Hacker News Reader

Comment by hairytrog | Hacker News Reader

Two interesting things to see here, of many. They are really planning the sites to be temporary compared to the solar/wind/ and other industrial power plants that have little or no plan for returning their sites to the original virgin soil. And two, they seem to use a molten salt loop like in concentrated solar to meet variable demand without changing the power level of the reactor, which you can't really do because of Xenon poisoning. So it's like a natural gas plant in its ability to deliver power, but clean and holds 20 years of fuel inside the reactor.

View on HN

And unlike a natural gas plant, it costs 3-10x what the equivalent generating capacity in renewables would cost.

I come to post this Same Thing every time this discussion comes up. It's not about safety. Everyone serious understand nuclear, even current technologies, is "safe enough" to be useful to build out.

Nuclear is outrageously expensive, though. And new reactor designs don't seem to be making much progress on that.

Look, I think nuclear is fine. But as a numerate and earnest environmentalist, I don't see where the case for it lies. Right now, we should be building out solar and wind as fast as we can, because they're good and safe and CHEAP. And when we get to that last 5-10% where we need buffering capacity that we're currently getting via gas? Yeah, maybe then we can talk about nuclear.

But even then... meh. What's the case on trying to finish the job on electrical generation vs. putting those same dollars into low hanging fruit in other areas of the energy puzzle?

Nuclear needs to make its case on a balance sheet before it makes sense to talk about.

Picking a nit here. You wrote:

> Nuclear is outrageously expensive, though. And new reactor designs don't seem to be making much progress on that.

Nuclear plants are expensive the way GE and Bechtel build them not inherently so. At some point I got to look at the Diablo Canyon cost-to-build breakdown and about 2/3ds of the cost were litigation, licensing, and specialized site and construction prep. Finding people who could do the kinds of welds they needed, build structures in the right way, inspect, inspect, inspect, and litigate dozens of lawsuits asserting on form of harm or another. It was pretty amazing.

This particular company can build the reactor in their factory. All their specialists are in one place, you can have permanent inspection equipment with costs amortized over all the reactors you build, and the safety systems this reactor design purports to have means no additional site prep (no backup generators, no double containment vessels, no borosilicon sandpits under the reactor to slag into glass while "catching" a meltdown.

If I were to guess, while the pitch is all about safety, this design screams that it is all about cost. When you factor in that the design in modular, so making a 100MW power station with 20 of these gets advantages of scale that multiplies the cost advantage.

When I read this I see those cost savings and recognize this could be both cheaper than renewables and way more reliable. A solid base load solution to kill the last of the coal plants.

My guess is that the pitch of safety is to the public (experts already know it is safe enough and that insurance more than covers expenses). And that modularity/size is a pitch to investors over cost.

From my understanding, small reactors haven't been popular in the past because they still have the same insurance premiums as large reactors. If I'm wrong, or this changes, then that's a big reduction in cost alongside the aspects you mentioned.

If this team convinces the insurance folks that their design really is safer, and so their premiums should be low enough to make small reactors competitive, it could be a huge breakthrough.

Nuclear has had 70+ years and untold zillions in government research money and subsidies poured into it[1] - if the same level of investment over time had gone into renewables, imagine where they'd be now.

[1] worldwide - germany, uk, canada, france, russia etc in addition to the us - everyone did build their experimental gas/molten salt/fuel recycling/pebble bed etc reactor research programs and reactors. not to mention all the free experiments and r&d experience from the military applications.

First of, I think you underestimate how much money have gone into developing wind and solar through subsidies.

More importantly, the way solar and wind has been funded has secured competition for lower prices and allowed parallel exploration of multiple technologies. On the other hand, nuclear has been funded like old-school space exploration: heavy on management, low on technical vision, and often with a reward system that favors inefficiency.

In the case of space-exploration, companies like spaceX has come along and shown how innovation can completely change the cost: just compare the crew dragon with the SLS-circus. I believe there is ample space for the same to happen in nuclear power.

And still, storing energy is not solved and renewables are not reliable. Nuclear vs renewable is apples to oranges. Look at California for an example what happens with your system (or course we ha to look at this like a system and not individual power plants).

Germany alone spends ~$1.2 trn (that would be ~4.8trn on a population adjusted basis for the USA) on the Energiewende, and Germany is far away from being 100% renewables.

The EU now plans to invest another trillion € till 2030. A lot of money is poured into renewables.

That's $1400 per capita per year since it passed in 2010, lower than many utility bills, so it would be surprising if it was able to have already transitioned the entire nation's electricity that quickly.

Yep, there is a lot being done today of course and progress has picked up in recent years. But knowing what we know, it's still underfunded.

And Germany has the most expensive electricity on the planet (for a firstworld nation and adjusted to cost of living). Funding is drying up because renewable subsidies make power unaffordable and thus politically unviable. Germany should be a big warning to everyone claiming renewables would be cheap or cheaper, the market here is gravely distorted by subsidies and yet, people point to Germany as an example...

NuScale claimed they could build their reactor in a factory too. Then that fell apart.

Reference? I know there was some push back on their "walk away safe" claim (something that these guys have a better story for) but I had not seen anywhere that Nuscale would not be able to build their reactors in a factory.

They are building in the factory. I heard there is some assembly on site, but it’s just bolting on valves and other appurtenances. There are no large pressure vessel welds in the field. That is what saves costs.

What? That's a pretty outrageous claim with no evidence

https://losalamosreporter.com/2020/08/22/get-out-of-the-fluo...

