antimongo

Well the grid needs the most help late afternoon. Which is right when solar starts to ramp down and when people get home and load starts to ramp up.

During solar hours, prices sometimes even turn negative. Literally paying people to take your energy, since solar is so plentiful.

The issue is those late afternoon, early evening hours.

And it’s actually more difficult on power plants. Solar is great when the sun is out, but when it goes away, you need all your power plants running. Issue is, a lot of power plants don’t like to turn on and off. They’d prefer to just run at one speed, all the time. But when the sun is out, we have to turn off power plants, since we’d make too much power. And turning them back on can be a long and expensive process.

And that’s where some of this rhetoric comes from. From a power plant perspective, we go from no-load in the afternoon (all solar), to full load in the late-afternoon/early-evening (no solar). The grid was never designed for this, and it’s having a hard time adapting so rapidly.

Batteries are totally a solution but the technology is super green and not really at a grid scale yet.

I’ve been thinking of replacing my Synology with one of these!

To be fair 10 hours is either a pretty old or pretty massive unit. 2 hours might be a little more reflective of modern gas turbines. Especially combined cycles. But depending on how big the peak is, you need every available unit, both old and new.

Ultimately the issue is it’s very hard to meet that peak when all of your gas units have to go from 0 to 100% output. Much easier (and more reliable) to take them from 10% to 100%. Which is what grid operators do currently.

Yea an affordable battery in every home would be a slam dunk. This is kinda already happening with vehicle2grid (v2g) electric car protocols. But not everyone has an EV yet. And operators are still working out the kinks using this in the grid.

Plus the lithium batteries in cars have their own supply/recycling issues.

Gravity energy storage doesn’t scale well. I’ve replied to other comments with more detail on this.

There are more feasible energy storage technologies out there, but these are super cutting edge and are not ready for grid-level deployment yet.

The future of grid level energy storage is almost certainly not going to be gravity based. At least not on a large scale.

You can’t have 100% of load be renewable/solar and have gas units online on top of that. That’s over generation. You have to match the supply exactly with the demand. If you mismatch, you destabilize the grid. Undersupply causes blackouts, oversupply melts power lines.

If a unit takes 10 hours to start, solar hours are from 6am to 6pm, and peak load is at 7pm with 0% solar; when do you recommend we start this unit? At the minimum, we’d have to order it on at 7am. Units have to run at a minimum load, let’s say 100MW for this unit. So now you can’t 100% solar from 7am to 6pm, you have to leave 100MW of room for this base loaded unit.

This doesn’t even factor in regulatory requirements like flex, spinning reserve, and other balancing and reliability requirements. Grids are required to have emergency units available at an instant to prevent mass destabilization if parts of the grid fail.

Yea, more control over the panels will help with the overgeneration issue.

But there’s other issues like ramping supply to meet peak demand and general generation during non-solar hours that still have to be addressed.

Each have interesting proposals on how to solve them, but they haven’t been developed to the point that they’re ready to be put onto the grid at a large scale.

Yes, pumped storage is definitely an existing technology that serves this need. I live near a massive one as well. However, large-hydro recently has not been considered as renewable form of generation due to the disruptive impact it has to local ecosystems.

I know in the US, new projects do not get approved due to permitting and water board issues. So I don’t think we’re going to see any new construction.

Piggybacking on your grid stability point, another issue I don’t see getting addressed here is ramp rate.

If we install enough solar where 100% of our daytime load is served by solar, that’s great. But what about when the solar starts to drop off later in the day?

A/Cs are still running while the sun is setting, the outside air is still hot. People are also getting home from work, and turning on their A/Cs to cool off the house, flipping on their lights, turning on the oven, etc.

Most grids have their peak power usage after solar has completely dropped off.

The issue then becomes: how can we serve that load? And you could say “just turn on some gas-fired units, at least most of the day was 100% renewable.”

But some gas units take literal hours to turn on. And if you’re 100% renewable during the day, you can’t have those gas units already online.

