Wholesale electricity markets pay for two distinct things, and conflating them is the most common source of confusion about how the grid is funded. The energy market pays a resource for the electricity it delivers, measured in megawatt-hours, in real time or day-ahead. A capacity market pays a resource for a different commodity entirely: a forward promise to be available to produce power when the system needs it, measured in megawatts of firm capability, regardless of whether it ends up running. The capacity payment is compensation for standing ready; the energy payment is compensation for actually generating.

The reason a separate capacity product exists is a problem grid planners call resource adequacy — having enough capacity on the system to meet the highest expected demand, with a reserve margin on top, at all times. Energy-market revenue alone can be too volatile and too thin in many hours to keep enough capacity online for the few hours of peak stress each year. A capacity market is one institutional answer: it creates a defined, revenue-generating product for availability, so resources can be paid to exist even when they are not dispatched. The Federal Energy Regulatory Commission framed the question this way in convening a technical conference on the subject:

The purpose of this technical conference is to discuss generic issues related to resource adequacy constructs, including the roles of capacity markets in the Regional Transmission Organization (RTO)/ Independent System Operator (ISO) regions that utilize them and alternative constructs in RTO/ISO regions without capacity markets.— Meeting the Challenge of Resource Adequacy in RTO and ISO Markets, source

How a capacity market works

In regions that run one, the grid operator first determines how much capacity the system will need for a future delivery period — typically a year or more ahead — based on a forecast of peak demand plus a target reserve margin. Resources then offer their availability into a capacity auction, and the auction clears at a price that procures the needed quantity. A resource that clears receives the capacity payment and takes on an obligation: it must be available during the delivery period, and it faces penalties if it fails to perform when called. This is why the construct is forward-looking — it secures capacity before the period it covers, giving developers a revenue signal to keep existing plants online and to build new ones.

The reserve margin is the quantity the construct is built around. A grid cannot plan to have capacity exactly equal to its forecast peak, because demand forecasts are uncertain, plants fail unexpectedly, and weather can push load above expectations; the reserve margin is the cushion of capacity above forecast peak that the system targets to keep the risk of falling short within an accepted reliability threshold. The capacity market exists to procure enough capacity to meet that target, and the demand for capacity in the auction is set administratively to reflect it rather than emerging from a willing buyer. The performance obligation is what gives the payment teeth: a resource paid to be available that then fails to show up when the system is stressed typically owes penalties, and in some markets those penalties are sharp, which is meant to ensure the capacity that was paid for is the capacity that actually performs.

The FERC notice makes the central distinction explicit: capacity markets exist in some RTO/ISO regions and not others. Regions without a capacity market still must solve resource adequacy, and they do so through what the Commission calls alternative constructs — for example, requiring load-serving entities to procure their own adequate capacity through bilateral contracts and obligations, or relying more heavily on the energy and reserve markets with administrative reliability requirements. There is no single national capacity market; the construct is region-specific, and the design choices — how far ahead capacity is procured, how performance is enforced, how new and intermittent resources are credited — differ from market to market.

Why it matters for storage and renewables

The capacity-market question has sharpened as the resource mix has shifted toward solar, wind, and battery storage. These resources contribute to capacity differently than a conventional plant: a solar facility's availability depends on the sun, and a battery's depends on its duration. Capacity markets address this through a capacity-accreditation process that assigns each resource a capacity value reflecting how much it can reliably contribute during the system's tightest hours, which for a four-hour battery is tied to how long the peak stress period lasts. A 100-megawatt, four-hour battery does not necessarily receive 100 megawatts of capacity credit; it receives a value reflecting how reliably four hours of discharge covers the hours when the grid is short, which can be less than nameplate if the stress period runs longer than the battery's duration. This is why duration is so often described as the central variable in storage economics — it affects both the energy a battery can shift and the capacity value it is accredited. For a storage developer, the capacity-market revenue stream — payment for being available — can be as important to a project's economics as the energy it sells, which is why the construct shows up in how storage projects are financed and described. A capacity market, in short, is the part of the grid's market design that pays for readiness, and resource adequacy is the reliability goal it is built to serve.