Solar and wind only cannibalise prices if you let them

What is it?

We define price cannibalisation as the reduction in market prices and thus value captured by a particular resource that occurs as more energy is added from that resource. It may be considered problematic when prices and revenues captured fall below those required to recover lifetime costs while, simultaneously, more of the resource would be of value to the system.

Some academics have argued cannibalisation may affect variable renewables as they “face a substantial difficulty in becoming economical at high market shares … [and thus] without fundamental technological breakthroughs, a deep decarbonisation of power systems will be hard to achieve based on wind and solar power alone.” Indeed theoretical analysis shows that the growth of variable renewable energy (VRE) in the power system can depress market prices when they are deployed and lead to prices of zero or below. Where for example wind capacity is already significant, holding other variables constant, adding more wind capacity means more production is concentrated in those hours when potential wind production is highest, disproportionately lowering the volume-weighted prices realised by both new and existing wind plants.

But price cannibalisation is not just a theoretical abstraction. Real world analyses have identified its presence. Analysis of wholesale prices in California from 2013-2017 finds both absolute and relative cannibalisation effects for both solar and wind, with these effects growing with penetration. A recent analysis of price cannibalisation in Germany over 2015 to early 2021 confirms the presence of price cannibalisation: more solar and wind production led to lower capture prices, with pronounced and non-linear effects (see Figure 1).

Figure 1. Market value (Euro per MWh) of wind and solar renewables against generation forecast, Germany, 2015-2021

Source: Liebensteiner, J., & Naumann, F. (2022, November). Can carbon pricing counteract renewable energies’ cannibalization problem? Energy Economics, 115.

Drivers

At first sight the driver is more renewables: more zero cost renewables equal more cannibalisation.

But recent contributions to the academic literature shed light on an important nuance — that the extent of price cannibalisation hinges crucially on the policy tool used to drive decarbonisation and deployment of renewables, as well as the role of flexibility, which we dive into later.

A distinction here is drawn between two policy tools:

• Carbon pricing. This can be achieved by imposing a carbon tax directly or by setting a price on carbon with the introduction of tradeable carbon allowances.
• Support. This might be in the form of a subsidy for renewables through a revenue stream additional to the electricity market, such as a feed-in tariff. But it could also be another policy that target a certain amount of renewables beyond what market prices would bear, such as a contract for difference.

To illustrate the role of each of these, the analysis that follows considers the extremes of relying on just one of these tools. This is not representative of the policy landscape, which in Europe encompasses both carbon pricing and renewable support, but it is helpful in disentangling the effect of each tool. We build on this in the final section to apply analysis to the real world.

Carbon pricing and renewable support compared

On the one hand, use of a carbon price to drive decarbonisation and deployment of renewables will not stimulate a price cannibalisation that impedes optimal deployment of renewables: the value that renewables capture in market prices will not dip below their levelised cost. Research shows this result holds in a power system model with high penetrations of solar and wind, such as above 80%. Why? Because carbon emissions are reduced to the point where the cost of reducing the last unit of carbon is equal to the carbon price. This stimulates market prices when energy is being generated from unabated fossil fuels, and the accompanying higher prices benefit renewable generators when they generate at the same time. It also stimulates demand — the price offered — for clean energy from renewable energy where it can be enabled through flexibility such as storage and new wires. Note: Europe already has instituted tightening carbon quota through the Emissions Trading Scheme, which should stimulate a growing carbon price trajectory.

On the other hand, relying alone on support that directly subsidises renewable energy — or that sets renewables generation technologies beyond a level that market prices would bear — will lead to price cannibalisation. By definition, market prices captured by VRE will not cover the levelised costs of renewables. Support policy can depress market prices through two mechanisms:

• When the subsidised renewable technology is generating, the larger share of this low marginal cost technology pushes down prices, even when the technology is not price setting by displacing more expensive resources.
• When a subsidy is linked to energy produced and the subsidised technology sets prices, the subsidy reduces the break-even point of the marginal producer by the extent of the subsidy and may stimulate a negative market price. (This might be dubbed the subsidy design price cannibalisation effect as it can be tempered by smart subsidy design, a point we return to later.)

Price cannibalisation is an outcome of the policy choice to use subsidies — whether explicit or implicit through setting a given level of renewables — as a substitute for carbon pricing to spur decarbonisation and renewables deployment

The first effect is counteracted by pricing carbon, which, for a given carbon quota, will rise until prices earned by renewables (in consort with flexible resources like storage) are sufficient to recover their levelised costs. The second effect is present neither with a carbon price nor with a well-designed renewable support.

