More detail in this writeup: climate.davis-hansson.com/p/price-sign...

If all of them instead had actually *participated* in the market, perhaps the price wouldn't have gone negative at all.

Anyway - we need these new protocols to adopt better signals than infinite-volume price lists, or we'll see much more of this kind of thing.

I can't say how many of those PV inverters were passive followers of the price signal vs how many were actually involved in sending wholesale market bids.. but I'd imagine most of them were passive followers.

This kind of "mass distribution of an infinite-volume price list" is how all the major contenders for distributed resource control expects *much* larger fleets of assets to be controlled: #openadr #matter #eebus #energyprotocol etc etc; it's all price lists with no volumes attached

..which would've been fine if the assets procured by the TSO to stabilize those oscillations had delivered as bid, which they didn't -> voltage spikes in distribution -> coordinated inverter shutoffs per regulation -> blackout

What’s nuts about these though is: Because there are no smart meters in DE apartments, this shows up as imbalances that need TSO intervention. If we were to account for this production we’d get in huge trouble; need to pretend it doesn’t exist and buy GWhs of gas production that isn’t needed

Nukes are primary driver of cost of standby balancing power for a reason, large single points of failure need large idle spare capacity.. but can’t have been just a lost nuke, that’s supposed to be fine at any time?

Aha, I think the graph itself is from the Economists “Sun Machines” feature: archive.is/ByJMh

The sources are in the graph, on the bottom. IEA, BNEF