"In answer to a question I posed to Nuscale at the town hall we have learned that the plan to save costs by fabricating the modules at a remote factory and shipping them to the Idaho site has been abandoned. The artful response to my question said that Nuscale engaged with approximately 40 … pressure vessel fabricators worldwide and … determined that Nuscale will use existing factories … in lieu of building its own factory.

The major module subcomponents will be manufactured at multiple manufacturer locations and shipped to a single location for assembly prior to installing into the facility.” This signifies the failure of one of the major cost-saving features of the Nuscale project, which was to forestall this exact scenario."

At least until somebody gets a design that isn't basically a coal power plant using different fuel, nuclear can not get cheaper than coal.

And guess what is being mass decommissioned right now because it's too expensive?

What an absurd thing to say.

Nuclear fuel is several orders of magnitude more energy dense than coal. A "coal power plant using different fuel" would produce energy so cheap it couldn't be metered.

No, he's right. Nuclear plants are still just conventional thermal power plants. They "burn" stuff to generate heat and use the heat to drive a turbine. They can't be cheaper because burning coal doesn't have ultra high demands on how the boiler is designed. Meanwhile nuclear reactors can be arbitrarily complex.

That isn't my understanding, would love to learn more about how you come to that conclusion. (Ideally based on an idea like the one in the article vs just your run of the mill designed in 1950 power plant :-))

Are they changing any relevant part of the power generator compared to a 1950 plant? Are they at a minimum heating the steam into a higher temperature than what a modern coal plant do? (Are they using water? Because coal works on the limits of what you can do with water already.)

This article seems to have no detail at all, but all the information there seems to be about the reactor. Guess what, you won't build a nuclear reactor that is cheaper than a coal furnace, thus if you don't change anything on the power generator, you can't get cheaper than coal.

Lots actually, the way it is fueled, completely different (as in you don't re-fuel it you replace it) so all of the re-fueling infrastructure is not needed. They replace water (low pressure or high pressure) with helium which gives them two benefits, one it carries more heat and two it doesn't become a high explosive[1] when exposed to extreme temperature. Another benefit of helium is that if it leaks from the system, it carries no fission byproducts(like Cesium) in solution and has no radioactive isotopes with half lives of more than a second. 6-Helium has a half life of .8 seconds and emits a beta particle.

So pretty much all different :-)

Most nuclear plants operate at lower steam temperatures than coal, because thermal efficiency isn't so critical and lower temperatures are less wear on the pipes and turbines.

WRT cost: Nuclear plants could be cheaper. Because, for instance, coal plants require scrubbers to meet particulate emissions rules, which are very expensive to install and operate. And coal is expensive: mining, transporation, cleaning.

Two questions: (1) Night time. (2) Quiet time.

Of course there are batteries! But factoring them in seems to increase the cost significantly. Power lines and cross-continent smart grids are fine, too — but also not free to build and operate.

So a self-contained unit that does not depend on weather, does not need refueling, and includes its own energy accumulator has some appeal.

>Night time

Why does electricity at night need to cost the same as electricity during the day? Would Western civilization collapse if it cost twice as much? Three times as much?

How much trouble do we need to go to so aluminum smelters can run at 100% capacity at 2 AM?

If you're going to shut down factories at night, that itself is a cost that should be accounted for. If you're shutting down for eight hours a day, you increase capital costs by a third.

That's worth paying if it's the only option to address climate change, but if the main argument against nuclear is cost, it's not a fair comparison if renewables externalize cost to the customers.

(Also, aluminum smelters in particular take damage if they shut down more than five hours or so.)

why don't they store the cheaper day time electricity by melting extra on site aluminum during the day? this aluminum could stay on site indefinitely?

The aluminum freezes in about six hours, that's why they take damage if they lose power.

its not a property of molten aluminum to freeze in 6 hours, it depends on the thermal insulation.

what I proposed was to have a secondary sealed thermal bath of aluminum which gets molten during cheap power and which returns heat during expensive power to maintain the temperature of the primary (work) batch of aluminum so that the work batch doesn't freeze

the heat can be transfered with an appropriate working fluid.

since the secondary thermal bath aluminum doesn't need to satisfy client demands (specific alloy etc) the alloy can be chosen so as to have a suitable melting point for such a setup

Right. So show me the balance sheet where building out reactors makes sense. Because as of right now it clearly doesn't, and the proof is that no one is building reactors but everyone is building windmills.

This is an incredibly weak argument, as most of the opposition to nuclear power even before widespread renewable deployment was political, not economic. Maybe the economics have changed with renewables, but that needs to be proved rather than can be inferred by the lack of US reactors being built.

Take Germany’s recent denuclearization, the efforts are completely political, there was no mention of balance sheet optimizations or concerns. China, India, Russia, among others are all building reactors - either we must agree they are fiscally irrational, or realize that there is an economic case to be made for nuclear power.

>most of the opposition to nuclear power even before widespread renewable deployment was political, not economic.

Most of the opposition to nuclear power was about a whole lot of things other than politics or economics. Decades of quantifiable failure after failure in the environmental, construction quality, maintenance, radioactive discharge, siting-safety, and waste-handling realms, to name just a few. Everywhere the problems were the same, whether they were covered-up or not. The problems were those of an arrogant energy industry primarily concerned with minimizing expenses. The facts weren't political ... although the many, many cover-ups certainly were.

To overlook all that history and suggest that nuclear is just a political football is absurd. Had it been a quantifiable success, little of the resistance would have evolved. And Karen Silkwood might not have died. There was a lot of money and power at stake, and little tolerance for realistic concerns. The industry earned the disrespect it continues to enjoy.

Remember the widespread promise "Too cheap to meter"? In what year was that promise kept?

