Hacker News Reader

Comment by Alexsky2 | Hacker News Reader

A bit unrelated to the Belgium story but I recently visited Diablo Canyon Nuclear Power Plant near San Luis Obispo, CA and learned a ton about the technical details, safety systems, and policy decisions that go into operating a nuclear power plant. When operating at full capacity, it provides up to 10% of California power! While there is certainly always more such facilities can do for safety and efficiency, my impression is that smart people are working hard to ensure the lessons of previous disasters and potential future ones are mitigated, and that nuclear energy, whether through next-gen small module reactors or legacy systems, will be an important aspect of our future energy grid, especially with the rapidly rising energy demand predicted over the next two decades. If you are interested in a tour, the form can be found here: https://www.pge.com/en/about/pge-systems/nuclear-power.html

View on HN

I did some research about that nuclear power plant. In 1985 dollars, the total construction cost was 5.6B USD. That is an astonishing amount of money. That is at least 16B USD in 2026 money. If you also include decomissioning costs of about 4-5B USD... how the fuck does nuclear power make any economic sense? PV solar plus batteries: ALL THE WAY. To be clear, I am not anti-nuclear power by any means. I think it is a terrific way to power our countries, but the ship has sailed. PV solar has won, and now we can add batteries (and some wind) to get reliability.

$5.6B actually sounds like a good deal. It outputs 2GW+ of power. While solar is definitely cheaper for 2GW of power, you still need batteries for when the sun is down. So you probably need approximately 30GWh of batteries to just replace this one power plant. The batteries alone would cost nearly $7B of grid-scale batteries that must be replaced every 20 years.

Ignoring the fact that the nuclear plant already exists, this still seems like the right way to go mostly because it's impossible to build this nuclear power plant for $16B in the US anymore (or so it seems).

Due to increased regulation etc you cannot just translate 1985 $, £ or Euro to a 2026 one. There is an actual example in the UK Hinkley Point C current estimate $43b, (£35b) where as sizewell B commissioned in 1987 was $3.2b billion (£2b) or about $7b in todays $. This is probably the worst example but makes the point.

    > $5.6B actually sounds like a good deal. It outputs 2GW+ of power.

I don't understand. Are you talking about 1985 dollars of 2026 dollars?

After some research, I learned that thermal powerplants (coal/gas/oil) completed in 1985 cost about 0.8B to 1.2B USD per GW. 5.6B USD in 1985 for 2GW sounds like a terrible price -- at least twice the cost.

Nuclear is high capex low opex. It needs such a miniscule amount of fissile material per year, whereas purchasing coal is an eternal ongoing cost.

Just to put some numbers on it, a 1GW conventional reactor consumes about 25 tonnes of enriched uranium per year, while a 1GW coal plant goes through 3.3 million tonnes of coal.

Price is not the only factor, paying double for energy that does not contribute to global warming and other health issues seems more reasonable.

> a terrible price -- at least twice the cost

I'd double my electricity bill if that means saving somewhere between 3 and 9 million lives per year[1], better health for myself and the people around me, and that's completely ignoring climate change benefits where prevention both saves money and reduces deaths/displacement/poverty in the long term

Either short-term solution is fine (nuclear or full renewable), but we're currently doing everything piecemeal. Plopping down a few big reactors in 20 years while people (in countries without salt planes, at least) are still trying to get permits for the remaining reasonable wind turbine and pumped hydro locations... it just feels like seven-mile boots for the energy transition

If we can make seven-mile steps by plopping down wind/solar plus the required storage in gigawatt quantities, all the better, but that hasn't been happening. We'll run out of uranium eventually but, for now, such reactors buy time. Of course, this discussion has been happening for so long that the "it takes too long to build" naysayers will get their way soon, even at the slow pace we're currently going full renewable at. It's now or never, we need to commit to an option, no matter which one

[1] https://en.wikipedia.org/wiki/Air_pollution#/media/File:How-...

7B for the first set of batts.

Then 7B in 2046 money which is probably $15 today.

assuming 300 days/year, 1c/kwh and ignoring opex that's $150m worth of electricity per year.

It makes economic sense because they require a large initial investment (CAPEX), but low cost per year to keep functioning for many decades (OPEX). In contrast to say wind or solar, which are smaller CAPEX but higher OPEX.

