Westinghouse sees a tech disrupter in its eVinci microreactor | Hacker News Reader

Westinghouse sees a tech disrupter in its eVinci microreactor

power-eng.com

353 pointsakeck277 comments

It's a liquid-sodium-cooled, graphite-core system with no need for pumps, aka 'passively cooled' via heat [edit] pipes. Output is:

> "The microreactor can generate 5 MW of electricity or 13 MW of heat from a 15 MW thermal core. Exhaust heat from the power conversion system can be used for district heating applications or low-temperature steam."

Control systems are kind of interesting:

> "The only moving or mechanical parts in the reactor system are reactivity control drums, which manage the power level and allow absorber material to passively turn inward toward the core if power demand is reduced or lost, and turn a reflector material toward the core if demand increases automatically. Hence the term “nuclear battery.”"

I'm generally not a nuclear advocate but if they've really managed to eliminate the need for active cooling, and have a robust system that can safely shut down with concerns about meltdown even without external power, that's a pretty big advance. Looks remarkably promising... keep your fingers crossed. (New nuclear tech hasn't had the greatest track record over the past several decades, i.e. pebble beds didn't work out etc.)

Not trying to be a pedant, but I think you might have a typo in the core of your point:

> …, and have a robust system that can safely shut down with concern about meltdown

Do you mean ‘without’?

I think it's 'with', as in 'regarding'. "With the rising concerns around orange peel ethics, we put focus on creating a humane peeling device for fruits."

I am intrigued by this concept and would like to learn more!

oops left out a 'no' between with and concern.

It uses TRISO fuel, which are fissionable materials enclosed in a carbon and ceramic shell that's extremely tough and can handle far higher temperatures than are present in a reactor without melting.

The shell prevents the release of radioactive materials, so TRISO fuel encapsulation would turn a melt down into a reactor damaging event rather than an emergency/crisis. A runaway reactor would get hot, melt its structure, and maybe drop parts inside its containment, but the fuel wouldn't melt and wouldn't travel far. Most likely the reaction would stop as soon as the structure of the core fell apart.

I wonder if that fuel is usable in a breeder after it's "spent". Usually "spent fuel" has spent a few percent of the usable fissionable material, but the products of decay (and partly synthesis) prevent it from working efficiently.

> It uses TRISO fuel, which are fissionable materials enclosed in a carbon and ceramic shell that's extremely tough and can handle far higher temperatures than are present in a reactor without melting.

I was going to make a top level comment along the lines of:

"How thick/expensive a bunker would one need to build around one of these to prevent it from spreading radioactive materials about when hit by say, an airplane?

As I understand it, large legacy reactors are hardened for this. Would the same level of safety still allow these smaller reactors to be economically viable?"

But now that I have read your comment, would a kinetic event happening to a TRISO fueled reactor be more of a non-event? Follow up, would non-fuel components of the reactor become radioactive over time as well?

Sorry if dumb questions, not very educated in this space.

For small reactors, with minimal requirements for support equipment and ~no need for active cooling during emergency shutdown, the design trend I've seen is "put it underground". Which makes sense - bomb shelters, military bunkers, etc. have been put underground for centuries because that's a relatively quick & cheap way to make something that's very hard to destroy.

The small size is a serious advantage for this kind of thing.

I could be wrong, I'm not an expert on the matter by any means, but I think the reason traditional nuclear reactors are so heavily fortified against attack is specifically because of the threat of a catastrophic meltdown if their cooling and control systems are destroyed. If these nuclear batteries can't melt down and rely on passive safety measures then the need for fortification is drastically reduced. Placing them in a fairly standard concrete building either underground or bermed with earth on all sides would likely be sufficient. You'd probably still want security around to make sure no one tried to manually tamper with the equipment.

It’s good to see the designers absorbed a dose of inspiration from the pioneering excursions in US sodium reactors.

Mechanical coolant pumps were the Achilles heel of the infamous Sodium Reactor Experiment (SRE) from 1957-64 at Santa Susana field laboratory. It had 50,000 lbs of liquid sodium in two coolant loops.

When 4 pints of tetralin leaked from a pump seal into molten sodium surrounding the SRE reactor core (500-950F), it fouled the fuel cladding with “brown stuff” and various fission products from the melted rods were found on both sides of the fence. [1]

A “small amount of sodium” in sealed heat pipes sounds pretty safe. But since Westinghouse filed all the eVinci’s NRC pre-application materials as “proprietary”, the actual design details aren’t available to the public for review.

“Disruption” I assume refers to hundreds of rule exemptions (licensing innovations) they filed as a non-LWR. Not judging, just stating the record. [2]

1. https://www.etec.energy.gov/Library/Main/Doc._No._1_SRE_Fuel...

2. https://www.nrc.gov/reactors/new-reactors/advanced/licensing...

> Westinghouse looks to off-grid applications like remote communities and mine sites as the entry market for eVinci.

Yeah, install a nuclear reactor at a remote mining site, probably without sufficiently trained personnel to take care of it. What could go wrong?

Why would you assume untrained personnel? These would still be regulated by the DoE.

What about security? In the UK nuclear installations are under 24/7 armed guard from a special police force. That would be expensive to maintain in a remote region, although I guess they already have security at diamond mines, or instance

Vastness of rural US is a security factor on its own. Plenty of military installations in the desert aren't even fenced.

They’re perhaps not fenced but presumably they’re full of armed military personnel?

You've never really been out in the American West, have you?

Just one example - the Nevada Test and Training Range alone is a bit over 4,500 square miles.

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

Vs. Wikipedia says the entire U.S. Army (active duty) has ~485,000 personnel.

Divide...and that's ~0.17 Army people per acre, even if the US Army had NO job except holding down the NTTR. (Which is actually US Air Force, BTW.)

[Edit: But +1 - because in the bigger picture you make a valid point. However vast the open spaces, at the spots where the cool & expensive stuff sits, there does have to be some sort of "real" security. If only so dodgy locals don't swing by with a pair of bolt cutters, and start helping themselves.]

Thanks for the perspective, you are indeed right that I’ve never been “out there”.

There was a German family once that might have paid with their lives for that presumption.

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

I guess that comparing it to the cost of setting up a super long power line and maintaining it (sabotages, fires, whatever), maybe it's interesting.

Reasonable Assumption: The "remote communities", "mining sites", etc. which both qualify for a nuclear power plant, and can afford it, will be fairly select & high-budget places.

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

Mines aren't run with pickaxes and donkeys anymore. Take some time to educate yourself.

I would be more concerned about security than trained personnel to run it. Obviously, there will be trained people on site. You don't put a small reactor onsite and tell Cletus and Bubba to press this button to start it and that button to stop it.

Until an explosion damages the heat-pipes and you get a runaway meltdown and explosion of the core after hydrogen gas builds up.

Very realistic scenario if the thing is used in a war zone.

Being sodium cooled, both a steam explosion and hydrogen gas generation via water is ac minimized. Lots of safety advantages to a coolant that is still liquid at operating core temperature with no added pressure. But of course liquid sodium does come with some other caveats that are of concern. Making it a sealed unit with minimum moving parts helps with some of those potential problems for sure. As does keeping the total thermal capacity relatively low, below the threshold where secondary fission products need active cooling to prevent meltdown even after the reactor is shutdown.

If the price is right and medium size communities could get their acts together, something like this could potentially disrupt the entire grid model. In California, regardless of what wholesale electricity costs, the retail cost is something like $200/MWh more than could be considered reasonable. Put another way, the utility (PG&E) is charging an immense premium. Normally, displacing PG&E would be impractical:

a. The actual transmission system is a phenomenally large capital investment developed over many decades. You can’t just VC up a new electric grid in a developed area. And the incumbent mostly owns the existing infrastructure.

b. Regulation, good and bad.

It’s possible to sell power to the utility for a reasonable price per MWh. But one can’t easily sell to the utility’s customers.

But this reactor is small! 5 MW could serve maybe 1000 expensive homes in an expensive area without an enormous transmission system. Anyone trying to disrupt the incumbent utility with something like this has $200/MWh of inefficiency to exploit. $24k per day of operation will offset a decent amount of capital cost and regulatory effort to get the electricity to customers.

Put another way, a wealthy community could buy a few of these, figure out local distribution, and ditch the incumbent utility. This could be fantastic.

I like the idea but not sure if wealthy homeowners would want to a nuclear reactor in the vicinity of their homes. It would have to be proven out over a couple of decades of incident-free operation before it becomes acceptable.

Los Altos Hills resident here. Sign me up for one in my yard. There's an empty lower acre begging to do something useful.

> There's an empty lower acre begging to do something useful.

How about building housing there? As long as house prices are as high as they are today, that should be the main priority.

Local zoning regulations don't allow it. But backup power systems? No problem.

