You’re out of free articles.
This Halloween, get 20% off using code ADVANCED
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Sign In or Create an Account.
By continuing, you agree to the Terms of Service and acknowledge our Privacy Policy
Welcome to Heatmap
Thank you for registering with Heatmap. Climate change is one of the greatest challenges of our lives, a force reshaping our economy, our politics, and our culture. We hope to be your trusted, friendly, and insightful guide to that transformation. Please enjoy your free articles. You can check your profile here .
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Subscribe to get unlimited Access
Hey, you are out of free articles but you are only a few clicks away from full access. Subscribe below and take advantage of our introductory offer.
subscribe to get Unlimited access
Offer for a Heatmap News Unlimited Access subscription; please note that your subscription will renew automatically unless you cancel prior to renewal. Cancellation takes effect at the end of your current billing period. We will let you know in advance of any price changes. Taxes may apply. Offer terms are subject to change.
Create Your Account
Please Enter Your Password
Forgot your password?
Please enter the email address you use for your account so we can send you a link to reset your password:
But there’s still plenty of room for regional grid operators to set their own rules.

Almost eight months have passed since the Federal Energy Regulatory Commission was tasked by the Trump administration with conjuring up with new rules to help speed up interconnection of large loads without increasing retail electricity costs. On Thursday, FERC finally responded with “major reforms,” in the words of Chair Laura Swett, putting the onus on America’s restructured electricity markets — PJM Interconnection, Midcontinent Independent System Operator, Southwest Power Pool, California Independent System Operator, ISO New England, and New York Independent System Operator — to figure out how to implement their suggested solutions.
Using what’s known as “show cause” orders, FERC presented those in charge of these electricity markets, known as regional transmission organizations and independent system operators, with what was essentially a menu of ideas that have been percolating in electricity policy circles since the rise of data-center-driven load growth has started putting pressure on the existing grid and told them to get to work. Secretary of Energy Chris Wright’s original “advance notice of proposed rulemaking,” published in late October, was more proscriptive and specific, whereas FERC essentially said to regional electricity markets, “do whatever you have to, just make it work.”
In a brief email, former FERC chair Neil Chatterjee described this as “a very FERC-y approach!” Or as Gretchen Kershaw, the chief operating officer of Grid Strategies and a former FERC legal advisor, explained to me that “it’s much faster to act on a region-specific basis instead of going through a full notice and comment rulemaking process.”
The commission’s proposed reforms fall into five categories:
1. The markets need “clear transmission service application and study rules” for large load customers seeking to connect to the grid, Swett said in her remarks. The commissioners specifically called out the use of “grid-enhancing technologies” to expand the capacity of America’s existing electricity infrastructure — things like reconductoring, which adds transmission capacity along existing wires, and dynamic line rating, which adjusts capacity based on local weather and conditions. “The cheapest transmission line is the one that already exists,” Commissioner David Rosner said, speaking after Swett at Thursday’s meeting.
2. The RTOs and ISOs will also have to show that they have “adequate safeguards against cost-shifting or take steps to create them,” Swett said. This will require “cost recovery agreements,” Rosner added, “which are designed to ensure that large loads pay their fair share of the costs incurred to serve them, regardless of whether the large load comes online as planned.” In other words, “If new infrastructure is built to accommodate a data center, and that data center doesn’t show up, residential customers are not left on the hook to pay the costs,” he said.
3. The third area that the electricity markets will have to address is co-location and behind-the-meter power, specifically coming up with rules that facilitate purpose-built generation facilities to support new large loads. This would allow data centers and big power users to be less of a burden on the grid, thus requiring less in the way of grid upgrades and additional costs that would be borne by all ratepayers.
4. The orders tells markets “to prove or develop new transmission services to reflect large load flexibility,” Swett said. Load flexibility is another idea designed to lower the system cost of data centers. Grids have to be built out to accommodate the peak demand of the system, but with flexibility, data centers could shave off how much power they demand during, say, a hot summer day, thus lowering that demand peak. To get there, however, they need to be properly incentivized. FERC is telling the RTOs and ISOs to come up with rules that would allow large loads to come online without necessarily requiring vast new buildouts of grid infrastructure and generation. “Legalizing flexible transmission service options for more large load customers can speed interconnection, avoid constructing unnecessary transmission upgrades, reduce strain on the grid, and make power bills cheaper for everyone,” Rosner said.
5. Finally, the orders will require the markets to come up with rules and procedures for generation that’s “proximate” to new load. This will encourage “bring your own new generation,” Rosner said. That stands in contrast to proposals requiring or encouraging new large sources of demand to place generation on their own premises. “Literal co-location is not the only way to facilitate faster, more efficient, and more cost-effective connections to the grid,” Rosner said.