So when you compare average cost per year over the complete expected lifetime of the plants, nuclear is good, but when you compare the up-front cost to build it, yeah it looks bad.

Another thing is that nuclear in the US is far more costly than in e.g. France. The key is that France standardized a few reactor designs that they kept building again and again, which made both construction and maintenance cheaper over time. While in the US, each nuclear plant is a unicorn, which can perhaps result in better individual designs but ends up more expensive.

Unfortunately France can no longer build nuclear plants cheaply either. All of the recent nuclear plants built by the French state owned company EDF in France, Finland, and the UK have seen enormous cost and time overruns.

Cumulative emissions matter. We simply don’t have the time to wait the 20 years it takes to build new nuclear plants.

Source please? The numbers I have seen of real opex paint a different picture. In general, nuclear plants close because of cost.

Googling I see numbers like $11/MWh for onshore wind [1] and $5-$10/MWh for solar [2] while nuclear is around $25/MWh [3].

[1] https://emp.lbl.gov/publications/benchmarking-wind-power-ope...

[2] https://emp.lbl.gov/publications/benchmarking-utility-scale-...

[3] https://www.nei.org/getContentAsset/47fa8caa-9b0d-4029-932c-...

Also wow, for Solar the property taxes are OpEx. So if there's more sunlight because of good weather these "Operating expenses" decrease because they're based on taking fixed costs like property taxes and just dividing them by power output that's unrelated.

I assume property taxes for a gas turbine are likewise OpEx but they just disappear in the noise of buying enormous amounts of methane as fuel.

It doesn't generate power by burning carbon and is a grid replacement for carbon sources. Grid cost rise sharply on 100% solar.

Taking china as an example they currently build solar, coal and nuclear. No country is building only solar/batteries.

Further if we build more nuclear we'd be better at it and it would be cheaper.

> 100% solar

100% solar is a straw man though, as much as the simplicity of it sounds nice.

> Further if we build more nuclear we'd be better at it and it would be cheaper.

This is far from being clear, nuclear is one technology that tends to have increased costs the more we do of it. Even in France!

The costs of the French nuclear scale-up: A case of negative learning by doing https://www.sciencedirect.com/science/article/abs/pii/S03014...

Human labor is very expensive, and every time we make humans more productive, that makes human labor more expensive, because their time becomes more valuable. Technological growth does that.

The cost of nuclear is primarily in labor and long-term financing, due to the very long lifetime and upfront labor cost. Until somebody has some sort of technological breathrough to decrease the labor cost of nuclear, it's not going to be able to compete. Even decades ago it had trouble, and now it's far worse.

You are talking only about the operations of the nuclear, and ignoring all the high energy process required to mine and process uranium before it can be used as a fuel, and after as waste. But let’s pass this problem to the next generation, they will know what to do :)

You underetimate the energy density of nuclear power. Yes. Uranium needs to be mined - slightly more 3xpensive if you extract it from sea water or recycle the fuel - but you need just one bathtub of fuel pellets to power a plant for 2 years. Solar and wind require more mining. https://ourworldindata.org/safest-sources-of-energy

Yes, but at what cost? The sun is free. Nuclear means dependence.

That's still essentially zero relative to the amount of energy we can get out of the uranium.

Yes, but at what cost? The sun is free. Nuclear means dependence.

Nuclear GHG are lowest per UNECE and NREL which do account a lot of factors. Nuclear requires least amount of mining vs any alternative so this argument makes little sense. Nuclear waste can be stored in facilities like onkalo or recycled like at la Hague(now) or Superphenix(in past)

Yes, but at what cost? The sun is free. Nuclear means dependence.

The energy density of uranium is such that the amount of energy required to mine and process uranium is trivial relative to the amount of power produced. The carbon intensity of nuclear power is lower than solar: http://large.stanford.edu/courses/2016/ph240/kountz1/

Surely you include the rare earths needed for solar panels as well in all of your comparisons. Nuclear fuel is incredibly energy dense.

Yes, but at what cost? The sun is free. Nuclear means dependence.

It's a large amount of money, but the plants have a long service life. And once a nuclear plant is built, it's operational costs are much lower than other forms of electricity generation.