What you're looking for is just a very small utility company. At 5MW of load you'll be needing distribution, substations, maintenance for all the above, planning, etc. The overhead you'd need at such a small scale of sales would probably surprise you.

And yet most of the world, and even most of the US, does this, including generation, for some $200/MWH less than PG&E. I’m not saying it’s cheap. I’m saying that the amount of money on the table is very large.

(Distribution cost really ought to scale roughly linearly in the length of the system. I don’t know if it does in practice. As a data point, Palo Alto manages to run their entire system for considerably less money than PG&E, and Palo Alto is not an inexpensive place to do anything.)

a. Yes, but you can easily put up solar+batteries such that the cost of retail energy to the consumer is entirely transmission line rental, at which point disruption is worth whatever the cost of line rental is minus the capital cost of grid construction.

> It’s possible to sell power to the utility for a reasonable price per MWh. But one can’t easily sell to the utility’s customers.

You might not be able to sell energy directly to them, but you can sell/lease energy generation to them which is essentially the same thing. You won't be able to charge the line rental that makes up the majority of the retail cost post installation of solar, but you can start with some fraction of the difference between they save with the solar system, which is probably at a better price point than wholesale cost of electricity.

Good thing everybody lives in Arizona and Sahara, and nobody lives in Sweden, Canada, or even in Seattle! /s

Sadly, sun is not universally abundant.

But HVDC lines are. The longest one in operation today is ~2200 km.

I wonder how their upkeep and bulging expense (especially across difficult terrain) compares to that of the small nuclear power plant described here. We could determine the distance when the HVDC line still makes more sense.

For true "wilderness" your total addressable market is like a million people. Sweden and Norway have abundant hydro power so they actually export power from north to south. So reduce that to tens of thousands.

For southern scandinavia it is already connected through HVDC to Central Europe and more are being built.

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

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

Nobody knows what this reactor is going to cost yet, the first one has not been completed ..

True, but consider that most of southern Asia, Oceania, Africa and central America the sun is abundant. That covers a lot of people.

Now we just need a bunch of these reactors to power the desalination plants to support all the folks living in the desert.

The Netherlands is about as sunny as Seattle, and it's possible to break even after a few years with a decent-sized solar array on your roof.

Jeesh, how expensive is power there? It would take me almost 15 years to get my money back on a solar/battery combo that is big enough to generally supply my electricity needs if I was at 100%.

Currently paying around €90 for ~250KWh per month. That is about 4 times what I used to pay in South Africa 5 years ago.

In Belgium we're at €0,50/kwh We're ordering 36 panels and 10kw battery tomorrow.

The only sad thing is that we can't utlize the battery when we lose netpower. Huawei has an addition, but it only gives an outlet on the battery. I wanted inline, like my ups.

Your payback would be a lot faster without a battery I think. At least that is the case where I live (US/East).

Could these be viable for shipping containers??

IIUC, 18,000 TEU ships have engines that generate ~60 mW - and use about 66,000 gallons of diesel per day.

That's ~$90M in fuel per year, which could be replaced by ~12 of there reactors.

That's $7.5M per year. If these can be purchased for less than $20M - with an 8-year life-span and $2M for fuel and operation per year - then maybe it could work??

If you do that the ships would go (dangerously ?) fast.

60 MW output from a combustion engine means rougly 3x the mechanical energy output for the same amount of Joules input with an electric engine.

On the plus side, container ship drag races would now be a thing.

Imagine how far they travel when trying to stop. I guess you would have the power to run the engines in reverse at full power too.

Power isn't the problem; vibration, the stress on the ship's hull, and risk of losing a propellor are the real problems with sustained running in reverse at full power.

Stop ? Just drift with all your cargo.

You'd get surprisingly little additional speed. An old rule of thumb is that if you double the power of a large ship, you get 4 additional knots of speed. Current average laden container ship speed is ~15 knots, down from about 20 knots from back before they cared about fuel costs and emissions. So you'd only get speeds up to what they were before.

I think the 60MW output is shaft horsepower so it already takes into account the low efficiency of converting fuel oil to movement (so in reality it's burning about 150MW of fuel oil).

It would require changes in regulations around the world. Many ports do not allow nuclear powered vessels to dock. Those regs were written with traditional nuclear power plants in mind but I think there would still have to be a lot of work to get acceptance.

Would these reactors take up less physical space than the fuel system they would displace?

Almost certainly. See picture [1]. According to Wikipedia, they're ~87 ft x ~87 ft x ~44 ft = 333k cuft [2]

These reactors are supposed to be a max of 4 containers in size.

So ~12 of them would be ~48 containers = 4x6x2 = ~80 ft x ~60 ft x ~30 ft = 144k cuft.

Though, I doubt you'd want to stack them like that, and I imagine you'd need more than just the reactors to replace the engine.

But more importantly - container ships need space and weight for ~66k gallons of diesel per day - about ~3M gallons total. That's an additional ~401k cuft for diesel, and ~13,500 tons... Each TEU is 22 tons - so you could probably carry an additional 600 TEU (or ~3.5% more).

[1] https://www.amusingplanet.com/2013/03/the-largest-and-most-p...

[2] https://en.wikipedia.org/wiki/W%C3%A4rtsil%C3%A4-Sulzer_RTA9...

> In California, regardless of what wholesale electricity costs, the retail cost is something like $200/MWh more than could be considered reasonable.

$200/MWh is $2/kWh. California electricity prices are around $0,27/kWh (https://www.energysage.com/local-data/electricity-cost/ca/). I guess you mean $200/GWh?

Also, why do you find PG&Ek’s pricing unreasonable, given that it required a “phenomenally large capital investment” (and, I expect, quite a bit of ongoing maintenance) to be able to provide customers with electricity?

> $200/MWh is $2/kWh.

$200/MWh is $0.20/kWh.

Oops. Thanks

PG&E pricing is higher than that, ranging from 37 to 49 cents per kWh on their default residential plan. That's more than three times the national average. It's unclear why the state expects folks to electrify when electricity is outrageously expensive.

https://www.pge.com/tariffs/assets/pdf/tariffbook/ELEC_SCHED...

There is the small issue of transmission and maintaining that. But surely it is better to also use the grid for when you peak over 5MW at dinner time when people turn on their ovens, lights and other appliances and also to be able to sell back to the grid at other times. Unless you have batteries.

But your argument is solid for a factory with constant 24/7 load where a single building might consume all that power.

The spreads are egregious, though. You can buy from PG&E for several hundred dollars per MWH. You can likely sell to them for a couple dollars per MWH. No, you do not get NEM for your neighborhood nuclear plant.

Or you can overprovision, discard unused output (or try to do something useful — even just heating up a big heat reservoir is useful for district heating if you feel like building out a system to deliver the heat), and still potentially win on overall price.

Large grids are economical when reasonably run. They are not economical when poorly run.

That is horrendous. Battery it is!

Residential solar plans in Sydney would buy power back for about 5c US / kwh so $50 / mwh

Although I think their hand might be forced to encourage solar. They used to pay crazy high amounts.

expensive houses require a lot more than 5 kw

I recall some California communities without PG&E have somewhat reasonable electric costs. Santa Clara has silicon valley power with rates of .11-.13/kwh which are significantly lower than PG&E at .28-.51/kwh

PG&E is expensive.

[Cries in European]

Northern communities in Canada are a perfect application of this tech. Currently, they mostly burn diesel for power.

This sort of technology is something the provincial governments are interested in.

https://www.cbc.ca/news/canada/calgary/nuclear-energy-albert...

Might also be good for concrete production, perhaps. Most cement plants burn loads of coal to heat up the limestone, producing loads of CO2 in the process. Not sure if this reactor could heat up the kiln to >1,000C. Seems like there's some interest in the UK, so perhaps this could be a good solution for on-site production in remote areas...

https://www.theconstructionindex.co.uk/news/view/hinkley-poi...

Sodium reactor outlet temperatures are around 550 C. There are other reactor concepts that come close to 1000 C, but it's a huge technical challenge.

Electric heating?

That works for some processes (making aluminum and more recently steel) but I'm not aware of any electrically heated processes for making concrete. Cheap, clean electricity could unlock a lot of things though.

Heating by electricity is not a problem. But both production and setting of concrete chemically produce CO2.

Hydraulic cement does not produce CO2 during setting. Non-hydraulic cement actually consumes CO2 while setting.

Doesn't it essentially just re-absorb the same co2 it emitted when it was made?

Yes. I was replying to parent's comment that concrete produces CO2 when setting, which is clearly not the case for hydraulic cement.

Probably, like for steel production, it would involve producing hydrogen then using that as a heat source.