I agree the take above was reductive, especially in truncating the historical aspects of nuclear power, so much so that in the general case it is incorrect. Collective amnesia is not a solution to yesterday’s nightmares.

I think there is hope that the new generation of nuclear scientists and companies have learned lessons, and I support giving them regulated room to prove it. America has a lot of bad energy policy (Fracking, mountaintop mining, to name a couple in addition to your comment’s nuclear perspective) - improving on that is important, but making it worse is not acceptable (Nor am I a fan of gambling with such large minimum bets).

Irrespective of my failures, the core of the original critique I think stands insofar as the argument it was responding to was not self-sustaining.

Edit: I don’t mean to age you, but I am honestly too young to remember “Too cheap to meter”. I can see where historical memory, or the lack thereof, has colored our respective perspectives, though, in important ways

And yet the alternative, burning dirty coal, has become an unmitigated disaster.

The only reason coal has been allowed to prosper is political support.

Thereby proving the resistance to nuclear power has been entirely political.

"Too cheap to meter" was never a promise. It is a quote taken out of context that was facetious and never meant to be taken literally.

About 50 power reactors are currently being constructed in 15 countries (see Table below), notably China, India, Russia and the United Arab Emirates.

   2020 Korea, KHNP Shin Hanul 1 APR1400 1400
   2020 Russia, Rosenergoatom Leningrad II-2 VVER-1200 1170
   2020 Slovakia, SE Mochovce 3 VVER-440 471

   2021 Argentina, CNEA Carem25 Carem 29
   2021 Belarus, BNPP Ostrovets 2 VVER-1200 1194
   2021 China, CNNC Fuqing 6 Hualong One 1150
   2021 China, CGN Hongyanhe 5 ACPR-1000 1080
   2021 China, CNNC Tianwan 6 ACPR-1000 1118
   2021 Finland, TVO Olkiluoto 3 EPR 1720
   2021 India, Bhavini Kalpakkam PFBR FBR 500
   2021 India, NPCIL Kakrapar 4 PHWR-700 700
   2021 Korea, KHNP Shin Hanul 2 APR1400 1400
   2021 Pakistan Karachi/KANUPP 2 ACP1000 1100
   2021 Slovakia, SE Mochovce 4 VVER-440 471
   2021 UAE, ENEC Barakah 2 APR1400 1400
   2021 USA, Southern Vogtle 3 AP1000 1250

   2022 China, CGN Fangchenggang 3 Hualong One 1180
   2022 China, CGN Fangchenggang 4 Hualong One 1180
   2022 China, CGN Hongyanhe 6 ACPR-1000 1080
   2022 India, NPCIL Rajasthan 7 PHWR-700 700
   2022 Pakistan Karachi/KANUPP 3 ACP1000 1100
   2022 Russia, Rosenergoatom Kursk II-1 VVER-TOI 1255
   2022 UAE, ENEC Barakah 3 APR1400 1400
   2022 USA, Southern Vogtle 4 AP1000 1250

   2023 Bangladesh Rooppur 1 VVER-1200 1200
   2023 China, CNNC Xiapu 1 CFR600 600
   2023 France, EDF Flamanville 3 EPR 1750
   2023 India, NPCIL Kudankulam 3 VVER-1000 1050
   2023 India, NPCIL Kudankulam 4 VVER-1000 1050
   2023 India, NPCIL Rajasthan 8 PHWR-700 700
   2023 Korea, KHNP Shin Kori 5 APR1400 1400
   2023 Russia, Rosenergoatom Kursk II-2 VVER-TOI 1255
   2023 Turkey Akkuyu 1 VVER-1200 1200
   2023 UAE, ENEC Barakah 4 APR1400 1400

   2024 Bangladesh Rooppur 2 VVER-1200 1200
   2024 China, Guodian & CNNC Zhangzhou 1 Hualong One 1150
   2024 Iran Bushehr 2 VVER-1000 1057
   2024 Korea, KHNP Shin Kori 6 APR1400 1400
   2024 Turkey Akkuyu 2 VVER-1200 1200

   2025 China, CGN Taipingling 1 Hualong One 1150
   2025 China, Guodian & CNNC Zhangzhou 2 Hualong One 1150
   2025 UK, EDF Hinkley Point C1 EPR 1720

   2026 UK, EDF Hinkley Point C2 EPR 1720

https://www.world-nuclear.org/information-library/current-an...

Everyone is building heavily subsidized windmill. And nobody is building reactors[1] mostly because it's really unpopular (which it pretty much deserves IMHO after decades of shady practices) but it's not a matter of technical merit (and it's never the case).

[1] except those who do, China leading the pack.

I'd suggest looking through Lazard's levelized cost of energy slide decks. The unsubsidized price of renewables is indeed substantially less than nuclear.

Next look at the time scales for building new nuclear. Then look at the trend lines in the cost of storage, particularly utility scale lithium batteries.

You don't have to be a genius to realize this makes new nuclear a bad bet purely on the finances. Even if you waved a magic wand to eliminate any environmental opposition, the only way new nuclear is getting funded is if government picks up the tab. And that's what we see globally.

One of the reasons NuScale and the linked company above are getting investment traction is exactly because they're attempting to reduce the capital costs and timelines. That's a bet investors are more willing to take, even if the technology is unproven yet.

> Then look at the trend lines in the cost of storage, particularly utility scale lithium batteries. You don't have to be a genius to realize this makes new nuclear a bad bet purely on the finances.

