Hacker News Reader

Comment by _Microft | Hacker News Reader

There has been ~ 180GW of NEW solar power installed in 2021. It's got a capacity factor of 10-25% but even that newly installed usable power is already a sizable portion of ALL nuclear power in current use (~400 GW with ~50GW of additional nuclear power capacity being built (how long does that take? 5-10 years per power plant? That'd make an extra ~ 5-10GW/yr then)). Photovoltaics are also still on the exponential part of their trajectory installation-wise and also on an exponentially decaying trajectory cost-wise.

I'd not bet against solar power with all these things considered. If anything, I'd start building additional HVDC power lines to geographically smooth out local variation of available power.

View on HN

Solar output is ridiculously low in Europe winter, Germany take exactly the path you describe, fully renewable. Except it don't provide a base load, so they must use gas.

You list installed capacity, solar available capacity rarely ever reach installed capacity in Europe.

> You list installed capacity, solar available capacity rarely ever reach installed capacity

That ratio is called "capacity factor" which I actually mentioned to be around 10-25%. There's a Wikipedia article on it if you want to know more.

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

Yes, 12% at winter noon, it's 0% at night, average in winter is way below your numbers: https://app.electricitymaps.com/zone/DE

CF for Solar in Germany is 11.6%.

Europe could use hydrogen for seasonal load leveling (as well as wind). The capacity for hydrogen storage in salt formations in Europe is many millions of GWh.

This would still be cheaper than nuclear.

> This would still be cheaper than nuclear.

Source needed. Germany electricity price is one of the highest price in europe while having invested heavily in renewables.

https://ec.europa.eu/eurostat/cache/infographs/energy_prices...

> Europe could use hydrogen for seasonal load leveling

You propose to use an unproven tech, never made at scale, instead of an existing, working one ?

Germany already have industrial scale Power2Gaz, which is a way to store energy, but they are still relying heavily on coal, and have one of the highest electricity price.

All the elements of hydrogen for grid leveling are demonstrated: electrolysers (now available at $300/kW in China), underground storage, combustion in turbines. It's really just a matter of integrating existing components. This is the easiest and surest kind of innovation. All that's left to be done is the regulatory background to push fossil fuels out of the way (sufficient CO2 taxes) and perhaps running some of these parts further down experience curves. The biggest uncertainty is whether there's an even cheaper way to deal with the problem of leveling renewables, which wouldn't help nuclear in the slightest.

Germany's residental rates are irrelevant; what matters is the cost at the wholesale level of providing the input energy to the grid. Germany soaks its citizens to subsidize power for its industry.

> All the elements of hydrogen for grid leveling are demonstrated:

It's not, the ultimate demonstration being it deployed at scale.

> sufficient CO2 taxes

Which will reduce solar competitiveness to nuclear, since solar have 4x the CO2 per kwh of nuclear electricity. It also doesn't solve the big problem solar have in Europe, since it produce less than 10% of it's installed capacity right now (last 30 days average for Germany), it's not even winter yet. &

> Germany's residental rates are irrelevant

Except being directly tied to energy production price.

> Germany soaks its citizens to subsidize power for its industry.

Which should make household price cheaper, not costlier.

The "at scale" argument is flaccid. It's feckless unless you can give a reason why the technology would not scale. For hydrogen, what would be that reason? We don't need to build bigger turbines or electrolysers; we just need to build more of them. That sort of thing scales wonderfully.

About the only thing that might not scale is storage capacity underground, but there are plenty of salt formations and spent gas fields to store hydrogen in. We already store a seasons' worth of natural gas underground in such formations.

If you want to see a GOOD example of "cannot scale", look at today's nuclear reactors. LWRs cannot scale to power the world because uranium runs out. This was the motivation for building breeder reactors. Unlike hydrogen burning turbines, hydrogen storage in underground caverns, and electrolysers, breeder reactors are still a technology under development, with serious issues (safety, economics, reliability, proliferation.)

> breeder reactors are still a technology under development

It's been 40 years we know how to build breeder reactors. And there are already been multiple commercial fast breeder reactor.

But like regular nuclear reactors, yes it is a technology under development. But we also have more than 100 years to figure out the rest.