I read an article about this recently and there's at least one company working on exactly such a process:

> Swedish green-tech firm SaltX Technology demonstrated that it can produce clinker with its Electric Arc Calciner: a proprietary system similar to the plasma torches widely used by automakers and other manufacturers for cutting metal. Plasma torches pass an electric current through a jet of inert gas, typically nitrogen or argon, which ionizes the gas and heats it to temperatures over 20,000 degrees Celsius. In June, SaltX announced a partnership with the Swedish limestone supplier SMA Mineral to accelerate commercialization of its technology.

Article: https://arstechnica.com/science/2022/11/the-road-to-low-carb...

Science writer David Roberts had a long conversation with Rebecca Dell from ClimateWorks about decarbonizing heavy industry back in February. Lots more detail here if that’s of interest to you: https://www.volts.wtf/p/volts-podcast-rebecca-dell-on-decarb...

(n.b. copy and paste of my comment from 24 days ago: https://news.ycombinator.com/item?id=33396422)

There will be surplus power, but that's not a bad thing - make bricks for housing with it; breakage (power not yet strictly needed) can go to indoor greenhouses, etc.

Desalinating seawater can consume huge amounts of surplus power, not necessarily stable.

(Extracting lithium, uranium, thorium, gold from the brine could be a bonus step.)

There aren't a lot of places in the north where you'd need desalination; streams suffice for thin populations. Elsewhere in the world, I can see that.

Indeed so! Closer to tropics, such places abound though. "Excess" solar generation can go there.

Unlike in the north, where there's no excess.

Honestly, I've been looking at just that in combination with agricultural/reforestation techs for a couple years now. Australia is way bigger than the maps make it look, Africa too. Find an area near a volcano you can rob and grind for soil and you'll do very well once energy is cheap enough. Due to climate change - esp in future - many locations will require heat control via shade (not too dense solar farms) and evaporation (of sea water on cardboard in one trial.) No doubt the current food shortage is temporary, the solutions emerge inevitably from cheaper energy (and AI farm equipment.)

all the communities that survive by selling diesel will see it as negative.

All communities in the world except maybe France (already nuclear) and some hydro-electric countries are a perfect application of this tech. Currently, most of the world burns diesel/coal/gas for most of their energy needs and politicians are actively preventing that from changing.

PS: oh hello Germany, looking at you specifically.

Sure, if you want to scrap all renewable energy and run your entire grid solely on nuclear, that can work.

But what you can't do is mix them, as you can't regulate the output from nuclear reactors fast enough to work together with renewable energy plants.

I recommend looking into base vs peak loading for the energy grid. Nuclear is good as base load power, batteries are good for peak loading and renewables are good for charging batteries.

It's fairly easy to mix nuclear and renewables technically, it just doesn't make much financial sense in most places, particularly if you have to build new nuclear or if you're not too near the poles.

you can't regulate the output from nuclear reactors fast enough

From the article:

"The company touts the microreactor’s solid core and advanced heat pipes, which enable passive cooling and also allow for autonomous operation and load following."

Load following to change your thermal output over minutes or hours isn't frequency correcting which isn't grid stability. Steam turbines can't change speed quickly. Some exotic closed CO2 turbine might.

Can we stop this rubbish? German by percentage uses as much natural gas as France (they use more in absolute numbers), so calling out Germany is just rubbish nuclear propaganda. The reliability of nuclear is a myth, France had to shut down >60% of their nuclear power plants this summer. If nuclear cost 3x as much as renewables why would I not buy 3 times the capacity in renewables. And consider g that this is completely unproven tech I seriously doubt it will be even within the same order of magnitude of current nuclear prices.

>> German by percentage uses as much natural gas as France

False. France uses around 28% non-nuclear (2018 numbers).

Germany uses on average ~65% non-"green" (last I checked) of which the vast majority is gas (especially recently now that they shut down their nuclear). There's no way that France is using equal or more. Not in relative numbers nor in absolute numbers. Source please?

Just to be clear: we're not talking installed potential energy conversion capacity. We're talking actually produced electricity.

>> The reliability of nuclear is a myth, France had to shut down >60% of their nuclear power plants this summer.

So because France fails to do proper maintenance for decades, that suddenly means that nuclear in general is unreliable?

France has had serious nuclear conversion running non-stop for over 40 years and because a lot of that is down for a few months (due to dumb delayed maintenance), that means nuclear is suddenly "unreliable"?

>> If nuclear cost 3x as much as renewables why would I not buy 3 times the capacity in renewables.

Because even with 10.000x the required capacity in renewables you would still need something for base-load. And that something needs to be able to convert ~100% of your energy need for the entire country when your renewables are doing ~0% (exaggerating to make a point). Which is the root-cause of the EU energy crisis (German base load = mostly gas).

Another reason is that "renewables" are not actually renewable at all and have a limited operational life, economic life and huge recycling problems (most solar panels installed now cannot be recycled. At all. Just to give an example).

>> I seriously doubt it will be even within the same order of magnitude of current nuclear prices.

Speculation.

> False. France uses around 28% non-nuclear (2018 numbers).

> Germany uses on average ~65% non-"green" (last I checked) of which the vast majority is gas (especially recently now that they shut down their nuclear). There's no way that France is using equal or more. Not in relative numbers nor in absolute numbers. Source please?

News flash, it's not 2018, and France's nuclear fleet has been having problems for a few years now. They also rely on importing coal and gas energy in winter even when it's actually working. France is better decarbonization wise, but germany has been hovering around 40-46% renewable electricity for a while, you also have to account for more electrification in France. It's about 22% vs 50% for primary energy, but the key take-home is the rate.

https://ourworldindata.org/grapher/primary-energy-source-bar...

> France has had serious nuclear conversion running non-stop for over 40 years and because a lot of that is down for a few months (due to dumb delayed maintenance), that means nuclear is suddenly "unreliable"?

Unplanned outages is a consistent pattern in nuclear everywhere except USA and China. Although if you correctly count overruns as an unplanned lack of generation, it's basically just China

> Because even with 10.000x the required capacity in renewables you would still need something for base-load. And that something needs to be able to convert ~100% of your energy need for the entire country when your renewables are doing ~0% (exaggerating to make a point). Which is the root-cause of the EU energy crisis (German base load = mostly gas).

Another myth. Most of the gas is for heating and other non-electric energy. Germany had to start up coal plants in large part to make up for france's massive shortfall. Uncorrelated renewables can provide a large fraction of power even with negligible storage or hydro. In Western Australia renewables hit 40% average recently. Interconnects, storage and dispatchable power like hydro increase it further (or rather make up for lower solar CF). France is still doing better than germany overall, but at vastly greater expense and Germany's renewable plans were hobbled by barvaria and a head of state who literally works for a Russian gas company.

> Another reason is that "renewables" are not actually renewable at all and have a limited operational life, economic life and huge recycling problems (most solar panels installed now cannot be recycled. At all. Just to give an example).

Another lie. All new PV in the EU must be recycled and the seller is responsible. The supply chains for this can handle any mono or poly silicon panel. Thin film are an obsolete tech, and the metals are safely encased in glass awaiting a time someone wants them. There is significantly more low level nuclear waste than total mass of pv for the same energy output, and orders of magnitude more mine tailings. Wind turbine blades are already finding second lives as building materials and structural elements, and even if they don't they're outmassed significantly by the low level and decomissioning waste of a nuclear reactor.

The only thing that comes even close to the uranium mine tailings in quantity and is not recyclableat a profit is the concrete foundations for wind, but they're not full of toxic heavy metals.

https://ourworldindata.org/grapher/carbon-intensity-electric...

With an ageing nuclear fleet built mostly in the 70s and 80s France has the cleanest electricity of any country in Europe (except those with abundant hydro).

I love seeing innovative tech like this spring out of the Pittsburgh region… Once the nation’s capitol of industry innovation, maybe it can be again.

the rust belt has good bones, as they say.

I had a couple trips to the Google building in the converted Nabisco factory there.

They'd refinished the hardwood floors, but they left in the giant gouges that the heavy machinery had put there, as a reminder of the past.

> maybe it can be again.

If we can get the Jones Act repealed or heavily amended, maybe some long overdue riverine infrastructure maintenance and upgrades and we could watch the whole region flourish.

This discussion has nothing to do with the Jones Act. Bot spam!

How would repealing the Jones Act help? By allowing more ships on the Allegheny River? If anything, I think Pittsburgh would have benefitted from a little protection before cheap foreign steel hollowed out the industrial base.

> How would repealing the Jones Act help?

By removing the restrictions that make riverine transport uncompetitive in many cases with trains and trucks. Without those restrictions sending goods by boat is far cheaper (and less carbon intensive) than the alternatives which would give industry in the entire Mississippi River system a boost in competitiveness. There’s a reason a lot of value add industries are located beside ports. The Jones Act effectively removed the likes of Pittsburgh from the list of port cities.

