Skip to content

A low-carbon quota for the Capacity Market

Bruno Prior

This is the third part of Bruno Prior’s review of Policy Exchange’s report on net zero. The first part was on PX’s proposals for local electricity pricing and the second was on their proposed changes to the contracts for difference regime. In this final article, Bruno looks at their proposals for a low-carbon Capacity Market and draws his conclusions on the whole of the PX report.

The Capacity Market is how we keep the lights on as increasing levels of intermittent generation make standby capacity both more important and less economic. People contract to have capacity available to generate (or not) when required. They are paid for being there, regardless of how much they are used, but can assume high prices if they run because they are only required when the market is short. So the ideal technologies have low capital and fixed costs, whereas operating costs (including the costs of fuel and carbon) are less important. Other technical factors are important, such as the ease with which they can be ramped up and down.

This all lends itself to gas- and oil-fired generation. It also suits existing, depreciated capacity over new. These types have dominated the capacity market. PX note that around 1/3 of capacity in the market is currently low-carbon, but the outlook is worse than that, because much of it is existing nuclear and coal-to-biomass conversion, some of which will not remain available for the Capacity Market indefinitely.

PX estimate that two-thirds of the capacity in the Capacity Market needs to be low-carbon by 2050, in order to reach Net Zero. It is not clear how they can know this. They must be assuming a certain level of utilisation of the remaining fossil-fired generation, and a matching level of negative emissions elsewhere. The uncertainties around that must make any forecasts practically worthless.

The Capacity Market included some provisions from the start to press down on the carbon emissions from the Capacity Market (the original Impact Assessment noted “a decarbonisation trajectory of 100gCO2/kWh in 2030 is assumed”). It has been amended recently to introduce further drivers to reduce the emissions. But PX presumably estimate that these measures will be insufficient.

Their solution is to break the Capacity Market’s technology agnosticism. Currently, all types bid against each other in a Dutch auction. This has kept the cost down (around £1/kW/year).

If we separate low-carbon from fossil-fuel capacity in the Capacity Market, and have a rising level for low-carbon capacity and a falling level for fossil-fuel capacity, we can expect a number of effects.

The falling level of fossil-fuel capacity will see the capacity reduce of gas-fired dispatchable power, which has kept our systems going in the face of increasing intermittency from other sources. This capacity was cheap because it mostly already existed and had low fixed costs.

It will make us dependent on a matching increase in dispatchable low-carbon generation and/or storage and/or interconnectors and/or “Demand-Side Response” (i.e. users switching off because there’s not enough electricity to go round). But dispatchable low-carbon generation is expensive for these purposes, partly because most of it would have to be built for the purpose, and partly because it has relatively high fixed costs. Similar points apply to the alternatives.

The clearing price in a Dutch auction for low-carbon generation (etc) where the level required exceeded the existing capacity would be orders of magnitude higher than the current Dutch auction including existing low-fixed-cost technologies where the required level is less than the existing capacity. For those with a long memory, the gaming opportunities for existing capacity from this arrangement would be similar to those that discredited the Pool.

To be fair to PX and Aurora, two things can be and probably are true:

  1. Something like this would be necessary to achieve Net Zero.
  2. The effects on costs would be as described above.

This is just one example of why Net Zero is economically incoherent. Diminishing marginal returns mean that the costs of reducing carbon rise as the benefits fall. There is some point, almost certainly above zero, where the cost exceeds the benefits. The efficient amount of an externality is almost never zero.

In this case, it means that we should not determine the amount of low-carbon and fossil-fired capacity available for standby on the basis of quotas, but on the basis of cost, settling at the level where the cost of fossil-fired standby capacity including the cost of its carbon emissions (quite small because of infrequent use in this context) is equal to the cost of low-carbon options. That would retain the technology agnosticism of the system and minimise the cost of ensuring system stability at whatever cost of carbon was thought appropriate.

In all likelihood, it will be more cost-effective to keep much of the existing fossil-fired capacity available but lightly-used, than to replace it with expensive and infrequently-used low-carbon capacity. But that should be discovered in a market, not rationed according to calculations by PX, me or anyone else.

Key beneficiaries of this proposal would be the operators of existing low-carbon generating capacity. They could game the system by pricing their capacity at the level required by new capacity, and should gain super profits as a result. The two main such technologies are nuclear and biomass power.

Interestingly, that is two of the three proposals in the PX paper that offer gaming opportunities for super profits to these technologies. The authors of the PX paper should probably declare if either of them has had a role previously representing the interests of either of these technologies.


Proposals can be right for the wrong reasons. Local pricing and cost-reflective pricing would be efficient, not because a model calculates that a certain version would save some money (obfuscating many other cost increases in the process), but because prices are information vectors. The more accurately they transmit information about supply from producers to consumers and about demand from consumers to producers, the more opportunities there are for creative solutions to respond to those conditions.

The question then should be why such pricing is not already widely available? It may be that the costs of doing so exceed any benefits, and it is therefore efficient to live with this market imperfection. But it may also be that the rules that govern the electricity market (and it is a market, like many others nowadays, that is laden to breaking point with regulations) constrain the opportunities for entrepreneurs to offer this kind of product.

The right response to noticing an apparent inefficiency in the market should be to investigate if there are institutional obstacles that encourage that inefficiency, and remove them. It is not to propose more rules to try to constrain the market to deliver what economists have calculated is the “right” outcome. Inevitably, those new rules will bring more perverse incentives and unintended consequences. It’s strange how often the beneficiaries of those unintended effects are the large corporations, while consumers, taxpayers and small businesses pick up the bill.