Simply saying "use PV plus batteries" really does not engage with the scale of storage required. The US uses 12,000 GWh of electricity per day. The world uses 60,000 GWh of electricity per day. Annual global battery production is around 1,500 GWh, and only ~300 GWh of that production is used for grid storage.

Even just provisioning enough batteries to satisfy the requirements for diurnal fluctuations of solar is far beyond the scale of what battery production can provide. Let alone fluctuations due to weather and seasonal output changes.

It’s not a choice between nuclear and PV. It’s a choice between nuclear and the other things that provide base load: gas and coal.

Or solar / wind (which mostly anticorrelate) + biomass + storage + interconnectors + smart demand.

The amount of baseload we technically need can be pretty slim.

Take Denmark: fossil powers just 9% of their electricity generation, the majority of it is wind and solar. Wind is strong in evenings/nights, solar during the day.

Then they have biomass (indirect solar) as a form of baseload, more sustainable than coal/gas.

Then there's interconnectors, they're close to Norway which can pump hydro, and Sweden, each day about 25% of the electricity is exchanged between these two countries, and that's a growing figure.

With more east/west interconnectors you could move surplus solar between countries. Import from the east in the morning before your own solar ramps up, export your midday surplus west before theirs peaks, and import from the west in the late afternoon as yours fades.

With interconnectors you can also share rather than independently build peaker capacity. Because a lot of peaker plants only run a small amount of time and therefore much of the cost is in the construction/maintenance, not the fuel.

And of course there's storage, which will take a while to build out but the trendlines are extremely strong. Just a fleet of EVs alone, an average EV has a 60 kWh battery, an average EU household uses 12 kWh per day so an average car holds 5 days worth of power a home uses.

And then finally there's smart demand. An average car is parked for more than 95% of the day, and driven 5% of the time. Further, the average car drives just 40km a day which you can charge in 3 minutes on say a Tesla. Given these numbers (EVs store 5 days of household use, can sit at a charger for 23 hours a day, and can smartly plan the 3 minutes a day of charging it actually needs to do) just programming cars to charge smartly, is a trivial social and technical problem in the coming 10-20 years.

Given this, baseload coal/gas can really be minimised the coming decades. It's not going to go away as a need, but I don't think it requires gas/coal or nuclear long-term going forward.

Lot of the biomass used in Denmark to form baseload power generation is imported.

"The utmost amount (46%) of wood pellets comes from the Baltic countries (Latvia and Estonia) and 30% from the USA, Canada and Russia.6 Estonia and Latvia have steadily been the primary exporters of biomass to Denmark, mainly in the form of wood pellets and wood chips."

https://noah.dk/Biomass-consumption-in-Denmark

https://www.eubioenergy.com/2025/03/13/no-smoke-without-fire...

So Denmark replaced lot of imported fossil fuels with imported wood.

Could we scale this form of energy generation to energy requirements of China, India?

No but every region has their own pros and cons. The idea Belgium has no other option than coal gas or nuclear is refuted, and biomass is just one of the reasons.

> "just programming cars to charge smartly, is a trivial social and technical problem in the coming 10-20 years."

One problem I've heard about this idea in the past is that cars and their batteries are expensive, and people won't want to run down the lifetime of their car battery more quickly by also using it as a home battery rather than just for driving.

Obviously this can be solved either by making it so cheap to replace car batteries that nobody cares, or by legislating that people have to use their cars this way. But is either of these solutions easy to happen any time soon?

I don’t think its a long term issue. The cost of battery storage is below 10c per kWh, whereas a peaker plant costs above 20c per kWh and runs 10% of the time.

So if you get paid double the value of your battery the incentives are there for an economic model to work. Today.

And batteries are only getting cheaper, gas is the opposite.

Plus batteries take surplus solar/wind, at these times they have a negative value. Add that and the economics are a no brainer. It’s a matter of time.

So why are the Danish and the Swiss working on Thorium?

https://interestingengineering.com/energy/danish-firm-molten...

The issue with "base load" is that people usually omit to mention how much GW they are talking about and for how long. Which makes it a bit of a bull shit argument.