The Jones Act was passed in 1920?

Wow... Walt's Mill. My mom worked there, I have a photo of her standing above a fuel storage well... It used to be a test reactor with a channel of water through the core that they would send boats through to test samples of materials for radiation exposure effects.

I'm sure these boats were quite small yet I'm imagining an underground river with rafts crewed by technicians in mopp suits.

Ever wonder why Oompa Loompas are orange?

Anti-radioactive-tan spray.

In 1960, there was a meltdown.

https://www.youtube.com/watch?v=PaGdJdXnagQ

https://en.wikipedia.org/wiki/Westinghouse_TR-2

Curiously, Russia has the largest number of small nuclear reactors in the world, albeit legacy tech.

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

https://defense-arab.com/vb/attachments/10536/

This boat is online and puts out ~70 MW of electricity gross:

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

Only one of those is operating. Most are "conceptual design."

table is outdated

Three kinds are operating - 3 EGP, 2 KLT-40 and I think 6 RITM-200

Wow that's a lot of power and a lot of heat. The most unexpected thing is it can produce 240,000 cubic meters of freshwater a day as well!

I feel that this reactor is too small for actual usage. It's about wind turbine sized. But better to err on this side, if you want the iteration speed up and start mass production. Future versions can always be scaled up.

The keyword in Small Modular Reactor is Modular. 5mw is not "too small for actual usage." You can connect many of them together in one plant to produce however much power you want. Just like wind turbines, a typical design includes more than one.

Seriously, 5MWe is what, a few thousand homes? Low thousand if you need to plug in local commons and you’ve got lots of EVs around maybe.

Alongside a few MWt to shed into district heating, that seems pretty nice in a distributed grid context.

> Just like wind turbines, a typical design includes more than one.

Wind farms are a thing because location is an issue, and there’s a lot of nimby-ism, so if you can plop down turbines you plop down a bunch.

Though I guess nimby would also affect SMRs, location is way less of an issue, if you have space for a farm you might as well use a classical nuclear plant.

Plus the capacity factor of nukes is way higher than turbines. Assuming SMRs follow the nuclear norm you don’t need to overbuild to compensate.

Wind farms are a thing because a wind turbine can only be so big: for more power, you put up more of them. They are so cheap because all the costs are transparent, there is noplace to bury wholly-legal graft. A big farm amortizes fixed project management cost.

Capacity factor of nukes is not so much more than of wind turbines, though if you have a bunch of nukes, it would be rare to have many of the nukes down at once other than for urgent retrofits. Steam turbines are down a lot, so nukes are always built with two or more.

But the main thing is that nukes cost far, far more than the wind farm that produces as much; or, a wind farm at the same price produces many times the power, with near zero lead time and possibly negative decommissioning cost. After their contribution to the grid gets large enough, you build out storage, which incrementally reduces the fraction of time you spend burning NG or, later, ammonia.

A farm of small nukes would cost quite a bit more to build than a big nuke of the same capacity, because all the systems are duplicated throughout. On the up side, the farm might be built incrementally with much less of the graft always attached to monster public works; you might save 75% vs a big nuke just on that basis. If you could get the first one going early, its revenue might help pay for subsequent units.

But whatever the heat source, anything with a steam turbine is just not competitive anymore. That is another reason why fusion is a dead end: it is just very hard to compete with zero opex.

Speaking of farms, 5 MW is enough to power LED grow lights equivalent to 5-10 acres of sunlight (napkin/google math) that could be packed in small-footprint building with hundreds of layers of hydroponics.

It still wouldn't be economical in the short term due to construction, water, nutrients, etc. But it's something solar can't possibly provide because there's only so much sunlight per acre. Long term, we could stop trying to farm every square inch of arable land on the planet, of which we're already farming about half.

The idea of reducing the farm footprint is pretty exciting to me.

> we could stop trying to farm every square inch of arable land on the planet,

It'll still be cheaper, thus it will still be done.

Mark my words, we're giving nothing back to nature.

>It'll still be cheaper,

No, it won't. One problem with arable land (i.e., land that's really good for growing crops) is that it's also usually land where people really want to live. Non-arable land is places like deserts and tundra, and almost no one wants to live there, for obvious reasons. So farmers are in competition with developers (and eventually property buyers) for using the best land.

I see you have not been to Iowa.

Iowa, like that whole part of the continent, has brutal winters. There's a reason people don't want to live in the interior of the country: the weather on the coasts is much better overall.

However, it is really good for growing certain crops, and that's why it's used for that, and also why America has historically had a huge advantage by being the "breadbasket for the world".

The other factor here is that there just isn't that much arable land in many places, so places that need to import food could instead grow their own using the facilities you mention here. For instance, here in Japan there really isn't much land for growing crops, for obvious reasons (plus there's a lot of typhoons which can be problematic for outdoor farming). But some big nuclear-powered hydroponic farms could very well be economically viable if the government provides some incentives because it wants to be more self-sufficient with food.

The soviets used nuclear power plants to build citrus greenhouses in the arctic; becoming a global-scale exporter of a crop that dies if it freezes and growing it in temperatures of -30.

Compare/contrast this question: electric car batteries are just big volumetric arrays of 18650 cells. If you know you need that much wattage — and you're manufacturing them yourself anyway — then why not make bigger cells, to reduce per-cell fixed-cost overheads and inefficiencies?

I assume there is a good answer to "why not" here; and it's one that's probably related to the "why" for SMR.

Cooling is one reason. Bigger batteries increase the surface area to volume ratio: heat produced in the active part of the battery has to be lost through the casing and into whatever cooling system you're system. Bigger have less surface area.

You also have packing issues: cylindrical cells are basically optimal, because they distribute any internal pressure equally. If you play with something like the prismatic Prius batteries, then despite being larger you have far less options to pack them because you need to counter-pressure them horizontally or they swell and then fail.

> Seriously, 5MWe is what, a few thousand homes?

Charging 20 teslas at maximum charge rate. Or 1k houses, provided it's general use and not electric heating. 5MW is 43.8GWh/year or an average annual power consumption of 8-12k people in the west (per capita consumption - so includes industry use). Accounting for peak vs average this is likely enough for 3-5k people.

Charging 20 Teslas at maximum rate sure, but that could also be viewed as fully charging 1600 Teslas per day, or steady state providing the energy for about 12k Teslas (assuming 10-15k miles per year per car). Likewise, based on my own electricity bills, 1k houses is closer to 3000-5000.

Obviously, peak load vs average load is very important so I wouldn’t expect the energy to go that far, but connected to a big battery… probably, right?

One of the use cases for SMRs is to install them in decommissioned coal and gas power plants. They have a roof, and are already connected to the grid and roads, and near enough people to supply workers but not so close to supply protesters. The locals are much more supportive, because some of the jobs remain but the pollution they have first hand experience with will go.

One of the major reasons for wind farms rather than isolated turbines is the economic and environmental cost of grid connections and roads needed for installation and maintenance.

There are no jobs, one goal of SMR is to be completely enclosed and self-safe. When the fuel runs out, you swap the SMR.

The use case for these nuclear batteries isn’t to power NYC. It’s to be able to replace industrial needs for energy and heat at the source. A glass or cement factory could have its own power source that’s carbon free. It also means that hydrogen production can be considered green. It’s also great for rural communities that might otherwise have expensive/dirty electricity.

The goal is for them to be autonomous for this reason: install it at some industrial site without any nuclear expertise. Have it shipped in fueled, and then shipped out to maintain or refuel.

These are absolutely the future.

Actually the capacity factor of offshore wind is around 60% that is similar to nuclear (~70%), not way higher. However the price of nuclear (in Germany) is 3x that of renewables, so it just does not make economic sense to build nuclear.

Wind turbines strive to be as big as they can, and they have grown a lot. Yet they are limited by physics and transport issues. Now you need to install thousands of turbines.

Engineering With Rosie has a truly excellent video that analyzes many of the scaling laws: https://www.youtube.com/watch?v=Ze-zaW3au9Q

GE gas turbines for example are 35 to 570 MW. https://www.ge.com/gas-power. They have stopped making smaller ones. There probably are even 100 kW gas turbines made by some companies but that's not used for major power generation (maybe as an airplane APU).

You probably won't have these small reactors in towns of 5000 people run by some local operators, because nuclear technology requires so much special training and is so risky because of the potential radiation hazard.

You might have 20 in one powerplant in a city, to provide 100 MW, as part of the energy mix (it's always good to have multiple sources).

There is a certain threshold for nuclear technology. It's not the same as a diesel generator that anyone can put in their back yard. The risks are too high.

