Specifically, both electricity and gas have robust futures trading markets. And gas can be converted to electricity.
So... In the ideal world, whenever the price of gas is higher than the price of electricity, gas generators should shut down - since there isn't money to be made.
However, looking at market data for the UK, that only happens sometimes. There are plenty of generators who, according to spot market prices, shouldn't be operating, yet are.
Obviously some gas generators might have purchased low priced gas futures, or have sold high priced electricity futures - making it profitable to operate. However, even gas generators in that position stand to profit more by shutting down and reselling the futures they have bought on the open market.
It appears that even big companies make suboptimal market decisions on quite a frequent basis.
I suspect the cause is general lack of business flexibility. If such inflexibility is widespread, it would be a good reason to disallow futures trading entirely - by forcing people to buy things as they use them, they are forced to notice that what they are about to do isn't profitable - and the futures market hides that from them.
Perhaps the gas generator company bought some gas for delivery today, and has some more arriving tomorrow. When they locked in those deliveries, forecast electricity prices for today were greater than gas prices.
Unlike futures, the gas in your storage tank isn't perfectly fungible (or even convertible at a cost) with gas next month or the gas in someone else's power station across the country.
Thats why specifically the UK is a good case study for a 'perfect futures market', because many of the real world market distortions don't exist.
But plants have thermal limits to what they can do, the heat they provide may be used elsewhere, they may provide services to the grid operator... a range of reasons why they'd run even when not profitable for that hour.
In general, plant operators are very good at maximising value (which translates to more revenue for them). And they will happily use all the flexibility at their disposal to do so.
This topic is covered with a focus on renewable (but applicable otherwise) in [1, pdf]. By accounting for cost of capital, volatility of the commodity, expected sale price, shutdown and startup costs, and a couple other variables, you can optimize the decision on how to act as a producer.
Having previously been responsible for forecasting natural gas consumption at one of the largest aluminum plants in the world, I can confirm inflexibility is a big component. Shutting one section of a plant down for a day is disruptive enough for planned maintenance. You need a really big incentive to shut it down on short notice, and you'll feel the effects your days to weeks in the future. Ideally you've sold product to book every machine pretty evenly and you're "sweating the assets", working them hard.
[1] An Exit and Entry Study of Renewable Power Producers: A Real Options Approach (PDF)
https://dr.lib.iastate.edu/bitstreams/7607fc3a-9acd-40a9-946...
Further, the two-settlement system means you are guaranteed the day ahead price which can be higher than spot rates. Furthermore, in the US you can trade the spread between day ahead and spot. Therefore, it isn't always rational to shut off when spot gas is higher than electricity times your heat rate.
That's why it's profitable to keep the gas turbines running. You must also consider the price of the turbine and surrounding infrastructure: if it can only run profitably 10% of the time it will be a much harder sell than running it 30%.
To some extent these higher variations in intraday or between-day prices should help make storage look more economically viable.
I have to say that we are implementing dynamic pricing and want to have fixed margin of 0.30/kw, I believe we will be the first medium sized operators who will have full price transparency. We are going to be publishing 'tomorrows' prices on our website, even.
If you look at consumers, only a tiny minority has an electricity contract that is directly based on day-ahead prices.
A PV installation can easily start and stop production as long as the sun is shining, right? Then it seems like PV producers should bid to produce as much as possible whenever the price is positive, and not produce anything whenever the price is negative. Is this how they operate, and if so, how does this contribute to negative prices and not just to bringing the price towards 0?
Households are usually not exposed to spot price signals, so residential PV will always produce electricity (network permitting).
Utility scale PV up to a few MW may not have the ability, or the operator may not have the option to curtail contracted.
50Hertz Transmission GmbH, formerly named Vattenfall Europe Transmission, is one of four transmission system operators for electricity in Germany, and is wholly owned by Eurogrid GmbH
