Last May, MIT researchers were the latest to raise alarm about the perceived limitations of solar PV’s continued growth. In short, these analysts propose that variable renewables will depress wholesale prices when they run, thereby limiting their own economic success.
But are these concerns really justified, or do they rely on outdated assumptions about the grid and about electricity markets?
We argue that these critiques, assuming a static grid and unchanging market mechanisms, can be used to make any innovation look bad. However, more integrative assessments of a least-cost, clean and reliable power system of the future will factor in high fractions of variable renewables, along with more-efficient markets (and usage) and new technologies to integrate these resources seamlessly and resiliently.
In this article, we argue that falling wholesale prices is a good problem to have, and that concerns about economic limitations ignore remedies available from supply-side evolution, demand-side resources, and updated market mechanisms.
Understanding the "problems"
There has been increasing concern that variable renewables such as wind and solar may face an upper limit to adoption in the U.S. grid. The argument is that large amounts of variable renewables will create excess supply concentrated at the particular times of day when they produce. The notorious “duck curve” is an example of this -- the duck-like shape of a particular daily demand curve modeled for California’s grid when the production of large amounts of solar photovoltaics (PV) is netted out.
Critics argue that this technical characteristic of variable renewables, specifically PV -- a daily generation pattern that is not perfectly matched with load -- can have economic consequences for all forms of generators, especially the renewable resources themselves. Large amounts of renewable resources can sell a glut of power when it’s available, offsetting production from higher-marginal-cost resources (like gas-fired power plants). Since power prices are generally set by the resources with the highest marginal cost that clear in the market, additional generation from renewables tends to lower market prices.
This “merit-order effect” often decreases revenues for fossil generators. This impact has been particularly dramatic in Europe, where generation from costly-to-run thermal plants during the daily solar peak was formerly very profitable for fossil generation owners. PV has decreased energy prices so much there that the top 10 EU utilities lost half their market capitalization.
However, the merit-order effect also means that variable renewables themselves may also earn lower profits as their adoption rises. A common conclusion is that variable renewables can play only a modest role in power production, marginalized by declining wholesale value at higher adoption levels.
The other half of the thought experiment
Analysts who have put forth these arguments have elaborated only the first half of a microeconomics thought experiment. The problems they hypothesize hinge upon the laws of supply and demand, but omit important aspects of both, drastically overstating the perceived “problems.” Let’s see how.
1) Supply is changing holistically, not incrementally
Many of these thought experiments consider adding just a single supply resource (often solar PV) without considering many of the other supply-side changes happening at the same time. In reality, solar PV, wind, and natural gas are all joining the supply mix in a big way at the same time; the first two are often complementary and the third is dispatchable, so together, they can do a lot to mitigate the “duck curve” often portrayed.
At the same time, retirements of uneconomic assets will provide a countervailing buoyancy to wholesale prices. For example, even though old, dirty plants often have low production costs, they may exit the market anyway due to high costs of compliance upgrades or other fixed costs that erode their profits. The resulting less-abundant supply can cause the marginal supply curve to contract in quantity, leading to higher prices and higher profits for renewables and remaining fossil generators -- unless demand drops too, as it’s doing in the industrialized world.
2) Demand is increasingly flexible, not fixed
Analysts arguing that renewables’ variability will limit their growth often assume perfectly efficient wholesale markets, but unchanged retail markets and fixed demand profiles. This incomplete and asymmetrical treatment ignores the emerging capability to harness the demand side of the equation. For example, people like and respond to time-varying pricing programs, and these programs are starting to roll out at scale.
The electricity demand of many appliances including electric water heaters and electric vehicles is inherently flexible without disrupting the service provided. Furthermore, new business models (from both utilities and third parties) are driving this convenient flexibility by providing seamless solutions, unobtrusively, conveniently, and without requiring customers to become part-time energy traders.
These factors together increase flexibility of demand, an important low-cost resource, and enable what is the most natural response to changing prices in an efficient market where consumers find ways to use and benefit from cheap electricity from wind and solar. In other words, as renewables reduce energy prices during certain times of day, demand flexibility allows customers to shift demand to those times, which will both reduce energy prices at other (peak) times and raise the price paid to renewables during times when they produce the most.