Yet, the current reactors like the Olkiluoto 3 EPR in Finland at 1600 MW electrical power are too big and unwieldy and risky. When it turns on or off, it causes some problems to the rest of the grid. One reason for such a huge plant was there was a legal process to provide permission for one reactor. So if you can only build one, of course you try to make it as big as possible. This doesn't make sense from engineering sense - it probably would make more sense to build something like power plants of 4x400 MW reactors (the seventies plants are 2x400 MW reactors).

So all in all, probably more optimal size for "small" reactors would be 20 to 200 MW. They're big enough that the radiation protection doesn't eat all the budget, yet they're small enough that you get to build many and can build a production line.

DARPA and NASA have some combined projects developing small reactors suitable for both mars and deep space exploration as well as replacing oil tanks for generators in remote military installations like in the arctic. This is almost certainly chasing that, not consumer generation.

From Robots In Plain English [1]:

The Curiosity and Perseverance rovers each carry an old kind of power supply: a nuclear power generator that runs on radioactive plutonium dioxide.

This generator has been used on many missions since the 1960s.

It produces a steady 110 watts of electricity. The decay of the radioactive material also emits heat, which helps keep the electronics onboard warm through the freezing nights on Mars.

Supplemented by rechargeable batteries, the generator provides enough power to let the rover pull all-nighters for years to come.

[1] https://www.robotsinplainenglish.com/e/2020-08-09-nuclear.ht...

I know what a RTG is.

I'm talking about this, a different topic entirely, as RTGs cannot practically meet these requirements: https://www.thedrive.com/the-war-zone/26152/the-u-s-military...

"Project Dilithium", heh ! (A Star Trek reference.)

Wide-ranging article, thanks !

RTGs are horrifically inefficient sources of power. They work well enough for providing a paltry 110W to a rover on another planet, but that's not going to scale up to providing power for millions of people. The amount of plutonium you'd need would be astronomical.

wind turbines only work when the wind blows.. not comparable

We all know thats for military camps.

A serious problem with this is scaling of operating costs.

A 5 MWe reactor, operating at about 90% capacity factor and selling power at wholesale prices (maybe $0.03/kWh) will earn $1.2M/year. You need at least four employees to operate it (3 shifts, with a spare), and probably many more.

I'm no expert here, but I wonder about the economics around the 13MW of heat that it outputs. There are certainly applications around the world where that could be hugely beneficial. Take Reykjavik[1] for example. How much are they paying for their system now on a yearly basis? How would they like an extra 5MW of energy along with it?

[1] https://vahterus.com/resources/cases/10-mw-heat-pump-system-...

I’d guess they have enough geothermal power for the time being.

TFA even mentions Alaska and Saskatchewan...

> I wonder about the economics around the 13MW of heat that it outputs.

I'll take a stab, if only to start a discussion:

13MW is about 450MBTU/hr.

Natural gas is about $7/MBTU. Assuming 80% efficiency of gas to heat conversion, you would need 562MBTU/hr, or a cost of about $80/hour.

If the heat from this nuclear device is sold at the same price per unit heat, it will make $80*8760=$700800/year.

If read correctly, the reactor can make either 5MW of electricity or 13MW of heat, but not both, so it wouldn't make a lot of sense to sell the 13MW of heat. At best sell some of the waste 8MW to make a few extra marginal bucks.

I imagine the heat could be captured and you do both, but you'd need a system to do that, which is probably more expensive.

Natural gas is only $7 / BTU if you are near an LNG terminal. "Remote areas" usually are not. Think about a military base, a relief effort camp after a major natural disaster, of a mining site in the middle of Canada or Siberia, stuff like that.

That Reykjavik doesn’t use energy to make heat; it uses it to pump it from A to B.

In addition, I guess the heat it pumps is geothermal (FTA: The new setup is expected to provide over 40 GWh/year of free ocean heat.)

A benefit of using this kind of reactor might be that the source of heat could be moved closer to where it’s needed, If so, but I don’t expect that to offset the advantage of the current heat source giving you heat for free.

From the article, they're planning on the reactors being unmanned (but monitored). So, no employees watching it.

That's going to require changes in licensing in the US. I don't think even remotely operated reactors are legal in the US.

How are they planning to handle security? The rest of the thread mentions they intend to use these in very remote regions (northern Canada and such). While that makes a lot of sense practically, doesn't it make security kind of tricky? I get that they'd be carefully remotely monitored, but it seems like a realistic possibility that a team of people could break in (the renders show these in shipping containers surrounded by a simple double layer of barbed wire fencing), steal the radioactive material, and escape before security arrives.

I get that we store actual nuclear bombs in unmanned remote sites (missile silos), but those are operated by the literal military and secure underground. These would realistically be operated by private companies, and are just shipping containers sitting on the surface.

Someone wrote in another thread:

Each grain of fuel is about the size of a poppy seed, with the majority of the seed being the ceramic shell. It's not a dense fuel. This stops issues of radioactive material leaching into groundwater, and also makes it quite difficult to use the fuel for nefarious purposes (like a dirty bomb). Someone would need to separate the fuel from the shell, not a trivial thing to do. All this to say that it's not super attractive for terrorist types.

Also, remember that whilst they talk about using them in remote areas, it's mostly about being "remote to other infrastructure", not "remote from all people". These things will be installed in small regional towns / mine sites etc. Places where a town sized number of people will be. So it's not like it will be hours away from any first responder.

Finally, these can be secured pretty well physically. They won't be easy to move, or cut into, or siphon out the material, at least not in a smash-and-grab type scenario.

https://www.energy.gov/ne/articles/triso-particles-most-robu...

edit: I wouldn't say they are risk free, just not as big a risk as it first sounds locating a nuclear reactor out in the middle of nowhere :-)

edit2: I'm by no means an expert, but I really liked this video on the topic on the benefits of Small Modular Reactors (SMRs). I really love the concept, especially their mass produced, regularly rotated/retired nature:

https://www.youtube.com/watch?v=TYnqJ4VnRM8

That makes sense. Thanks!

It's shit like this:

> In 2009, this improved TRISO fuel set an international record by achieving a 19% maximum burnup during a three-year test at Idaho National Laboratory (INL). This is nearly double the previous mark set by the Germans in the 1980s and is three times the burnup that current light-water fuels can achieve—demonstrating its long-life capability.

That makes the entire industry seem like a bunch of liars and grifters.

Tripling the burnup when you're increasing the amount of U235 6x isn't an improvement, it's a step backwards. It uses 2x the uranium and almost 3x the enrichment.

Just present the fuel honestly on its actual merits rather than telling five lies and half-truths and one real advantage.

Sorry, I'm not quite following you. Could you please ELIA5? (No snark intended)

They're implying it resulted in 3x the fuel economy (20% burnup) when in reality it is half by hiding the important figures (20% enrichment rather than 3.5%, and it adds a lot of other mass to the fuel that is difficult to remove when you need to store it). So that you'll assume that the bit they didn't mention is normal rather than very unusual.

An analogue would be advertising a new car that gets 150mpg, but not mentioning the fuel it uses is a special exotic fuel not made in many existing refineries that requires 6x as much oil and you need 9 gallons of non-fuel to run through the engine for every gallon rather than 0.2.

Every single statistic nuclear proponents cite that can be easily checked in a few minutes is either technically true but designed to mislead like this one, or an outright lie. It makes it very hard to believe the things that cannot be easily checked.

In all likelihood one or more of the SMR concepts around is safe and economically viable enough to fill an important niche, but when all information they mention turns out to be lies it's kinda hard to trust.

Thank you. Yes, I think that there is still some ways to go with making these cost efficient vs other tech. I have heard people mention similar issues that you are rightly calling out. I believe the hope is the viability of these designs relies on cost reductions that would come from mass production/ economies of scale. I don't know how realistic that is, but it does tend to happen when things get produced in greater quantities, smart people find many optimisations to reduce cost.

New Uranium mining is incredibly destructive and limited. Dropping the breeding ratio to effectively zero is a non-starter.

Plus noone really know how you might make TRISO pellets not cost $40-600/MWh (or rather it costs well over $600/MWh and they think it might come down maybe) or cost $20/MWh to handle and store at the back end of the cycle (although this one might be solvable by burying the whole reactor I guess?), so it doesn't really matter how cheap the reactor is if it uses that fuel.

> quite difficult to use the fuel for nefarious purposes (like a dirty bomb). Someone would need to separate the fuel from the shell, not a trivial thing to do. All this to say that it's not super attractive for terrorist types.

Those grains are fine as-they-are for a dirty bomb. Sure, one of the safest, cleanest dirty bombs you could have the pleasure to meet, but still a dirty bomb. Maybe not attractive to someone looking to deny territory through contamination, but great for someone looking to incite fear.

I won't speculate on how much fear one can incite. Anything with the word nuclear in it could most lilely be enough to srir up feelings of panic. Hell, D&D was enough at one stage. So you're probably right.