For the case of renewables support without carbon pricing, as renewables penetration grows so, too, does price cannibalisation. Modelling suggests that at 50% penetration, the market value captured by variable renewables can fall to zero, such that subsidies become their only revenue source. (This finding draws on a simplified model of Germany, Poland, France, Netherlands and Belgium.)

Thus, how pronounced any price cannibalisation effect is for variable renewables depends on the carbon price: the lower the carbon price, the lower will be the level of penetration that price cannibalisation will kick in at for carbon-free resources such as variable renewables.

Other factors

A number of additional factors are important here in facilitating or impeding the extent of the cannibalisation effect.

The first is that as more flexible resources — storage, conveyance through wires, demand response — are added to the system, the renewables penetration rate at which market value declines to zero increases. In this way, flexibility slows down the speed of price cannibalisation. We return to flexibility in answering question 2: what policymakers can do.

Other factors identified as having further facilitated price cannibalisation in Germany in recent years include absence of demand growth and low or falling input prices, such as the price of fossil gas. The wider literature identifies other drivers, including fossil fuel subsidies. Finally, note the investment-chilling impact of a given market price reduction is lessened as the levelised cost of renewables falls. We treat these variables as constant for the remainder of the paper in order to focus on levers more directly under the influence of policymakers.

A renewables thing?

When carbon pricing is not aligned with, or trails behind, the desired decarbonisation trajectory and support framework for renewables, the variable nature of some renewables make cannibalisation effects particularly troublesome. This is because variability deprives these resources access to high prices during system scarcity — scarcity will increasingly occur when they are not generating.

In time, investments in flexibility can address decarbonisation effects, but in absence of a carbon price consistent with the required decarbonisation trajectory, it is not evident there will be a signal for the full extent of flexibility required by the system, in turn necessitating subsidy for flexibility.

In sum, price cannibalisation happens when support is used directly to substitute for market mechanisms (carbon pricing) to jam in more renewables and is exacerbated by system inflexibility.

Effects

Absence of a fit and proper price on carbon has been identified as a key driver of price cannibalisation. The main drawbacks and risks of price cannibalisation link to complexity, expense, the crowding out of merchant RES investment, necessity to provide support for flexibility, impeding demand response and misdiagnosis. We go through each of these.

Support policies increase the share of low-emission technologies. When they are added to a system with fossil-fuelled energy, they displace the most expensive resources — thus adding more reduces their average market revenue. Furthermore, absence of a carbon price allows fossil-fuelled alternatives to remain competitive and limits the value available to investments (such as through storage) that can erode the share of dirty energy through the flexing of abundant clean energy. As a result, these flexible technologies need to be compensated outside the market, and the necessity of support becomes self-perpetuating. All of this risks killing off merchant investment in RES.

Reliance on subsidy contributes to complexity, which will grow as innovative new technologies must be accommodated and as other sectors are integrated. This can be compared with carbon policies that raise market prices when fossil-fuelled generators are running, encouraging low-emission generators into the market, leaving investment to the market and thus decentralising the energy transition.

With subsidies, support becomes self-perpetuating — subsidy necessitates more subsidy.

What about expense? Both approaches entail cost: consumers (or taxpayers) finance renewables subsidies and they also pay for carbon pricing. Nevertheless, analysis suggests using subsidy support as a simple substitute for a carbon price makes the transition more expensive because a carbon price drives implementation of the cheapest measures to reduce carbon emissions. This might include efficiency measures in consumption, shifting from coal to fossil gas and a multitude of other decentralised decisions that renewables subsidies alone do not stimulate. Furthermore, mechanisms to finance these subsidies are prone to lobbying and special interests and can lead to outcomes that might be considered unfair. For instance, Germany’s EEG renewables surcharge (abolished in 2022) targeted smaller consumers to subsidise big corporate energy consumers.

Perhaps the biggest concern is policymakers potentially concluding that price cannibalisation means variable renewable energy is hitting fundamental integration limits. This conclusion would misdiagnose the causes of price cannibalisation and would risk derailing the energy transition. Policymakers should not see market value decline under variable renewable support policies as an indication that variable renewable energy is hitting fundamental integration limits.

Policymakers should not see market value decline under variable renewable support policies as an indication that variable renewable energy is hitting fundamental integration limits.