As an insurance against unspecified lack (how much for how long?) of wind and solar (and batteries, cable capacity, hydro, etc.) base load is supposed to swoop in and save the day when those temporarily fail locally. So, it's a valid question to ask how much insurance we need against that. Nobody seems to really know. There are loose estimates of course. And people seem to assume it's months and that renewables are going to 100% be offline throughout that very very long period. In reality in most connected energy markets, we have a short gap of a few weeks or so in winter at higher latitudes of reduced output that we already manage to cover with flexible generation.

It's more constructive to think in terms of dispatchable power rather than base load. When the sun doesn't shine or there is no wind, it's nice if you can quickly bring online additional generation, tap into battery reserves, or bring in power from elsewhere (via cables). That favors flexible power, not inflexible power. Nuclear and older coal plants are a bit inflexible. Shutting down and starting up a nuclear plant is really slow and expensive and requires a lot of planning. And especially older coal plants need quite a bit of time to bring their boilers up to temperature such that they build up enough steam pressure to generate power. Until then, they are just blowing smoke out of the chimney. Modern coal plants are a bit better on that front. Same with gas plants.

The modern ones only need about 10-20 minutes or so. Still quite slow but something you can plan to do. Slow here means expensive as well. Because shutting them down when there is a surplus of renewables (which is a very common thing now) is really inconvenient. Which means consumers have to pay extra for perfectly good electricity from renewables to be curtailed. That happens by the GW in some markets and keeps consumer prices higher than they should be because they have to pay for gas/coal that is technically not actually needed.

Batteries have a much lower LCOE than gas or coal plants (never mind nuclear) and it's being produced by the TWH per year now. A lot of markets are serving much of their peak demand using batteries now. Australia and China are good examples. Even in the US, you see batteries being deployed at a large scale now. That's starting to push gas and coal out of the market. A gas peaker plant that rarely runs is just really expensive.

A requirement for base load is a fallacy promulgated by fossil fuel preservation lobbying

When it comes to residential/consumer use base load is irrelevant - but when it comes to business (especially industrial) use base load is a strict necessity. The proportional requirements of base load are fading but it is still something that needs to be considered carefully.

Do fossil fuel companies overstate the importance and scale of base load to justify additional fuel subsidies? Indubitably - but don't let their bullshit hide the truth within it that actually is a critical requirement for our power grid.

No, you need to match the demand curve at all times.

This is a confusing thing to say, can you explain?

What you need - the only thing you need - is dispatchable power. That is power supply that can rise and fall to meet demand. That is not what baseload is. It's also not what wind/solar provide.

What baseload is is electricity supply which is only economical if you use it all the time. Nuclear falls into this category because of its very high capital cost and low op-ex. If it's cheaper than dispatchable power (nuclear isn't) it's nice to have as much of it as the minimum demand that you see on the grid, to lower costs. If it's as expensive, or more expensive, than dispatchable power, that's fine, you just don't need it at all and can replace it entirely with dispatchable power.

It's similar to wind and solar in this, which also aren't dispatchable (though there supply curve looks different than the constant supply curve which "base load" is used to mean). Except wind and solar actually are cheaper than dispatchable power so they make economic sense.

The term is half marketing term and half a theory that constant supply non-dispatchable power would be significantly cheaper than dispatchable power so we should organize the grid around it. That theory didn't really pan out (apart from some places with non-storable hydro, and a few with geothermal).

have a read through this: https://cleanenergyreview.io/p/baseload-is-a-myth

basically, base load means the lowest point of demand on the grid. And you matched that with slow-to-respond thermal power plants (coal mainly, also nukes). Because those are slow to respond and are most profitable running at 100%, so you tried to keep them there. So called base load generation.

But note there is no rule of the universe that says you have to meet the base load demand with some static constant power source, you can get it from anywhere. And now, since renewables and batteries are cheaper than this base load generation, it knocks them off the grid rendering it unprofitable. So the whole concept of base load supply is obsolete. Anyway, the linked blog explains it better.

You don't need to run coal power plant close to 100% to be profitable. You want to run nuclear power plant close to 100% because fuel is cheap and you want pay back CAPEX as early as possible.

The article you send is perfect example why it's not economic to build new coal or nuclear power plants in US. The reasons are: very cheap natural gas and no CO2 tax. In US natural gas + solar is the cheapest way to generate electricity.

In Europe the situation is very different.