You are off by more than one order of magnitude! For non-hydro renewable to be self sufficient, you need between 200h and 600h of storage[1]. So to replace a 1300MW reactor, you need 3000MW of renewable power (assuming 33% load factor, which is a good one for renewable) AND at least 200GWh of storage! The storage cost just dwarfs the cost of everything else (and because it won't last 40 years, you'd need to pay for it at least twice!).

Solar and wind power are financially interesting because: the grid handles the load variations, subsidies, and the financial markets financing short-term projects easily compared to bigger one expected to run for 60 years.

Non-hydro renewable are nice when you want to reduce fossil fuel consumption in a grid where most electricity comes from fossil fuel (and ideally not too much coal), but on a purely technical standpoint, they are no match for nuclear. The thing are never purely technical though, and overall I'm sceptical about the future of nuclear.

[1]: https://bourrasque.info/articles/20180116-moulins-%C3%A0-ven... (in French)

> For non-hydro renewable to be self-sufficient, you need [8-25 days] of storage.

If you're adamant on generation following load, low transport, probably no overgeneration, no hydro and absolutely no backup biomass, gas or similar, then yes, you probably need 25 days of storage.

It would be good to have at least a couple of days or weeks of energy buffer in the world's supply chain, but that can take many forms. Electricity and lithium batteries seem a bad choice for the bulk of it.

And NuScale appears to be failing on that, with the UAMPS buuld being delayed one year per year for the last four years, and still only 30% subscribed. In the meantime, costs have escalated 70%.

>it costs 3-10x what the equivalent generating capacity in renewables would cost

Except renewables can't actually replace fossil fuels. They need baseload provided by fossil fuels to be viable. They are also defuse energy sources, and so require huge surface area and lots of materials for collectors - not great for a growing world, both in population and per capita energy needs. Oh, they are variable across days, seasons and inter-year periods, but there is no battery technology coming that is able to store even enough power to a moderately sized city for a few minutes, much less the weeks it would need.

>And when we get to that last 5-10%

Try 50%.

> Except renewables can't actually replace fossil fuels. They need baseload provided by fossil fuels to be viable.

False. Renewables could get to 100% of the grid. A key is using hydrogen for the last 10% or so. This is not currently competitive with fossil fuel, but then neither is nuclear.

I'm not an expert in this, but isn't this just a really extra lossy type of battery? Hydrogen takes energy to create, so if you're creating hydrogen from renewable energy sources during peak load, you have to go energy -> hydrogen -> energy with efficiency losses both ways. I remain unimpressed by hydrogen, as it seems to only be practical if you don't any other options for existing stores of energy. For instance a large scale battery system, or even some mechanical energy storage systems.

The benefit of power to gas is that most countries already have an existing system that can both store and burn it. The efficiency is low but the necessary infrastructure investments are also incredibly low compared to building battery based energy storage. It is also an unavoidable step if you want to invest into carbon capture in the future. Excess methane can be exported. Suddenly even a resource poor nation like Germany could become an energy exporter.

When you consider that renewable share in % per day in Germany follows a normal distribution then you realize that there are only a few days that actually need to utilize hydrogen or methane to generate power.

Yes. The advantage of hydrogen is the extremely low cost per unit of stored energy. That is the cost you want to minimize for a storage case with few charge-discharge cycles, like rainy day or seasonal storage.

For diurnal storage, batteries or other more efficient storage technologies would make more sense, since there would be more charge-discharge cycles over which to amortize the cost of the system.

Do you have a source on that low cost?

https://pv-magazine-usa.com/2020/07/03/nrel-study-backs-hydr...

>A National Renewable Energy Laboratory (NREL) study concludes that by 2050, hydrogen storage lasting for two weeks “is expected to be cost-effective.”

2050!!!

By 2050 we'll also have Fusion and other mythical power sources.

To be fair, hydrogen is one of those "mythical" power sources. It just happens to be the one that is becoming viable in the short-term and is not dependent on some future discovery.

It also gave costs for 2025, which are still very low.

So, what? Just use hydrogen like a gigantic grid sized battery?

Also, nuclear is only not competitive due to strangling regulation, not fundamentals.

Burn the hydrogen in combustion turbines, just like natural gas.

You repeat an excuse, but nuclear builds have failed, and continue to fail, not because of regulation, but because of the high unforgiving complexity of nuclear power plants. The post mortems at huge cost overruns point to management failures, not new regulations sprung out of nowhere.

>Burn the hydrogen in combustion turbines, just like natural gas.

That simple? Anyone in the world actually using hydrogen in this way?

You can dissmis nuclear all you want, but doesn't change the fact that renewables (outside of hydro/geothermal for which you need special geography) do not work. There is no nation on this planet that is powered by renewables. There is no nation on this planet that is planning to be powered by renewables. Germany believes in renewables so much they are signing multi-decade contracts to ship gas from Russia and building new pipelines!

> That simple? Anyone in the world actually using hydrogen in this way?

https://www.ge.com/power/gas/fuel-capability/hydrogen-fueled...

"Our turbines have nearly 30 years of experience operating on a variety of fuels that contain hydrogen, totaling over 6 million operating hours as hydrogen-fueled turbines using concentrations ranging from 5% to 95% (by volume)."

> You can dissmis nuclear all you want, but doesn't change the fact that renewables (outside of hydro/geothermal for which you need special geography) do not work. There is no nation on this planet that is powered by renewables.

Ah yes, the old "nothing can ever happen for the first time" argument. Mindless reactionary nonsense. No nation is powered by renewables, therefore no nation can ever be powered by renewables, technical arguments be damned.

That hydrogen can be used as fuel is not controversial. Of course it can. What I mean is that nobody is actually using hydrogen as part of a renewable life-cycle (i.e. generating hydrogen from excess renewable energy, and using as energy store).