But I will say that if you spread this fuel over a large area with a bomb, it will ve orders of magnitude easoer to clean up, and result in far less long te issues. It wont leach into groundwater. If you swallowed/inhaled it, whilst it wpuldnt be gpod, it wont be absorbed ibto your tissue. Same with lifestock/crops. The cleanup problem goes from almost impossible, to pretty involved.

Again, I'm no expert, so take whatever with a grain of radioactive salt :-)

The grains would also be fine for a terrorist reactor. Setting off an unshielded reactor in a dense city core could cause much exposure to radiation. The miscreants wouldn't even have to guarantee it wouldn't go prompt critical.

What is a terrorist reactor? And prompt critical?

A terrorist reactor would be a reactor used for terrorist purposes. Imagine something like the infamous demon core, but set off in a populated area.

Prompt criticality is when a nuclear reactor is critical on prompt neutrons alone. This is to be avoided (in most cases) at all costs, as the doubling time of neutrons becomes very fast, a small fraction of a second. In a normally operating reactor, the core is subcritical on prompt neutrons, but critical on prompt + delayed neutrons. Delayed neutrons are emitted after the beta decay of certain fission products, and this slows the doubling of the neutron population enough that feedback control can keep the reactor's power steady.

Prompt criticality is what happened at Chernobyl.

You would still need big equipment to do that, you can't just break in and "take the fuel", it's deadly taken straight out of active reactor. Probably easier to just truck the whole container.

You may mix it with other public services which need some security already - remote airport, forest guard house, polar lab etc.

They don't have to be on the surface. Infrastructure is often buried underground even for non-security related reasons.

I assume the intent is to have a bunch of them replacing the boilers in a large coal plant. Capex and security wise it seems pretty good, and I'd expect capacity factors going on 100% for the heat generating part as it's incredibly simple and you just swap the whole thing once a decade.

But it's nuclear so there's always a scam if you look at the other hand instead of where they're directing your attention.

The card in the other hand today is it uses TRISO fuel. This produces >10x the high level waste which can't be reprocessed. It uses twice as much uranium, triple the enrichment (at levels not possible with most current enrichment facilities). And it is fabricated using a process that is estimated to cost anywhere from $40 to $600 per MWh.

More than likely it also needs a much higher quantity of hafnium or iridium or silver for control rods as well, and odds are you can't make the heat pipes out of non-exotic materials.

It uses reflectors instead of control rods.

Made of? If you say beryllium there's another non-starter. Graphite would be workable though.

They don't say.

> The microreactor can generate 5 MW of electricity or 13 MW of heat from a 15 MW thermal core. Exhaust heat from the power conversion system can be used for district heating applications or low-temperature steam.

They are aiming 8 years planned service life, and one novel thing is the use of heatpipes (like your CPU cooler) using liquid metal as a working fluid.

They actually don't say how big it is, I guess still quite sizeable given the heat output. Definitely not a single-family home device.

It's a shipping container, same type of footprint and install as the tesla battery packs. They have pictures and videos if you google it.

4 trucks for the whole setup

Just imagine the ragne of those trucks running on that "battery".

My understanding is on-shore windmills cost about ~$1.5M and $50k in maintenance per year - last for 20 years - for 1MW.

I'm skeptical this can come anywhere close to that.

But if it's within an order of magnitude - it could replace old coal powerplants as they're decommissioned.

Apples & oranges: wind is an intermittent generator versus eVinci for base generation

bingo! it's crazy that people keep comparing this to wind turbines.

A problem with on-shore wind is that it is already getting hard to find more space for it in some European countries. You also can't put windmills in densely populated areas where most power is needed so you have to take energy transmission cost into account as well.

If it can be demonstrated that these reactors are safe, you could put them almost anywhere.

the number one problem with onshore wind is a capacity factor around 30% and on top of that you could have long periods with no wind at all.

Periods where wind operates below 5% capacity in Poland are really common. I've seen it once below 1%. When I write this it's 7,5% - below 10% for 14 hours already.

Nobody in the developed world will allow a nuclear reactor to be built anywhere near their house...

So, however safe these are, they'll still be relegated to 100+ miles from any big city. So the transmission costs will be similar to wind/solar.

While the operational safety sounds excellent, I am worried about what a generalization of this kind of reactor would entail. Your typical reactor is under heavy surveillance and 200 (67?) times more powerful : we are very unlikely to "lose track" of the fuel.

Let's say this becomes moderately popular and we soon have a million of them across the world (enough for about 10% of our today's energy needs)... who is going to keep track of all of them ?!

The High-Assay Low-Enriched Uranium required here is enriched between 5% and 20% compared to a conventional reactor using between 3% and 5% fuel. Because enriching from the natural 0.7% to 20% is 90% of the effort, 20% is considered to be the line separating civilian and military uses. And we're much closer to that limit with these reactors. I am baffled that this is not being even mentioned in the article, nor the pdf ? (I guess they are not particularly willing to disclose just how many of those would need to be gathered for the minimum 25 kg of 90% enriched Uranium for a bomb ?)

And as safe as these might be under normal operation, how many decades for a forgotten abandoned one, exposed to the weather, to leak ? How much and what kind of radiation release can we expect ? (Carried by water to the closest steam I assume?)

You need over 90% enriched for practical nuclear weapon use. It is not "most of the effort" to get to 20%: in fact, if you're stealing it then it's none of the effort: you don't have a centrifuge operation, so whether it's natural or 20% or 70% it is all equally useless.

The difficulty of enrichment is the plant and logistics of doing it, none of which you can steal.

I should have been more clear that I am assuming an hostile nation-state with some minimal capabilities : the biggest difficulty is probably to be able to deliver the bomb to a target once you have one, which gets even worse if you want to use it as an open threat for deterrence rather than a (frankly suicidal) surprise attack : you pretty much need an ICBM (see all the North Korean failures before they (supposedly) stole SSR-Ukrainian tech).

Difficulty of enrichment seems to be heavily proportional to time : the whole "game" around Iran's enrichment seems to be about not letting them get it high enough that they can go from weapons-useless uranium to a ready bomb in a short enough amount of time for the other countries not being able to react.

(There's also a possibility of a much smaller actor to make a dirty bomb from that mildly enriched uranium, but I have even less ideas about how likely that is.)

Uh, you can't run life saving medical equipment or refrigerator service on wind turbines. That means that all the bodies of people that die without ventilators etc, during no wind days would have no chilled morgue to sit in, and would rot

>Uh, you can't run life saving medical equipment or refrigerator service on wind turbines

Man! If only there was some device that stored energy for later use!

You gotta include it in the cost figures, though.

Wind has the distinct advantage compared to this tech in that it actually exists. I can buy a turbine tomorrow and have it online in a month or two.

Sure, but that's a different and unrelated argument to the question of theoretical cost which is what was being discussed.

Sure. This costs infinity because it doesn't exist and 100% wind based power generation would cost less than that, but more than a wind turbine by itself to include storage.

There’s such a thing as energy storage and long distance DC transmission. Why does the old “intermittent power” argument crop up every time when there’s clear engineering solutions to those problems? Next thing you know you’ll be blaming regulations on lack of new nuclear and talking about how we’re running out of lithium, both of which are false arguments as well.

It crops up because people give these absurdly misleading cost estimates for adding incremental wind power to the grid. You can’t have wind power without spending money on an equivalent amount of reliable power, and no one wants to include those required costs in their estimates. This tends to dramatically understate the actual economics of wind.

There are plenty of studies of near-100% renewable energy grids and the electricity comes out similar to long-term wholesale trends (pre-Ukraine), even taking storage/transmission into account.

Let’s see one for wind power, then. OP implies wind is an order of magnitude cheaper.

How about this one:

https://web.stanford.edu/group/efmh/jacobson/Articles/I/21-U...

With those 2 things combined into a solution, you're gambling that we will always have average good wind conditions long enough to keep the grid batteries from going empty. Sadly, under your proposed regime, as time approaches infinity the chance that the grid will run out of power approaches 100%.

If you could avoid straw-manning me about nuclear and lithium in the future that would be super cool. I've played enough factorio to know that green power and batteries are part of the best solution, but you need high energy density generation systems for fast advancement too! If we turn our backs on that tech we know works we will lose future opportunities, or at least delay them

If your local energy storage or grid isn't reliable enough, you can always have a backup diesel generator. Or hydrogen powered generator if you insist this is green. Or some solar cells on your roof. And as much fuel as your risk analysis demands. But the reality is, hospitals can rely on local emergency services for support, and it is likely more cost effective run the risk of spoilage in the rare event of a multi-day power outage than invest in local generation for your refrigeration. My guess would be local equipment failure is more of a risk than a multi-day renewable energy outage, even in latitudes where solar is useless.