The markets will have to come back in a month to explain how they “intend to ensure that adequate generation will be available to serve existing and new large loads,” a FERC staffer explained at Thursday’s meeting, then again a month later to explain either how their existing rules conform to the new requirements or how they plan to charge their rules to do so.
The commission’s decision is not a formal rulemaking. Instead, the commissioners argued that tasking each RTO and ISO with specific orders would result in a more tailored set of reforms. “Today we’re engaging those to act with more speed, more durability, and more precision than we would get with our proposed rulemaking,” Commissioner David LaCerte said.
The action was strikingly bipartisan, with Democratic and Republican commissioners approving it in a 5-0 vote. It also won plaudits from clean energy and environmental groups. The Sierra Club said in a statement the action was “responsive to Sierra Club’s requests on several fronts,” while the clean energy trade group Advanced Energy United lauded the orders as “potentially creating much-welcome regulatory certainty and transparency, as well as some safeguards to ensure that co-location won’t negatively impact the electric rates and system reliability of all other customers.”
Federal energy regulators have been mulling these reforms as the Trump administration and state and local government officials have grown increasingly restless with rising electricity prices, utilities, and data center developers. Swett herself has scolded America’s largest electricity market, PJM Interconnection, for its inability to meet its own preferred level of excess capacity to ensure it can maintain continuous service, as well as continual high capacity costs, which have translated into tens of billions of dollars of added costs for electricity customers in the mid-Atlantic. Swett has even gone so far to suggest that PJM “ simply has grown too big to function,” leading some market observers to speculate that a forced breakup may be nigh.
Electricity prices nationwide have risen 5.3% in the last year, according to the Bureau of Labor Statistics, while overall prices were up 4.2% — a number that includes gasoline price increases stemming from the war in Iran. In PJM territories like New Jersey, average bills have increased from about $91 to $140 over the past five years, while prices are up some 52%, according to the Heatmap-MIT Electricity Price Hub.
The existing rules, Swett said, are “unjust and unreasonable because they do not adequately address how to integrate large and co-located loads onto the transmission system.”
“Free-riding on other customers is not an option,” she added.
This Halloween, get 20% off using code ADVANCED
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
SoftBank’s error and Nvidia’s scandal suggest data centers may have already peaked as an economic and energy trend.
This is Heatmap Daily, a weekday digest written by our executive editor. Subscribe here.
Covering energy and climate change sometimes means seeing certain stories in a different light than others. And a particularly vivid hue illuminated The Wall Street Journal’s report today that the Japanese conglomerate SoftBank Group is taking a $3.7 billion loss on several gigawatts’ worth of natural gas turbines.
For the Journal, the big news is that SoftBank and its billionaire founder, Masayoshi Son, are failing to capitalize on the artificial intelligence boom. But at Heatmap, we’re thinking about the turbines.
The data center boom and the return of electricity load growth has depleted the world’s supply of natural gas turbines. SoftBank had purchased reservations for about 13 gigawatts of turbines, according to the Journal. That’s a staggering amount of electricity — it could meet New York City’s maximum power draw on the hottest part of a summer’s day. The conglomerate now seems set to dump reservations for about four gigawatts’ worth. Keeping a reservation, it turns out, really is harder than making it.
Most of those rotors were reportedly from GE Vernova, one of the world’s largest producers of gas turbines. GE Vernova’s order list for gas turbines has been filled for nearly two years and the waiting list now stretches into 2031. That has led data center developers, hungry to get power as soon as possible, to reach for ever dirtier forms of electricity generation, including diesel generators and even old jet engines.
On that front, SoftBank’s failure to sell the turbines to a higher bidder suggests (albeit weakly) that demand for energy-producing assets might not be as strong as the AI boom’s backers once thought. SoftBank was already set to send about 9 gigawatts of turbines to the enormous Trump-backed OpenAI data center in Pike County, Ohio. The roughly remaining four gigawatts will be released back their manufacturers. (One thing this news can’t quite tell us, alas, is what type of turbine set-up will be used in the Pike County facility, which will be the world’s largest gas-fired power plant if it is build to completion. Recent public filings have suggested the facility isn’t planning on using any turbines in the more efficient combined-cycle configuration, which would mean the plant’s carbon intensity will be worse than most other new natural gas plants. But we’d love to hear any tips on this front.)
To zoom out for a second, I’m struck by how the AI bubble has been replaced by an AI muddle. As I wrote this, news emerged that Nvidia is preparing to fire its CEO, Jensen Huang, for his affair with a Chinese woman who is alleged to be a spy. (He had at least one more affair with another Nvidia executive, according to The New York Times.) Under Huang’s leadership, Nvidia has invested in a sprawling and sometimes circular network of energy-related companies, including Emerald AI, a startup that helps data centers flex their electricity consumption, and Redwood Materials, a battery recycler and producer led by the former Tesla executive J.B. Straubel. Huang has used Nvidia’s enormous balance sheet to buy down the risk of other AI and AI-related companies, propelling the overall AI trade to become the biggest story in American markets.
