Competitive wholesale electricity markets are caught between a rapidly decarbonizing resource mix spurred by falling clean energy prices and market structures designed around fossil fuels. Markets are becoming ill-equipped to handle large amounts of cheap renewable energy and flexible distributed energy resources.
A research paper series led by Energy Innovation tackles the critical question facing grid managers and regulators: What wholesale market design provides the best framework for reliably integrating clean resources and decarbonizing the power system at least cost?
This two-part opinion series outlines underlying questions about wholesale market reform, and introduces two pathways for markets to evolve. In part one, Rob Gramlich of Grid Strategies LLC and Mike Hogan of the Regulatory Assistance Project propose a solution with a central spot market and active longer-term bilateral contracting between wholesale buyers and sellers outside of the centralized market.
In part two, analysts Steven Corneli and Brendan Pierpont, along with Eric Gimon of Energy Innovation, propose a hybrid solution that adds long-term markets to today's short-term spot markets to support clean energy investment and reliably decarbonize the power system at least cost.
Effective wholesale electricity markets are critical for rapid and affordable decarbonization, due to their track record of efficient and rapid investment in new technologies. But electricity markets will only support massive clean energy investments if they are able to send efficient price signals as decarbonization accelerates.
The scale and pace of investment needed to stop the climate crisis means wind and solar energy, due to their low costs and speed of deployment, are virtually certain to play a major role in future power systems. But the variable output and minimal marginal costs of wind, solar and other forms of variable renewable energy (VRE) threaten the ability of current market designs to send the needed price signals.
For this reason, we propose adding long-term markets, working in tandem with evolving short-term spot markets, to support efficient investment in VRE resources and the other, complementary resources needed to deeply decarbonize power in the next 30 years.
Why high levels of VRE are challenging for today's spot market designs.
The challenges posed by high levels of VRE output and minimal marginal costs are due to the most basic features of today's electric systems and the spot markets designed to work with them.
Maintaining reliability requires the electric system to continually balance, in real time, the amount of electricity generated with demand. Spot markets do this by dispatching the resources with the lowest short-run marginal cost (SRMC) to meet the lowest levels of demand first.
As demand increases, they then dispatch additional resources with higher SRMC on top of those with lower costs, so their combined output matches consumption, and vice-versa as demand falls.
Short-term market prices, in turn, are based on the SRMC of the highest cost resource in operation at any given time — including the very high marginal cost of scarcity when demand does occasionally exceed supply. This allows resources with lower SRMC to receive prices above their own operating costs at least some of the time they are selling power.
The revenues from such periodic high prices are precisely what allow investors and owners of these resources to recover their costs, either directly from the spot market, or through voluntary bilateral hedges (e.g., power purchase agreements, forward sales and swaps) made against spot market prices.
Both these revenue paths depend on short-term market prices and revenues for energy sold at those prices being high enough, over time, to cover costs and provide a return.
Such price levels depend on the electric system itself not having a persistent oversupply of power resources relative to demand, that would eliminate scarcity prices and tend to reduce prices in other hours.
In theory, oversupply with low prices should lead excess resources to retire, and undersupply with high prices should induce new resource entry. The resulting equilibrium, theory suggests, should over time result in an optimal balance.
However, experience shows that market entry occurs with expected prices at full costs, but retirement requires persistent prices below the much lower "going-forward" costs. As a result, oversupplies can last for decades.
Spot market prices that can support cost recovery will be squeezed into fewer and fewer "scarcity hours," when many VRE resources will have no energy to sell.
High levels of VRE interact with these system and market dynamics in new and complex ways.
In hours when VRE is setting market prices — i.e., when deliverable VRE production exceeds demand — prices will be at or near zero. In other hours, when demand exceeds total VRE output, and wind and solar are in good supply, the excess load will need to be met by clean flexible resources such as hydro, battery storage, and by the ability to directly manage load levels through controllable end-use devices (e.g., smart thermostats, EV chargers, and water heaters).
All these resources generally have low SRMC, so prices in these additional hours will be low to moderate.
This combination of near-zero energy price levels and moderate energy price levels in many hours will squeeze most opportunities for high prices and high revenues into the relatively few remaining hours, when total demand exceeds available VRE and moderate cost supply, and must instead be met by much higher marginal cost supply and scarcity (avoided high-value consumption).