Grid operators have to leave their gas units online, running as low as they can, while the sun is out. So that when the peak hits, they can ramp up their grid to peak output, without any help from solar.

There are definitely some interesting solutions to this problem, energy storage, load shifting, and energy efficiency, but these are still in development.

People expect the lights to turn on when they flip the switch, and wouldn’t be very happy if that wasn’t the case. Grid operators are unable to provide that currently without dispatchable units.

I think a more feasible potential technology for the grid are flow batteries.

They work through some kind of ion-exchange. Where they have two liquids, one charged and one not. By running power through a catalyzer, they move charges into one tank. Then you can apply a load across the catalyzer, and remove the charge as power.

I’m by no means an expert, but these are already pretty popular in Japan, and have started to make their way into the US.

Still definitely an expensive technology, but I’m hopeful that scale and investment can drive the cost down.

One of their biggest advantages over other technologies like Li-Ion is that their duration is independent of their capacity. Because the duration is only determined by the size of your tanks and the amount of liquid you have.

Meaning that you can take an existing 50MW, 4 hour plant and upgrade it to an 8 hour plant by doubling the size of the tanks and filling them up with the electrolyte. All without having to upgrade the catalyzer.

Edit: also worth mentioning they don’t have the same supply/environmental/recyclability concerns that lithium batteries do. I believe the electrolyte is relatively inert and does not degrade over time.

I hesitate on

that work on the scale needed to support large sections of electrical grid

That first link is for a 10MW, 8 hour battery. 10MW is on the smaller end of generators, you’d need quite a few of these to start making an impact. For example, a small gas turbine is like 50MW, a large one is over 250MW.

And you could say “just build a lot of them” but the capacity per unit of area tends to be pretty low for these types of technologies.

Building them where we have ample space is okay. But now this power has to be transmitted, and we are already having a lot of problems with transmission line congestion as-is. The real advantage of energy storage is when it’s done local, no need for transmission lines.

Plus there’s permitting/stability issues as well. These wouldn’t work if the area was prone to earthquakes or other natural events.

I’m adjacent to this problem, so I have a little context, but am not an expert at all.

To my knowledge, we don’t have granular control over panels. So we can shut off legs of a plant, but that’s a lot of power to be moving all at once.

Instead, prices are set to encourage commercial customers to intake more power incrementally. This has a smoother result on the grid, less chance of destabilizing.

A customer like a data center could wait to perform defragmentation or a backup or something until the price of power hits a cheap or negative number.

I really like your response. Right behind you about energy storage.

Whoever cracks that nut is an instant billionaire in my opinion. The first cheap, effective, and practical storage technology is going to change the world. But we’re not there just yet.

I’m curious on your statement about nuclear. While I do think nuclear is a great energy source, I’m not sure I agree on the on-demand part.

Our current nuclear plants take hours or even days to start up and wouldn’t provide enough reactivity for a highly renewable grid. Are you referring to a future Small Modular Reactor technology? One with a significantly faster startup and ramp rate?

There have been proposals for technology like this. Putting a motor above an abandoned mineshaft and suspending a weight. Charged by raising the weight, discharges by lowering against a load.

The issues is the capacity ends up being pretty tiny, not really at a grid level.

You’d need a TON of motors to get to something a grid could actually use to stabilize, and by then the economics don’t work out. Let alone the actual space requirements of that many motors

Additionally, a lot of the advantages of batteries come from local storage, where you don’t need to transmit the energy long distances anymore, and these “natural” batteries tend to take up a lot of space.

A better and more accessible form of “natural” energy storage are already in most homes. Heat pump water heaters in homes could do things like make the water extra hot during solar hours, when power is cheap, so they can make it until the next morning without turning back on.

Or with better building envelopes (insulation) we could run more cooling during solar, maybe even make a ton of ice. Then later in the day, when solar drops and the grid load peaks, you can still cool the building with ice.