Hi there,

We got a peek into the new interconnection queue for the largest power grid operator in the US. That’s right, I’m diving into PJM’s just-announced queue intake and what it reveals below.

This is Sightline’s weekly newsletter on the moves and motives shaping the load growth era. Not a client yet?

The IRA's cliff-hanging in PJM queue

Late last month, PJM's reformed interconnection queue opened for the first time in years, drawing 811 projects totaling 220GW. This was the first major intake under its new first-ready, first-served cycle approach.

The last queue intake was mostly renewable: 56% solar, 25% wind, 3% gas. This one is 48% gas, and renewables down to 9% combined. 

That's what a tax credit deadline does to a queue. Wind and solar developers have until July 4, 2026 to break ground or lose IRA tax credit eligibility entirely. So this queue tells you what developers are willing to build, and some recent deals hint at who'll own it.

What happened

PJM reopened its interconnection queue with new rules. Under PJM's old process, anyone could file anything, and many did. Projects enter the queue to get studied, but historically most would withdraw before even signing an interconnection agreement. The reformed cycle raises the bar to entry, requiring site control and upfront financial commitments before you get in the door. 

So this time, 811 projects and 220GW got in line, already a filtered number. (But even then, it's seven times oversubscribed against PJM's own 32GW load growth forecast through 2030.) For those projects, here’s the breakdown:

The gas dominance in the queue gives us the first view of the IRA tax credit deadline. The OBBBA, signed 4 July 2025, compressed begin-construction deadlines for full ITC/PTC eligibility on wind and solar to 4 July 2026. After that, the safe-harbor window closes. In a region with weaker solar resources relative to other regions, slower interconnection than other markets like ERCOT, and the largest data center load growth in the country, gas is better positioned.

Mark's take

I keep thinking about two deals.

The first is Constellation's acquisition of Calpine back in January. The second is GIP & EQT's take-private of AES in March.

Together they frame what's happening in the PJM queue, and what to expect over the coming months.

Let's look at the first. Calpine, mostly concentrated in gas generation with some geothermal in California, was taken private by a group of PEs in 2018 for $5.6bn. Then, Constellation bought it in January 2026 for $26.6bn, a 5x exit for the consortium.

Why was it worth it to Constellation to pay this premium? 24/7 generation in PJM is gold for hyperscalers, so Constellation can use the fleet to attract hyperscalers to build in their area and secure above-market PPAs. Having the firm assets is the way to be more attractive than your peers – the play seems to be nuclear in the long-term, but right now it's gas, so snap-up whatever you can. We're seeing this same type of play by Vistra in Texas with the $4.7bn Cogentrix acquisition. And where it's not inorganic growth, it's showing up in the PJM queue – those that want to build, want to build gas, hence the 157 gas projects. And while the identities of the applicants are still confidential, I'd bet the list leans the familiar PJM gas-heavy IPPs (Vistra, NRG, Talen).

But the AES take-private has the same goal – capture above-market PPAs for the AI power buildout – but a different approach. When the deal was announced, I made the chart below to see what GIP/EQT were buying. It all comes back to the safe-harbored pipeline.

Right now, the market is pricing renewables developers as if the IRA cliff has already come. But safe-harbored projects already cleared the deadline. They aren't affected, and just like gas, can command an above-market PPA. Unlike greenfield gas, they come online a lot faster: 2-4 years versus 6+ years once you stack turbine queues, interconnection, and construction. So for PE, the trade is to find undervalued developers, buy them, and get to new revenues ASAP as the portfolio comes online.

For GIP/EQT, the AES deal allows it to get ~15GW of wind, solar, storage online over the next few years, capturing the premium.

But this explains the shape of the PJM queue, with strategics taking the long-term bet on gas generation jostling for position. And those wanting to grab the speed-to-market via renewables are not jostling at all for post-IRA projects, but looking back in the queue and buying up the safe harbored ones.

Over the next few months, expect more deals on both sides. PE will keep surfing the safe-harbored pipeline. IPPs will keep buying up the gas one.

Read only the queue, you see "gas wins." Read the deals, you see who's writing the next chapter, and which side of the cliff they're standing on.

Who this helps:

  • Strategics with hyperscaler relationships and firm-power fleets. Constellation post-Calpine, Vistra post-Cogentrix, Talen post-Cornerstone. These platforms monetize the data-center premium.

  • PE infrastructure funds rolling up safe-harbored renewables. GIP, Brookfield, EQT, ECP, CDPQ. Buying mispriced pipelines at 30-80% take-out premiums.

  • Gas turbine OEMs. GE Vernova, Siemens Energy, Mitsubishi Power. 106GW of intake validates the backlog thesis. 

Who should be nervous:

  • Public renewables developers without safe-harbored pipelines. Multiples have already compressed, and the cliff date is 60 days out.

  • Solar-only and wind-only developers in PJM. 15GW of solar and 5GW of wind against 106GW of gas. The math doesn't pencil without IRA economics.

  • Coal generators in PJM coal-heavy states. 106GW of new gas in the queue accelerates the retirement curve in PA, OH, WV, and IN.

Meter reading (1 May - 7 May)

A quick read on the numbers shaping the market. The capex, the contracts, the regs, all anchored in the so-what.