"Europe is in the opposite spot. The continent's main gas point, the TTF benchmark, nearly doubled to over €60/MWh by mid-March."

https://www.briefs.co/news/u-s-natural-gas-just-hit-a-record...

Renewable + battery is already the cheapest and fastest way to build new power in many domains + geographies, and the number of and range keeps expanding as the price keeps dropping.

It's always a peculiar response that outright ignores certain power combos, and it always seems to come in nuclear discussions.

You mean solar + batteries. Batteries can be used to flatten the day/night cycle of a solar power plant on many days. But it's depending on location and it's not cheap.

For cost of $100 / MWh, you can do it in sunny regions like Spain, Mexico, middle East.

https://www.carbonbrief.org/guest-post-how-solar-panels-and-...

It doesn't make economic sense to combine batteries with wind, because periods of low wind output can be multiple weeks long. The betterie costs would be ridiculous.

How do you cover the last 10-20% of the missing solar+wind+batteries output, at what cost?

Usually its a combo of solar+wind+batteries+fosil , costs depend gas cost and CO2 costs.

Five years ago, skeptics were saying a few hours of grid on battery was "impossible" and coal would be displaced only by gas. Now we're talking days of grid supply from batteries even from non-grid optimal lithium tech. Batteries are following the manufacturing optimization S-curve and its not even on the fastest part of the curve yet. Lithium is dominant only because its early in the pipe but multiple types of other battery techs (flow, iron+ etc) will probably become viable unlocking weeks -> months -> seasons of output. And probably this will happen before the first nuclear plant from this Iran adventure energy squeeze is even ready to light up - as its taking the west 20+ years to plan and commission a nuclear plant.

It took decades to develop, comercialize lithium batteries, starting in 1970s. Only after 2016 we started to see a large growth.

https://en.wikipedia.org/wiki/History_of_the_lithium-ion_bat...

Cheap gas has displaced lot of coal in US, staring from 2007, fracking gas revolution. In China not much.

https://ourworldindata.org/grapher/share-energy-source-sub?t...

Currently, the Iran energy squeeze is not much felt in the west, just slight increase in prices. It would take real pain, something like 1973 oil crisis, for west to reconsider nuclear power generation.

> How do you cover the last 10-20% of the missing solar+wind+batteries output, at what cost?

First of all, we're very far from this being a problem. If you "only" move 90% of the electrical grid to solar and use fossil fuels to make up the remaining 10% it's a ridiculously huge win anyways. The person you are talking to is just talking about "new power", not "replacing all existing power"... so unless the grid is growing by 5 to 10x your objection here is utterly irrelevant.

Secondly, that whitepaper shows you can do this with incredibly unfavourable assumptions. Namely that they're

1. Ignoring transmission, in reality we can and will move power around from sunny to shady areas. The paper is assuming a single off grid facility. Because different areas are cloudy at different times this greatly reduces the peak amount of batteries needed.

2. Ignoring other sources of energy, like wind, hydro, etc. Because their failures are uncorrelated with solars failures, they greatly reduce the amount of storage needed to hit reliability targets. It's a lot more likely that you'll have a cloudy week than a cloudy and windless week.

This is also why pairing wind with batteries makes a ton of sense. You aren't just pairing wind with batteries, you're pairing wind with a mix of other electricity and batteries. The more uncorrelated sources of electricity you have the less batteries you need to paper over outages.

so what should europe do? gas being expensive doesnt make nuclear economics better for the role of variable backstop of an increasingly renewable grid. Its still a fatal economic equation for nuclear.

Btw battery is rapidly changing the math on > US natural gas + solar is the cheapest way to generate electricity

california went from 45% gas in 2022 to 25% gas in 2025 almost entirely because of batteries (and more solar), and they're just getting started. I know its not generally true across the US, but very soon batteries are going to be pushing a huge amount of gas off the grid.

Europe should:

1. Think very hard how to reduce reliance on imports (US, Russia, Middle East fossil fuels), (China solar, batteries, consumer goods).

2. Think very hard about the decarbinization strategy. The strategy was: we will replace fossil fuels with renewables, we will be successfull, everybody else will see how see how successfull we are and everybody will follow our example. This strategy is not working, world-wide CO2 emmisions are increasing every year, industry from Europe is moving outside Europe because high energy prices, GDP is stagnating. Europe should think how to push other major CO2 emmiters China, India, US, to strongly decrease CO2. Europe alone will fail to fix climate change.

