You’re out of free articles.
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:
CoreWeave signed a deal for a new facility in New Jersey, behind-the-meter power on the side.
The cloud computing company CoreWeave announced Monday that it is leasing a former medical research facility and turning it into a data center. Along with it comes a 25-megawatt power plant that once provided power and steam directly to the former Merck headquarters in Kenilworth, New Jersey, but began to sell more and more power to the grid, the plant’s owner said in a filing with the Federal Energy Regulatory Commission. In 2023, the facility was purchased by Onyx, a real estate firm, and Machine Investment Group, with the intention to market the site to another life sciences or biotechnology company.
Then the AI revolution happened.
CoreWeave, which started as a miner of cryptocurrency, is now raising and spending billions of dollars to acquire and install the chips necessary to train and run artificial intelligence systems for companies that rent out access to them. According to the deal announcement, the company plans to pour $1.2 billion of investment into the 280,000 square foot facility, along with electrical upgrades from the utility PSE&G and investments from Onyx. The power plant will stop serving the grid and go “behind the meter,” the plant’s owner Atlantic Power said in a letter to PJM Interconnection, the regional electricity market, in September.
The deal confirms that when it comes to power, data centers will take what they can get — and that the long timelines necessary to bring on new power in much of the country may end up benefiting existing owners of generation, especially natural gas.
Data centers require both large amounts of power — sometimes 100 megawatts or more — and the ability to surge up and down quickly. “Renewable power generation is well placed to capture mounting demand from data centers and AI in the long term,” analysts at BNEF wrote in a report in September, “but time constraints for grid interconnection and intermittency issues could support natural gas-fired output.”
Goldman Sachs analysts expect data center power demand to rise from about 3% of the U.S. total to 8% by 2030, with growth running at 15% annually. They assume that capacity will be met mostly by natural gas, but actually finding — let alone building — new natural gas generation is a challenge.
“The hyperscalers are evolving from single data centers dependent on 60 to 100 megawatts to starting to look at multiple gigawatt-size data center parks that support a number of data centers in one location,” GE Vernova chief executive Scott Strazik said on a recent earnings call with analysts.
Building a new natural gas plant on the grid — and especially the transmission infrastructure to serve it — can be a prospect well beyond the build-it-now timelines of big technology companies with a desperate need for computing power.
“Thanks to 10-year delays in permitting for new transmission lines and connecting generation capacity to the grid, the most viable near-term option is behind-the-meter,” Tim Fist and Arnab Datta wrote in a report for the Institute for Progress, a technology and science policy think tank. In other words, one way to get around grid interconnection and intermittency issues is to have your own power plant.
“The economics of developing the power on site don't really hurt the data center economics that much. These things are just really profitable,” Carson Kearl, an analyst at Enverus, told me.
Some data centers have developed their own natural gas generation on site, such as XAi’s cluster in Memphis, Tennessee, which ispowered by gas generators.
CoreWeave, meanwhile, is one of the most talked-about and well-funded companies in cloud computing, with access to a huge number of chips made by Nvidia, the leading designer of high-end processors, and which is also an investor in CoreWeave. But the chips can only perform when they’re powered, turning the data center business into a hunt for electricity wherever it can be found.
“Access to reliable power could be a roadblock towards the timely buildout” for a data center, Francois Poirier, the chief executive of TC Energy, the Canadian pipeline company, told analysts on an earnings call in August. “We’re seeing a shift in siting preferences from regions where big telecom infrastructure is in place to regions where energy and supply infrastructure is in place.”
CoreWeave, PSE&G, Onyx, and Atlantic Power’s owner, I Squared Capital, did not respond to requests for comment.
This situation has not come about for lack of effort on the part of the several electricity markets that have been trying to get new natural gas generation on the grid. PJM, for example, has been working to entice new supply, but even following a record auction for power capacity that paid out billions to natural gas plants, few producers have indicated their willingness to make large new investments. Texas has established a multibillion-dollar loan fund to provide low-cost financing to new natural gas plants.