>Ah yes, the old "nothing can ever happen for the first time" argument. Mindless reactionary nonsense.

The problem for you is that renewables have been around for years so the fact that they aren't powering any economy needs an explanation. Furthermore, even conceptually, you haven't explained HOW they would power an economy. Renewables have well known limitations. They are diffuse power sources, require huge surface areas covered with high-tech collectors, and are highly variable. The only way we can get them to work is by attaching them to a grid with natural gas or coal - because we have no way to store excess energy enough to bridge their variability. You can deny this, but it is an actual fact and the fact that you cannot point me to a region that has solved this should be quite telling.

> The problem for you is that renewables have been around for years so the fact that they aren't powering any economy needs an explanation.

Renewables have only recently become competitive (or more than competitive). This has happened so fast that existing generating capacity is still largely the old technology. Those old technologies will only be ripped out when their OPERATING costs are greater than the full cost of installing renewables to replace them.

That this old technology is still there doesn't mean it's competitive on a clean sheet basis, it just means it's not worth ripping it out yet.

The rapid decline in renewable prices leads to some interesting contrasts. In the UAE, for example, they are now bringing some Korean reactors online that were green lighted about a decade ago. They will produce power for somewhere around $.08/kWh, perhaps a bit higher. At the same time, contracts have been signed for a large PV field there that will sell power for $0.013/kWh. If they had waited on those nuclear plants and just built PV now they would have come out ahead.

What you want to look at is where new money is going, when new generating capacity is needed (and remember, power demand has been flat in the US for a decade). Renewables are taking a large share of that, and would take a larger share (in the US) with nonzero CO2 taxes.

>A key is using hydrogen for the last 10% or so

Where is this 10% coming from? Solar doesn't work at all between sundown and sunrise. Similarly wind also doesn't align with energy demand either.

Hydrogen isn't the answer either. It is incredibly inefficient to produce via electrolysis, meaning you would have to massively over-provison your collectors.

>This is not currently competitive with fossil fuel, but then neither is nuclear.

Why are you focusing on price? Renewables don't work. They could be free and you'd still be building natural gas plants. This why no nation is actually powered by renewables. When an article claims a nation has reached 100% renewable energy it's always geothermal or hydro - which require the right geography.

The last 10% comes from rare dark/calm periods. I requires a storage technology beyond mere diurnal storage (for which there are plenty of options and where continued cost decline will almost certainly address your negativity.)

> Why are you focusing on price? Renewables don't work. They could be free and you'd still be building natural gas plants.

If price is no object, then obviously renewables can be made to work anywhere. After all, one could dump heat into underground thermal stores and use that as artificial geothermal. The thermal time constant for a several hundred meter chunk of bedrock is measured in centuries.

Baseload is the last thing renewables need. Baseload displaces renewables because you cannot regulate baseload fast enough. Just take a look at Germany and their coal fueled baseload. It has slowed down the Energiewende and made electricity expensive because consumers have to pay for curtailed renewables via the EEG surcharge. Imagine building a renewable energy infrastructure and then throwing out the electricity it generates because coal buddies get priority treatment. If anything Germany needs to get rid of any legacy baseload plants and take advantage of more flexible power plants like natural gas plants.

> Oh, they are variable across days, seasons and inter-year periods, but there is no battery technology coming that is able to store even enough power to a moderately sized city for a few minutes, much less the weeks it would need.

There is no need for "battery technology". This strawman keeps getting repeated just like crappy quadrocopters being used as killer weapons. It's because docile consumers can't see past their own little bubble and imagine that industrial giants also buy all their stuff at the supermarket.

No, you use an entire mix of different technologies and strategies to solve this problem. I don't want to repeat myself but seasonal differences are usually solved by curtailment. If winter needs more power you build enough plants for winter and then curtail the excess energy in summer. Medium term storage needs are trivially met with power to gas (both hydrogen and methane), short term storage needs can be solved via batteries, compressed air or thermal storage. All of these technologies have been available for a long time. The reality is that renewables can easily reach 80% generation without any storage investments at all so practically no country on earth has bothered to invest into additional storage, not because technologies are missing. It's just not a practical concern for the next 20 years.

>you use an entire mix of different technologies and strategies to solve this problem.

Like what? What's the mix?

And why is nobody doing it?

>It's just not a practical concern for the next 20 years.

Because they use coal, natural gas, nuclear, hydro or geothermal ... You know power sources that can actually power a modern economy.

>Medium term storage needs are trivially met with power to gas (both hydrogen and methane), short term storage needs can be solved via batteries, compressed air or thermal storage.

You keep using words like 'trivial' when no nation is actually building this kind of infrastructure. No nation even has plans to build this infrastructure.

>All of these technologies have been available for a long

Yes. Therefore it should make you question why they aren't being used. Perhaps they aren't because they don't work at grid-scale?

I think this is artificially restricting the solution space.

We have fixed price and supply of electricity over the day due to historical reasons, but it's not the future. And we can make the change gradually using variable pricing and speeding up the transition with the tools of regulation.

And of course other developments will help counter the price spikes - manufacturing and cooling systems adapting their power usage patterns, hvdc lines, energy storage, houses getting more energy efficient to cool/heat by using insulation and heat/cold recovery in ventilation etc etc. Energy is currently just so incredibly cheap that most obvious improvements are left on the table or progressing at glacial speeds.

Only residential has fixed prices for electricity over the day. Industrial electricity has always been sold in variable prices, e.g. furnaces and other large consumers run when power is cheap and go into hold mode when it is expensive. There even is a large discount if you allow the power company to switch your consumption on/off. All that is already a reality and has been for decades.