I suggest you read a power system engineering textbook.

it actually is a choice between nuclear and PV, because base load supply is an obsolete concept. Because actually nuclear is terrible in a grid increasingly full of nearly-free variable sources (solar&wind). The nukes need to stay at 100% all the time selling their power at a high fixed price to have any remote chance of being economical. Cheap variables push nuke's expensive power off the grid during the day, and increasingly into the evenings with batteries. This is unavoidable in an open energy market, and is fatal to the economics of nuclear. You cannot make them work without massive state subsidies.

Gas is far better suited economically to backstop a variable grid. I wish it werent true, because i dont hate nukes, but it is just economics.

I will also point out that california is down to 25% fossil sourced power in 2025, from 45% in 2022. Due to renewables and batteries, and there's far more coming. The amount left to backstop on gas in a few years could plausibly be 10%! which is amazing.

No country in the world has 100% solar+wind power supply. Even tiny island countries have to use expensive diesel to supply guaranteed power.

And once you have diesel generators, it turns out that batteries are more expensive than just buying a bit more fuel.

In the long term nations will likely end up with whatever renewables work for them (hydro/solar/wind/thermal) plus the appropriate amount of batteries and expensive, low use (stored) gas/oil plants for "emergencies" where the renewables do not deliver and the batteries might be exhausted. Some nukes will be in the mix obviously but they will not be widespread globally.

The future is all about sovereign power generation and distributed reliability.

Maybe there is a discussion to be had about WHY it needs to make economic sense? Power is a natural monopoly, maybe it doesn't need to be a part of the economy?

It was not a good idea for Germany (and certain other parts of the EU) to be so dependent on Russian gas. It was also not a good idea to become dependent on LNG from Qatar or the US. Spain uses natural gas from Algeria (via Morocco), also not great. Italy also gets some from Algeria/Tunesia, still not great. Inside of Europe, we are far too dependent on Norway. Not because Norway is likely to turn on us (or we on them), but because the pipelines are relatively easy to disrupt.

The transition from coal to gas gave us cleaner air (and less CO2) but it definitely also had costs, some of them in the form of many thousands of dead Ukrainians, some of them in the form of concessions to the US.

And $ cost is a poor metric to chase when what you really care about includes a lot more-- exposure to the whims of geopolitical forces you can't foresee or control, which have both $ cost and more.

I agree to an extent… but a state forcing a nuclear share and locking the populace into higher power prices for 30+ years is going to politically very unpopular. Short term economic concerns dominate today.

Tough luck. That's the point of representative government: look out for the interests of the nation and sell it to your populace. If you can't sell it, be prepared to be voted out, but do the right thing.

you're talking about "should". Im talking about the world we live in. They are unfortunately not the same.

Higher energy prices is something the population notices when they come.

But when higher prices stick around industries close or never opened.

Yes, and pursing state sponsored nuclear means higher prices are guaranteed to stick around

Power distribution is a natural monopoly, power production is commodified/competitive business.

Because if a thing is valued by thing-consumers at x and you set the price to <x, then you are incentivizing people to use more of the thing than they need, even to waste the thing. This thus requires more infra than is actually needed or wanted.

This doesn't go away under socialism/communism/collectivism. If you set the price too low, you either have to build far more production capacity at public expense than needed, or you cope with regular blackouts.

It no doubt made sense in 1985 - solar was rubbish then.

> PV solar plus batteries: ALL THE WAY.

how much this would cost for the same guaranteed power output?

would it be more or less than 21B?

how it would look like in areas that have winter with snow?

Wow - nearly 20% of the California bullet train! Almost double the wildlife crossing!

decomissioning is embedded in opex cost and fairly cheap www.kkg.ch/de/uns/geschaefts-nachhaltigkeitsberichte.html

The complexity now is doing it without delays. China shows that it can be built very cheap and fast with good supply chain

> China shows that it can be built very cheap and fast with good supply chain

I mean, thank you, the USSR already showed this, no more is needed.

it's not about ussr, what a nonsense. It's about having good supply chain. Like France had during messmer or Korea now (albeit far from china). China is building the same ap1000 copycat much faster and cheaper

> how the fuck does nuclear power make any economic sense

Because these plants run for 80+ years (some countries are now considering 100) while most renewables run for 25 at most. And also because `plus batteries` doesn't exist. The world battery capacity isn't enough to power California for a single week. Large scale battery technology isn't even in its infancy, it just doesn't exist.