> The "at scale" argument is flaccid. It's feckless unless you can give a reason why the technology would not scale.

Cost, and cost.

Power2Gaz already exists, you never showed how it was inferior to your tech.

But in both case, there is an enormous problem you never mentioned: the efficiency of power2gas is currently of 60% at most, hydrogen included.

Without including the equipement cost needed, stored solar electricity is suddendly costlier than nuclear.

There have been no successful commercial fast breeder reactors (that is, competitive with existing burner reactors, especially when the cost of reprocessing is included.) France has mothballed their entire fast reactor effort, not even building another prototype.

Fast breeders have grave safety issues inherent in their physics. The delayed neutron fraction of plutonium (or U-233) is about half that of U-235. This means any breeder is skating closer to prompt supercriticality. If a fast reactor goes prompt supercritical, it will explode more violently than Chernobyl, since the time for the neutron population to double (for a given value of k) will be orders of magnitude shorter. Worst case, it would be an actual nuclear explosion.

Fast breeders also present proliferation concerns. To get a good breeding ratio, a blanket of fertile material is placed around the fissioning core, so the neutrons escaping the core are efficiently utilized. If this is U-238, the Pu produced in the blanket will be "superweapon grade", with very low concentrations of higher Pu isotopes. Thorium breeders also have proliferation concerns, from the need to separate Pa-231 and allow it to decay before. This produces essentially pure U-233 outside the core.

> cost and cost

Let's look at a simple model to see how using hydrogen affects cost.

https://model.energy/

Try Germany using 2011 or 2012 weather data and 2030 cost assumptions (note the cost of electrolysers in China is already half the number assumed there.) The overall cost of providing a steady 100 MW output is not bad, lower than what one would pay from a newly constructed nuclear plant including the cost of construction and financing. Hydrogen helps reduce the cost here, not increase it. Even very slight amounts of hydrogen noticeably help, covering for rare prolonged dark/calm periods.

> France has mothballed their entire fast reactor effort, not even building another prototype.

Yes due to public pressure/antinuclear lobby.

> Fast breeders have grave safety issues inherent in their physics.

Yet there has been a dozen of fast breeders reactors, and somes are still operating today.

I trust way more the IPCC forecast on the subjects, than yours. The fully renewable path was rated "very uncertain, very risky" for the exact point I told you.

The fully nuclear path for France was also uncertain and risky, but for another reason, they are uncertain we can build nuclear plant fast enough, which is understandable because they recommand a 2x increase in electricity production.

France abandoned their fast breeder effort because of serious technical and economic problems. There was no good prospect that anyone would want one, including France.

BTW, this shows that for all the talk, France doesn't see nuclear as the global solution to CO2 emissions. Nor does Japan (whose fast breeder program was even more a fiasco) or the US (where we abandoned fast breeders even earlier, after massive cost overruns.) Fast reactors are like ordinary nuclear, only with even more ludicrousness.

As for safety, the person to listen to is Edward Teller. He famously and publicly (in 1967) brought up the safety issue I mentioned. Because of that safety issue, I seriously doubt if any heterogenous fast reactor (with conventional fuel elements) would be licensed in the West today. That leaves molten salt fast reactors, but no one has ever built on of those, and they'd have serious problems with neutron damage to reactor structural materials.

> Which will reduce solar competitiveness to nuclear, since solar have 4x the CO2 per kwh of nuclear electricity

This is incredibly disingenuous. Current gen PV produces less CO2 than nuclear, and even a $1/kg a CO2 tax would only add a fraction of a cent per kWh to either.

IPCC 2018 report list solar PV with 48g eqCO2/kwh (median):

https://www.ipcc.ch/site/assets/uploads/2018/02/ipcc_wg3_ar5...

Nuclear is 12g eqCO2/kwh (median)

The levelized cost of electricity is also listed as cheaper for nuclear.

A cost based on high yield mines that are almost empty and 8 year old PV tech that hasn't been used in any new project since 2017 is worthless.

As is comparing electricity that costs $100/MWh after publicly guaranteed loans and after the public takes on any accident liability in excess of the cost of a quarter of one reactor to unsubsidized electricity actually sold for $15/MWh

Oh look. The more up to date version of that report:

https://www.ipcc.ch/report/ar6/wg3/downloads/report/IPCC_AR6...