Now that balance sheet is in peril: Nvidia’s stock is already down 11% over the past five days, and it has fallen another 8% in after-hours trading since news broke of Huang’s ouster. If the chip designer’s financial network began to fray — or even if the next CEO had a lower risk appetite than Huang — then its demise could impact a much larger set of firms.
Even before the Huang scandal broke, companies at the center of the AI trade were showing signs of weakness. Alphabet, Meta, and Microsoft are all down on the month. Nvidia was already lagging by about 10% on the month before its hebdomas horribilis began, too. The memory and chip makers Micron and Applied Materials are both down more than 25% in about the same period, and GE Vernova and Oklo have fallen nearly 20%. Although the data center company Coreweave rebounded on an earnings beat today, it’s still down 24% on the month and nearly 50% since May.
The market rout isn’t total — Anthropic seems likely to enjoy one of the largest IPOs ever in the next week (although perhaps not with quite as large a valuation as its investors once imagined). And the underlying technology of AI seems to be getting worryingly better all the time. But even though the politics of data centers and AI seem set to dominate the new Washington, I’m wondering if data centers as a mainstream economic trend might have already peaked. Americans voted for a more subdued AI boom in the midterms. Now it seems like they might get it ahead of schedule.
Lunar Energy’s latest AI-powered system aims to integrate EVs into the automated demand response mix.
In theory, it’s straightforward to assemble the smart energy home: Solar panels, home backup batteries, and electric vehicles, plus smart thermostats, heat pumps, and water heaters — all the individual pieces can be had for the homeowner who can afford the up-front cost. The challenge is getting all those pieces to talk to one another.
Bay Area-based Lunar Energy is one of the tech firms that’s been trying to crack this problem. Today, it announced an upgraded version of its artificial intelligence-powered home energy control system, which not only integrates all those separate systems but also responds tactically to energy markets. CEO Kunal Girotra told me Lunar AI saves its users money by monitoring their home 24/7 and making decisions to charge, not to charge, or discharge so that they save the most money.
“Utilities are all moving to almost-real-time changes of rates,” he said. “You need software to handle that for you. I think that the race is on for who has the best AI machine learning algorithms to save the most amount of money. And so far, we seem to have a head start.”
Lunar’s first-generation system focused on generation, storage, and control. In other words, it connected elements like a home’s rooftop solar panels with the backup battery the residents might keep in the garage, using the AI to decide when to meet the home’s real-time energy needs with solar versus grid energy versus stored power. What it didn’t have was the ability to work with electric vehicles to manage charging and use the car’s battery as a bonus source of energy.
To fix this, Lunar has built its own bidirectional EV charger — one that can charge up the battery or discharge the car’s battery for a secondary use, like backing up the home’s power supply. Girotra, formerly the head of Tesla Energy, told me the problem with that functionality to this point has been the connection point between the car and the home electrical system. Carmakers and other companies in the space didn’t have the deep experience with electrical certifications to make the process seamless, he tells me. That led to setups like the original Ford Charge Station Pro, which required several different wall-mounted electrical boxes to connect the F-150 Lightning EV to the home.
Lunar’s charger is built to connect to its customers’ existing home Lunar energy system, which includes a microgrid interconnect device they call Lunar Bridge that islands the home in case of a power outage on the grid. In that way, the power stored in the EV battery is ready to back up the home in case of a blackout without any additional work — or to be optimized as part of a fully smart home that’s making money as part of a virtual power plant.
The company’s setup allows for another fun feature: “slow” charging that’s not slow at all, but actually two or three times as fast as the Level 2 charging most EV owners use to recharge their batteries overnight. Electric cars in the U.S. can typically take on only about 11 kilowatts of AC power from one’s home because of the limitations of their on-board hardware. Because Lunar’s system is fully DC, however, Girotra said the company decided to build a DC home charger, one that — like a Tesla Supercharger or other DC fast chargers out in the world — isn’t limited by that power cap. Lunar said it could use a combination of power from the grid, home solar, or home batteries to charge an EV at up to 25 kilowatts. That’s enough to recharge an EV in three to four hours rather than six to eight.
“If the standards get high, and if everybody sees cars can get charged in three hours, then all of a sudden people are like, ‘Wow, this is how I want it,’” he said. “Especially for 25% of the users who are like me, who always are running low on charge and they don’t mind a boost every now and then.”