Don't forget, you've paid for the nuclear power plant once. You will pay for a new set of renewable capabilities every 25 years in <current-year + 25> dollars.

25 year replacement for solar is a myth. [1] . They may degrade to ~80% but they keep on working and producing, so far it seems almost indefinitely.

[1] https://www.ecoticias.com/en/goodbye-to-the-idea-that-solar-...

So in essence they studied all of six (6) panels in a non-dusty region with a mild climate and without notable saline corrosion AND the article even mentions that most other studies are well in the 25-30 year range. Also the study clearly outlines that older silicone panels can't be compared against modern mass produced variants. Finally the study only examines modules that are still working, failed or removed systems are not in the dataset so you have heavy hidden distortion.

The sample size is extremely limited. Six systems are not at all robust enough for global conclusions. This popsci article of yours doesn't hold up to scrutiny and neither it nor the study are enough to make sweeping generalizations like declaring the common 25 year lifecycle a myth.

The “common 25 year lifecycle” was never a drop dead expectation. That’s usually just what they’re warrantied for. It’s always been a common misunderstanding

Edit: If you don’t trust my source , please show one of your own that proves they need to be replaced at 25yrs

I'm so glad we saved Diablo. It was VERY close to being shut down the same year we were having rolling blackouts.

So close - big save indeed.

The problem has never been the lack of smart people for Chernobyl or Fukushima. Rather the fact that dumb, short sighted people were in power and drove the smart people away.

And unfortunately, it doesn’t look like this is going to stop any time soon.

I really wish the same could be said for San Onofre. To say nothing of its value as a landmark -- it will live on in our memories as the great San Onofre boobies

One upside -- is that SONGS being decommissioned gave the energy storage market the ability to level up in a big way back then. They filled part of the gap with some large MW procurements. Allowed BESS to be part of the collective energy solution. Nuclear + Solar + BESS + some small amounts of NG is a dream team.

"Ironically, what originally motivated pumped storage installations was the inflexibility of nuclear power. Nuclear plants’ large steam turbines run best at full power. Pumped storage can defer surplus nuclear power generated overnight (when consumption is low) to help meet the next day’s demand peak."

https://spectrum.ieee.org/a-pumped-hydro-energystorage-renai...

Main challenge with pumped storage is its geographically limited, always a custom project, and large scale deployment.

Yeah, nuclear provides a steady base load, so the percentage goes up or down depending on overall grid utilization. Right now its doing 2.28 MW [0], which is more than what Wikipedia claims as its "Nameplace capacity" of 2.256 MW [1].

0. https://www.gridstatus.io/live/caiso

1. https://en.wikipedia.org/wiki/Diablo_Canyon_Power_Plant

The correct unit is GW.

To use the strongest plausible interpretation, the writer could be from Belgium, where the meaning of 2.256 depends on whether they wrote the number in French, Dutch or German.

The Belgians apparently typically invert the meaning of . and , in numbers (from how they are used in the US).

  To make large numbers readable, Belgians use either a period (.) or a non-breaking space. Example: Two thousand thirty-six is written as 2.036 or 2 036. In formal Belgian French, the space is increasingly preferred over the period to avoid confusion with the Anglo-American system, but the period remains very common in Belgian Dutch and everyday shorthand.

I would guess Europeans tend to be better at SI units than people from the US. And let's not mention the the cancer of changing the value of G depending on context.

diablo canyon has 2 reactors that can make 1.1MW per hour continuously. about 2.2MW/hr they both aren’t always on but that’s the goal. It’s closer to 2MW/hr actual

the largest solar plant in california is Ivanpah. It made 85GW/year. Thats 97MW/hr.

It would take 20 clones of Ivanpah to match one diablo canyon. Ivanpah took 4 years to build, cost 2.5B and was in discussions to close because it’s not cost effective.

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

The math in this comment is all over the place.

