The imperfect LCOE and CfD

J.P. Morgan has recently stated in their Eye on the Market annual energy paper, 13th Edition that the Levelized Cost of Electricity or LCOE is practically irrelevant for wind and solar energy and is a misleading basis for estimating total system costs to governments, electricity consumers, and taxpayers. In another recent piece, the author Dan Lee is arguing that focusing on the capacity factor (or the energy produced per MW installed, which is a close concept to LCOE) for photovoltaics is not enough to compare the value of different projects. Nevertheless, it is common to use LCOE as a proxy to estimate the cheapest way to generate electricity but, as we will argue in this post, we should be careful to draw conclusions on total system costs based solely on LCOE. And for similar reasons, this observation could be extended to the generalization of support schemes such as Contract-for-Difference (CfD).

LCOE

The levelized Cost of Electricity is a measure of the average net present cost of a generation asset over its lifetime. The concept is well defined on Wikipedia. LCOE is used widely to compare various different technologies, see for example the reports from Lazard or IRENA. With the recent energy crisis, the LCOE of renewable generations became much better than the ones of fossil fuel generation as shown in the figure below from IRENA. We can observe that wind and solar are up to 6 times cheaper than fossil gas in terms of LCOE. Of course, such differences will reduce with a lower price for natural gas (and they have already).

A point that is not considered in LCOE calculation is that a kWh has not the same value over time. This is a critical point as we increase the share of weather-dependent sources. When the power systems were mainly composed of fossil fuel assets, the main driver for the value of a kWh was the demand. Nowadays, wind and solar power are slowly taking a greater role in the electricity mix. When the wind is blowing and the sun is shining, the prices drop drastically, meaning that an additional capacity would only bring very low additional value (even negative on some occasions).

Does it matter already?

Devaluation of the market value captured by renewables is already happening in some countries, as we can see in the graph below from Energy-Charts (prices were lower than 5 EUR/MWh for 8 straight hours from 11 AM in Spain on 13 and 14 April).

Spain is somehow impacted faster as the Iberian peninsula is poorly interconnected with the rest of Europe. Nevertheless, the projected growth of renewables, especially solar, would be important across Europe. Support for renewables, such as CfD, is generally warranted for a decade or longer, and therefore, projections on the dynamics of the future power market are important.

Going beyond LCOE

As the value of electricity is very dependent on the moment of generation, the concept of LCOE with its complete abstraction of this characteristic does not provide a good indicator concerning the overall system cost. Different proposals have been suggested in order to tackle this limitation: value-adjusted LCOE, marginal system LCOE, Levelized Value of Electricity, System LCOE, etc. These new metrics try to answer mainly two questions:

• Does the generation profile match the favorable market prices? If the generation is added on top of an existing surplus (low value of electricity), there would be a need for storage, demand flexibility, etc.

• Does an additional renewable capacity add any system cost? Other system costs might be the need for additional balancing capacity, increased transmission grid, etc.

Contract-for-difference: a focus on LCOE

With the proposed market reform of March 2023, the EU Commission is pushing for the introduction of contract-for-difference (CfD) as a support mechanism for renewables. The basic principle is that the renewable producer would receive (or pay) the difference between the market price and a strike price. An auction mechanism will select the cheapest bidders, generally in terms of the lowest strike price, independently of the generation profile.

The direct consequence of the CfD is that the producer does not face the prices of the market (this is actually similar to a Feed-In Tariff for that matter). Of course, this system has its merits as it ensures a foreseeable revenue stream for the producers rendering their projects more bankable. Nevertheless, it does not give any incentive to produce when prices are more favorable. The only objective is to produce as much as possible independently of the market prices. As soon as price cannibalization becomes important, such lack of incentive might be detrimental from a system perspective, and consequently from the entity responsible to give the difference (which is generally supported at the end by the consumers). We can also note that some authors, such as David Newbery and Lion Hirth, have suggested some improvements.

Potential innovations with a market price incentive

There are various potential innovations that most probably degrade the LCOE but improve the value of the electricity provided in case there exists a market price incentive (non-exhaustive list):

• Co-location with storage. This is the most obvious case. Adding a storage system would increase the LCOE but it would provide a service to the power grid.

• Undersizing the inverter and the AC connection to the grid. The peak generation is likely to be the one with the less economic value, while the transmission grid is often a constraint.

• Technology adaptation. Some projects are being developed with bifacial PV and a vertical orientation East-West in order to flatten out the generation output throughout the day.

Moreover, renewable producers would be incentivized to perform their regular maintenance during hours when the electricity prices are lower.

No simple way

The design of support programs for renewables should always be done with care and realistic assumptions with regard to the value of electricity and the total system cost. In particular, we must always keep in mind that LCOE is only a metric but does not provide the whole picture to assess the total system cost. Similarly, having an abundance of CfD contracts with low strike prices would not lead systematically to a low total system cost.