Can resource adequacy be attained without defining what is enough?
Texans suffered a catastrophe in February 2021. It occurred for many reasons, including insufficient cold-weather preparations of all systems—electricity, natural gas, water and transportation—for the extreme temperatures experienced. The design of the Electric Reliability Council of Texas (ERCOT) wholesale electricity market was not one of the reasons, but in the aftermath of the catastrophe, with a heightened awareness of all things ERCOT, there has been intense scrutiny of market design. A central question to be addressed is whether ERCOT’s unique energy-only market can provide sufficient capacity to serve Texas’ growing population and its increasing demand for electricity.
Outside of Texas, resource adequacy has typically been accomplished either through regulatory or market mechanisms making sure there was enough capacity to meet the highest annual customer demand, thereby providing confidence of sufficiency at all other times. However, ERCOT has never required any market participant to contribute to resource adequacy or system sufficiency. Load-serving entities (LSEs) are certainly motivated to have supply to serve their load, but this motivation is merely a financial trade-off between the costs to supply their load versus the costs of being short.
Further, the changing economics and preferences for no-carbon, variable energy resources are leading to an increasingly complex and uncertain determination of resource adequacy. Resource adequacy can no longer be focused on the single point in time when customer demand is the highest. The hourly and annual variability of wind and solar output must now be part of the calculation. The problem is further complicated when storage technologies can be used to adjust temporal patterns of both demand and supply.
The theory behind ERCOT’s energy-only market design is that by allowing real-time energy prices to rise to high levels during scarcity conditions, supply will be motivated to make more resources available at times of highest need. If there’s insufficient supply, the system will more frequently be in scarcity conditions with commensurate high prices, incentivizing new supply to be added. It is important to note that in this context, scarcity does NOT mean insufficient supply to meet customers’ electricity needs. It means that operating reserves are at levels lower than desired. Operating reserves are extra capacity that provide protection against unexpected events that result in an inability to meet the demand for electricity. Just like one might choose to wear a belt to keep their pants from falling down, typical operating reserve quantities equate to the extra security of wearing both a belt and suspenders. Having lower operating reserves than targeted does not mean that customers are in imminent danger of having their electricity curtailed. It is cause for market incentives through higher real-time wholesale prices, not societal panic.
Three conditions must exist for an energy-only market to be successful. First, politicians and regulators must accept the high energy prices needed to support revenue sufficiency. Second, the independent system operator must tolerate occasions of scarcity conditions, which are needed to produce high energy prices. Third, market participants must be properly hedged against the necessary occurrences of high prices during scarcity conditions to avoid financial failure.
Unfortunately, all three conditions have not held up in the aftermath of the February catastrophe. The Public Utility Commission of Texas (PUCT) reduced the system-wide offer cap from $9,000 to $5,000 per megawatt hour. The colossal financial failure of the oldest cooperative utility in ERCOT remains unresolved with nearly $2 billion owed to ERCOT. More troubling has been ERCOT’s highly touted implementation of “conservative operations.” The PUCT Chair frequently expresses that ERCOT is no longer relying on “crisis-based operations,” indicating no tolerance for scarcity conditions, which are one of the foundations on which ERCOT’s energy-only model is based. Unfortunately, these conservative operations result in ERCOT taking preemptive action eliminating the very scarcity conditions the energy-only market relies on. More evidence of the PUCT’s reluctance to allow any scarcity to occur in ERCOT is in this quote from Chairman Peter Lake at the Aug. 15, 2022 meeting of the ERCOT Board of Directors.
But of course, as the grid operator, knowing that our job is to keep the lights on, we’re never going to forego a generator and leave a perfectly good generator sitting on the sideline in order to preserve the pure economics of the marketplace, just for the sake of economic theory and allow the lights to go out.
In addition to ERCOT’s new activist operations ensuring fewer scarcity events, the changing supply mix—where 30 percent of energy is now being provided from resources with zero energy costs—creates further pressure on the energy-only market design.
With reduced potential to earn revenue during scarcity events, an energy-only market cannot be expected to provide sufficient revenues to support the new supply necessary to serve the growing demand for electricity reliably. Unless Texas recommits to satisfying the conditions for the energy-only market design to work, it is time to explore alternative market structures.
Any change to ERCOT’s energy-only market design will need to include some type of payment for capacity. Capacity payments could take the form of more robust additional ancillary services, either through the creation of new services or by securing more capacity within the existing services. Ancillary services are simply daily capacity auctions for specialized reserve services, centrally cleared by ERCOT, where ERCOT determines the amount of capacity to be procured for each service. Ancillary services come in multiple forms, varying in the amount of time within which their capacity is to be turned into energy.
If energy and ancillary service revenues are insufficient to achieve resource adequacy, Texas could introduce a centrally cleared, forward capacity auction to ensure sufficient resources are procured to meet a reliability target. Capacity markets come in different shapes and sizes, which carry different performance implications. As with all forms of centralized capacity planning, capacity markets make many administrative assumptions, such as capacity accreditation for different resources, which affect performance. However, capacity markets are highly advantageous in terms of cost, risk and innovation when compared to cost-of-service capacity procurement.
Texas is not alone in the quest for resource adequacy with an evolving mix of generation sources. The designs of ERCOT’s peers in the Eastern United States suppress energy and ancillary service revenues substantially, resulting in a stronger reliance on capacity markets to make up the “missing money.” Many Eastern stakeholders are working on efforts to “up-size” the role of energy and ancillary services markets and “right-size” the role of capacity markets. By comparison, ERCOT is starting from a position of relatively robust energy and ancillary service markets, which means a robust capacity market need not be the only option to solve the missing money in Texas.
A minimalist option for a capacity market is to procure capacity to serve in a backstop role. Backstop capacity would be procured and not used for energy unless absolutely required. Once the desired amount of capacity is determined, it could be procured by ERCOT either via a centrally cleared auction or by selecting the best proposals submitted in response to a request from ERCOT. Of note, a form of backstop capacity will be procured before this next winter. ERCOT is soliciting proposals to provide Firm Fuel Supply Service. The budget for this service is $54 million to procure 3,000-4,000 megawatts of natural gas generation with backup fuel (oil) capability for four months.
The challenge with procuring what could be significant amounts of backstop capacity is whether customers, regulators and politicians will tolerate any amount of scarcity and resulting scarcity pricing while keeping the backstop capacity on the sidelines. If the backstop capacity is used to produce energy at a price lower than that of any other non-backstop capacity, earnings for the non-backstop capacity will be impaired, increasing the likelihood of it leaving the market and resulting in an increased need for backstop capacity.
The third mechanism to improve resource adequacy is to impose resource adequacy requirements on LSEs. This type of mechanism could decentralize procurement activity for all but the largest LSEs. As long as the minimum reliability requirement is met, LSEs would be free to select how to best meet their customers’ needs, allowing for yet another way for competitive LSEs to differentiate themselves.
None of these market design options are new or even that innovative. All were described a decade ago in a report prepared for ERCOT in the aftermath of the last curtailment of firm load, resulting in a lengthy debate about how to ensure resource adequacy.
All U.S. states except Texas include capacity planning requirements that pre-determine a minimum amount of capacity required to ensure resource adequacy. Reliability criteria, including a target capacity reserve margin, are inexplicably missing from the current discussion of ERCOT market design.
As the PUCT evaluates the variety of ways to modify ERCOT’s market design, they should be clear about what level of reliability they are trying to achieve. If they are unwilling or unable to state a reliability objective for the region, the evaluation of each alternative should explicitly describe the reliability differences along with the system costs. Only a careful evaluation of the trade-offs will result in the right outcomes for Texans.