Low-Energy Fridays: Everything You Wanted to Know About PJM’s Capacity Auction But Were Afraid to Ask (Part 2)
Last week, we reviewed some of the background to help us understand PJM’s recent 2028/2029 electricity generation capacity auction. Today we get to the meat of the matter. For those readers who loved math in school, congratulations. For those who, like Chevy Chase, feel somewhat differently, we will try to be gentle.
The capacity auction had two main results of note. First, the auction “cleared” at the system price cap of $325/megawatt (MW)-day (without the cap, power demands would have resulted in higher prices for average ratepayers). Second, there was a 6,831 MW shortfall against PJM’s “reliability requirement” (meaning that PJM has less in reserve power than it would like). In addition, the Federal Energy Regulatory Commission (FERC) convened a technical conference yesterday with the purpose of evaluating potential reforms to how PJM governs itself. Each of these deserves to be unpacked.
Let’s start with the clearing price, and the cap we referenced in last week’s LEF. As economists have told us, when there is a shortage of any good, you should expect prices to go up. After decades of flat electricity demand throughout the United States, and specifically in PJM, the 2025/2026 auction saw demand finally reflect increasing power needs. As a result, the PJM auction cleared nearly 10 times higher than the previous year (going from $28.92/MW-day to $269.92/MW-day).
Every auction from the 2026/2027 through the 2028/2029 has hit the price cap championed by Gov. Josh Shapiro (D-Pa.). With every capped auction, PJM’s reliability shortfall has widened, which instead of lowering costs, has masked economic reality and broken the market mechanism designed to bring new power generation online. This is because without higher prices, the auction is failing to signal that demand is high enough to incentivize much-needed investment to meet growing power needs.
PJM’s Independent Market Monitor has determined that, over the last four auctions (2025/2026 through 2028/2029), data centers were responsible for 46.2 percent of the increase. The remaining increase resulted from electrification (heating and transportation), reindustrialization (electric arc furnaces), and organic load growth (new home construction). All these new developments require more generation resources to maintain reliability in the future.
Now onto the capacity shortfall: PJM procured less capacity in this auction than its reliability requirement calls for. A reliability requirement is the mandatory amount of excess capacity a grid operator must keep on hand. That means when the system faces extreme conditions, the lights stay on. By falling below the reliability requirement, less capacity will be available in case of an emergency. And context matters: PJM was seeking to have 20 percent reserves after last week’s auction. Instead, they got 14.7 percent (coincidentally this amount was what they were aiming for in the 2024/2025 auction and they exceeded it). The difference between what they wanted (20 percent) and what they got (14.7 percent) is 6,831 MW—which is roughly six or seven conventional nuclear reactors. Which is to say, this is not something that can be built within the 23 months between the time the capacity is purchased and the time it’s expected to be used.
However, this doesn’t mean the sky is falling. To explain, I need to introduce two concepts: installed capacity (ICAP) and unforced capacity (UCAP). For the sake of simplicity, think of ICAP as the maximum output a power plant can make, and UCAP as the statistically likely output of a power plant based on historical performance of similar facilities. For example, a 1,000 MW ICAP nuclear plant equates to a 960 MW UCAP. This is because, while the plant is capable of producing 1,000 MW under ideal conditions, things happen, and it won’t realistically operate at all times. So while PJM does have a capacity shortage and it is evident more capacity is needed, there is or will be more than 200,000 MW ICAP connected to PJM’s system in 2028 to keep the lights on.
Despite the issues raised by last week’s auction, PJM is often seen as the model for wholesale electric markets and, with that, comes criticism of PJM’s governance model. At yesterday’s FERC technical conference (which included insights from R Street’s Kent Chandler and was relevant to prior analysis from Michael Giberson), attendees discussed potential reforms that could improve PJM’s ability to effectively address its perceived shortcomings. The central takeaway was that PJM’s governance issues do not require top-down intervention from FERC.
As Kent noted, previous FERC interventions have only made it more difficult for PJM to address key issues. And, as Kent explained, if FERC does intervene, it should “… take efforts to ensure any governance amendments, formal or suggestive, that arise from this technical conference serve to enhance RTO decision-making and market efficiency, rather than diminish them.” However, the PJM Board of Managers, the PJM states, and the PJM stakeholders must take action to submit durable, self-driven reforms to the commission. This will maintain PJM’s leadership as a Regional Transmission Organization (RTO).
The price for capacity in the 2028/2029 auction may be higher than before, but it is also being driven by real load growth, which is good because it signals a growing and prosperous economy. But it is also driven by limits holding back new supplies, which isn’t good. The capacity shortfall is a problem, but all markets require time to respond to deficiencies, and it would be a mistake to return to vertically integrated utilities, which would only lead to inefficiencies and harm consumers at a time when the country is grappling with affordability.
Instead, PJM can continue to facilitate the interconnection of new capacity backed by independent capital that stands ready to meet the challenge. Yes, there are governance challenges to overcome, but the PJM community can achieve those without outside interference. Let the market work.