High Power Prices
This post offers some thoughts on why electricity prices in Great Britain are so high, with a few reflections about what should be done about it.
I signed up back in late 2018 with Octopus Energy’s Agile tariff, which reflects the half hourly wholesale prices. My reason for signing up for this experimental tariff was that I was interested in keeping a closer eye on electricity prices, and learning how prices can incentivise more sustainable consumption.
Since then, I’ve experienced a particularly volatile period, as shown in the chart below. The orange line shows the average system price (sometimes referred to as the imbalance price) each month, while the blue and green lines show the 90% percentile each month and the 10% percentile price. Since the start of 2019, prices fell steadily until the middle of 2020, before rising rapidly over the past year. Mean and high prices are now higher than they were two years ago, though interestingly the low prices are still lower, though at £28/MWh in June 2021 much higher than the £0.8/MWh they were in June 2020.
How are GB electricity prices set?
Before answering why prices have behaved this way, I need to explain how power prices are set in the GB market. Roughly (there are complexities I won’t go into here), generators work out a price for each hour or half hour, above which they are happy to generate. Whatever price is needed to ensure meet the demand for that half hour sets the price. This process is known as the generation ‘stack’.
Generators typically work out a price that is enough to cover their variable cost of operation. This will be zero in the case of wind and solar, low prices for nuclear, moderate prices for coal, biomass and efficient gas turbines (typically CCGT), and high prices for less efficient gas turbines (typically OCGT/peakers), diesel and storage (pumped hydro and batteries).
This means that even if there is a lot of wind available to meet demand, as long as some gas is needed (which is currently always the case), the price won’t be below the operating costs of a gas generator (based on its generation efficiency, gas prices and carbon prices). The less gas needed, the more likely it can be met by more efficient turbines, but we still shouldn’t expect cheap prices.
In practice, generators almost always sell their capacity in advance, sometimes in the day ahead auctions but often much further in advance. This has the advantage of reducing their exposure to price uncertainty (and also allowing electricity retailers to buy in advance).
In theory this shouldn’t make a difference to generator activity or wholesale prices, but in practice it probably does: for example, a generator that has pre-sold its capacity more likely to run than one that hasn’t.
Some generators (mostly nuclear and coal) are pretty inflexible, so can’t switch on and off. They will typically decide to run all day or switch off for the day. They can still decide to run even if the prices are low, especially if they’ve presold their capacity, but this can puts further downwards pressure on prices.
Storage sites play into the mix as both generators and consumers. They will try and buy power whenever the price is low, and sell whenever the price is high.
Finally, we have interconnectors to Ireland, France, Belgium and Netherlands (there is one opening up later this year to Norway). Market participants can buy the right to flow power along the interconnector, and if they think the price in GB will be higher, they will schedule this flow, adding to the GB supply. (The day ahead scheduling used to work more efficiently before Brexit, but since power flows to GB most of the time, this is less of an issue.)
So, why have electricity prices behaved like this?
Now that is out of the way, we can start looking at why we’ve seen these price changes. This is often a question without an answer, but I’ll provide some possible reasons. Firstly, and I would argue most importantly, the electricity prices mirror the gas prices over the past two and a half years, as shown in the chart below (my source of day ahead NBP gas prices isn’t exact, but it gives the right picture).
The UK Gas price is strongly linked to European and International gas prices (through our ability to transport gas through pipelines and as LNG), so most of this fall and rise is unlikely to be specific to UK (I don’t have data handy for other gas markets to verify this, feel free to comment if this isn’t true).
Covid is often used as a justification for the fall in prices in the first half of 2020 — due to the reduction in industrial demand. This is likely to have some effect, but it is important to recognise that most of the reduction happened in 2019.
Next, emissions prices (shown here in EUR/MT from Sandbag) fell a bit from mid 2019 to mid 2020, but have subsequently had a huge increase (which has continued since April 2021). This may be due to political pressure, as people expect the EU to reduce the number of allowances.
Next I looked at generation by source, taken from UK Government Statistics, table ET5.4. A lot of the reasons given for the increase in prices is the decrease in coal and nuclear generation, but I’m not really seeing that in these figures. There is quite an increase in wind generation in Jan-March 2020, and a fall from January 2021 onwards, but I’m not convinced this was a major driver of prices, given that the fall and subsequent rise in prices have been much more sustained.