Which doesn't include mining from ores that are 1/10th concentration that would be required to expand nuclear enough to solve the problem. But it does show you intentionally picking out of date information to lie about LCOE.

When proponents of the nuclear consistently lie about everything, it makes it very hard to believe the industry hasn't done things like forge the paperwork on essential safety parts.

Luckily people like you are most likely just lying to prolong the transition from fossil fuels, and aren't part of the industry.

You need a backup a little your claims if you deny IPCC numbers, IPCC is considered very serious, and I said to another commenter, I have way more trust in IPCC numbers and planning than random internet commenters.

You are saying "with CO2 taxes it will be profitable"

The thing is, nuclear don't need CO2 taxes to be profitable, and is low carbon, if your solution was that cheap and safe, it would be more invested than nuclear is.

Then use the 2022 IPCC numbers I linked rather than the 2018 ones.

> You are saying "with CO2 taxes it will be profitable"

Nowhere did I say this. There are entirely private PV installations that sell their energy for under $15/MWh, less than a tenth of recent nuclear plants like Vogtle even including massive subsidy

Nuclear absolutely does need CO2 taxes to be profitable in the US vs. fossil fuels. Quite large ones, considerably larger than the "cost of carbon" in the IRA, otherwise fracked gas and combined cycle power plants blow it out of the water.

A CC power plant costs maybe $1/W; a nuclear plant costs 10x that. With cheap gas and no CO2 charges nuclear doesn't have a prayer of a chance to compete with that.

This is why the so-called "nuclear renaissance" in the US collapsed even before renewables got cheap.

The claim that solar produces more CO2 than nuclear is quite dishonest.

It may produce more right now, with the existing supply chain for materials. But in a fossil-fuel free economy, there is no CO2 from fossil fuel combustion. Where are the CO2 emissions coming from? They could only be from manufacture of cement for concrete (calcination of limestone to calcium oxide and CO2.) PV doesn't have to use any concrete -- it can be mounted to the ground with steel earth anchors. Nuclear uses a lot of concrete.

The near term emissions of CO2 in construction of solar or nuclear are irrelevant, since they are swamped by the CO2 emission avoided as fossil fuels are displaced. Here, what matters is how rapidly the non-fossil alternatives can be rolled out, not what dribble of CO2 they produce directly.

As for levelized cost, the table lists the min/median/max cost of installed utility scale PV as 1700/3200/4300 $/kW. But the cost of utility scale PV in the US is already well below that minimum number. The footnotes give their sources as no later than 2013. PV prices have fallen greatly since that time (module prices per W down by about a factor of 4 in constant dollars.)

The min value for nuclear given is also absurd vs. actual experience with recent nuclear builds. The footnote says the data was from no later than 2013, before the EPR and APR1400 efforts went so disastrously wrong, when the "nuclear renaissance" had not fallen so flatly on its face. It also says (about nuclear):

"Limited recent data and/or original data are available in the published literature. More recent, (grey literature) sources provide investment cost and LCOE estimates that are considerably higher than the ones shown here (Brandão et al., 2012)."

> It may produce more right now, with the existing supply chain for materials.

This hasn't been true for some time. There's simply not enough money. The solar module in a good location costs 0.7 cents per kWh installed. Although that can buy you a couple hundred grams of coal, doing literally anything with it uses up all the money.

A 450W monocrystalline panel is under a kg of silicon, 50g of copper an a gram or so of silver.

The neat thing about solar panel factories is they have a bunch of solar panels, and so tend to produce 30-90% of their own electricity, so the carbon footprint is almost entirely racking, shipping 20kg, and the 11kg or so per panel from the raw silicon (which uses 70% of the money).

Meanwhile Uranium ore concentration is decreasing slightly faster than mining is getting more efficient.

Good to know. I think the CO2 emission will be mostly from making the glass and aluminum, not the cells themselves.

BTW, last I checked, reduction of silica to silicon is done with charcoal as the carbon source. Something about the porous nature of charcoal leading to better reaction with silicon monoxide vapor.

Reply on news.ycombinator.com