Comment by XorNot | Hacker News Reader

Comment by XorNot | Hacker News Reader

Parent Full thread

Upvoting this because this is where I end up. Nobody proposing solar, wind and storage ever wants to back it up with hard numbers. They back it up with soft numbers - lies of omission.

"Look how expensive nuclear is compared to solar!" they say, ignoring capacity factors, reliability or storage. "Look how many gigawatts of batteries were installed last year" while ignoring that the energy-storage of those batteries is maybe 3 hours and the total power of batteries scales alongside energy.

"We'll just overbuild" they say, ignoring that the 4x the solar is now within a stone's throw of being the same capex as equivalent nuclear, but still needs storage.

Everyone has some just-so narrative which omits numbers. Because you can make green hydrogen they assume it must be cost-effective to do so. Because someone said "solar is cheap" they assume it must be too cheap to meter. As though you can wave away multiple 50% energy loss steps as insignificant.

I'm open to being convinced I'm wrong, but show me the numbers which prove it. Show me that you're talking about a today technology, not "by 2028 it's expected costs will reduce..." while saying nuclear will "take too long to build". Don't tell me power-to-methane exists, show me the numbers saying it works and include the costs of storing it (stored methane is superb on energy density for the grid, but about 30% efficient panels-to-power plant and has a cost per kWh in storage...and has a suspicious paucity of information surrounding operating costs of plants versus "future projections").

Focusing on batteries is thinking too long-term IMO. If you have 50% renewables or less in the grid, you basically don't need batteries yet, and our deadline isn't time here but greenhouse gases emitted - getting to 50% renewables will give us more time to tackle the rest of our emissions. I expect nuclear to easily beat solar/wind+batteries for the last 10% of the grid, but frankly the last 10% of the grid almost doesn't matter - if we've reduce our emissions by 90%, then we have 10x the time to tackle the rest.

Of course, I've been wildly conflating electricity with all emissions, but I think the low-hanging fruit argument still applies.

But you can't run the grid on 50% renewables without solving storage. It's just not possible: the grid has to run. Peak usage isn't in the middle of the day, it's around 6-8pm when people are home. In winter that's pitch black, in summer depending where you are there's still very little solar output, so you better hope the wind is blowing. And it is...mostly, but not on demand.

Germany's power grid has started to struggle because they're at 30+% solar capacity, but not reliability - the grid can't sustain losing large swathes of generation capacity suddenly because a cloud passed overhead, or if a region is overcast that day then you've got to make up that shortfall - which you can do so long as you have basically the same installed capacity of gas, coal and oil on standby where they can just turn on an extra-turbine or boiler on demand.

But if you want to actually decarbonize, then you need to solve that problem. Otherwise you end up with the perverse situation that increased renewable investment necessitates new fossil fuel power plants, because you still need to backstop the intermittent sources.

The NSW Australian power grid has this problem right now: the reality is that no one wants to invest in fossil fuel power plants. Great! You'd say. But the grid actually does need a new reliable generator and power price pressure is a big political issue. Getting reliable watt-hours onto the grid - not more intermittents - needs to happen. The catch-22 is people rightly don't want to subsidize more fossil fuels, but the available renewable options are being installed anyway - but they won't solve the actual availability problem we have (and in a country with bush fires it's a big one - 3 months of summer with overcast skies along the east coast from smoke means no one's solar panels were working well, and grid-interconnects have limited capacity).

> But you can't run the grid on 50% renewables without solving storage.

Counter-example: Brazil's national grid routinely runs on over 85% renewables with essentially zero storage (there might be a couple of experimental battery storage units, but if any they're much less than 1%; and all it has of pumped hydro storage is one partial preliminary study of possible locations).

65% of Brazil's energy is coming from hydropower, which is extremely reliable, but geography dependent. When people make that statement, they're usually referring to wind + solar, which are intermittent.

This right here is an example of a lie of omission: over what time frame? With what renewable?

Oh right - hydroelectricity. Providing 65% of it's total electricity.[1] Wind is 11% and solar is 2.5%.

And they're also building a new nuclear reactor at their existing plant.[2]

[1] https://en.m.wikipedia.org/wiki/Renewable_energy_in_Brazil

[2] https://en.m.wikipedia.org/wiki/Angra_Nuclear_Power_Plant

> This right here is an example of a lie of omission: over what time frame?

Sorry about that. It's the average over a whole week (I looked at the weekly bulletin of a random week from last month, but they're all similar).

> With what renewable?

Around 62% hydro, 22% wind, and 1% solar (we're in winter right now, so we have less solar and more wind; on summer we'd have more solar and less wind).

> And they're also building a new nuclear reactor at their existing plant.

And that reactor is the main reason I'm very skeptical of nuclear power: it's been under construction for nearly 40 years, and it's still not near done. All the parts have already been bought and have been stored on site for a long time; just the upkeep to make sure none of these parts degrades while under storage is already a high cost.

> Nobody proposing solar, wind and storage ever wants to back it up with hard numbers.

For hard numbers, see real-world case studies like Denmark:

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

They get 50% of their electricity from just wind. So that's the lower bound on what you can achieve in the real world today. Here's how you increase that without needing any storage technology:

(1) Be a very large land mass like the United States, where wind variability is more likely to cancel out.

(2) Add solar instead of only relying on wind. Solar cancels out wind variability because it's more windy when it isn't sunny.

(3) Overbuild by 2-3x, so that the troughs in energy aren't so bad. Even if you overbuild by this much, it's still a lot cheaper and faster to build out than nuclear.

(4) Import solar/wind from neighboring countries, losing only 10-15% due to resistivity.

I don't know exactly what % renewables you can achieve by picking all this low hanging fruit while still being cheaper than Nuclear. 80%? 99%? If you add solar and you overbuild and you have access to a large landmass like the US, you can probably get very, very high, with the last few % being powered by natural gas peaker plants as a temporary stop-gap until storage technology becomes good enough.

As a bonus, if you overbuild, you can sell surplus energy to Mexico and make money!

Denmark's electricity is almost three times more expensive than in France. Solar+wind does not cancel out variability. Backup is needed. LCOE does not mean lower prices. Higher shares of renewables drives up costs for the network to rebalance. Producing a lot of energy when it is not needed means producing low value energy. It really depends on the electricity mix.

I disagree with this 100%, the only people using numbers to mode future carbon-free grids are ending up on massive amounts of solar, wind, and storage, with a tiny bit of nuclear. But that nuclear is so difficult to build that the models that end up adding nuclear also have unrealistic costs and deployment timelines associated with it.

We have ~100 reactors in the US, nearly all are close to end of life. We would need to start building ~100 reactors today in order to keep the same amount of nuclear in the future. We have one, just one, nearing completion, and no more that have any chance of getting funded and starting construction within a few years. We are going to go through a long period with far less nuclear than we have today, if we ever get to a period with more nuclear than today. I doubt we ever will.

If you want numbers, check out Christopher Clack's modeling. It is by far the most detailed, and the most complete, and the most innovative.

But to say that nuclear advocates run the numbers is just false. They can't even run the numbers on their own construction plans, much less the numbers on the system level of what it would take to keep nuclear a significant part of our grid. Nuclear advocates are the least numerate folks in the energy business, again and again, since the big burst of reactors in the 70s.

Reply on news.ycombinator.com