This is where the variability of VRE becomes critical.
In a high VRE system, most of those high price periods are likely to occur when there are extended periods of low wind and sunlight. But if wind and solar resources have no energy to sell at the only time prices are high, the market will either fail to incent their development or will incent their early retirement.
This potential market malfunction alone may be sufficient cause to develop an alternative market-based revenue stream to support the high levels of VRE likely needed for rapid, low-cost decarbonization.
Efficient spot market solutions to such imbalances are unlikely due to the complex interactions of VRE and key complementary resources.
Could a spot market's equilibrium process, described above, avoid this problem? If wind and solar aren't getting paid enough during scarcity events, wouldn't the market signal be for more storage to shift excess wind production to higher price scarcity periods with dark, calm weather?
Perhaps, although fossil fuel technologies may be even more attractive for those periods.
But with high levels of VRE, adding enough storage to buy the excess VRE output and resell it during scarcity events could easily oversupply scarcity demand levels and completely eliminate the scarcity prices. Since both storage and VRE have low going-forward costs, such a low-price disequilibrium could last a long time — quenching investor trust and stalling the rapid decarbonization called for by the climate crisis.
The urgency of this crisis calls for alternatives that avoid the risk of both these market malfunctions:
The mismatch of scarcity revenues with VRE production;
Failing to solve the increasingly complex problem of getting the right quantity and mix of VRE resources and key complementary technologies — storage, controllable load, existing clean energy resources, and new controllable clean resources, along with the new transmission and distributed energy resource management tools they need.
Rapidly improving tools to analyze optimal electric system design offer a variety of approaches to design long-term markets that would co-exist with today's short-term markets, avoiding both critical malfunctions. These tools can identify efficient decarbonization pathways, made up of least cost combinations of VREs and complementary clean resources, to augment and replace existing resources over time, while meeting balancing requirements and emissions reduction targets.
Studies increasingly suggest efficient combinations of such resources can achieve high levels of decarbonization at very reasonable costs — potentially at costs well below scenarios that retain large amounts of fossil technologies and deploy less VRE.
Core elements of long-term market designs to avoid these problems.
Each co-author of this article has developed a distinct proposal for such a long-term market.
Each proposal addresses key problems: the risk of prices too low to support high levels of VRE; the risk of insufficient support for the massive investment needed to decarbonize quickly; and the risk of failing to channel that investment into an efficient, complementary resource mix that evolves over time as technologies and costs change.
While each proposal is different, they all have the following core elements:
Long-term contracts with creditworthy central market clearing entities, for clean energy resources whose financing is threatened by spot market price risk.
Ensures ample, low-cost financing opportunities for the entire volume of new, capital-intensive resources needed for rapid decarbonization.
Competitive project selection with voluntary participation by sellers.
Properly allocates risk to developers with strong incentives for risk management, cost management, and project performance.
Coordination with the short-term market.
Additional incentives for new, flexible, and less capital-intensive resources; efficient operation and utilization.
Dynamically scales the long-term market to just the financing and efficient resource mix problems actually encountered by short-term market.
Avoidance of path-dependence and technology lock-in, with strong support for innovation.
Periodic market rounds (e.g., every three years) to "ladder" in new technologies and adapt to emerging ones.
Convergence between long-term market and clean energy/climate policy.
Emerging clean energy system optimization tools help align and inform lowest-cost decarbonization pathways and policies.
Incremental market development and implementation.
Proposals grow organically out of existing regulatory and market practices, without rewriting laws and major short-term electricity market software.
Continued analysis on the nature of market challenges and their solutions is urgently needed.
The three proposals also have some important differences, such as the nature of the products bought and sold in the long-term market, and whether the emerging system analysis tools are used as part of the long-term market, or before it. These and other differences leave ample room for further work, as does their basic agreement on core design elements.
The commonalities and differences point out two fundamental questions that merit serious additional research:
First, are the concerns articulated here about the risks and limitations of short-term markets robust enough to warrant adding a long-term market approach? And, if so, which of the common and different design elements discussed here should be included in that market?