Which goes to show: there still is baseload to consider, and there always will be. There is a green electricity ceiling that can only be circumvented with storage.

But still "base load" is no law of nature. Half the market is on fixed price, and regulation has so far dictated that there has to be a lot of "base load" type power production, and without co2 externalities priced in that's been profitable for producers too. But it's all rules and tech we invented and can be changed.

The amount of supply following the current industrial users are incentivised to do, and the requisite investments, will be much higher once the fixed price system is dismantled from the remaining portion of the market.

>But still "base load" is no law of nature.

It's a term that means we can guarantee a set amount of power regardless of environmental conditions. Renewables are sensitive to environmental conditions, and we don't have a battery technology to bridge renewable variablity and hence the need for 'base-load'. In that context, 'base load' is a law of nature.

"base load" is also a statistical law of nature. Your consumers might want or need to consume your product regardless of the price. I will run my heating before I freeze to death, and I will thaw a meal before I starve. Somebody will switch on their lights in the dead of night. The hospital will run its MRI when an emergency arises.

All these things are rather randomly occuring things, some correlated, some uncorrelated. All these create a "ground noise" of consumption. All these things lead to a need for baseload power, which can only be removed by switching off the grid.

There are quite some MWh batteries coming online in California. Storage is progressing

Sure. And those are great for temporary grid management.

If you scale that up by several orders of magnitude, you may just provide enough to power a small city for a few hours.

The growth numbers currently show that for every extra watt of renewable energy consumed globally, 10 more watts of fossil fuels get consumed. (https://ourworldindata.org/energy#energy-consumption-by-sour...)

Looking at the multiple data sources at https://en.wikipedia.org/wiki/Cost_of_electricity_by_source, there is several patterns that should be fairly clear. Gas and oil is very cheap to build, and the market price they can get out per generated MWh is higher.

> Look, I think nuclear is fine. But as a numerate and earnest environmentalist, I don't see where the case for it lies

Lets put this is in numerical estimated numbers.

An investor builds a wind farm. On average they produce a MWh that costed $35. For the period which the energy was produced they managed to sell it for $85 netting them a profit of $50.

An other investor builds a natural gas power plant. On average they produce a MWh that costed $45. Since they can choose when to produce it they managed to get an average price of $150, netting them a profit of $105.

As a balance sheet, $35 is a cheaper price than $45 when producing 1 MWh. $50 is however much less than $105, making the more expensive energy source the more profitable choice of investment.

Cost is just half the picture in any commercial venture. The daily energy price for 1 MWh varies heavily based on demand and supply. $50 per MWh one day could be $500 a few days later when supply is low and demands is high.

I would also describe myself as numerate and earnest environmentalist, and my view is similar to your but with a clear distinction. Right now we must stop burning fossil fuels. If it cost $45 to produce and they can earn $150, investors who only care about money will continue to invest in fossil fuels. That must stop. The climate will have won a partial victory when the investment into fossil fueled power plants are a proven poor investment, and then we can move on to the transport sector.

The downvotes demonstrate quite clearly what earnest environmentalist think about continuing the existing trend of fossil fueled energy.

I don't think they're necessarily chasing grid power at first anyway. They seem positioned to tackle mass market electricity last, after addressing smaller higher price markets.

> the equivalent generating capacity in renewables would cost

But most of the time, the “cost” of renewable is undefined, because you simply can't use it: you need the grid to compensate for the intermittent generation. No battery won't be enough unless you are in a really specific situation:

- hydroelectric source available, in which case, it's a no-brainer, but most of those sites are already exploited (hydroelectric power was historically the first to come).

- if tpu want to go for solar power, you need tropical or subtropical area, where you have the same amount of sun during the whole year. Otherwise you either need to have batteries able to sustain the whole winter, or dimension your system for winter (which dramatically increases the cost and you end up with a lot of unused power during summertimes).

- for wind power, you need a regularly windy area (the top of a hill, the middle of the sea) and batteries because the world best wind power sites still don't produce every days (storms means shutdown for instance).

If you don't have the perfect spot for one of those, the batteries aren't even a solution[1], because you'd need something like two to four weeks of power in terms of storage…

Or, you could go for a mix between fossil fuel and renewable, which is why the oil and gas industry is making a big push towards renewables …

you missed judrogen electrolysis which comes out comparable in expense to nuclear when deployed on top of renewables, amd gives the same level of stability?

Yield is a problem, as well as rhe ability to store hydrogen safely (injecting it in the natural gas grid isn't possible when we're talking about off-the-grid scenarios).

And we where supposed to talk about renewable “3 to 10 times cheaper than nuclear”, so bringing a storage technology with an higher investment per Watt than nuclear pretty much defeats the argument.

Think of smaller nuclear reactors as an option for places where renewables might not make sense, like in mining. Building power generation that costs 10x some renewable source will never make sense. However, building small scale nuclear as replacement for onsite diesel generators could make a whole lot of sense. Illinois Energy Prof has a good video on this => https://youtu.be/7gtog_gOaGQ

> But as a numerate and earnest environmentalist, I don't see where the case for it lies.