Some of the other improvements in the second-gen system are more incremental. An inverter that’s 40% more powerful helps Lunar get its system up to enterprise scale. That’s crucial, as the world moves toward a place where more homes and businesses employ battery backup systems. Lunar’s new batteries come in modular 5 kilowatt-hour blocks ranging from 10 kilowatt-hours to 30 kilowatt-hours in a single unit and scale up to 300 kilowatt-hours for home or small commercial use; there’s also new pad-mounted configurations with capacity up to 1 megawatt-hour.
Batteries, though, are the simple part. Everyone from EV makers like Rivian to utilities like Octopus Energy is racing to make a person’s or a business’s batteries and solar panels and electric cars work for them, using the power of AI to account for moment-to-moment changes in energy prices and supply. Lunar’s new tech makes the case that the most important element is a single brain to integrate every last smart, electrified thing in your house.
The fundraise comes on the heels of a $50 million grant from the Department of Energy.
Almost all lithium-ion batteries on the market rely on graphite for their anodes, more than 90% of which is processed in China. That concentration has created a chokepoint in the Pentagon’s supply chain as the military increasingly turns to batteries to help its transportation network operate more quietly and reduce reliance on complex refueling logistics.
So when Congress passed the latest National Defense Authorization Act at the end of 2025, barring Chinese-linked battery materials from military use starting in 2028, Coreshell saw an opening. While the California-based silicon-anode battery startup had previously focused on automakers as its beachhead customers, it’s now shipping its first battery cells to the military for use on Navy ships and Army ground vehicles. On Wednesday, Coreshell announced a $30 million Series B to scale up manufacturing operations to meet the needs of this growing customer base.
“We’re entering a new stage where we’re no longer, I would say, demand limited,” Coreshell’s CEO Jonathan Tan told me. “This year we’ve had to put on second and third shifts, work nights and weekends, just to get the product out the door.”
Coreshell sets itself apart by going a step further than the NDAA requires. The law doesn’t make suppliers trace battery materials all the way back to the mine, which would be exceedingly difficult given China’s dominance across the battery supply chain. But Coreshell has nevertheless eliminated Chinese sourcing from its anodes by replacing graphite with silicon mined and refined in the U.S. by Ferroglobe, one of its Series A investors.
Ferroglobe and global investment firm Access Industries led the latest round, with defense contractor ADS also joining as a new strategic investor. The financing comes on the heels of a $50 million Department of Energy grant that Coreshell secured in August as part of a program to strengthen the domestic battery supply chain. That funding will support the company’s first commercial-scale facility, a 2-gigawatt-hour factory.
That’s about 20 times the size of the 100-ish-megawatt-hour plant the startup had originally envisioned, as the DOE grant has allowed Coreshell to “expand much further than we were originally expecting,” according to Tan. The company is now evaluating several potential U.S. sites and expects to announce its selection early next year. For now, it relies on unnamed contract manufactures, along with its small pilot line in San Leandro, to produce its cells. But once the gigafactory gets up and running, the startup plans to bring electrode production in house by 2028, followed by full cell assembly in 2029.
Building this plant will ultimately require significant additional capital. But Tan believes this latest vote of confidence from the federal government will help Coreshell to diversify its capital stack as it continues to fundraise. “That’s one of the great things about having such a notable DOE award is that it also allows us to attract multiple forms of capital,” he told me.
Silicon has long been recognized as a promising alternative material for battery anodes, as it can store far more energy than graphite. The catch is that it tends to swell during charging, causing it to break down quickly and ultimately leading to a much shorter cycle life. Coreshell says it’s extended its anode’s lifespan by giving it a proprietary coating, which Tan told me back in April essentially means “painting the inside of your battery” with an ultra-thin protective layer.
Other companies, such as Sila Nanotechnologies and Group14 Technologies, are also working to commercialize silicon anodes. But Coreshell says many of its competitors rely on more highly engineered, expensive forms of silicon, such as silane gas. By contrast, Coreshell uses metallurgical silicon, a cheap, raw commodity already widely used in aluminum alloys. That, Tan argues, gives the company a cost advantage not only over its competitors, but potentially Chinese graphite too.
Coreshell hasn’t yet demonstrated that its cells can meet the performance demands of the automotive industry, however. When the company raised its Series A in early 2025, it said it expected its automotive cells to exceed 1,000 charge cycles while retaining 80% of their original capacity. But Tan told me many defense applications don’t require such demanding performance targets, as Navy ships and ground vehicles are “not intending to have as many miles or missions or use cases on that same battery” as a consumer EV would.
That doesn’t mean the company is abandoning its automotive ambitions. As Tan pointed out, electric vehicles remain by far the largest source of global and domestic battery demand. “It’s natural for us to obviously utilize the security supply chain first,” he told me. “That allowed us to really get to commercialization, revenue, and production faster, because automotive qualification is this longer journey. But ultimately, we want to do both of those and more.”