While several large technology companies have announced their intention to buy nuclear power from refurbished or new plants, those deals will take at least several years to actually get any new electrons on the grid.
That leads data center developers like CoreWeave scrambling to find what power they can. In interviews, the company’s chief strategy officer Brian Venturo told Wired that they are avoiding Northern Virginia’s “data center alley” precisely because it’s “a food fight to get power.”
“There's a lot of growing backlash in that market around power usage,” he told Bloomberg. “We're kind of siting our plants and markets where our data centers and markets where we think the grid infrastructure is capable of handling it.”
And what better place than where the power already is.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
Key projects for the Energy Department’s hydrogen hubs are dropping like flies. And it’s really not obvious why.
Three hubs DOE selected for potential federal support have lost projects that were linchpins. Industrial giant Fortescue is no longer publicly committing to a hydro-powered hydrogen production plant proposed in Washington state that was key to the Pacific Northwest hub. News of a pause at the project was previously reported, but the company notably declined to even say the project was still getting built when asked about it this week.
“While Fortescue will continue to maintain a portfolio of other projects for the future, our financial discipline always comes first. We will never do projects that are not currently economically viable,” the company said in a statement provided to me this morning.
Meanwhile CNX, a natural gas company, has indefinitely put the kibosh on a blue hydrogen ammonia plant in West Virginia crucial to the Appalachian hydrogen hub known as ARCH2. Marathon Petroleum’s midstream subsidiary MPLX also confirmed to me they’ve canceled a hydrogen storage facility planned for that hub, and Chemours is no longer involved with the hub either.
Another blue hydrogen ammonia plant in North Dakota crucial to a different hub – known as the Heartland hub – has been canceled by Marathon and TC Energy.
In other words: a year after the Biden administration made a big announcement about the seven hubs that could potentially receive billions of dollars in government funding, almost half of them are running into serious trouble.
The companies that have quietly pulled out or paused projects are laying blame on implementation of the federal hydrogen production tax credit, claiming rules enforcing the “three pillars” and carbon intensity requirements are too onerous. Meanwhile critics of the hydrogen hubs are seizing on project cancellations and delays to argue against their construction outright; the Ohio River Valley Institute, an environmental group opposed to the ARCH2 hydrogen hub, has received a lot of press in recent days for a report claiming the hub is “coming apart.”
I’m already hearing whispers from industry insiders in D.C. who are trying to spin these cancellations as evidence the credit implementation has been too favorable to climate activists and is constraining growth in the nascent hydrogen space.
But what’s really going on?
Conversations with experts and stakeholders indicate to me this could be evidence of broader macroeconomic issues hitting the hydrogen industry, from inflation pushing up the price of electrolyzers to the stubbornly low price of natural gas. We saw this with the Plug Power project in New York, which we were first to report problems with. These market issues may be overpowering the subsidies and demand-side benefits of the bipartisan infrastructure law and Inflation Reduction Act.
These hiccups may also be a calm before a storm of hydrogen investment and a reshuffling of capital that’ll become more evident after the IRA’s production tax credit is fully implemented with final regulations. Perhaps it’ll take final rules to see the companies supportive of the “three pillars” move more projects forward.
It could also be a mixture of these things and other factors, like issues with the specific sites companies had selected for their plants.
No matter the cause for these hubs stuttering, these projects falling out of the fold is a shock to no one, especially supporters of the “three pillars” approach to the tax credit. Though it may indicate flaws with a disorganized approach to the energy transition.
“I’m not surprised if at the end of the day some of the many projects supported by DOE are not viable in the end,” said Jesse Jenkins, an assistant professor at Princeton University and expert in energy systems engineering. In addition to co-hosting Heatmap’s Shift Key podcast, Jenkins leads the REPEAT Project, which produced influential policy analysis supporting the “three pillars” approach to Treasury’s implementation of the hydrogen production tax credit.