> you can't run life saving medical equipment or refrigerator service on wind turbines

There are countries running on 95%+ renewable electricity right now[1]. Including wind, hydro, solar, geothermal and biomass. Humanity keeps neglecting what is already proved to work. Because politics? Coal and gas are cheaper options? Environment is not considered in the equation?

Edit: source [1]: https://en.m.wikipedia.org/wiki/List_of_countries_by_renewab...

> There are countries running on 95%+ renewable electricity right now[1].

That list ignores that countries trade electricity. If, say, Germany produces a lot of electricity by renewables it can sell the excess to France (which has lots of nuclear). If Germany is low on electricity, it can buy from is neighbors. The system as a whole is nowhere near 90%.

The one exception are probably countries that have all hydro and biomass.

You created a straw man. There are 10 countries that fit the 95%+ criteria, all of them probably have excess of production and have 0 interest in buying power. But yeah, what you say could theorically happen.

All large hospitals have emergency generators. Of course they are supposed to be for emergencies, not for dealing with regularly intermittent power. That is typically dealt with advertised in advance power shutdowns to residential areas, and in richest countries, by not sending the signal that remotely turns on your water balloon heater during cheap hours.

you can run life saving medical equipment or refrigerator service on batteries.

I think it's a standard thing for vital infrastructure to be backed by generators.

big industrial consumers can factor their own specialised requirements for storage, backup.

Current wholesale price of electricity in EU is 110-120 USD / MWh.

5 * 110 Euro * 24 hours * 365 days * 8 years planned service life = $38,533,000. If they can offer a price under 5M USD per MW they won't have problem finding buyers.

This doesn't even account for savings for businesses that can consume most of that output - they save another 20-30% on transmission fees + there are several MW of waste heat as low temperature steam that can be used directly or indirectly in many industrial processes.

Yeah dedicated units for large industry would be a huge buyer here: if you can hit the right price point, aluminium smelters in particular would be very interested customers I suspect.

This thing only outputs hot water, you need a lot more equipment and people to turn that into useful power on the grid.

Fuel costs, maintenance, and operation would be over $2M.

Normal energy prices are about half of what you quoted. It's TBD if they come back down to normal in the near future in the EU.

You're looking at $18M in revenue. This thing will probably cost more than $10M to purchase, and then you need land, a grid-tie in, and permits.

It would still work - but the numbers aren't going to look as good as wind & especially solar.

Comparing the price of unreliable energy to reliable energy is like comparing the price of a junk bond with a treasury and wondering why the latter costs so much more. It costs more because when you turn it on you get a known and reliable source of power, day or night, cloudy or sunny, windy or quiet. That's why it costs more.

I wonder what the cost of decommissioning is. That's easy to ignore just like it's easy to ignore the cost of mitigating the CO2 emissions from gas...

If they're modular then they'll be intended to run more than one. You get the land, grid tie-in, permits and staff and then... stack 4 of these instead of 1. You use the same benefits from economies of scale to have as many of these as can be safely crammed into your building and go from there.

> I guess still quite sizeable given the heat output. Definitely not a single-family home device.

I’d think the 5MWe would be a hint. Standard residential service drop is like 40kW.

"Big" there might be referring to physical size.

I'm wondering, they say using sodium, and having tested heat pipes up to 800°C... but their schema shows sodium vapor... but sodium only boils at 900°C ??

So I guess they underpressurize these pipes ?

That would make sense, since in a traditional liquid metal reactor, the boiling of the working fluid is instead a failure mode to be avoided at all costs (since boiling dramatically increases the pressure resulting in burst pipes), especially with sodium that burns on contact with the air, explodes on contact with water, oh and also is highly radioactive at that point (for a short time).

They seem very sure of themselves, but I still wonder what kind of stresses are involved, and what does this mean for heat pipe longevity ? (and calling heat pipes with a phase changing working fluid "passive" still seems kind of wrong ?)

EDIT : The relatively small amount of sodium probably matters a lot here ?

Static vs passive cooling, I think is what you're looking for here.

Heat pipes are used in CPU coolers and work on the same principle. A small amount of liquid is in the pipe, at reduced pressure, which makes heat transfer very efficient.

I suck at chemistry, but wiki clearly discusses advantages of sodium for heat transfer in nuclear reactors: https://en.wikipedia.org/wiki/Sodium-cooled_fast_reactor (tl;dr: safety margin from the large range of temps at which sodium is liquid; doesn't like to absorb neutrons, and isotopes aren't that problematic when it does).

Yes, but they use water. (Also reduced pressure means reduced density, which should in theory make heat transfer less efficient... except when phase changes are involved !)

That reactor is a radically different design, operating at a much lower temperature, where sodium doesn't evaporate.

But the whole point of a heat pipe is that phase changes -are- involved! So, water’s fine for temperatures in a “normal” range, like a CPU, but perhaps not great for higher temperature ranges.

And then, this is a nuclear reactor. Whatever’s in the heat pipe in a reactor will be bombarded with neutrons. Water is a neutron moderator, so you wind up with heavy water in the heat pipes and not-very-nice fission products in the reactor due to slower neutrons.

I never said they should use water, I'm just wondering what kinds of stresses (including chemical ones ?) are involved here...

They say they envisage delivery of the whole system on four trucks.

Not single family home (unsurprisingly, the economics of very small scale nuclear are awful even ignoring the safety concerns), but night and day compared with conventional large nuclear reactors.

If you can get the heat out, no reason it couldn't be tiny. A hayabusa motorcycle engine can be tortured into producing 1MW mechanical from about 5MW thermal.

The heat engine required to meet efficiency would likely be the limit (although I think TEGs might be getting cheaper).

The main issue is unless you have a big pile of them, you need 1% of your population to be nuclear engineers and security.

There is no way in the world this is going to be economically competitive to connect to big electricity grids any time soon.

Even taking into account the cost of extra transmission and storage (and to a certain extent, building extra storage, which is politically easier, reduces the amount of transmission you have to build), wind and solar are ridiculously cheap and trending cheaper long term, and you don't really run into the requirements for massive amounts of storage until the fraction of your power from VRE gets very, very high.

Remote parts of Alaska, Canada etc face very very high power costs and solar isn't viable there (though wind might be depending on the site), so I can absolutely see the initial applications there.

Not to mention a need for heating!

The article does not mention grid applications. In fact, it explicitly Alaska and Canada, and off-grid applications, like mines. So, don't think it's trying to promote the idea of cheap wholesale electricity in normal places, but, instead, reliable on-premise electricity in places where it's not so very easy to get.

Yep, absolutely. But there are plenty of other people, notably here on HN, who get very excited about grid-connected SMRs.

It's also worth noting that the majority of people living without access to grid electricity are in sub-Saharan Africa and South Asia, places with much better (and much less seasonal) solar resources, and far less capacity to safely manage nuclear power.

https://www.visualcapitalist.com/mapped-billion-people-witho...

> and far less capacity to safely manage nuclear power.

A common point of design (or at least hope) for SMR is for the design itself to be sealed and fail-safe, no operators and no management, when fuel runs your you replace the reactor and send to to the factory for refurb.

Think smartphone, not lamps-based computer.

Could it be safe even with actively malicious people?

..maybe. For a given definition of malicious. The flip side of the eye-wateringly expensive and inefficient fuel they're proposing is it's very hard to get the radioactive stuff out of the SiC shell, even with explosives. And getting it to go into thermal runaway would require some exotic materials that can out-last the SiC as it melts.

Spreading tiny fuel pellets everywhere would still be horrible, but far more contained.

The dangers of this would be more diffuse. Large amounts of high level waste compared to a conventional reactor, double the mining, fuel fabrication facilities that are one hopper jam away from going chernobyl. If it were accepted, the industry would get complacent about cracked pellets and Kr-85 or other hard to contain fission byproducts would slowly accumulate everywhere and we'd have to spend 50 years fighting misinformation about how radiation is good for you while it builds up irreversibly.

AIUI, microreactors are not SMRs.

The modularity of SMRs comes from being able to build the different modules in factories, but the SMR would still be assembled on-site inside a building/facility.

Source: https://omegataupodcast.net/359-modern-fission-reactors/

Westinghouse's current SMR design is IRIS, which is a PWR. They cancelled their previous design.

There are remote parts of Canada where solar is viable, and can save on lots of diesel fuel, but doesn’t do much in the winter.

e.g. https://www.yukon-news.com/local-news/diesel-generators-go-q...

Sure. I should have been more precise and differentiated between "viable replacement for diesel generators" and "viable supplement to diesel generators".

I wasn’t trying to be pedantic, I just think that solar project is super cool!