As further steps for analysing these prices I would be interested in seeing how much the UK gas prices have varied from global gas indices (eg TTF in Netherlands and Henry Hub in the US), and also how much of the power price movement can be explained by the UK gas and emissions price movement (I suspect most of it).
Reflection 1: Questioning the relationship between short term and long term prices
As wholesale spot (daily) prices have gone up, longer term prices (including fixed term retail contracts) have also gone up, as you’d expect. But the longer term prices haven’t gone up by as much. This has made Octopus Agile worse value (unless you are a net exporter) than many fixed contracts. It is worth questioning why.
The economics textbook answer is that fixed prices reflect expected spot prices. If spot prices are well above fixed prices, we should expect spot prices later in the year to be much lower. I’m sceptical of this: I don’t see spot prices falling in the next year.
One answer I’ve heard a bit is that the wholesale spot market itself is broken. This answer doesn’t make sense to me. Parties can still trade between the two markets, so if the half hourly prices were always higher, it feels to me like someone would step in and profit.
A second answer is inertia. Because suppliers buy in advance, this smooths things out. When spot prices are really low, the fixed prices don’t fall as much. When they rise, likewise. I suspect there’s a bit of truth to this, especially as suppliers don’t want to lose customers if they raise prices before their customers.
Next, you can get market mismatch. If the renewable generators are more desperate to presell their generation than retailers are to pre-purchase, they may accept a discount to the expected spot prices.
Finally, a theory that I haven’t seen described elsewhere. Generators may be worried about spot prices falling, not in the next year, but 3–10 years down the line, once we get a lot more renewables on the system. They might therefore be willing to presell on longer-term contracts, at a significant discount. If the term is long enough, the generators might be right to do this, and retailers might be left with exposure. If the term is short, the retailers are probably benefiting.
My guess is that there’s a mix of generators selling long term at a discount, retailers buying further out than they are selling (hence getting cheaper prices at the cost of some term mismatch), and retailer inertia.
Reflection 2: The current prices aren’t a disaster, but there are still actions the government should take
I hear a lot of people claiming that the high prices are proof that:
- our system doesn’t work, or is rigged
- we don’t have enough generation
- things were in the 1960s when the electricity system was nationalised
- we would be better off without interconnectors to other markets
- we can’t afford to decarbonise
I’m not convinced by these claims. The current prices aren’t extraordinarily high, and are helping the business case for the things we need (more wind and solar capacity, and more flexible capacity). If anything, I see the current situation as more sustainable than the low prices of 2020, or the low day time prices seen in some markets with lots of solar capacity.
That said, there are some actions I would like to see the government take.
First, and most importantly, the government should stop artificially increasing retail electricity prices relative to gas. For example, carbon prices increase the retail electricity price, but not the gas price. This is counterproductive to efforts to encourage the electrification of heating.
Second, as power prices rise, the cost to the government of some of the environmental policies like Contracts for Difference go down. I believe the government could reduce the environmental levies on electricity prices at such times, since the customers are already paying the price (and similarly, increase the environmental levies if wholesale prices subsequently fall).
Third, energy poverty is a problem, and high electricity prices make it worse. I wouldn’t want to see the government subsidise electricity use, but I am sure there are programmes that could help address poverty more generally, and facilitate improvements in energy efficiency.
Reflection 3: Wholesale markets still aren’t liquid/complete enough
I mentioned above my suspicion that generators want to sell longer term fixed contracts than retailers need to buy, and it isn’t clear (or transparent) what the right prices for different terms are. For example, I have a theory that prices will remain reasonably high for the next 5 years before falling a bit, but it is hard to see if that is reflected in the market. This is made harder to compare by the fact that a lot is bought in the form of Power Purchase Agreements, which all refer to different (and uncertain) volume profiles.
I believe the market would be well served by liquid and transparent forward markets, going out as much as 20 years, that would allow counterparties to correctly price and hedge different kinds of exposures. I wrote about this in my most recent blog post.