$424.30/MW-day // MISO's summer capacity clearing price for its North and Central regions, down 36% from $666.50/MW-day a year ago, as new supply outpaced demand growth in its 2026-27 Planning Resource Auction. Solar drove the increase with 12.2GW of accredited solar cleared, up 59% YoY, and capacity reserve margins are above target. After two years of scarcity pricing and reliability anxiety, new supply is catching up. The data center load wave also hasn't fully hit MISO's queue the way it has PJM's.

+33% // Entergy's 2026-2029 capex surge after its Meta data center agreement that we covered last month. The company's industrial deliveries grew 15% YoY in Q1, with a pipeline of 7-12GW of potential data center load. Entergy simultaneously filed with the Louisiana PSC for seven new combined cycle units totaling 5.2GW, plus transmission and battery storage -- the largest single generation filing by a regulated utility in years. This is what happens when a single hyperscaler deal reshapes an entire utility's resource plan.

2.5GW // Blue Energy and GE Vernova’s planned hybrid gas-plus-nuclear power station in Texas, pairing GE Vernova's 7HA.02 gas turbines with GE Hitachi's BWRX-300 SMRs. The model is simple: gas delivers ~1GW by 2030 to generate revenue and serve load growth immediately, then ~1.5GW of nuclear set to phase in from 2032. This is the most concrete "gas bridge to nuclear" project announced to date -- GE Hitachi's BWRX-300 is the furthest-along boiling water SMR in NRC review, and pairing it with proven heavy-duty gas turbines from the same corporate family reduces supply chain and integration risk.

$6.5bn // Fervo Energy's IPO target valuation, filing to raise $1.3bn on Nasdaq at $21-$24/share under ticker FRVO. It reflects what the Cape Station project financing, the Turboden turbine deal, and the Vallourec pipe supply agreement have done for the credibility of the EGS buildout story. The valuation math is aggressive, but it’s a bet on Fervo, the team, and EGS as the advanced geothermal tech to beat.

11% // xAI's Model Flops Utilization (MFU) rate at its Memphis data center supercluster, powered by behind-the-meter mobile gas turbines with Tesla batteries and mobile chillers. (MFU is the measure of how much computing power it gets from its chips.) The temporary solutions have contributed to outages at the site, which pose a problem amid xAI and Anthropic’s new deal for compute. Three of Q1's five biggest data center project announcements say they’re skipping the grid, but off-grid power isn't exactly the easy way out.

$5.7bn // Coatue's debut through Next Frontier, a new powered-land venture that buys sites near large power sources and builds to suit for AI operators. First project: a 430MW campus in New Lebanon, Indiana, with a 15-year Fluidstack lease and Google guaranteeing the downside if Fluidstack exits. Another extension of Google's strategy of backstopping power infrastructure, but a pivot for Coatue, a firm built on software bets.

Explore more Signals on Sightline here.

On the docket

The policies, rulings, and company moves worth watching.

CAISO's Extended Day-Ahead Market (EDAM) goes live with PacifiCorp. The new market coordinates resource commitments across Western utility footprints in the day-ahead timeframe, when most electricity is scheduled, closing the gap that lets excess solar in Nevada go to waste while Oregon’s short. Portland General Electric joins in October, and LADWP, SMUD, and PNM for 2027, bringing the market to ~42% of Western demand. Watch whether NV Energy and Idaho Power also join, bringing together the California core and the Rocky Mountains.

Pennsylvania becomes the first state with a model tariff for large load interconnection. The PUC voted May 1 to approve nonbinding guidance defining large load customers as 50MW+ individually or 100MW+ in aggregate, with cost-causation pricing, tiered collateral, clustered study timelines, and interruptible service options built in. It's nonbinding and utilities still file their own tariffs, but it could be the template, even echoing FERC’s future federal framework. 

NERC's Level 3 alert on sudden data center load losses. The alert responds to a series of "widespread and unexpected" load reductions in 2024 and 2025, where 1,000MW or more dropped off the bulk power system at once. NERC told states to take action, It's the first time NERC has treated large computational loads as a distinct stability problem.

New & upcoming at Sightline Climate

The latest research, features, and data drops on the Sightline Climate platform.

Sightline Climate's Q2 Data Center Outlook is live on the platform. Announced capacity is up by more than 50%. Delays are also way up, mostly because of the trouble at Fermi. The Q1 Outlook was our most sought-after report last quarter, cited in Bloomberg, Axios, and Semafor. Clients can read the update here.

Events

Where the market is meeting, and where to find us

📅 DTECH Data Centers & AI // Scottsdale, AZ, 12-14 May, 2026 // A conference for utilities, data center operators, and technology providers tackling the engineering and operational challenges of rapid MW-to-GW-scale load growth, grid integration, and AI-driven infrastructure. Shout if you’ll be there.

📅 EEI 2026 // Las Vegas, NV, 2-4 June, 2026 // The flagship conference of Edison Electric Institute. Join utilities, regulators, and value chain players for discussions about the swiftly evolving US power sector. Join us on site, and we’ll be hosting a dinner for Sightline clients and friends on 2 June.

📅 SightLive London // London, 23 June, 2026 // Held during LCAW, join us for our flagship London event. This year we’ll focus the discussion on the data center buildout (what else?!) and the pathways to the fastest MW. We’ll send out the official invite in a couple weeks, but mark your calendars!

Attending an event? Connect with our team.

Interested in diving deeper? Talk to our team and leverage the tactical intelligence that hundreds of energy and investment decision-makers like Southern Company, Tokyo Gas, Jefferies, Galvanize, B Capital and others use to stay ahead in the energy transition. 

Keep Reading