still unsolved.


Is disposal really a concern when we're facing the collapse of entire continent-sized ecosystems?

Compared to runaway global warming, even widespread radiation poisoning seems trivial. And such poisoning is both hypothetical and extremely unlikely to be widespread.


We need more than "clean" energy.

We need a change in a number of habits. We need actual leadership, and not only gov types. Etc.

Nukes are the easy short term way out. The issue is what that means over the long term and how it impacts the other vectors in entity. of this problem.

Without a wake up call we'll continue to mindlessly zombie on, as we have been. Nukes __might__ slow that (but only if they don't also create a false sense of overconfidence). However, they're not going to alter the course. They alone are not going to change the end.


Engineering wise the disposal problem is solved. Making it politically palatable isn't.

Australia has long debated taking the world's nuclear waste, numerous commissions have found it feasible and safe. The community is overwhelmingly against it though.


99% of nuclear waste, and basically all the long-term stuff, is transuranics. Those are fuel, for fast reactors like this one. Consequently they produce 1% as much waste as conventional reactors, and it goes back to the radioactivity of uranium ore in about 300 years.

Another consequence is that these reactors can use most of our existing stockpile of nuclear waste as fuel. It's the only technology that can actually reduce our waste stockpile instead of just storing it somewhere.

Storage for 300 years is generally considered a solved problem. I think the basic idea is to encase the waste in blocks of glass and bury them somewhere.


I was referring specifically to the parent's concerns. Fuel disposal is certainly a problem in nuclear in general. I'm not sure about this specific reactor type; it's possible it produces less-troublesome waste, but I don't know. Most power generation is exceedingly bad for the planet in one way or another, but nuclear is certainly a whole lot better than fossil fuels.


According to many politicians, the Simpsons, the general public, and the anti-nuclear community, yes. But not according to geologists and engineers! The Finns have a deep geologic repository for long term waste storage under constructions [1].

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


Safe storage is not even remotely a technical problem. It's a political issue. NIMBY concerns over transport are the only barrier. The government already has a completely viable storage facility plan at Yucca Mountain.


> I don't think we're facing any critical electricity shortage

Shortage? No. But energy could always be cheaper.


Energy is a BIG, HUGE industry (about 5-10% of world GDP). It will always be a mix of technologies.

Cost wise, however, i agree nuclear is late to the party.

In 2018, we had large contracts for solar(+storage!) at 3cent/kw (in the US, Mexico and Saudi Arabia).

European new nuclear came in at 15c/kw.


That's a lot closer than you'd think then. While it varies from location to location, the capacity factor of solar seems to be about 20%. That means - sustained - a solar system generates 20% of it's maximum power rating, because the amount of power it produces is lessened during the winter, cloudy days, and of course at night.

The nuclear capacity factor, on the other hand, is rather high. Upwards of 90%, and sometimes even upwards of 100% (due to the powerplants being able to produce more energy than they were designed for).

If you factor that in, a nuclear plant has about a 4.5x capacity compared to a solar plant, meaning you'd have to spend 13.5c/kw for solar with the same capacity.

Of course, this varies greatly based on where you are. In sunny climates, solar's capacity is more like 30%, while in northern climates, it can be 15%. And I'd imagine the cost of new nuclear very much depends on regulations - in the US, it might as well be infinite.


I think you're applying an imaginary capacity factor to a number that already accounts for solar's intermittent yield. Solar is simply way cheaper than nuclear.


It depends if the $/kWh is cost to build or price paid for the energy.

I believe the .03 figure for the solar plants was $/kW and therefore had no bearing on total energy generated.

I think you would need a solar plant that incorporated energy storage to deliver nameplate power overnight for 40 years to be making an apples to apples comparison.


Does that account as well for the longevity of it? The current nuclear reactors lasted around 40 years already without much issues, I wouldn't bet the solar installations to last that long.


When you hear "nuclear reactors last for 40 years" you should instead hear "nuclear technology improves at such a slow pace that it's reasonable to talk about 40 year lifespans".

Almost no one cares if your desktop computer can last 40 years. And PV is improving so fast that making it last 40 years would be nearly as pointless.


Cost of products are averaged for their lifespan, if the solar panels only last 10 years, they need to be 4 times cheaper than nuclear to be at the same price per year.