The decisive thing is always going to be cost. Thus far, reducing capacity has made nuclear plants faster to build, but the outcome was more expensive per MWh [1]. This then forced future plants to again increase in scale, but it again made them too slow and expensive to build, with huge budget overruns. We're now back to small modular, but if the electricity they produce is again even more expensive, it will offset any other benefit.

[1] https://www.researchgate.net/publication/275673955_The_forgo...

After ethics reviews, send some prototypes to Ukraine this winter.

Sounds like something you don't want Russians to target thought.

Ukraine has multiple traditional nuclear power plants running.

The attacks of russians on Ukraine's electricity grid revealed a safety feature of the convenient nuclear power plants (at least those operated in Ukraine) - they cannot produce electricity while being alone in the grid (or in isolated pieces of the grid). If inflow of electricity fails the nuclear power plant switches to diesel generators and gradually decreases its power output going effectively into hibernation. Attacks of November 23 caused exactly that, all three nuclear power plants (excluding ZPP which is on temporarily occupied territory) were shut down since they lost incoming power [1].

1. https://www.theguardian.com/world/2022/nov/23/russia-ukraine...

> Commercial deployment is targeted for 2027.

(Sounds like the new fuel might be available sooner, but probably not for this winter?)

- "the power conversion unit (an open-air Brayton Cycle system)"

Isn't that essentially a jet engine?

The advantage of small reactors is that you can make them in factories (mass production).

The disadvantages are that you lose all on site scale and you are very vulnerable to a single part failure or design floor. Also miniaturising things is non-trivial.

Then you still have all the other issues around nuclear (waste and perception)

Given the cost of nuclear, you need that advantage to be much much bigger than the disadvantages. If someone can do that, they'll make (serious) money. But it's a >100bnUSD project. So I don't think there are many corps who will take that sort of risk given the uncertainty.

What a bizarre title. Either it is, and then "sees" makes no sense, it's a proven disruptor, or it's not, and it's just armchair self-congratulatory PR speak.

Or it’s indeterminate, which it is. They think it might be a disruptor, but they’re not sure yet. They see that potential, but they don’t have it yet.

Gets more clicks than "They designed this thing and made all these claims but hit us back in 5 years when they actually run one fueled-up with enriched uranium."

It makes sense if you substitute "recognizes" for "sees". Westinghouse (well, Cameco/Brookfield Renewable) recognizes what a disruptive tech this could be, whereas others have overlooked it, maybe?

The only option is to put many of these to replace old plant. Any other is not feasible due to its nuclear.

Even a small amount of bombing it can be used to threaten local or even there is anyway to hack dirty bomb threat to a city. Once security concern is in not just regulation, back to the hole. But if it is safer than mild one even if more expensive, it can still have some use.

Using TRISO fuel, which is (best case) half as fuel efficient as a PWR and costs $40-500/MWh just for the fuel

https://web.mit.edu/nse/pdf/researchstaff/forsberg/FHR%20Poi...

Passive cooling sounds super turbo safe which is what we always seek in reactors. Yeah that would disrupt the market.

This device is mostly aimed at the military and maybe the industrial market (think about a chemical plant or large factory).

When our military invade yet another country they now don't need to haul diesel fuel to keep the lights on.

I don't know why people are enthused about "passive heat removal" for a reactor that's operating and generating electricity (as opposed to post-shutdown decay heat removal). That requirement rules out every possible type of fission reactor there is, other than extremely small ones with very low power densities (like OP).

Forced convection is how you fit gigawatts of power in the volume of a small room; it's what makes nuclear fission practical at scale.

Gigawatt reactors do not fit in a small room. Even the reactor itself is the size of a large room. Then there is a huge amount of infrastructure around them, then enormous cooling towers. Forget about power density when you have sites measured in tens or hundreds of acres with 300 foot towers--ultimately heat has to be dissipated to the environment, and that takes space.

The entire heat source, the active part of the core, very easily fits in a small room (at a mean power density somewhere around 100 MW/m^3 for water-cooled reactors).

I'm talking about the difficultly of heat removal, not trying to minimize size or complexity or anything.

Sure, but you gotta consider the entire system, so adding the heat pumping system actually may decrease the overall power generated per unit of volume; at least, certainly increase cost and decrease reliability. I wonder if thousands of SMRs spread over an area wouldn't actually generate more power per acre.

"I want to start a business that builds garden lights, fridges and nuclear microreactors".

It would only take ~2 of those to power a decent sized charging site for electric semi trucks.

I get that it makes and moves heat... What process is it using to create the electricity?

Heat water, make steam, spin turbine.

In this case it is boil sodium, condense sodium to heat air, run air through turbine.

Neat, TIL

A bit off topic but Factorio buffs will find this microreactor eerily familiar.

Seems perfect for mining... crypto mining that is!

I am not very knowledgeable about nuclear reactors but wouldn't it be concerning to have tiny nuclear reactors everywhere in the city? Doesn't this increase the potential dangers?

These are not for cities and the "danger" of nuclear power is overstated.

The "danger" of nuclear power is not their day-to-day operations, it is the nature of the waste product (both in terms of its decay to safe levels and the ongoing storage requirements) and the use of said waste materials in the relatively simple production of "dirty bombs" that just require standard explosives to distribute said radioactive waste around a large area.

So the danger is not overstated, it's that the promoters of nuclear energy refuse to accept the complete lifecycle of nuclear power, including both the fuel generation (uranium mining is environmentally damaging, leaving tailings and polluted water), fuel management (moving radioactive fuel around the community/environment), and fuel disposal (unsolved).

Really cool to see this close to completion

At 2:34 camera captured reactor failure

Finally some real innovation on here.

how micro is micro? how big is this thing?

4 shipping containers

> They envision the plant being delivered in four truckloads: the reactor container, the power conversion unit (an open-air Brayton Cycle system), an instrumentation and controls container and miscellaneous support and equipment (like a heat exchanger to capture waste heat for district heating).

I guess Elon Musk will be interested in taking one of these things to Mars. He is almost getting the transport capability part done.

Probably after he is done jerking around with twitter.

Reminder: Microreactors mean more radioactive waste, not less. No country has solved the safe long term storage radioactive waste. No, not even Finland, even though they are close. Expanding nuclear power, especially through microreactors means expanding access to extremely toxic radioactive material. And make sure the government is stable (for the next few hundred years) and there's no war in the area.

Please don't spoil the party. We have been lazy for so long and absolutely don't want to cut back on luxury or use the cleaner solutions that producing nuclear waste with expensive subsidies and all the other downsides is now cool again and full green! Big innovaton here, that counts!

Reminder: solid radioactive waste is many times more safe than pumping waste into the air we all breathe. Fossil fuels kill many people per year by their waste and radioactive waste is an easy to solve problem comparatively speaking. Additionally some SMRs are designed to be thrown out wholesale because they are already the container for the nuclear waste. Finally, radioactive waste can be used as a power source.

Yes, "and everybody clapped". The reality is that widespread nuclear reactors are not safely manageable in today's society.

First find a solution to prevent all wars and social unrest now and it the future, then we can talk.

The Russians found that when they put their artillery pieces near a nuclear power plant, the Ukrainians don't shoot back. Funny that.

And almost every time there is a cruise missile attack the nuclear power plants have to go offline as a precaution (not even because of them getting targeted though, which also may happen in the future).

The idea that radioactive waste is safe is at best misleading, if not outright false. You ignore that fossil fuel is not the only alternative to nuclear and vice versa. Then you hand-wave over the amounts of waste and climate gases.

If nuclear waste disposal is so easy to solve: Why has nobody ever solved it? You talk about some hypothetical SMR designs? Yeah, in theory, future nuclear technology is safe. Unfortunately even current technology, nuclear power projects always run over time and over budget on top of an already unwieldy upfront budget. Using radioactive waste as a powersource sounds nice and smart, but turns out to be unfeasible so far and at least one or two decades in the future.

Next, the danger from nuclear power, both during and after the runtime is impossible to calculate. Which is were those "impossible accidents" come from, which happen a bit too often. Yeah, in the theory from the nuclear power lobby, nuclear power is totally safe. But reality has to deal with things unforeseen, and on top of that, with bad intentions. Just that nobody tried (or succeeded) yet in building dirty bombs or destroying another country's nuclear plants doesn't mean that won't ever happen. Very small risks with catastrophic consequences are impossible to calculate and should be approached about as carefully as radioactive material itself.

Reminder that reprocessing (in addition to being the dirtiest part of the fuel cycle and incredibly expensive) doesn't really produce much energy.

All you are doing is getting the tiny leftover fissile scraps, 95% of the material is irrelevant U238 and 4% is fission products (the really bad waste) which just gets leaked everywhere. It adds 15% to your energy output and quadruples the fuel cost (bringing it close to the total cost of solar).

Reply on news.ycombinator.com