Hacker News Reader

Comment by EricE | Hacker News Reader

Probably because it ignores the fact that solar isn’t predictable (nighttime? clouds? dust?) and nuclear is.

Solar is NOT a replacement for nuclear. It absolutely can augment power generation, but it - especially solar cells - is not a reliable base load provider.

No amount of hand waving by solar cell proponents will change that. If we ever discover a way to reasonably store electrical power that’s far more reliable and cost effective than chemical batteries then that equation will change - but we aren’t there today. Nuclear is here now and there are more than enough designs that are inherently safe and cost effective that never get discussed because everyone emotionally freaks out as soon as the word “nuclear” is uttered :p

View on HN

Solar + hydrogen is just as predictable as nuclear. It costs roughly twice as much (because of 50% loss on electrolysis and back), but ends up providing a stable power output independent of weather.

It’s conparable in place to nuclear if I’m not mistaken, and it’s fault scenarios are less severe.

Especially in US, with vast amounts of deserts, it seems like a no-brainer.

(having said that, climate change is so urgent that I’d say do whatever it takes to get to zero emmissions - be it nuclear or solar)

In support of your point, I’d add that when doing cost comparisons with nuclear, one has to be wary of the accounting tricks they play such as not factoring in the costs of failure scenarios. Or even end of life decommissioning in some cases.

Except if you are pitching it as a solution for the base load problem - where is anyone doing solar+hydrogen at scale? Where is anyone committing to doing it at scale - let alone in the near future? How do we know it will remain economical at scale? All kinds of fun things can happen between theory/demonstration and implementation in the real world.

We do nuclear at scale today. The vast majority of blockers from doing more nuclear at scale today are political; not technical.

That’s the other facet of this that always seems to get overlooked.

Do you have an example of a specific scaling problem which applies to solar+hydrogen and probably can't easily be surmounted?

Or is your argument just "We're not doing it yet, therefore it can never be done"?

It’s not a bug, it’s a feature.

You don’t need one huge plant to manage the baseload when we’re talking about renewables.

You can just as well have the same wattage distributed across multiple places in the country, removing a single point of failure and lowering transmission costs (although requiring a redesign of the energy grid).

Solar coupled with modern batteries, obviously.

No hand waving required. It’s fine for other sources (even nuclear) to fill in the gaps. Also no need for accusations of emotional freak outs. I think the nuclear side has its fair share of deluded people who for some reason pretend that both solar and batteries are not improving year by year. And who pretend that solar advocates demand solar-only solutions. It’s just not the case.

Where are you going to get enough batteries?

If batteries were viable Elon wouldn’t just be talking about them.

Batteries are no where near the cost efficiency of, say, pumped hydro. Nor can batteries touch the overall capacity of pumped hydro. And pumped hydro can’t happen just anywhere - you need the right physical environment for it to be effective.

So yeah, unless you have something more concrete than “batteries, obviously” it is hand waving.

> Elon wouldn’t just be talking about them.

Just talking? He’s making them at scale.

Batteries aren't going to cut it. The last I checked batteries are an expensive source of power even if you can charge them for free. The cost of the battery / cycles it can deliver is already more than the cost of electricity in most places. We need a major breakthrough in storage before renewables are more than a way to reduce fossil fuel use in powerplants. (You still need the plants, just not as much fuel for them.)

There is also somewhat of a use case for renewables for powering things that can make do with highly unreliable power. Consider, for example, a desalination plant. The heart of the plant is pumping water through membranes--but do you really need to do that? Lets build our plant differently, build a storage system high enough up that gravity provides the pressure. When you have power you run the pumps to fill the storage system, when you don't have power you don't run the pumps. You need bigger pumps and you need a big storage tank but it can be done.

Or flip the scenario--don't pump the seawater in the first place. Place your desalination plant on the ocean floor and use your pumps to extract the fresh water. In this case you need fresh water storage underwater rather than salt water storage up high.

Let's look at results from a model.

https://model.energy/

In the 2030 cost model they use, batteries are 142 euro/kWh of capacity and 160 euro/kW of power. This (using 2011 weather data, along with various other assumptions, and also includes use of hydrogen) gives "synthetic baseload" in the US for 53.7 euro/MWh, cheaper than nuclear. The LCOS in batteries and (even more so) in hydrogen is higher than the cost of electricity, but that's ok.

(If you look at other countries, or specific states in the US, the results vary but remain broadly similar. The mix of solar vs. wind can change a lot though. Places like Poland are the worst for renewables in this model.)

The demand for battery materials could eventually exceed the pace that we can mine them, so the world's grids will need significantly more types of non-battery energy storage like pumped hydro and flywheels. This will ensure that the market for grid storage materials doesn't have to compete with the car market's growing demand for batteries.

There are a few new flywheel plants that have popped up in the US in recent years, and it'll be interesting to compare the material costs and benefits over the coming decades.

Reply on news.ycombinator.com