And nuclear does not improve in a "slow paced", I don't know where you got that impression, there's been a lot of innovations on the last 10 years.


Nuclear technology has improved in cost in the past decade by a factor of 6, like PV has? If not, I will call its rate of improvement "slow".


that's not how percentages are working... There's clearly some math issues here... And the latest generation of plants are already 20% better than the previous ones.


Solar installations are generally rated for a 30 year life, so ¾ of nuclear isn’t bad, especially when you add in decommissioning costs which are going to be a lot lower for solar


UK discounted provision for decommission costs:

£100+ billion in 2013 https://www.theguardian.com/environment/2013/jun/23/britain-...

£161 billion in 2017 https://web.archive.org/web/20170516093449/https://www.gov.u...

£234 billion in 2018 https://www.gov.uk/government/publications/nuclear-provision...

... and counting.


> While it varies from location to location, the capacity factor of solar seems to be about 20%.

Another case for the importance of specifying solar variant.

Solar thermal plants with storage (molten salt, not battery) are above 70% CF already. eg the Torresol in Spain.


Average capacity factor of new utility-scale PV installations in the US southwest is above 30% now, I think. Single axis trackers (rotating on a north-south axis) have become the norm.


Is it still setting birds on fire?


Plus: the value of the power (energy) is variable throughout time: solar produces energy synchronously and tends to drive power prices down due to a very low marginal cost (hence negative power prices when there's a lot of wind and solar in Germany e.g.). When there's no wind and no solar, combined with a high demand (e.g. low temperatures in Europe), prices will be much much higher. So the remaining 70% of your capacity factor is worth more than the first 20%.


I have corrected the the PP

This is 3cent/khw and 15 c/kwh

This means it is cost per final energy produced i.e. after taking into account capacity factors, day/night, etc


EDIT UNIT CORRECTION: this is 3cent/khw and 15 c/kwh

This means it is cost per final energy produced i.e. after taking into account capacity factors, day/night, etc


I understand mostly because it has been regulated to death


I don't know man?

If you're familiar with dispatch stacks, you know that 3c/kw is REALLY attractive. Regulation or no regulation, I don't think you're gonna get down to 2c or 1c with a new nuclear plant.

Maybe I'm wrong, and one day someone will figure out how to build SUPER cheap nuclear plants?

But I doubt it.

Best I would think they could do is some sort of hybrid system that uses nuclear generation, and then storage via a conventional hydro storage system or something maybe? I'd have to do the math on that. But yeah... 2c? That's very hard to do.


Moltex is claiming they can get below $0.04/kWh (for N-th of a kind plants; FOAK would be a bit more expensive.)

They can claim this plausibly for two reasons: first, they can build a MUCH smaller containment building, due to lack of water or other volatiles in the nuclear island. Second, their design isolates the turbines from the nuclear island with a molten salt "thermal battery". This allows some dispatchability in their output, but more importantly it means turbine trips have no effect on reactor stability, which means non-nuclear-grade turbines can be used. This is apparently a very large cost savings.


Might be talking about European regulations. 15c per kw versus 3 in usa - thats 5x magnitude.

Doesnt suprise - gallon of gas in most european countries (sold by liters and funny thing - you first pump then pay!) is about 40% more than same amount in USA. it has been like this forever and doesnt look like will ever change.


I have doubts you’re going to be able to remove enough regulation to get nuclear near the same cost as storage backed renewables (which will continue to decline in cost each year even further).

Without storage, renewables are already close to or below 2 cents/kWh (unsubsidized). Can nuclear startups make generating units as fast as automated fabs can spit out solar panels?


A big argument I have heard in support of nuclear is that it is good for maintaining some amount of base load. As I understand this is particularly important in the current scenario of renewable energy without good storage, since we can't necessarily maintain our energy requirements with solely other clean (relatively speaking) forms of energy.


The race will be to see who can win the day: modular commodity nuclear, or utility scale battery storage (Flow batteries, lithium ion/polymer, etc).

Presumably, both get cheaper as you make more, but anyone can make batteries and ship them around the world with little notice. Tesla’s Hornsdale Power Reserve system was built in 90 days. I don’t know of any nuclear plants that get built faster than 10 years.


> I don’t know of any nuclear plants that get built faster than 10 years.

And the Brits are demonstrating just how woefully optimistic costs, build quality, and schedules can be even with that kind of protracted delivery expectation :

https://jeremyleggett.net/2018/08/22/why-the-uk-nuclear-rena...


