Socialised losses, private profits
Policy Exchange's proposals for a floor price are fundamentally flawed
This is the second part of Bruno Prior's commentary on the recent paper on energy reforms from Policy Exchange (PX). The first part can be seen here.
Contracts for Differences guarantee a generator a “strike price”. If the market price is below the strike price, the price to the generator is made up to the strike price. If the market price is above the strike price, the generator gives back the additional value above the strike price. This is a simplification of the process (which does not happen instantaneously), but it is the effect.
CfDs de-risk investments for those who secure them. The trade-off is that the investor has to sacrifice any upside in order to avoid any downside. De-risking minimises the finance cost, and auctions for CfDs are supposed to ensure competitive strike prices.
In a world of (a) increasing levels of intermittent generation, (b) increasing electrification of heat and transport, and (c) subsidy structures like CfDs that insulate low-carbon generation from the level of demand, we can expect much greater wholesale-price volatility. So the worst case (insulated by CfDs) will be worse, and the best-case (taken back by CfDs) will be better.
Into this world, PX propose to introduce a modified CfD called a “floor-price CfD”. Generators are protected from prices below the strike price, but can sell at the market price where prices are above the strike price.
Even semantically, it is absurd. A CfD isn’t a CfD if it only affects the downside. That’s just a straight subsidy.
The consequences might seem to be minimal, but if so, what’s the point? People may be fooled into thinking they are minimal by observing that high prices typically occur when intermittent output is low, in which case they have little opportunity to profit from the high prices because they are not generating.
That forgets, as the proponents hope we will forget, certain ways in which the market may vary from that example.
With increased electrification, especially of heat, network stresses may occur as much from an excess of demand as from a shortage of supply. Too few people appreciate the discrepancy in the scale of the electricity and heat markets. Annually, twice as much energy is consumed for heat as electricity. Electricity is also less seasonal. Peak heat demand is over three times higher (170 GW) than peak electricity demand (50 GW), while heat demand falls below electricity demand at its lowest points. The scale of this seasonality will create supply pressures sometimes even when there is material output from intermittent generators. They would like to be rewarded for generating (by luck, through no means within their own control) during these occasions without taking the consequences of the times when they are generating more than we need.
Not all low-carbon generation is intermittent. Two generation technologies stand to benefit particularly from “floor-price CfDs”: biomass and nuclear power. These technologies could assist with balancing the network by producing mainly when demand exceeds intermittent supply. But that would hurt their returns, because they (especially nuclear) have significant fixed costs. Lost output means fewer MWh to spread their costs over. Ramping up and down is also an operational nuisance. They would rather run flat out (“baseload” as we used to call it). A “floor-price CfD” means they could make acceptable profits running even when there’s no shortage of power, and super-profits running when we really do need it. Heads I win, tails you lose; or as we came to know and love it during the 2008 Financial Crisis, “privatised profits, socialised losses”.
Proponents will object that the auction will take care of these criticisms. The clearing floor price will be lower than the clearing strike price of current CfDs. Access to higher prices in peak periods will allow lower subsidies in off-peak periods.
How much competition will there be for nuclear CfDs (for example)? It’s unlikely that the scheme will be technology-agnostic, given its history, the impact that a single nuclear plant could have on a cross-technology market, and the different contract periods that are likely to apply for different technologies. Assuming separate pots remain, a nuclear bidder can be confident that they are too significant to decarbonisation plans to reject, and bid conservatively, expressing pleasant surprise when peak prices yield higher profits than anticipated.
That is where the true purpose of “floor-price CfDs” comes in. As the floor price will be lower than the old strike price, the various interest groups will claim that the cost has fallen. Ce qu’on voit et ce qu’on ne voit pas. People will see the lower contract price. Few will notice the annual accounts that show that the average price achieved is higher than under the old CfDs. Government will use the “falling cost” as a shield to justify the award of the contracts, and to extend the narrative that the costs of low-carbon technologies are falling. Retail prices and total system spend will continue to rise, but that will somehow be unrelated to the “falling costs” of the technologies used to produce the power.
Aurora almost give the game away. A footnote tells us that their model included “CfDs with a strike price of £70/MWh (real 2012£) for new nuclear post Hinkley Point C”. That would be a dramatic (approx. 25%) reduction. But their calculation of total system costs is significantly (>20%) higher than today (though they don’t acknowledge it). Either (a) that £70/MWh is the projected strike price under a floor-price CfD worth a lot more than £70/MWh in the round, or (b) choosing an improbable cost-reduction for nuclear has failed to offset even larger, unidentified cost increases elsewhere. Occam’s Razor says (a).