Ivanpah is solar thermal. Nobody is advocating for solar thermal, photovoltaic has decisively won.

mount signal, the largest PV plant in california makes 1,200GW/hrs per year. it would still take ~15 copies of mount signal for a single diablo canyon https://en.wikipedia.org/wiki/Mount_Signal_Solar

my whole point is solar is great, but the insane scale it requires to get reasonable output is really underestimated. you would need solar fields 100sqmi big. probably many of them. solar alone won’t be the future of humanities energy needs because it’s not efficient enough. we should still keep building solar. but if we aren’t building nuclear too its not enough growth

The other day I calculated what it would take to run my entire country on pure solar, assuming magical infinite storage capacity. Even here in Central Europe, the required area for all the panels was a pretty insignificant number that, even if built as a single huge circle, would easily fit in many different places.

Did you ever calculate the cost for a hypothetical battery that could keep solar power available whenever the sun does not shine? This is where nuclear, well, shines

100 square miles isn't big, the US is 3.5 million square miles. It's also the size of a nuclear plants emergency planning zone, so sizes are comparable.

What truly are humanity’s energy needs, though?

Do we need Facebook? Do we need Instagram? Do we need deepfakes and AI music?

> diablo canyon has 2 reactors that can make 1.1MW per hour continuously

MW/hr is a nonsense unit for generation capacity. The 2 reactors at Diablo Canyon each generate around 1.1GW of electricity (not MW, and not “per hour”, watts are already energy/time.)

> the largest solar plant in california is Ivanpah. It made 85GW/year. Thats 97MW/hr.

Ivanpah is a badly designed plant that isn't representative of CA’s solar generation (which is largely distributed, not large utility-scale plants) and is being shut down, but also these numbers are both nonsense units and unrelated to the actual stats.

Ivanpah’s peak output capacity is 397MW, it was intended to produce around 1TW-h per year, and it has actually produced an average of 732GW-h per year (equivalent to an average output of around 84MW).

There is so much misinformation in here, so densely packed.

Ivanpah is is not the largest solar power plant in California. It's an experimental solar-thermal plant. Talking about megawatts per year is not a meaningful term (megawatt-years would be). Ivanpah despite its much talked about failures delivers between 350 and 850GWh per year.

The largest solar plant in California is Edwards Sandborn, producing somewhere around 2500GWh per year (it's newer so numbers are less published).

Diablo Canyon produces around 18000GWh/year, which is huge.

But with all costs combined, Diablo's price per MWh is close to ONE HUNDRED AND TWENTY DOLLARS off of a massive initial capex. Modern solar battery installs trend towards $30-60 for the same output.

So I'm sure your tour guide had some neat numbers but you should be careful not to repeat them verbatim (or unremembered).

Watt contains time already so watt per hour does not make sense. You might mean MWhr/hr which is the same as MW

Diablo Canyon can output 2.2 GW, if you assume 50% (1.1 GW) for the sustained output, I come up with 9636 GWh per year, or ~19,200 GWh per year if it was able to run at 100%

What does 2.2 MW/hour mean?

It doesn't. Watts were a mistake by whatever committee it was that standardized unit names. Power should not have been given a unit; it should have been left as ∆energy/time just as velocity is distance/time.

Joule is a derived unit, it is kg*m^2/s^2. There are lots of derived units, like hertz and newton, because they useful than writing out the whole thing. Electronics would be really annoying if had to write out volt, ohm, and watts (ampere is base unit, coulomb is derived).

Don’t put words in my mouth. I only said that power should be J/s instead of watts. The “per second” part of that is what is most important thing about power. It’s the rate at which energy is accumulating or being used up.

It's shorthand for a Joule (unit of energy) per second (unit of time). Watt is the problem with that?

The real problem is the widespread usage of Wh as a unit of energy

It would make way more sense to use J and J/h instead

I regard that as a downstream effect of giving power a unit in the first place, but yes. We should have just stuck to J and J/s. It would have prevented the kWh and also abominations like the mAh “capacity” ratings you see on batteries.

I gotta start describing distances as mph-hours

So joules per second per hour?

Using watts is fine for anyone who deals with energy and power all the time. The problem comes when the lay person tries to reason about power. If power were written as J/s then they could use the same reasoning that they are already familiar with from dealing with speed and position, or with flow rate and volume.

Reply on news.ycombinator.com