I like regulating things that are radioactive or can potentially go boom. Especially both.

[And yes, I know some designs are theoretically not able to melt/blow up]


My main concern (I'm not a nuclear engineer) is that smaller nuclear plants means lots of radioactive material (fuel/waste) is being moved and stored in a piece-meal fashion, enabling "small scale" accidents or even theft.


I've always thought if the reactor was so small then why can't we just by default build the reactors underground? Isn't the underground a more stable area to begin with? Even during earthquakes? Reactor waste could be stored at the location permanently and if the reactor gets the 1 in a million accident then the whole place is already basically built in a cement/rock containment vessel. If a serious meltdown occurs then it would be easy to just start pouring lead or cement into the facility. Even if there was a small nuclear explosion we already know the impacts of underground nuclear explosions.


> why can't we just by default build the reactors underground?

It would appear to carry an increased risk of hidden problems and groundwater poisoning. There might also be issues with corrosion and accessibility for maintenance.

That said, I'm not ruling it out. A lesson these new approaches to nuclear engineering is that we never figured out the right defaults the last time around. We were too focussed on making bombs and submarines.

> we already know the impacts of underground nuclear explosions

Nuclear plants should never fail by way of nuclear explosion. If this is a possibility it is a bad reactor design.


> Nuclear plants should never fail by way of nuclear explosion. If this is a possibility it is a bad reactor design.

Fast reactors have always had this potential problem, since they require a much higher density of high enrichment material. A larger fast reactor might have a ton of plutonium in it; in an accident where fuel is melting and moving around unpredictably it's hard to prove that a prompt supercritical mass of Pu won't assemble somewhere. Edward Teller famously warned about this in 1967.

This is one advantage of MSRs as fast reactors: the fuel is already dispersed in salt, so such rearrangement is less plausible.


"Fast reactors have always had this potential problem"

- Not really. Check out the dramatic experiments at EBR-II (fast reactor) demonstrating unprotected loss of coolant at full power. It shut itself down. Without damage. And started back up later that day, and operated that year, as a research reactor, with better uptime (capacity factor) than the fleet at the time.


That just shows that in that particular accident scenario the situation would not arise.


Many of the current small reactor concepts are indeed designed to be installed below grade.


Most modern designs are specifically designed to not create materials that are easily weaponizable, which reduces the value of them (and thus theft). Accidents are an issue, and what do do with the waste is a huge question that will decide whether or not nuclear plants are economically feasible moving forward.


Radioactivity is invisible, and not detectable short of a Geiger counter. It doesn't even hurt at a dose that would eventually kill people. Also, depending on the radiation, radioactivity can make other materials dangerous.

And yes, the fuel they are talking about is a lot more radioactive than a banana or whatever.

I just worry that somebody with ill will or even just bad luck will defeat security measures "en detail", if the material is moving around in smaller parcels. Also, more people "legitimately" possessing a reactor and radioactive material will hardly decrease the risk of intentional abuse.


Just because one cant make a nuclear bomb from the material does not mean it cannot be weaponised.. Anything radioactive that can be spread into the food chain is an issue.


There's a lot there that depends on the specifics of the substances. Radioactivity doesn't mean dangerous or weaponizable. Bananas are radioactive. Granite countertops are radioactive.


I assume they are not fueling nuclear plants with bananas and granite.


Anything radioactive that can be spread into the food chain is an issue.

or mixed with a suicide belt's "materials"


If Terrapower or other molten chloride fast reactors work out, then waste is a solved problem.


No reason a bunch of small reactors can't be put together at one site. The reason for building small is just to reduce construction time risk, and maybe allow factory mass production.


Also not a nuclear engineer, but "mini-nuclear" anything doesn't conjure up anything good for me either. Hopefully, unlike in the tech world, the nuclear world thinks about and implements safety first rather than as an after thought.


I'm sure Uber is working on this.


very edgy, but this isn't reddit.


I heard Bill Gates on a ted talk a long time about talking about some sort of nuclear power that would burn waste and was like a candle. I think this isn't that.. but anyway, it'll be nice if we have excess power.


That's the traveling wave reactor which is just a breeder that doesn't need reprocessing. The first will be sodium metal cooled (not salt). This is a fluid fuel reactor that has similar reprocessing reducing goals.


