[tialaramex]

Because it's the most expensive used generation almost always?

But I'd actually be interested in seeing the underlying market data, even say for a week ago. Actually seeing the underlying offers really helps you to properly understand what the future looks like.

For example, right now wholesale is about £70 per MWh. What did the underlying bids and offers look like to hit that price ? How much CCGT was available for less than £50 ? 10MW ? 100 MW? 1 GW ?

[mhh__]

If it was the most expensive it wouldn't transact at that price, the price clears roughly to minimise excess demand i.e. gas is often the energy you can actually at a given time.

[tialaramex]

No, by definition the most expensive used generation did transact at that price because it's a clearing auction. You probably skimmed over what I wrote because it didn't conform to how you'd imagined this works and you mistakenly thought you understand.

I expect that you've modelled this as a sort of "All or nothing" situation but of course that's not how it works for electricity, which is why the underlying data would be interesting. That "guts" of a market help you understand the meaning of the surface numbers like a wholesale spot price or a week-in-advance price.

For example, you can see in the released data that spot prices can go negative without all the gas generation switching off. Some modest amount of CCGT will apparently pay money to the grid to stay running, sometimes for hours, despite lack of demand. Most of the gas goes away at negative prices, but some will still be there, so what's that about exactly ?

My guess (which the underlying data would help confirm) is that generators bid their "tick over" generation at a low price, perhaps even a negative price, so as to rarely need to switch off the generator and pay the considerable shutdown/ startup prices, but bid a lot more at a high price which covers their fuel and other expenses. Do they also have a further tier of pricing for running the generator at some sort of "red line" less efficient but more productive rate? I have no idea, and without the raw data or an inside knowledge of their companies there is no way to know.