There's two main benefits to nuclear:

1) The big case is the lack of good power storage. While areas like the South West US have plenty of sunshine to go around, most other places don't have this luxury. Therefore you need substantially more storage. I'll give you an example, I've been sitting in Oregon for the past week and sun levels have been pretty abysmal, the past week. Dust and ash are covering panels, making things worse, and this is the sunniest time of the year. There's plenty of times I don't see the sun for a few days straight. Same with wind. Because of this we need to mix up our power sources a bit. I mean you also don't want to depend on only two power sources. I think most of us in the nuclear community agree on: wind + solar + hydro + nuclear as the model. Without good enough batteries nuclear is a great option for baseloads. Remember that you need enough battery storage for rare events where there isn't much sun and wind. Current solutions aren't quite there yet without massive footprints, CO2eq costs, and a high price tag. The unfortunate fact is that wind and solar can't stand up by themselves. They need support. People talk about how you can do it with batteries, water storage, whatever, but until it is implemented I'm not holding my breath.

2) Nuclear has a small physical footprint. This means less disruption of local wildlife, homes, etc. This is a big environmental factor to me.

As to what these people are doing, there's a few factors they are banking on (bets). It'll be interesting to see how that pans out and if their bets can pay off.

- With small modular designs they can get the benefits from economies of scale. This hasn't ever really been a thing for nuclear in the past. This has been a big driver for the huge costs.

- Small reactors should be cheaper to insure (this still needs to be resolved from what I'm aware of and has historically made small reactors unattractive).

- Small and modular enables the ability to better fit the local environment and meet the specific needs. Southern California? Probably don't need any. Alaska? Few would be nice. The idea behind these is that you don't have to transmit energy far (lower loss due to transmission), can add independence from the grid (to corporations or universities), and you don't need to put them in areas that can better rely on wind and solar, i.e. better scaling than conventional nuclear plants.

Of course, this is still a bet and has been one people have been talking about for well over a decade, but why dismiss it before they place their chips? If the bets pay off we should welcome them with open arms. If they don't pay off, well we're on a forum hosted by a startup company where we all know most startups fail. Glad someone at least tried. We're not talking about enough investment money that it would sway the scales and it is good to spread out your bets.

Independent of energy production, I support any nuclear program that reduces our stockpiles, on any time scale. Bonus yay if we also manage to get some economic benefit.

So, why did so many countries/companies originally build nuclear reactors, if the economics inherently don’t make sense? Has something changed?

To make bombs. And what's changed now is we have enough bombs.

Seriously, even the reactors that aren't expressly breeding fuel were designed with an eye to the technology and as part of programs in the 50's-70's to build out expertise in this critical technology area of national security.

Countries wanted bombs. To make bombs you need a nuclear industry. Ergo, everyone who wanted bombs built civilian reactors, without exception.

This argument doesn’t hold. There are several countries such as Finland that have operated nuclear power plants for decades and do not have ambitions for nuclear bombs.

If you want to go with that argument: cite Japan, not Finland. But it doesn't have to be 100% true to be largely true. The overwhelming majority of nuclear capacity has been built out by nuclear powers or aspiring nuclear powers. And with few exceptions the end of their warhead buildout happens to correlate with the end of their megawatt buildout.

Japan has a stockpile of separated reactor grade plutonium sufficient to make thousands of bombs. This is a deniable kind of proliferation. If push came to shove they could weaponize that material without having to make new plutonium.

(And, yes, reactor grade Pu CAN be used in weapons, with proper design.)

Explain Japan's nuclear power plants.

I think a lot of countries built reactors when the economics of Nuclear reactors as well as the economics of renewables were unknown with respect what they would look like in scaled deployment. Modern looks at the operating costs vs time to build, vs modern construction practices from the US, France, China, et al seems say the economics of Nuclear reactors have stalled or increased, while the economics of renewables continue to make accelerating improvements.

Yes we have more stringent safety requirements on nuclear, driving up costs

renewables were 10-100x times more expensive back then

But we also didn't know about climate change yet (for the most part), right? So nuclear was competing with oil.

Producing plutonium, I suppose.

Yes. Regulations.

The economics are what they are do to the legal burden and regulatory overhead more than the technology itself, at least here in the US.

And renewables are cheap because peaker plants exist, which allows the price of storage to be externalized to the peaker plants.

Well, you also have to consider that many gas plants emit less CO2 than coal plats. You're going to need them either way because it's the fastest way to reduce the CO2 footprint of your nation's electricity mix.

I really hope *all" natural gas plants emit less CO2 than coal. Joule for Joule, burning C produces more CO2 than burning CH4.

If it can be combined with desalination, and profitably create medical isotopes, that could help balance the books. But yea, you are right. And solar keeps getting cheaper.

The cheapest technology for desalination is reverse osmosis, which uses mechanical power, not heat, to drive things.

Well, I am still hoping for ITER.

If its gross fusion power were converted to electrical power at 40% efficiency, it will have cost $100/W(e) to build, an order of magnitude more expensive than fission power plants.

(Its net power output would be negative, btw, and there won't be enough tritium to run it for more than a few weeks total, as it's not going to breed its own tritium.)

Not to mention the waste which needs an armed guard almost forever, making the total cost of the energy almost infinite.

> an armed guard almost forever

Things that have long half-lives are not very dangerous, as in order to have a long half-life it must be emitting radiation at correspondingly lower levels.

For example, something with a half-life of 100,000 years is emitting 1/100,000 of the radiation that one with a half-life of a year is.

There's an easy way to remember this with a little knowledge of nuclear. Radiation means the material is losing mass (yes, even gamma emitters. Remember Einstein). If it is putting out a lot of radiation, it is losing a lot of mass. Therefore, logically, it can't last that long.

Essentially it is like having a well. If you attach a firehose to it you'll drain it overnight. If you attach a dripper to it, it'll take centuries to drain.

