© forrest9/iStock/Getty Images Plus
We wrote yesterday about proposed capacity market reform within PJM, the regional transmission operator (RTO) that serves 13 states and Washington, DC. PJM recently submitted a proposal to the Federal Energy Regulatory Commission (FERC) to alter its capacity markets in order to address low clearing prices in the presence of state subsidized resources. This proposal raises several concerns, including substantially higher costs for capacity through higher clearing prices and additional payments to unnecessary units that did not clear the market (and therefore do not provide capacity) but are used to determine the clearing price.
The suggested changes would create incentives for building and maintaining more capacity in a system that is already flooded with excess capacity. PJM’s Summer 2018 Reliability Assessment reported that the 2018 forecasted reserve margin was 28.7% despite a required reserve margin of only 16.1%. In particular, the proposal gives us pause because it attempts to provide a patchwork, temporary solution to a longer-term challenge: integrating renewable projects into wholesale markets. In no way does PJM’s proposal actually get the market closer to what the long-term solution should look like.
FERC claims that out-of-market subsidies threaten the “integrity” of the competition of the capacity market. Yet most of the state policies behind these out-of-market incentives are used to correct for environmental externalities and encourage carbon-free electricity generation, either through Renewable Portfolio Standards (RPS) or Zero Emission Credit (ZEC) programs keeping nuclear plants online. Arguably, in the absence of federal or state carbon pricing, these policies are an attempt to internalize the environmental costs associated with certain types of generation and are thus addressing an important market shortcoming. Although they are certainly not a perfect substitute for pricing carbon, these efforts theoretically help move the market toward a more economically efficient outcome.
The issue PJM faces may therefore not be that out-of-market subsidies are distorting the market, but that the capacity market as it currently exists is not designed to properly accommodate intermittent generation units that have low marginal cost. Capacity markets used to be comprised of nearly identical products (non-intermittent resources like coal, nuclear, gas, and hydro) competing to provide the same service. Today, the units competing in the capacity market are not identical. Renewables offer value in different ways, including environmental benefits and in some instances modularity that allows generation to be located close to load. But they do not offer the same level of reliability assurance that fossil-fueled plants do, and their variable nature poses challenges for determining their capacity value (which can vary with the season and level of penetration in the market). Solar technologies especially have a tendency to become less valuable as penetration increases because their peak production occurs only during certain hours with diminishing marginal value (see NREL’s brief and a piece by James Bushnell for more information).
These concerns typically lead to a highly conservative approach to valuing existing renewable capacity, especially for wind. PJM currently calculates the effective load carrying capability (ELCC) of wind energy on off-season production (i.e., summer production for wind)—a practice that can vastly underestimate its value. The RTO also recently reduced its ELCC for wind from 13% to 8%. This means that only 8% of the region’s nameplate wind capacity is counted in the auction, even though the actual capacity provided is much higher. For example, in 2017 PJM accounted for only a little over 1100 MW of wind capacity in the auction even though wind power actually contributed 3100 MW during the “Bomb Cyclone,” an extremely powerful winter storm.
Seasonal capacity markets are one way to improve the accounting of renewable capacity by accommodating their strong seasonal fluctuations in production and availability. PJM has partly begun trying to do so by enacting the Seasonal Capacity Performance product as of the 2020/2021 delivery year, which will allow summer-only resources to pair with winter-only resources and collectively bid into the capacity market. A recent report by the Brattle Group analyzing this option, however, claims that PJM undervalues and excludes a high percentage of the seasonal resources available (such as solar, wind, energy efficiency, and demand response) by requiring resource matching in both seasons despite the summer peak being far greater than winter.
Under this construct, solar capacity that produces more in the summer cannot provide summer seasonal capacity unless bidding with a corresponding amount of winter capacity. Consequently, the Brattle Group report suggests that many available resources do not bid or will only bid into the annual markets where they are valued according to their off-season availability at peak hours, paradoxically leading to undervaluation of resources in a market that is still attempting to prop up existing, uneconomic generators in the name of supposed capacity needs. Taking full advantage of these existing capacity resources could also reduce ratepayer costs for procuring capacity.
Even if the value of renewables is accurately represented, however, their presence in capacity markets as currently structured presents a disconnect between the short- and long-run market incentives. In the short run, low marginal cost renewables supported by state RPS policies competing in the capacity market drive down capacity prices. These low prices will discourage capacity build-out, which is beneficial in the short-term since the region is already flooded with capacity. But looking far into the future, we will likely still need long-term investments to maintain a reliable grid. In the long term, these low capacity prices could potentially retire too many plants and ultimately deter long-term investment because—unless capacity prices rise sufficiently to reflect increased scarcity—potential merchant plant investors will not have incentives to build. PJM, like all RTOs, will face the challenge of maintaining a sufficient level of long-term investment and resource reliability in a flawed capacity market, and the solution may be to find alternative ways of doing so.
These issues reveal ways wholesale markets might evolve in order to capture the value of state-sponsored resources and still maintain a reliable grid. A clear distinction exists between capacity needed to meet regional energy needs reliably and resource flexibility. Existing renewable energy generators do contribute significantly to the capacity needs of the region and are underrepresented in this value, especially on a seasonal basis. These resources are typically unable to provide ramping service, however (the ability to instantaneously adjust electricity output when called upon). As penetration of intermittent sources grows in capacity markets, ramping could become an increasingly important ancillary service. Rewarding that service could be an effective method of maintaining system reliability and could help encourage long-term investment.
In California, the state’s grid operator has attempted to do just that by creating the Flexible Ramping Product (FRP) in 2016. The FRP procures and compensates units that can ramp up or down on a “5-minute Real Time Dispatch” market for being available to do so. California notably does not have a capacity market and uses this product along with resource adequacy requirements for utilities as methods of managing their very high penetration of intermittent capacity and encouraging reliability.
The evolving US electric grid will require changes in market design to best account for the value that new resources offer. The US Energy Information Administration’s Annual Energy Outlook predicts that solar and wind will grow collectively by 147 GW between 2020 and 2050 in the reference case (notably without a carbon policy), a two- to three-fold increase, and markets will need to find a way to accommodate these and other resources.
How this can or should be done in the current construct is unknown. Will updates to the existing capacity market, such as better designed seasonal markets, be sufficient to integrate these resources? Or will major changes be needed—such as eliminating the capacity market altogether and relying solely on markets for energy and expanded ancillary services, or, alternatively, focusing on long-term contracts? RTOs will need to explore new options in order to continue procuring necessary capacity while simultaneously efficiently incorporating zero-carbon sources of electricity.