[aksss]

Right, I think hydro is about as close to at-scale 24/7 always-on generation as you can get with renewables right now. Even with that, everywhere I’ve lived with hydro power had backup fossil fuel generation, which was periodically used, though rarely.

Arguments for all other forms of renewables at scale right now seem to depend on having/maintaining fossil fuel power generation just to satisfy daily demand reliably, so you end up with two sets of power generation infrastructure designed to meet peak capacity (so effectively double capacity of what’s needed), which is extraordinarily expensive.

Advocacy for non-hydro renewables requires heaping doses of “hope” that new technologies will be developed one day to deprecate the side-by-side ff power-generation.

That hope and risk has to be priced in to the comparisons with nuclear for supplying electric grids, or the debate isn’t “nuclear vs renewables” it’s “nuclear vs. (renewables + fossil fuel generation)”. At that point it even makes sense to think of “nuclear vs. (renewables + nuclear)”.

If you want the cost of CO2 and its global impact priced into using fossil fuels, perhaps that’s what the cost of nuclear should be compared to.

[bryanlarsen]

If the world was completely solar, we'd need 30TWh of batteries, which would cost about $3T. [0]

$3T would buy 200 GW of nuclear power plants, about 3% of what's needed.

0: https://caseyhandmer.wordpress.com/2019/06/21/is-nuclear-pow...

[akvadrako]

Current all-in battery costs are about $200/kWh. Daily global energy use is about 400 TWh. That would be an 80 trillion dollar battery for just one day of backup.

That just isn't feasible to build near term and it isn't even enough to prevent blackouts. Plus the current methods probably don't scale that far, so the cost is even higher.

[bryanlarsen]

We need a lot less than a day of backup.

To deal with the capriciousness of renewable energy, in order of preference:

- time of use pricing - use a mix of renewable energy sources - demand reduction incentives - interconnect regions - overprovision - storage

If the first 5 things are done well, not much storage is needed.

[bdauvergne]

Previous generation of nuclear reactors in France were built for less than 2 billion euros (constant euro) per GW (sorry it's in french, you can see the numbers here http://i-tese.cea.fr/_files/LettreItese18/ECLAIRAGES/REP.pdf on page 10). There were 50 reactors of three types (900, 1300 then 1450 GW). If we had kept this technology instead of losing the know-how and developing a new program, we could certainly build 1500 GW for those 3T. Don't know if any other country ever achieved such economy scale on its civil nuclear program, maybe South-Korea as like France they have only one nuclear company, and even more reactors per sites of few different models.

[Ericson2314]

Finally and (slightly) anti-nuclear piece worth reading. Yes, solar is easy for capitalism, but I'm not sure building all those batteries is. Building smaller pre-fab nukes vs building larger battery arrays seem pretty close in terms of challenges.

You don't have a citation on "$3T would buy 200 GW of nuclear power plants", and I very much doubt that extrapolation, because if any sane planner got the $3T budget, the first thing they would do is invest heavily in pre-fab.

[bryanlarsen]

It's all relative. Yes, a $3T order for nuclear would reduce costs. But a $3T order for batteries would also reduce their cost. Which would reduce more? I posit the batteries.

[mattashii]

> Right, I think hydro is about as close to at-scale 24/7 always-on generation as you can get with renewables right now.

There are quite some solar-thermal installations of >100MW, which store solar energy in a thermal buffer to later convert this to electrical power, creating an effective 24/7 stable supply of power, weather permitting.

[Gibbon1]

Couple of years ago I tried comparing the cost of a solar thermal plant with PV solar. Claimed advantage of solar thermal is the ability to store energy as molten salt. The cost differential though means it'd be cheaper to use PV Solar to heat phase change materials.

[aksss]

https://spectrum.ieee.org/energywise/green-tech/solar/china-...

[pedrocr]

> Even with that, everywhere I’ve lived with hydro power had backup fossil fuel generation, which was periodically used, though rarely.

> Arguments for all other forms of renewables at scale right now seem to depend on having/maintaining fossil fuel power generation just to satisfy daily demand reliably, so you end up with two sets of power generation infrastructure designed to meet peak capacity (so effectively double capacity of what’s needed), which is extraordinarily expensive.

I'm by no means an expert on this but I've been trying to learn more about it recently. From what I gather the end-state can be 100% renewables in different ways:

1) Have enough hydro that can be turned into pumped storage that a complete grid mix can be done with just hydro+solar+wind with very little overbuild. Portugal has very good conditions for this and back of the envelope calculations tell me it's possible with just 15% overbuild. All of the extra is solar which is very cheap[1]. With a better interconnected European grid it may be possible to do that across the whole continent.

2) Overbuild solar and wind by a large amount since they're so cheap, supplement that with some expensive batteries, and allow energy prices to go to zero and even negative at times to see if anyone has a use for the excess energy. We're talking something like 3-5x overbuild and then having so much excess energy that you start disrupting other fossil fuel usage[2].

If you have a seasonal storage breakthrough you're back at 1) with just another technology in place of hydro. Financing hydrogen generation seems to me like picking winners within 2). Both scenarios can be supplemented if shaping demand can be made at scale. Things like heating buildings and charging EVs can be shifted a few hours during the day without much inconvenience and possibly allow shaving off some important peaks.

Most of this discussion would be avoided if we just put a steadily increasing price on carbon, remove all other subsidies, and let the market shake things out.

[1] https://docs.google.com/spreadsheets/d/1UTUjhrBF04MP38b4WlQx...

[2] https://static1.squarespace.com/static/585c3439be65942f022bb...

[aksss]

If the issue is cutting CO2 with a power generating solution that’s highly, highly reliable, renewables like solar+wind have an uphill battle to fight. Their load factor is abysmally low, and storage tech at scale just isn’t here to sustain daily demand for 72 hrs. Supplementing with hydro instead of gas/coal to pick up the slack probably implies some massive “three-gorges” style damming projects. How feasible that is varies by region. What do you think about the feasibility of building that much hydro capacity in a way that can sustain Europe’s daily need for three days running? If you had that, why even bother with solar+wind except maybe as localized off-grid solutions? Somewhere out there is also a question of hardening generating capacity against kinetic attack, but that’s a whole separate thing. I only think of it because dams seem pretty vulnerable, especially massive ones.