Will this type of reactor produce material that can be used in nuclear bombs and produce waste we will have deal with for millions? Of years. Is it "just" a smaller version of the current types of reactors?


I've never understood this. If the waste product is radioactive for millions of years then doesn't that mean it's not very radioactive?

Wikipedia has this to say:

Since radioactive decay follows the half-life rule, the rate of decay is inversely proportional to the duration of decay. In other words, the radiation from a long-lived isotope like iodine-129 will be much less intense than that of a short-lived isotope like iodine-131.¹

So my understanding is radioactive products are dangerous because body takes up the elements and they're either chemically toxic, dangerous because of radioactive decay, or both.

So isn't this solved by vitrifying the waste and burying it? Glasses tend to be extremely chemically inert.

I'm not convinced the spectre of nuclear waste warrants the paranoia it receives.

1. https://en.wikipedia.org/wiki/Radioactive_waste#Physics


Material will eventually be decaying in the environment at the same rate it is produced, in steady state. This is independent of the halflife.

What a long halflife does is increase the time until that steady state is reached. If we operate a nuclear economy for a million years then a hell of a lot of that long lived isotope will have accumulated.


The only waste from molten salt fast reactors is fission products, i.e. atoms that already fissioned. Encase them in glass and they'll be back to the radioactivity of uranium ore in 300 years.

We could use these reactors to eat the nuclear waste we have right now.


Sometimes it fractionates me, how we manage to move from 1 to 1000x the computation power and yet Nuclear technologies seems to move at a very very slow pace.

I would also like to know why no one invest in Nuclear Fusion?


People have started investing in fusion. Tokamak Energy and a new company spun out of MIT's fusion program have both gotten tens of millions, for fairly mainstream designs. There are also companies using more speculative designs, the biggest being Tri Alpha, which started in 1999 I think and has about $500 million invested. There's also Helion, which got a modest investment from YCombinator.

One reason fission moves slowly in the U.S. is the NRC. A couple years ago I got to sit in a meeting between a bunch of U.S. reactor startups and a former head of the NRC. The reactor people's main complaint was that the NRC required detailed blueprints before it would even look at a design. It would take several hundred million dollars to get to that point, and then the NRC would give a flat yes or no. With a no you were done, and with a yes you still have nothing but a paper reactor. It's a very difficult environment for investors.


Good news for Japan if this gets developed successfully. We have a ton of nuclear reactors on some of the most seismically active land on the planet - only a matter of time for another Fukushima.


By Energy Department, they mean the Department of Energy (DoE), right?


According to the article, yes


Is part of the benefit to these that the risk can be more accurately calculated and insured against?


Wasn't there a YC backed company doing something similar with nuclear a while back?


There are like 50 advanced reactor companies out there. But I think the one you're referring to is Oklo.

https://www.thirdway.org/infographic/the-advanced-nuclear-in...


Or possibly Transatomic, which rumor has it has recently shuttered its doors.


What could possibly go wrong?

Radioactive material everywhere in private hands.


Smart people like Bill Gates are interested in nuclear because of the low footprint associated with the unbelievably high energy density of nuclear fuel. 1.5 soda cans of the stuff fissioned in a conventional reactor releases enough energy to power an average american's entire life (including transportation, heating, and electricity). The waste created is about 2 soda cans of toxic solid stuff that can be easily, safely, and practically stored in deep crystalline bedrock like the Finns are doing. Other energy sources dump their (more dilute) waste into the air causing lung disease (worldwide fossil fuels are estimated to kill 3 million per year this way) and global warming (death toll still unknown). Intermittent renewables are energy harvesters with free fuel but they require vast swaths of land, concrete, steel, fiberglass, storage, transmission, rare-earths, lithium, etc. etc. Nuclear has tiny land, material, waste, carbon, etc. footprints due to its energy density. The risks associated with radiation are just overblown by almost everyone. Nuclear reactors have net saved 1.8 million lives already, and counting [1].

[1] https://pubs.acs.org/doi/abs/10.1021/es3051197


I see you have never had to desl with radioactive material and activated tools in private companies.

It is not a pretty sight. But hey, keep advertising nuclear. How much do they pay you?


