Hacker News new | ask | show | jobs
by ZeroGravitas 1398 days ago
If your peak is during the day, then you probably want solar, then you can use that to fill the pumped storage for the evenings.
3 comments

Even though it sounds like the original design was to store energy at night for use during the day, it looks like now that solar is such a large part of the grid they've just flipped the schedule around. From the CAISO website, hydro and battery go up around 4pm as solar output starts to go down and they peak around 8pm (fossil fuel generation follow the same trend, roughly). Although I don't know if there's the ability to check generation at a per-site level.
Nuclear paired with pumped hydro is very flexible, being able to handle peaks at any time. It's also very reliable, since the vast majority of outages are scheduled, and the plant's output can be increased or decreased to ensure there is enough surplus to store enough energy to meet expected peak demand. It also uses much less land and material resources, and results in less waste, substantially reducing damage to the environment.
Most of this is false, if you're trying to imply solar plus hydro is worse than nuclear and hydro. Taking them in order:

Demand is higher during the day. solar matches this better than the flat production of nuclear, so less storage per watt of production is needed for cheap solar energy when used to pump hydro storage, and even non pumped hydro can be varied to complement solar production and demand needs (within certain limits). And seasonally hydro and solar complement each other.

The land usage of pumped hydro isn't great, but since that's the same between both we'll ignore it for now.

Since the solar can be placed on top of the hydro dam water, or on the top of buildings, or used to shelter crops, the land use can and should be negative.

I very much doubt nuclear takes less material resources, but I don't have any good numbers for that off hand. Generally, cost is a reasonable proxy for energy and material input though (in the absence of large externalities) and solar costs less.

Again, hydro isn't super good for the environment, substiantially worse than nuclear and solar, but if you're going to build it to help nuclear deal with it's unhelpful power supply timing then you might as well use it for cheap solar.

> I very much doubt nuclear takes less material resources.

In the case of an existing nuclear power plant it does. Because if you shut it down, the plant is still going to be there, except that it won't produce anything.

California has been building solar like crazy for the last 5 years if not more. Nobody is saying stop building solar. Nobody is even saying build nuclear. The only thing this proposal is saying is don't shut down a very large existing nuclear power plant, at least not until we get to carbon neutral.

If you are suggesting that somehow by shutting down this power plant you can get faster to net zero, then we're all very eager to hear the details.

A lot of people say that, actually.
My points were that nuclear is flexible and reliable. Your point seems to be that solar's output is closer to average demand and has a lower price. These are totally separate design goals. My view is that flexibility and reliability are core design goals because they reduce resource use by decreasing the need for storage and long-range transmission, and because they reduce the risk of energy shortages, which have serious social consequences. Why do you think matching average demand and reducing monetary price are core design goals?

Your claims on land use and material use are empirically false.

Unfortunately, peak electricity demand is in the evening right as the sun sets.
This isn't true. The peak demand that remains after solar has lowered the actual peak, is in the evening.

And that's when you use hydro and batteries.

Here's the historical yearly peaks: https://www.caiso.com/documents/californiaisopeakloadhistory...

Note, in recent years this ignores home based solar, which also reduce the grid seen demand during the day, but the new peak is still before sunset.

The local grid refers to this post solar/wind demand as 'net demand peak':

> Net demand is the total electricity demand minus utility-scale solar and wind generation at a given time, and the net demand peak (the “net peak” for short) typically occurs later in the evening than the total demand peak.

https://www.energy.ca.gov/data-reports/energy-insights/peek-...

> Initially, net demand peaked around the same time as total demand. Wind, which tends to pick up in the evenings, was responsible for most of the renewable generation. In 2013, solar began to eclipse wind. By 2016, the average timing of the net demand peak shifted from before 5:00 p.m. to around 7:30 p.m., where it has remained. (The total demand peak has also moved later in the evening to a lesser extent, driven by customer-owned solar.) Grid operators can’t turn to solar after the sun sets to meet the resulting net peak.

Luckily California has big plans for offshore wind, but that'll take about a decade. In the meantime solar and demand management is probably still the low hanging fruit.

Energy use peaks at around 6pm, which is well after solar's peak production: https://www.eia.gov/todayinenergy/detail.php?id=42915

Your own sources show this, with energy peaking around 1700 or 1730 (which is 5-5:30 pm). "Demand management" is a euphemism for blackouts. You can mail notices and flyers, but energy users don't change their consumption patterns.