Inside PJM’s $555/MW-day Reliability Backstop Plan for Data Centers

In this episode, we dive into the labyrinth of PJM's latest regulatory moves to address the insatiable electricity demand from data centers and the looming generation supply shortfall. We break down the newly approved Reliability Backstop Procurement (RBP) plan, which will allow utilities and large loads to secure capacity directly from suppliers through a one-time bilateral auction starting in the 2028-2029 delivery year.

Key topics covered in this video:

The New Auction Rules: Learn how load-serving entities will secure contracts spanning 2 to 15 years with a volume-weighted price cap set at $555 per megawatt-day.

The Price Jump: Discover why this new price cap is considerably higher than the current $325/MW-day limit, and how the sequential auction formula is designed to keep generators from getting too greedy and gaining immense leverage.

The Supply Chain Reality Check: We explore PJM's recent white paper, "Powering Reliability Through Market Design," which reveals that the estimated Cost of New Entry (CONE) for new gas-fired turbines has nearly doubled due to overwhelmed global supply chains.

The Risk of Capital Flight: Could capping prices at $555 simply push energy developers to invest in supply resources in other states or countries with better revenue opportunities?

The Future of Energy Markets: We discuss whether this crisis will eventually push PJM toward a Texas-style approach with no capacity market at all, and examine the irony of calling today's highly complex energy landscape "deregulated"

Transcript

Why This Regulatory Fight Matters

SPEAKER_00 0:02

Okay, I'm afraid I'm going to do it to you again. Walk you through a labyrinthian regulatory thing that you probably want to ignore, but you probably shouldn't, because it may eventually affect you in this insane era of insatiable data center electricity demand, mobile capital, and finite energy supplies.

PJM’s Reliability Backstop Proposal

SPEAKER_00 0:21

Last week, PGM stakeholders, including the electro distribution companies and the data center coalition, OK'd a reliability backdrop procurement plan to address growing data center demand and an anticipated generation supply shortfall in two years. This approach, if approved by the PGM board and then subsequently by the Federal Energy Regulatory Commission, would see utilities, other load-serving entities, and perhaps the data centers themselves requesting PGM to access a specific quantity of capacity in a one-time auction. The load-serving entities would then bill the large loads for that capacity with average weighted prices capped in at $555 per megawatt day. The new approach would start during the 2028-2029 delivery year with contracts, bilateral contracts from 2 to 15 years. PGM had created these so-called critical issue fast path groups to consider options for this procurement process. So this is now the fruit of that. By the way, the normal BRA, base residual capacity auctions, the one for 2028-2029, ends July 7th, which happens to be today, and results will be out shortly. But this data center capacity issue has been at the core of concerns for quite some

Who Registers Load And Pays

SPEAKER_00 1:40

time. And this proposed reliability backstop procurement package, the RBP, includes these bilateral auctions that allow large loads and suppliers to directly connect for capacity with specific negotiated terms and conditions to be decided among the parties. It also ensures that any related costs, quote, are assigned to the cost-causing parties, unquote, so that other ratepayers aren't negatively impacted. Since the load-serving entities, generally utilities, are legally and contractually obligated to supply these new large loads, they're responsible for finalizing procurement targets and bearing related financial and credit obligations. To develop the procurement target quantities, a large load registry will be established. Either those large loads or the associated energy providers would report the identity in megawatts associated with each one of these large loads.

How The $555 Price Cap Works

SPEAKER_00 2:33

The pricing is where it gets weird. With a proposed volume-weighted average cap that uses the estimated cost of new entry that's been administratively decided by PGM for the 2028-2029 delivery year. In other words, what would it cost to bring new gas for its supply on in about two years? And then there's a cap that, with a little bit of a formula, comes up with a $555 per megawatt day on that. Ding ding-ding. Note that this number is considerably higher than the administratively capped price of $325 per megawatt day that was put in place after Governor Josh Sapiro got involved with PGM and then PGM got approval with a FERC. That happened for the past two auctions. It's happening for the one that's taking place for 2028 that's ending this week. And it will also impact the following option for 2029 and 2030 delivery year. After that, all bets are off. This $555 per megawatt day one, as noted, is based on a formula that uses PGM estimates for that new supply. And this is where it also gets a little strange. PGM will commit the resources by ranking the offers by volume-weighted prices in dollars per megawatt day, average across the contracted terms based on the resources in service dates. It would then accept each forthcoming offer sequentially until the procurement capacity target is met or that 555 per megawatt day price cap is exceeded.

Walking Through The Auction Example

SPEAKER_00 4:00

So PGM uses an example of 1,000 megawatts and says, okay, let's assume each project is the same duration. So offer one comes in for 250 megawatts at $475 per megawatt day. Offer two comes in at 250 megawatts at $500. Offer three, 250 megawatts at 525 per megawatt day, and offer four, 250 megawatts at $600. That last offer would then lead to a weighted average of $525 per megawatt day, so all resources would be accepted. But what, in my example, if the last offer were to come at $725? Well, that would then result in a weighted average price of $556 per megawatt day, which would exceed the $555 cap. So the auction would then exclude that final offer and it would end. The reason for this price cap is to keep developers from obtaining immense leverage in this one-time auction and pushing prices even higher. As such, it's meant to enforce some competitive discipline so generators don't get too greedy. That matters in a world in which estimates of potential revenue for a gigawatt of data center compute capability are as

The Risk Of Not Enough Offers

SPEAKER_00 5:13

high as $15 billion. At the same time, though, as I've argued before, caps may function like a billionaire's tax in which the rich simply get up and leave a state for a location with a better tax regime. If there's no cap and higher perceived revenue opportunities in another state or country, why not simply invest in your generation there? And here's another challenge. PGM's recently issued 70-page white paper entitled Powering Reliability Through Market Design explicitly notes that the administratively established cone, the one we're talking about and upon which these price formulas are based, is significantly below the empirically observed prices for new turbines. And PGM notes that a recent report of actual projects under development show combined cycle prices, for example, have as much as doubled over the past five years. So a project delivering in 2026 versus a project delivering in 2031 can be twice as much in the latter years. In large part, that's because the tsunami of rapidly growing forecasts of data center demand has simply overwhelmed global supply chains that can't keep up. Since the reliability backstore procurement projects will come online in 2028-29, the RBP prices and caps may not yield many suppliers willing to offer capacity at those prices. With GE Vernova and other manufacturers reporting turbines in oversold situations through 2030 in some cases and even beyond, there may be no magic bullet here, at least not at those $555 per megawatt

Signals For Future Capacity Markets

SPEAKER_00 6:42

day cap prices. But this exercise will provide a very useful real-world example of what's out there and at what price. And it'll send a very strong example to PGM ratepayers as to what capacity prices may look like in the years after the current $325 per megawatt day cap expires, if indeed there is a future capacity market at all. One thing the PGM white paper sets out for consideration is the adoption of a Texas-style approach under which there is no capacity market at all, and energy prices send the scarcity signals and are allowed to soar.

Are These Markets Really Deregulated

SPEAKER_00 7:15

The PGM board will issue a letter to its members about the RBP with details of the final plans before filing a proposal at FERC. This stuff clearly gets more complex with each day as market operators and regulators put one complex and detail laden patch on top of the next, which makes me wonder why on earth anybody continues to call these markets deregulated. They are, in fact, more complex and regulated than ever before. Well, thanks for watching, and we'll see you again soon.