Could you please improve your commenting? We're looking for civil and substantive posts that are informative.

https://news.ycombinator.com/newsguidelines.html


I think this is a great idea. That said, this article seems like a puff piece for the DoE under Trump. Nor does the article linked really support the headline either.

Consider the source.

"When Anschutz first started the Examiner in its daily newspaper format, he envisioned creating a competitor to The Washington Post with a conservative editorial line. According to Politico, "When it came to the editorial page, Anschutz's instructions were explicit—he 'wanted nothing but conservative columns and conservative op-ed writers,' said one former employee." The Examiner's writers have included Michael Barone, Tim Cavanaugh, David Freddoso, Tara Palmeri, Rudy Takala, and Byron York."

That said, don't take my word for it, look at their home page (https://www.washingtonexaminer.com/) and judge for yourself.


I myself is against idea of small scale nuclear. Economics is not of its side. Every single nuclear facility is has a lot of fixed expenses to keep it running. Think of the amount of qualified cadres needed to run it: every nuclear power station is effectively a small science institute on its own.

Multi-gigawatt facilities are the ones where nuclear has biggest the payout.

The few nuclear power stations that are being completed these days are single reactor designs and effectively experimental plants. This is why digits on economics got unfavourable recently.

Fuel costs are negligible in comparison to every other operating expenses. This way, simple PWRs and BWRs on multi-gigawatt scales are the only designs making sense economically. Next gen CANDUs make great sense for fuel economy (and economic costs of refuelling), but even then, frequent refuelling needs are not as dramatic as the higher costs of CANDUs. Heavy water price has only been going up - and even something like an international initiative to setup a global heavy water "bank" will help little.

4th Generation reactors - the first few reactors are assuredly can't run at profit. Their sole purpose are to be research facilities. Any nation constructing them should have full realisation that constructing gen 4 only servers the purpose of advancing its science and industrial competence, not and making money in any immediate future.


I used to feel this way: that large nukes are the answer because they provide megacity-scale energy for an increasingly urban world. But as I've studied the nuclear industry a few things have eroded away at this belief.

First, iteration. You can iterate on the design and supply chain of a small niche nuclear reactor much faster than a large baseload one. And you can do so without bankrupting the likes of Westinghouse or Areva. Once you figure out a new design at small scale, you can scale up. We did with with the current reactor designs and we should do it again for the newer fancy ones. I think this is what Oklo and similar reactor companies are planning on.

Second, and more novel, is that you can imagine large-scale nuclear farms of small reactors. Build small reactors that are sized to handle their own decay heat without fancy safety systems and manufacture them in a factory. Put them out there on shared security infrastructure and siting. There's more energy produced from internal combustion engines than all baseload plants in China, for instance. "You want scale? I'll give you scale!"

And finally if you still need to go big, there's also the potential of large modular reactors. If you build huge power plants in shipyards on floating platforms you can get economies of scale AND economies of mass production, while improving safety by being intimately coupled to the ultimate heat sink: the ocean. Wild idea politically but technically very intriguing. Would require much more remote operation than current plants because a crew of nuclear rough necks would have high salaries. But totally doable, and could probably decarbonize the world very rapidly.


> Every single nuclear facility is has a lot of fixed expenses to keep it running

Counterintuitively, economies of scale are more frequently found in small products which can be mass manufactured than big projects which must be custom built. This isn't an ironclad law. But neither is "nuclear must be big."

> the first few reactors are assuredly can't run at profit

"Assuredly" based on what? This article is about a commercial reactor project. One of many. Their investors have less of an interest in "advancing [their nations'] science and industrial competence" than in turning a profit.


>Counterintuitively, economies of scale are more frequently found in small products which can be mass manufactured than big projects which must be custom built. This isn't an ironclad law. But neither is "nuclear must be big."

Nuclear power stations are not a mass manufactured product. The most pass produced ones still amount to decade long projects. There are no nuclear reactors that amount to "just add water" replicate designs, and there wouldn't be any in foreseeable future.

Just look at small coal firing powerplants, even smallest ones suitable for use in utilities are horrifically inefficient, horrifically expensive per kW/h, and still involve tons of on-site and custom engineering solutions every time. All and every vendor of "turnkey" powerplants market them as such, but just ask any power engineer if it is actually so: coal feed systems have to be optimised for type of coal used, flue solids capture systems have to be customised, foundations have to be custom engineered to accommodate heavy machinery (imagine a turbine shaft of many tonnes to get few mm bend due to soil subduction), and of course the geometry of the plant has to accommodate position of coal storage, flyash storage, transformer sites, administrative buildings and etc.