3) Storage makes renewables dispatchable, not variable
Diverse supply and flexible demand will play a big role in easing renewable integration concerns but, to the extent that issues remain, the continuing decline in battery prices and the range of values available from batteries mean that remaining variability issues can probably be addressed at modest incremental costs.
At the retail level, this can lead to increasing self-balancing of distributed generation (we’ve already seen this in Germany and Australia, and it may affect utility business models in the U.S.). At the wholesale level, as variable resources begin to saturate the market, high-priced hours will incentivize developers to begin to look at storage. Already, storage is seen as a near-term replacement for peaking generation, and batteries installed for peaking capacity can also be used to smooth the economic impact of renewables on power prices.
Storage is already a common feature of concentrating solar power (via molten salt), and becoming an increasingly common feature of solar PV. For example, the all-renewable winning bids in the latest Chilean auction for unsubsidized electricity included not just solar power as low as $65 per megawatt-hour in the daytime, but also nighttime solar power -- via thermal or electrical storage -- for $97 per megawatt-hour at night.
With storage, variable renewables become dispatchable, and dispatchable renewables do not have nearly the same merit-order effect as variable ones. To be sure, our recent demonstration that 13 kinds of benefits of behind-the-meter distributed storage can make batteries cost-effective does not necessarily make them competitive with the many other ways to achieve grid flexibility, but similar reasoning suggests an abundant range of options for averting the problems that narrowly constrained models imply.
Whole-system thinking illuminates a path toward least-cost outcomes
Analysts arguing that renewables will economically limit their own continuing adoption generally leave out the considerations listed above -- and more importantly, these arguments are built on incremental thinking, assuming that today’s grid and markets are fixed and only one thing changes (e.g., PV or wind-energy market share). A more holistic, integrative, and accurate analysis would start with the ultimate objectives (reliable, resilient, and least-cost energy services), and promote a whole-system design to get there promptly.
With this perspective in mind, the characteristics of renewable energy that have caused so much hand-wringing -- variable output and near-zero marginal costs of production -- simply add to the list of design considerations for a market design that rewards efficient investment. Given supply diversity, demand flexibility, and emerging technologies like storage, variable renewables are unlikely to face any practical limit to growth even under current grid paradigms and market structures.
Nothing sacred about existing markets
But even if renewables do face adoption limits in current markets, there is no reason we have to keep these markets the way they are. Wholesale power markets are largely a product of historical coincidence, formed out of the paradigms of the last century in which thermal power plants competed only with each other.
Modern market design that reflects the realities and changing resource mix of the 21st century grid, being pioneered in Germany already, can go a long way toward aligning incentives for least-cost resource mixes. Particularly, incorporating behind-the-meter distributed energy resources and flexible loads into energy markets -- as is being done in California and New York -- can bring new capabilities and a refined level of control to the grid.
An integration challenge?
Evolving supply, flexible demand, storage, and updated markets can remove the limits to increasing renewable energy on the grid. These levers can address the common concerns -- and misunderstandings -- about “integration costs” of renewable energy.
For example, a much-hyped recent paper claims that high-penetration renewables must incur steeply rising integration costs. But that turns out to be an artifact of extremely restrictive assumptions in the models used, combined with an assertion that competitive harm to thermal-plant incumbents is an economic cost of the renewables that beat them.
Renewables are here to stay
The “problems” with renewables often brought up by analysts may be real in isolation, but are overstated when the full range of options is considered. Indeed, these are good problems to have: they’re the natural forces of supply and demand acting to send signals to market participants to diversify resource choice, incentivize demand flexibility, and invest in storage and other emerging technologies.
Arguments against wind and solar PV conclude that these resources will need greater subsidies to survive in the “duck curve” era. But instead, we can tap the latent power of supply diversity, demand flexibility, storage, and market design to level the playing field for all resources, rather than clinging to the premises of the 20th century grid.
Protecting the old system is far inferior to enabling the new one so that innovation can flourish, entrepreneurs can thrive, and all options can compete fully and fairly.
Mark Dyson is a manager with Rocky Mountain Institute's electricity practice; James Mandel is a principal at RMI's electricity practice; Amory Lovins is RMI's chief scientist.
This article was originally published on RMI's Outlet blog and was reprinted with permission.