What do we need to do to meet "net zero"?
To meet the 2050 net zero target, and the nearer target of 2035 to decarbonise the GB grid, renewable generation capacity needs to grow even faster. Currently we have 54GW of renewables, but we need 140-174GW installed by 2035: a 160-220% increase in the next 11 years. DESNZ's modelling is based on CfD-backed generation making up the majority of the GB electricity market by 2035. This will impact wholesale electricity price formation, since generators are paid a fixed strike price for their electricity, protecting consumers from fluctuating prices.
The second REMA consultation poses a series of challenges, as we detailed in our first Insight. Challenge 2 is Investing to create a renewables-based system at pace. DESNZ express this as How to de-risk investment in renewables while increasing operational risk exposure to deliver lowest overall system cost. So they want to make it easier to invest in new renewable assets but also to change the way that they are incentivised to generate once they are built, so they are more responsive to what the system needs and aren't just generating as much as they can, whether the grid can take it or not.
What's the problem with the current CfD?
In the current CfD regime, renewable generators bid for a fixed strike price for their electricity for 15 years. This provides stability and encourages more investment in renewable generation. It has been very successful in this to date. There are, however, some drawbacks.
The CfD gives generators no protection against volume risk – a risk which is likely to increase in the future. This is where on sunny/windy days, supply of power could exceed demand so renewable generators may struggle to find a buyer for their power and have to stop generating. This means they would not receive CfD payments, as these are paid based on metered volume so if they don't generate, they don’t get paid.
The design of the current CfD incentivises maximum power production, even if this will overload the system. There is no incentive for generators to provide ancillary services that keep the grid stable, unless the value of these is higher than the strike price the asset would get for generating power.
The CfD can also create price thresholds, leading to "herding" where assets turn on/off at the same time, causing prices to go negative – generators end up paying customers to use their electricity rather than the other way round.
Finally, the CfD price is based on the day-ahead Intermittent Market Reference Price, which disincentivises forward trading, making it more difficult for suppliers to hedge.
What are the options for reform?
There are five options being considered:
1. Existing CfD and ongoing reforms – continuing with the current CfD and the current reforms. The Government has been consulting on potential CfD reforms for AR7 (next year's allocation round) and beyond, including expanding scope of CfD to support repowering of existing projects and introducing new hybrid metering arrangements to help generation that is co-located with storage. However, DESNZ have stressed in industry workshops that these reforms will not address all the problems and this option is effectively the counterfactual.
2. Deemed CfD – generators assets are paid a subsidy based on how much power their assets could have produced each hour, rather than what they actually produce. This encourages assets to operate on merchant terms. They would still have an auction where they bid for a strike price. Assets with a deemed CfD could have the negative price rule (where CfD payments are stopped if prices go negative) removed as they would no longer be incentivised to generate when prices are below their short run marginal cost; it would be better to switch off in that situation.
However, paying a CfD on a deemed output introduces more risk. If the actual output is higher or lower than the deemed output, the generator will retain more or less revenue than their strike price suggests they should. The deeming methodology will need to be carefully designed to make it fair.
DESNZ need to think about how a deemed CfD would link with locational pricing. A basic deeming model is likely to shield generators from locational risk. If an asset is located in a highly constrained area where they might regularly have to turn down their power supply, if it has a deemed CfD it would still be paid for its deemed output rather than its actual (lower) output. DESNZ are considering how to deal with this. For instance, a CfD subsidy calculation could use a system average reference price, but assets would sell their generation at the price in their zone. This would mean assets in areas with lower prices would get lower revenues than those in areas with higher prices.
3. Capacity-based CfD (this is a new option) – generators get paid a regular fixed amount based on their assets installed renewable capacity, independent of how much they actually generate (although potentially adjusted by an "availability factor", a measure of the proportion of time an asset is capable of generating). Under this model generators operate on merchant terms and are exposed to both volume and price risk, but also have some revenue certainty. They would bid for a capacity payment on a £/MW basis. A capacity-based CfD also fits better with zonal pricing.
DESNZ are seeing the capacity-based CfD as a two-way subsidy, where the generator pays back gainshare above a certain price point. There is more detail on this option in Appendix 2 of the consultation.
In parallel to, or instead of, the three reforms to the payment structure of the CfD, the consultation also moots:
4. Partial CfD (this is a new option) – only a percentage of an asset's total capacity would get a CfD. The asset would be split in two, each independently metered. One part gets the strike price, the other gets the market price.
The benefits to this are that assets dispatch more efficiently, there is less herding, and more incentive to provide ancillary services.
Potential risks are an increased strike price (as bids will be priced to cover the increased risk), higher costs of capital, lower investor confidence, and it is not clear how generators would operate the CfD-backed asset versus the merchant asset: DESNZ see potential for gaming the system.
5. Reference price reform: instead of using the day-ahead Intermittent Market Reference Price, the reference price could be based on a longer reference price period or an average of the previous day-ahead prices.
Two options from the first consultation have been discounted: a CfD with a strike price range; and a revenue cap and floor.
How will this affect existing CfDs?
Most of these reforms will only apply to future CfDs, likely to be from Allocation Round 9 (2027). As mentioned in our previous Insight, if locational pricing is introduced, existing (and future) CfD contracts would likely be amended to use a local reference price, to shield investors from the effect of locational pricing.
Comment
The aim of these reforms is to put more volume risk and operational risk onto investors. The trick will be to do this in such a way that it does not increase overall costs. Any increase in subsidy (as investors are going to price in these risks into what they bid) ought to be offset by a reduction in system costs, as the renewable assets are operated in a way that benefits the electricity system as a whole.
Investors should see this as an opportunity as well as a threat. A deemed CfD or a capacity CfD would enable an asset to operate more on a merchant basis and be structured to take advantage of ancillary service payments in a way that they cannot do at present. We expect to see more co-location of storage with generation; and some of these reforms (e.g. the capacity-based CfD) will benefit assets that have lower capital costs, even if they cannot deliver as much power as comparable projects that cost more to build.