>"Assuredly" based on what?

Based to watching nation scale efforts of gen 4 reactor commercialisation failing every time. All of them wanted money from day one, and all were sure that "this time we got it right."

>Their investors have less of an interest in "advancing [their nations'] science and industrial competence" than in turning a profit.

Given the known level of intellectual ability of a successful finance professional (aka "Pro Investor") in a Western country, this doesn't surprise me at all.


> Nuclear power stations are not a mass manufactured product. The most pass produced ones still amount to decade long projects

You're arguing a car is impossible because horses only run so fast. Coal is a bad model for nukes because coal has a low power density per unit of fuel. You have to burn lots of fuel to get a meaningful amount of power, and burning lots of fuel takes lots of space. You naturally get a large plant.

There is no similar fundamental restriction on the size of a nuclear reactor. RTGs, for example, are tiny (albeit inefficient) nuclear reactors [1]. Here we have commercial efforts to miniaturise nuclear power plants so they can be mass manufactured.

"Past projects were X so it will always be X" doesn't make sense in a barely-explored and novel technological domain.

> the level of intellectual ability of a successful finance professional (aka "Pro Investor") in a Western country

"Pro investors" (which I take to mean generic fund managers) aren't backing nuclear start-ups. VCs with domain expertise, laboratories, engineers and utilities (together with the public sector) are.

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


>"Past projects were X so it will always be X" doesn't make sense in a barely-explored and novel technological domain.

Nuclear engineering is by far not a barely-explored and novel technological domain, but an industry with near 70 year history with extremely high barriers to entry.

For you, it should be reasonable to believe that engineers and physicists with multiple postdocs, people of far greater intellectual achievements that you, me, and probably most of this website's demographic, were banging their heads non-stop for the last 70 years on the problem of making nuclear power stations economical. Nothing what came from years of their work indicate that anything but big and simple PWRs and BWRs makes sense economically.

It took decades to shave everything that can be shaved off the construction of gen 3 plus reactors to arrive at their current designs. Anything that will be employed in a design claiming bigger efficiencies has to be using less than a bare minimum used by them and that's not a lot:

- Nuclear safe superalloys - check

- Large scale forging - check

- Refueling infrastructure - check

- No mechanisation beyond control rods and pumps - check

- Advanced sensing - check

- Fuel assemblies - ridiculously cheap in relation to everything else

- On-line chemical filtration and corrosion control - check

- Steam power plant - check

Really, the only opening for lower costs are reduction in operation expenses - less scientists on site, less consumables, less or not manual labour intensive servicing, less frequent and cheaper refuelling. Besides refuelling costs, that's not a lot.


> it took decades to shave everything that can be shaved off the construction of gen 3 plus reactors to arrive at their current designs

One of the game changers are new superconductors. These are smaller, lighter and cheaper than anything we've had before. They're a very recent product of American and European colliders. Their use is common across many compact designs.

> Anything that will be employed in a design claiming bigger efficiencies has to be using less than a bare minimum used by them

Not necessarily. For example, a low-efficiency design that is simple to mass manufacture could be cheaper to build and operate than an expensive, efficient design.

Furthermore, for each element you list as being necessary there are designs which do without them. For example, Tri-Alpha uses direct energy capture (no steam power plant). Several companies eschew refueling in favor of capsulation. Large-scale forging is an odd thing to list when we're talking about minitiarrisation. Et cetera.

A small amount of research would show many arguments you present as established fact to be untrue or unfounded.


At least two of those points can arguably be greatly improved or avoided in MSRs. Since the reactor is not pressurized in an MSR, no large heavy forged pressure vessel is needed. And the steam cycle of a MSR can operate with supercritical steam w. reheat, like a conventional fossil plant's steam cycle, reducing the mass flow needed by as much as as factor of two.

The other big savings for an MSR is a very large reduction in the size of the containment building. This is because there isn't water in there that would turn to large volumes of steam during an accident and that would have to be contained by a large pressure-bearing structure. Moltex claims their containment would be a factor of five smaller per unit of power output, compared to a LWR. This would be a huge cost advantage.




Consider applying for YC's W25 batch! Applications are open till Nov 12.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: