Solar is so prevalent in Australia that power prices regularly go negative as supply exceeds demand. One interesting thing that is starting to happen is companies like Tesla build huge battery banks and buy cheap or negative price power and then sell it back a few hours later when it is much more expensive.
Putting further downwards pressure on prices, small scale solar generators (eg homes) generally have a fixed price for solar exports. The spot price could be -$1000 but they will still be getting paid $100 under their retail contract. So they have no incentive to stop generating, even when the price is negative. The market regulator is attempting to change this, but consumers are resistant to having to pay to generate power or allow their system to be remotely switched off.
Coal generators won't switch off either as they take a while to ramp up and down, and are sometimes are directed (and paid) to remain running to provide system stability.
As for the price going negative - this isn’t a super new phenomenon. I remember learning about coal power plants in West Virginia having negative power prices in the late night. This is simply because it’s hard to spin down and scale up these loads. It’s more cost effective to pay people to take the energy for a short time than it is to shut down the plant and potentially destabilize the grid in the future.
As for sinking it into the ground - there’s lots of research there. Tesla is pushing utility scale batteries. Other efforts move large rocks up and down hills, convert water to hydrogen and oxygen, or simply pump water up hill. Each of these has some losses and hardware costs to get going and efficiency isn’t great.
At a home you can get a set of Tesla Powerwalls and essentially do this - but you’ll find they cost about $6k each and eat about $7/mo in electric losses (at the $0.23/kWh price I pay - thanks, Eversource).
For home users this isn't such an issue today (everything in my home with a motor, has a three phase motor, so there is an inverter inside to convert that from 2 phase AC -> DC -> 3 phase AC, so the frequency doesn't matter), but for industry it is as a lot of equipment needs 60/50Hz and will be damaged with something outside it's operating range.
At multiple places in the power grid there are systems that will disconnect if the grid frequency falls or rises outside it's operating range. This is what caused the cascading failures and power outage over most of the country in the UK in 2019:
https://www.ofgem.gov.uk/publications/investigation-9-august...
Units start to trip below 58 and above 62 Hz ish, and turbines should respond to 0.01 Hz change in frequency within 200 ms according to the IEEE 125 guideline, so even a 0.5 Hz deviation is huge.
A 1 Hz deviation is a major event that would result from a mismatch on the order of gigawatts
I assume in this situation, the voltage also goes below the target, because people talk about "brown-outs" which I believe means that you get less than 120V from an outlet.
I can speculate that when supply exceeds demand, this creates a similar problem, but in the opposite direction.
Can confirm, did some work in a power station. The frequency is displayed in very large font front and centre of the control room. It's really a leading indicator of the health of the network, so any fluctuations are monitored very closely and the operators would sometimes jump on the phone with nearby generators if it starts doing something unexpected.
Another interesting fact, in the control room there was also an indicator showing the frequency adjusted clock time. Older clocks (like many of those bedside alarms) used the 50/60Hz as a clock rather than using a built-in crystal, so it was important to make sure that over time the frequency did average out to 50/60Hz to keep clocks in sync.
Hypothetical scenario: We have a persistent -8 cent spot market price. Industrial companies get paid for consuming electricity. However, retail doesn't benefit from the negative pricing. If the spot market is -8 cent and the feed in tariff is 8 cent then the EEG has to subsidize 16 cents. Thus electricity for retail gets 16 cents more expensive to get back to the original 8 cent. However, this isn't the whole story. We still have to pay for industrial consumption. If 33% of energy usage is industry and is exempt then every retail kWh has to pay 0.5 industrial kWH. So we now end up at 24 cents for electricity. If you were to add taxes (yes you pay them on top of the EEG surcharge) and grid maintenance costs, etc you would probably end up with 40 cents per kWH.
This was just an extreme example but it explains around 16% of Germany's extremely high electricity price. Getting rid of the various taxes on electricity would lower the costs by 20% without changing anything about the grid or making it less profitable.
You don't. Usually there's some external reason that selling renewables at a slight loss is still profitable. Like tax incentives or production quotas or fixed-rate contracts.
Then coal plants get dragged along because they can't change output quickly.
> Why not just sink it into the ground or something?
That takes equipment, which takes money. They won't install it if sporadic negative prices are cheaper.