Hacker News Reader

Some things that are easy in theory are not as easy in practice.

In theory it's so easy to build lots of huge solar power plants. We all know that solar is dirt cheap, because we've all seen the Lazard report with the levelized costs of energy.

In practice, here's the list of solar power stations in California [1]. None is close to 1 GW. Only 3 are above half a gigawatt nameplate capacity. But their capacity factor is on average less than 30%.

Or to put it differently, 20% of the electricity generated in California comes from solar, and 9% from nuclear (i.e. from Diablo). If you want to replace nuclear with solar, you need to increase the current solar generation by 50%. That's doable, but it's not something you do overnight, it will very likely take a few years. Meanwhile, guess what you'd use to replace Diablo? Of course, it would be natural gas, which, by the way, accounts for 47% of the current generation.

We've heard the story of nuclear being replaced with renewable in my state, New York. Two reactors, with a combined capacity of 2.1 GW were shut down in 2020 and 2021. But instead of renewables, they were replaced with natural gas [2].

  > As a result of the permanent shutdown of the plant, three new natural-gas fired power plants: Bayonne Energy Center, CPV Valley Energy Center, and Cricket Valley Energy Center were built, with a total capacity of 1.8 GW, replacing 90% of the 2.0 GW of carbon-free electricity previously generated by the plant.[6] As a consequence, New York is expected to struggle to meet its climate goals

[1] https://en.wikipedia.org/wiki/List_of_power_stations_in_Cali...

[2] https://en.wikipedia.org/wiki/Indian_Point_Energy_Center

The solar doesn't need to be in one large install, and that would actually be a very bad idea. Big central points of failure are less reliable than widely spread resources. So I'm not sure what the point is of saying the biggest single site is X MW.

And if the idea is that "oh no we are going to see an emissions spike when DC shuts down" that's going to happen anytime before 2035, while there's still gas on the generation network. If it happens now or in five years is less important than the total emissions over time. If we bit that bullet in five years, but didn't spend the $6B on more permanent solutions that bring us to our long term transition, it's money wasted.

Sure, but I think the GP's point is that this doesn't just happen overnight. Let's say the same people who planned out keeping this nuclear plant open another 5 years today said "ok, let's plan out a same-price solar build and get it started".

How long would you expect the planning process to go, buying up the land, permitting, the construction process, connecting it to the grid, testing, and then finally energizing it? A couple years? More? (Also remember that this is California, where people will put up stupid roadblocks even for construction out in the desert, away from anyone's home.)

Now, they absolutely should be doing that in parallel! It's dumb that they aren't. But this nuclear plant is running today, and will continue to provide much-needed electricity, and that'd still be needed if we got this (unfortunately only hypothetical) solar project going right now.

> Sure, but I think the GP's point is that this doesn't just happen overnight.

The article says Diablo Canyon generates 2,250 megawatts.

https://en.wikipedia.org/wiki/Solar_power_in_California says "The Solar Energy Industries Association predicts that California will increase its solar capacity by over 27,000 MW over the next five years".

According to the chart titled "California Solar PV Capacity by Year", they added two Diablo Canyons worth (4,779 MW) of PV capacity last year alone.

Keep in mind you need to multiply those power numbers by capacity factor. Solar in California is at ~29% capacity factor (much higher than I expected honestly), while nuclear is generally around 92%.

That is, Diablo Canyon actually puts out 2,070MW of power from it's nameplate 2,250MW. While 4,779MW of PV in California will generate 1,385MW of power.

So, they really only added half of a Diablo Canyons worth of capacity last year. (Still an impressive amount of power, but not as much as it seems).

It seems that people do not understand that capacity factor is the #1 criteria when building an electricity grid.

Electricity grids do not have the luxury of being able to shut down, they must meet the demand every day of the year, every minute of the year

Technologies with an higher capacity factor are inherently more valuable than the ones who don't and technologies with an unpredictable capacity factor are worth even less than those two.

There are lots of gas generators with super low capacity factors, but the capacity factor is not the key part of the generator, it's dispatachability, the ability to generate when asked to by the grid.

Neither nuclear nor coal are very dispatchable, as they can not respond to changes in demand in cost effective ways. Gas turbines can easily, as can batteries. While the sun is shining, solar can be dispatchable, because it's easy to shut off on demand or turn on demand. But for 60%-75% of the day, solar is not dispatchable.

In the modern grid, dispatchability is a far more useful attribute than capacity factor. Capacity factor is just an incidental.

In today's world ran by fossil fuels, there's not a need to turn off the non fossil fuel production, there's always more parts of the economy willing to get cheap electricity.

What matters then is the reliability and this is why the capacity factor and especially its variation are critical. See what happened in the EU when the russian gas turned off.

Batteries are a low scale solution at best at the moment anyways and aren't working at the scale of a grid.

More importantly, Nuclear generates the base load capacity that renewables struggle with - the 27,000 MW is great, but its use it or lose it, as we dont have anywhere near the capacity in place to store the excess - nor will I think we will build it in anything resembling a timely manner.

If that's the reasoning, it's just a variant of the sunk cost fallacy.

The question is what is the most effective use of our limited resources right now. Diablo Canyon will be shut off at some point, causing a momentary blip in emissions rising. Is $6B on a short extension a good use of our resources? It's certainly difficult to see much benefit.

Instead, the benefit is mostly to assert that we still are open to using nuclear. Which, fine, is a good goal, but instead what if we just spent $3B each on an SMR, with the expectation they would probably fail? I think that would have a bigger positive impact towards more nuclear in the future.

Check the markets. Everyone’s losing money on solar. If there’s a sunk cost at play, it’s probably there.

You must be looking at very different market data versus the rest of the economy and players.

Solar projects take about 6 months. Solar projects are one of the only types of major infrastructure project that regularly come in ahead of schedule and under budget. (I couldn't quickly find the cite for this, but if I find it again I'll edit it in).

If you show me a 1GW solar plant that took 6 months to build, I’ll eat my hard hat.

California added ~5GW in solar last year.

Doesn't have to be all in one project.

And how many years ago did those start planning?

You're just moving goalposts. Solar is by far the cheapest and fastest to put up and plan for.

No they asked for 1GW of power in 6 months. You responded with the amount of capacity added in 1 year implying that amortized over a 6 month period. They asked how long it took from the start and you accused of moving the goal posts but it just looks like you’re talking about different things - throughout vs latency.

Im moving them back. The original statement was that a solar plants go in in 6 months, which is a total disconnect from reality.

California already has a lot of solar installed in the desert. They could just expand that. Nuclear is a base load though, so they would need batteries or pumped storage to add to that.