On Contracts for Difference

Cast your minds back - or possibly don't, it wasn't a fun time - to the Great Toilet Paper Shortage of 2020, early on in the pandemic. The supply of toilet paper was fine, but demand skyrocketed - and some enterprising folk decided to exacerbate this by buying in enormous bulk with the intention to resell. Almost everywhere, this was condemned as price gouging - which it was - and in many cases, the oppressed rolls of the people were liberated by law enforcement, no doubt to go to a better place.

"Aha", free-market fundamentalists cry, "but this is why you have a shortage in the first place! Maintaining a stockpile costs Real Money(TM). Therefore, if you want a stockpile to exist, you need a way for someone to profit from having one". And, it's a valid point, except that the empirical evidence is that in the real world, this sort of stockpiling - actually useful stockpiling, which occurs before it is necessary - doesn't happen much regardless. You're asking investors to put Real Money into something that is hopefully never used, and therefore expires worthless. This causes me deep concern, because a lot of businesses do buy large amounts of insurance. Money, on its own, cannot magic toilet paper into being, at least not unless it's already paper money. So where, if not in warehouses, is the money going?

---

One thing the UK should, across party lines, be proud of today is the degree to which renewables have penetrated the National Grid. Last year 23 % was generated from renewable sources - in a country with little sun, no less (for the record, nuclear at about 21 % - I'm counting the contribution from heavily nuclearised France). Part of the reason for this success is the North Sea, which provides both easy wind resource and - thanks to oil - a workforce already skilled in offshore installations - you're just changing what goes on top.

The second part is the Contract for Difference, or "strike price", mechanism, whereby the government promises to buy a MWh of electricity from a low-carbon source - irrespective of the market price. This way, investors in low-carbon power can be certain their investment will, assuming their other calculations and judgements are correct, pay off. For Hornsea wind farm this strike price is as low as ￡41 / MWh, for Hinckley it was a - less good - ￡92 / MWh. This isn't a pure subsidy mind, it works both ways - so it's the government, not private investors, raking it in if prices are short. At least, that's how I understand it.

The other half of that figure is what's known as the Levelised Cost of Energy (LCoE) - which is simply "how much does a plant cost over its lifetime, divided by how many MWh does it deliver". Via Econ 101 we know the total cost will be an overhead cost - the cost required to deliver the first MWh - plus the marginal cost of each MWh multiplied by the number of MWh delivered. Most power plants are overhead-heavy, low-carbon ones even more so - Hinckley, for example, will cost ￡20 bn over 60 years lifetime generating a steady 3 GW. The fuel, the bit that's actually consumed, is not even 5% of that. Wind and solar use no fuel at all, though bearings in wind turbines probably wear a bit. You can calculate the levelised cost of other energy-related things too - Levelised Cost of Hydrogen, Levelised Cost of Storage, etc.

Unlike toilet paper, renewable energy does face regular supply-side crisis. I've written this before, but for an entire week in late March this year, wind output?- I want to be very clear about this, across the entire country, so no more diversification is possible without relying on external suppliers - was essentially 0. That's not even that unusual, maybe a 95th-percentile event, certainly not a 99th. Electricity demand, by comparison, almost always stays within a factor of 2 of the mean, so for a quick n' dirty calculation we can simply ignore variation in inherent demand. If prices are high, it's because the wind is not blowing and the sun is not showing and therefore there is not enough supply. Vice versa if prices are low.

So, the government - the counterparty of the Contract for Difference - nominally makes money if power is short. But realistically, the government already owns the consequences of insufficient supply. Let's be clear on that - a TWh shortfall might cost a supplier ￡41M in lost revenue. If that shortfall isn't covered somehow, people have died by the time the turbines spin up again.

And, as we have seen, nobody in the private sector is building something like Hinckley on the bet that electricity prices will skyrocket, just as they do not build toilet paper warehouses, or pork futures warehouses for that matter. Even if, erm, electricity prices have skyrocketed. The way to profitably build low-carbon power that works when the wind doesn't blow - whether batteries, hydro, hydrogen, or nuclear - is to convince the counterparty that your more expensive MWh are worth it.

(As an aside: hydrogen is probably the best storage play here. Batteries currently cost ￡100 per kWh - per kilowatt-hour - which is just too much for anything more than "buy us enough time to spin up a gas turbine", and no honest individual is predicting that cost to decrease nearly enough. Hydro is limited by geography, and empirically, has killed more than nuclear, for vastly fewer MWh. However, I've written about the amount of hydrogen necessary before - and well there's no way around it, a TWh is a lot of energy, that hydrogen can turn into heat and pressure extremely quickly. The term "megatonne" is not inappropriate. That said, hydrogen does like to move outwards, and upwards, quickly. So it's possible that Buncefeld-type incidents are unlikely. I'd love to see the COMAH on that one.)

There is another option, that I'm not necessarily endorsing: the government could simply abandon Contracts for Difference and impose Service Level Agreements (SLAs). I believe this is what Virtual Power Plants are designed for - the supplier must cover its intermittency somehow, and the buyer doesn't have to. Now it gets complicated: we need to start considering the Levelised Cost of Power (how much does it cost to provide a MW?) instead. The LCoE for the cover is (slightly) less relevant, since we want to rarely use it. Bad news: all the low-carbon backup options have a horrible LCoP. Unsurprisingly, therefore, we're burning gas, and storing it.

I do have to say, it feels extremely odd that the few hundred tonnes total of solid waste produced by nuclear plants a year is "difficult" to safely store, but the 63 million tonnes of gas or supercritical fluid generated by gas plants is not.

For now, the government remains the insurer and therefore the counterparty whose job it is to cover low-wind periods. For the rest of us, the question is whether the counterparty has spent enough on actual warehouses.