Irrespective of the reasons, it’s important to remember that on some level both industry and the Biden administration stumbled into this mess. That’s because Congress passed the bipartisan infrastructure law mandating the creation and financing of these hubs before the IRA was even introduced. The infrastructure law itself required DOE to start soliciting proposals for hub funding mere months after it was enacted. This means the hub program was crafted independent of a tax subsidy boosting supply.
The hubs may be lobbying for a specific version of the hydrogen production credit to be implemented, as many D.C. lobbyists like to point out, but the program wasn’t referenced in the tax credit’s statute either.
As Jenkins put it, any conflict between the hubs and tax credit provisions is evidence “that reflects that many of the projects [selected] are not compliant.”
Biden administration officials spoke to me for a half hour this morning about the canceled projects on the condition of anonymity to candidly discuss the tax credit and hubs. To them, this can be explained as the process working as intended, and they emphasized how the credit and hub are independent programs. They also expect more capital to be unleashed after the credit is finalized, as companies who’ve supported the “three pillars” get certainty to make final investment decisions.
The administration’s view sounded akin to the optimistic vision relayed to me by Clean Air Task Force’s Conrad Schneider: “This is what progress looks like. It’s slow, it’s steady. It’s not [a] steady state though.”
My take? This is further proof we live in a disorganized energy transition. So far in The Fight, we’ve covered the struggles to get projects built because of opposing forces at a grassroots level. That same dynamic applies to the federal climate programs incentivizing a switch from carbon-intensive business practices. And sometimes, there’ll be tug-of-war competing interests between the climate programs themselves.
And more of the week’s most important conflicts around renewable energy.
1. Douglas County, Kansas – A legal headache is consuming Kansas Sky Energy Center, a 159-megawatt solar project proposed by Savion and Invenergy… and showcasing how “agri-voltaics” may not be the community engagement panacea some in industry are praying for, according to legal filings reviewed by Heatmap.
2. Worcester County, Maryland – We finally get to see the contours of the legal strategy against the Maryland Offshore Wind Project, after Ocean City and surrounding local business and government officials filed their lawsuit last week.
3. Barnstable County, Massachusetts – Another blow to offshore wind came Friday in the coastal town of Barnstable where leaders voted to oppose cable landings for Avangrid’s New England Wind 2 project.
3. Somewhere near Houston, Texas – The small city of Katy rejected a 500-megawatt battery storage project proposed by Ochoa Energy despite the community struggling with regular blackouts.
Here’s what else I’m keeping tabs on…
In Iowa, a county fighting the Worthwhile Wind farm proposed by Invenergy is asking outside legal counsel for help after a federal judge ruled the project could resume construction.
In Kentucky, the mayor of Lexington city Linda Gorton testified against construction of a 40-megawatt solar farm proposed by the East Kentucky Power Cooperative.
In Massachusetts, a private non-profit that operates historic sites in Nantucket has withdrawn from the Vineyard Wind good neighbor agreement.
In Michigan, the tiny town of Groveland Township is trying to get a restrictive battery storage ordinance in place before a new state law curbing local control comes into effect. (Sorry you’re dealing with this one, Vesper Energy.)
In Virginia, Dominion Energy has sold a non-controlling 50% stake in Coastal Virginia Offshore Wind as public comments resume on the proposed project (which haven’t always been favorable).And more of the week’s top policy news around renewable energy.
1. Offshore wind lease win – Two companies, Avangrid and Invenergy, purchased four of the eight leases up for grabs yesterday at the first floating offshore wind sale in the Gulf of Maine, according to the Bureau of Ocean Energy Management.
2. Community benefit plans – The Energy Department’s Loan Programs Office is letting the public in on its community benefit agreements, publishing three plans for a wire harness plant in Texas, a solar-plus-storage project on tribal lands in California, and the revived Holtec Palisades nuclear plant in Michigan.
3. Big ports money pour – The EPA yesterday debuted nearly $3 billion in IRA funding to port decarbonization projects ranging from direct acquisitions of zero-emission tech to internal emissions planning.
Here’s what else I’m watching right now…