I actually have no idea how long waste needs to be kept out of the hands of criminals, but I assume it's "a long time". My comment was mostly facetious, but the point I was making was serious. We never really take the full cost of disposing waste and pollution into consideration.

I think, when you buy a new solar panel, you should also pay the full cost to pull it apart and dispose of it as well.

The way to do that is to make the manufacturer responsible for the mandatory recycling.

That works to make the manufacturer price in that externality, but it also likely leads to lease arrangements instead of outright purchase.

That’s a carbon tax right? How much should a gallon of gas cost?

Not just carbon, but plastics and all kinds of pollution. Even the clean up of recyclables like glass and metal should be paid for upfront rather thank asking society to pickup the tab after the fact.

At this point, probably $20-30. Then we'll see how "expensive" nuclear is.

Nuclear waste only lasts "forever" when it is not recycled. And even then it is not dangerous and doesn't need an "armed guard" after a few weeks when all the most radioactive material has decayed away.

let us weigh the damage of secured monitored nuclear waste leaking vs the uncollected nuclear isotopes into the open environment released today by the carbon burning industry.

https://www.scientificamerican.com/article/coal-ash-is-more-...

also we don't need to prevent 100% of the waste from leaking just enough to keep from rising the natural ambient background that would be in the environment naturally before we extracted it. Secondly we have the means disposing of in newer reactor designs that use waste from older reactors as fuel and their waste has a much lower half life (down from 100,000 to 500 years)

I’m okay with costs being 3-10x for something with zero emissions. And the more we build the cheaper it gets. We have to invest in the future technology. Solar is a good 20th century technology but it is poisonous and degrades over time. Wind is useless.

There are quite a few ways to generate electricity from solar, and not all of them degrade over time or are "poisonous".

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

> which you can't really do because of Xenon poisoning

Can you clarify this part? You mean that channging the power level of the reactor would cause Xenon poisoning?

I highly recommend HBO's Chernobyl. Explains a lot this stuff. Basically, as a nuclear reactor operates, it is turning the Uranium into lots of other stuff. One component of this other stuff is Xenon-135, which absorbs neutrons and basically gets in the way of nuclear reactions. This is not an issue if you keep the reactor on because there are enough neutrons to keep things going and overcome the Xenon absorption. But if you turn the reactor off, you now have no neutrons from the fission reactions and lots of neutron absorbers. There is so much absorption that it's impossible to start up the reactor for a couple of days. That's why nuclear reactors operate at constant power and only turn off to refuel. This MMR also operates at constant power, which would also have benefits for cycling issues, and uses the molten salt loop to store heat and then use it when it's needed.

Scott Manley made a more condensed form of the explanation of the nuclear physics involved here: https://www.youtube.com/watch?v=q3d3rzFTrLg (~21 minutes). It's an excellent explanation of the processes.

The HBO show is also an excellent drama and worth a watch.

When a reactor is "poisoned" with Xenon, the Xenon absorbs a large fraction of neutrons that otherwise would cause fission.

If you don't want to wait for the Xenon to decay you could construct a reactor that has a larger control swing than you would need otherwise -- for instance you could put in more and denser control rods. In that case, however, you need to have a lot of "excess reactivity" and also a lot of neutrons lost in the control rods under normal use, which in turn means the "neutron efficiency" is worse. (If you're not planning to breed Pu239 or U233 maybe you don't care)

Note the Xenon concentration can vary in different parts of the reactor so you have to manage the "oscillations" in space just as you do in time. Not a catastrophe, but definitely a hassle.

You can't shut it off on a whim, but lots of reactors can drop to half output or one third output without much trouble. It probably doesn't make a big difference on the current grid, but if you have a significant fraction of nuclear power then the ability becomes useful. (I don't know if those mechanisms are applicable to this design.)

Ah, the the molten salt loop is like a flywheel?

My read is that it's somehow the inverse of a flywheel. A flywheel means consistent output for variable input whereas this is consistent input leading to variable output.

Or maybe I'm way off.

That's just one way you can use a flywheel.

A flywheel is the equivalent of a battery/capacitor. It can smooth inputs, or it can smooth outputs, or it can simply save power for later.

That's my understanding as well. Rather than ramping the power output up or down with demand, you keep it at a constant power level and basically use the molten salt loop as a battery -> store energy when you're outputting more than is required, draw energy from it when you need more that you are producing.

The multi-gigawatt nuclear installations are typically supplemented by natural gas "peaker" plants to deal with higher demand than the reactor output, but lose any extra power produced.

If I recall correctly, the naval reactors (submarine and surface vessels) are the only ones that are designed to rapidly shift their power output.

> The multi-gigawatt nuclear installations are typically supplemented by natural gas "peaker" plants to deal with higher demand than the reactor output, but lose any extra power produced.

In France, the load variation is directly handled by the nuclear plants themselves[1], even if there are also some natural gas running (and most of the time their output is more regular than the one of the nuclear plants, so their must be drawbacks to big load variations on a gas plant as well).

[1] https://www.rte-france.com/eco2mix

That's an excellent explanation, thanks!

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

Excuse me, could you be more specific about why you think Xenon poisoning is relevant here? I think I just missed your point.

In a situation where you want to rapidly ramp a generator's output up or down, Xenon poisoning narrows the window of reactor control system authority. This prevents the reactor from achieving full power on a reasonable time scale.

Xenon is both a fission product and a neutron absorber. So if your reactivity dips too low you won't have enough neutrons buzzing around to sustain the reaction and will have to wait for the xenon to decay before you are able to start up again.

I know very well what is Xenon poisoning, I just fail to see the point they were making.

Great point, power density should be calculated based on land usage.

Reply on news.ycombinator.com