Blog post

It looks, waddles and quacks – will we see Californian duck curves in Europe any time soon?

Energy Blog, 14 February 2024

The US state of California leads the world in the transition to renewable generation but this has led to some unexpected negative consequences which may be a foretaste of problems in Europe

One of my friends is a twitcher and one is a trainspotter. To explain, a twitcher is somebody who collects sightings of birds. Sometimes stormy weathers blows unfortunate birds across the Atlantic and twitchers will travel huge distances to spot a miserable foreign bird looking a bit lost in a big wet field by the sea.  A trainspotter is somebody who collects locomotive identification numbers and will also travel long distances to see an obscure locomotive in a remote railway siding though trains tend not to be blown in by storms. What’s more interesting is that despite their apparent similarities, my two friends can’t stand to be in the same room as each other. But the underlying history of their mutual antipathy is probably another blog all in itself.

The inevitable segue way into renewables is the possibility of California duck curves appearing in Europe. This won’t appeal to my twitcher friend as a Californian duck curve is not a real duck  but a metaphorical term used to describe an unusual pattern in  electricity demand/price markets. To those of us in the renewable energy industry it is equally exciting.

A California duck curve describes the phenomenon of the demand for non-intermittent and typically fossil fueled electricity  being high in the early hours of the morning, falling to a plateau during the day then climbing steeply in the evening given rise to the classic duck profile shown below.



The shape of the curve is a function of both the demand for electricity and the supply over a 24 hour period.

In pretty much every country in the world, demand for electricity falls dramatically during the night when everyone is asleep, rises in the morning when we get up, plateaus during the day then peaks in the early evening when the population is commuting home, cooking etc. This has always led to a slightly lumpy demand curve and this has translated into differences in peak and off peak pricing.

What the California duck curve illustrates is how up of solar is significantly affecting the demand (and hence the price) for non renewable sources of energy. The  curve has become much pronounced over time because of the increased day time supply of solar generated electricity over the last ten years. Thanks to a winning combination of legislative carrots (subsidies for domestic solar) and sticks (planning regulations requiring solar on new builds), solar capacity in California has grown from 2,550 MW in 2012 to 39,729 MW in 2022, an astonishing 15 fold increase in ten years. Solar now accounts for 27% of the State’s electricity generating capacity which, as we discuss below, may be a useful metric for Europe.

Although they are great for consumers, duck curves are not all good news and cause problems both for incumbent non renewable and renewable generators The obvious problem for incumbent fossil fuel generator is they make what was already a lumpy demand curve much more lumpy. In California demand for fossil fuel generation drops during the day but rockets back up when the sun comes down and everyone gets back to their no doubt sweltering houses and ramps up the air conditioning. It is very hard to rapidly turn on large power stations so this rapid ramp up means the grid operator CAISO (Californian Independent  System Operator) either has to pay  power stations to stay on all day despite the lack of demand or pay renewable operators not to generate (curtailment)

The tendency to date has been to keep fossil fuel fired power stations running at less than 50% capacity for most of the time and this has created problems of oversupply and  high frequencies on the system. More recently CAISO has transitioned to being a net exporter of cheap day time electricity and a net importer at nights. However this creates the surreal situation where as a direct result of pursuing a goal of carbon free generation, CAISO is both running gas fired power stations during the middle of summer and, to avoid a subsequent over supply in the system, at the same time exporting the subsequent dirty electricity to the rest of America thereby inhibiting the uptake of renewables elsewhere.

The pattern of demand as recently as last Saturday, 10th February nicely illustrates all of these phenomenon,

You will see from the graph that solar  generation peaked at 18 GW at around lunchtime whereas gas fired generation fell to about 3.5 GW. at the same time. However by 7pm, solar had all but gone whereas gas generation was back up to 10 GW Two other trends were also evident on Saturday,  day time exports to other States of about 4 GWs per hour (nicely mirroring the day time gas production as per above) and the State becoming a net importer of up to 5 GW at nights.

As well as exporting gas fired power, CAISO has also been busily curtailing large scale renewable generators and in 2023, they paid for 2.659 million MWh of electricity not to be put into the system because of over supply. To put this into context, total domestic solar capacity in the State is about 2.436 million MWh per year so the equivalent of over 100% of domestic solar capacity is being curtailed.   

The curve is also problematic for renewable generators as in the times they are not being curtailed, the power price they achieve during the day is far less than the base load power price charged in the evening peak. This is the phenomena of ‘sub optimal capture pricing’ whereby the price paid to or ‘captured ‘by  day time solar generator over a 24 hour period may be significantly less than the average paid to their fossil fuel counterparties picking up the evening demand.

As more solar is added to the grid, the day time price drops further and the capture price shortfall increases for all solar generators.  On some days in 2023, the Californian capture price differential between day time and evening pricing was as much as 90%. In theory the market should reach an equilibrium where capture prices are so low as to deter new investment in solar generation but we seem to be some way short of that point to date

Although duck curves are a real and problematic issue for the California grid, they don’t seem to have made it across the Atlantic to Europe just yet though it is probably  just as question of time. As I write Europe has 208 GW of solar capacity and this increased by 56 GW in 2023. Total electrical capacity in Europe is circa 1,000 GW so solar is now over 20% of the total capacity. A further 62GW is planned for 2024 and some more excitable commentators are predicting  650 GW by 2030. Our California friends have shown us that once solar exceeds 25% of the total generating capacity, some form of duck curve will most probably start to show up in the summer energy statistics.

The key point to bear in mind here is that Western Europe is relatively quite a small place and the installed solar capacity  is concentrated in an even smaller number of countries; Spain, Germany, Nederland and Italy so when it gets sunny in one it tends to get sunny in the other countries at the same time. This correlation of generation is much more pronounced than it is for wind and means the main European solar generators will all put power into the interconnected system at roughly the same time in the summer months. As we see with California, this will probably mean very low capture prices compared to evening peaks and the same issues for grid operators of oversupply and frequency imbalance.

It also has serious consequences for the European nuclear and gas generation fleet which will be asked to scale up and down in much quicker periods than they are designed to do. The one thing I learnt from  the TV program Chernobyl is that you really don’t want to try and turn a nuclear power station on or off very quickly.

So if duck curves are not all they are cracked up to be, what can be done to mitigate the adverse effects. Storage will inevitably be part of the solution as it will enable solar generators to time shift exports to the more lucrative evening periods and improve capture price %s. It can also help avoid the issues of frequency imbalance by absorbing excess power from the system in the summer months. In the California example above we see around 5 GW of stored energy being put into the system at around 18 pm.

Increasing interconnectivity throughout Europe will help to shift the supply to less sunny places and the big shift to electric cars should go a long way to absorb day time excess providing everyone remembers to plug in their cars.

Ultimately low capture prices for day time generation should act as a brake on the juggernaut of solar installation but at the moment Europe seems to be speeding up rather than slowing down. The near vertical fall in solar panel prices over the last ten years is adding to the momentum.

In conclusion, I am going to put my head above the parapet and predict we will see some sort of duck curve in summer European power markets in the coming decade. It may be more of a Crested Shelden duck curve (which my friend says is the rarest duck in the world with only 49 left alive) than a Peking duck (the most common with over 20 million) but some sort of duck all the same. This will be good news for consumers, less so for existing generators as every sunny day the European grid will be massively overloaded with super cheap power which will significantly drive down returns for both renewable and non renewable generators. In theory  batteries and the take up of electric cars should step in and help mop up the excess but as the old adage says, markets can remain irrational a lot longer than you can remain solvent.