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:
A new report from the American Council for an Energy-Efficient Economy has some exciting data for anyone attempting to retrofit a multifamily building.

By now there’s plenty of evidence showing why heat pumps are such a promising solution for getting buildings off fossil fuels. But most of that research has focused on single-family homes. Larger apartment buildings with steam or hot water heating systems — i.e. most of the apartment buildings in the Northeast — are more difficult and expensive to retrofit.
A new report from the nonprofit American Council for an Energy-Efficient Economy, however, assesses a handful of new technologies designed to make that transition easier and finds they have the potential to significantly lower the cost of decarbonizing large buildings.
“Several new options make decarbonizing existing commercial and multifamily buildings much more feasible than a few years ago,” Steven Nadel, ACEEE’s executive director and one of the authors, told me. “The best option may vary from building to building, but there are some exciting new options.”
To date, big, multifamily buildings have generally had two flavors of heat pumps to consider. They can install a large central heat pump system that delivers heating and cooling throughout the structure, or they can go with a series of “mini-split” systems designed to serve each apartment individually. (Yes, there are geothermal heat pumps, too, but those are often even more expensive and complicated to install, especially in urban areas.)
While these options have proven to work, they often require a fair amount of construction work, including upgrading electrical systems, mounting equipment on interior and exterior walls, and running new refrigerant lines throughout the building. That means they cost a lot more than a simple boiler replacement, and that the retrofit process can be disruptive to residents.
In 2022, the New York City Housing Authority launched a contest to try and solve these problems by challenging manufacturers to develop heat pumps that can sit in a window just like an air conditioner. New designs from the two winners, Gradient Comfort and Midea, are just starting to come to market. But another emerging solution, central air-to-water heat pumps, also presents an appealing alternative. These systems avoid major construction because they can integrate with existing radiators or baseboard heaters in buildings that currently use hot water boilers. Instead of burning natural gas or oil to produce hot water, the heat pump warms the water using electricity.
The ACEEE report takes the cost and performance data for these emerging solutions and compares it to results from mini-splits, central heat pumps, geothermal heat pumps, packaged terminal heat pumps — all-in-one devices that sit inside a sleeve in the wall, commonly used in hotels — and traditional boilers fed by biogas or biodiesel.
While data on the newer technologies is limited, so far the results are extremely promising. The report found that window heat pumps are the most cost-effective of the bunch to fully decarbonize large apartment buildings, with an average installation cost of $9,300 per apartment. That’s significantly higher than the estimated $1,200 per apartment cost of a new boiler, but much lower than the $14,000 to $20,000 per apartment price tag of the other heat pump variations, although air-to-water heat pumps came in second. The report also found that window heat pumps could turn out to be the cheapest to operate, with a life cycle cost of about $14,500, compared to $22,000 to $30,000 for boilers using biodiesel or biogas or other heat pump options.
As someone who has followed this industry for several years with a keen interest in new solutions for boiler-heated buildings in the Northeast — where I grew up and currently reside — I was especially wowed by how well the new window heat pumps have performed. New York City installed units from both Midea and Gradient in 24 public housing apartments, placing one in each bedroom and living room, and monitored the results for a full heating season.
Preliminary data shows the units performed swimmingly on every metric.
On ease of installation: It took a total of eight days for maintenance workers to install the units in all 24 apartments, compared to about 10 days per apartment when the Housing Authority put split heat pump systems in another building.
On performance: During the winter, while other apartments in the building were baking in 90-degree Fahrenheit heat from the steam system, the window unit-heated apartments maintained a comfortable 75 to 80 degree range, even as outdoor temperatures dropped to as low as 20 degrees.
On energy and cost: The window unit-heated apartments used a whopping 87% less energy than the rest of the building’s steam-heated apartments did, cutting energy costs per household in half.
On customer satisfaction: A survey of 72 residents returned overwhelmingly positive feedback, with 93% reporting that the temperature was “just right” and 100% reporting they were either “neutral” or “satisfied” with the new units.
The Housing Authority found that the units also lowered energy used for cooling in peak summer since they were more efficient than the older window ACs residents had been using. Next, the agency plans to expand the pilot to two full buildings before deploying the units across its portfolio. The pilot was so successful that utilities in Massachusetts, Vermont, and elsewhere are purchasing units to do their own testing.
The ACEEE report looked at a handful of air-to-water heat pump projects in New York and Massachusetts, as well, only two of which have been completed. The average installation cost per apartment was around $13,500, with each of the buildings retaining a natural gas boiler as a backup, but none had published performance data yet.
Air-to-water heat pumps have only recently come to market in the U.S. after having taken off in Europe, and they don’t yet fit seamlessly into the housing stock here. Existing technology can only heat water to 130 to 140 degrees, which is hot enough for the more efficient hot water radiators common in Europe but too cold for the U.S. market, where hot water systems are designed to carry 160- to 180-degree water, or even steam.
These heat pumps can still work in U.S. buildings, but they require either new radiators to be installed or supplemental heat from a conventional boiler or electric resistance unit. The other downside to an air-to-water system is that it can’t provide cooling unless the building is already equipped with compatible air conditioning units.
One strength of these systems over the window units, however, is that they don’t push costs onto tenants in buildings where the landlord has historically paid for heat. They also may be cheaper to operate than more traditional heat pump options, although data is still extremely limited and depends on the use of supplemental heat.
It’s probably too soon to draw any major conclusions about air-to-water systems, anyway, because new, potentially more effective options are on the way. In 2023, New York State launched a contest challenging manufacturers to develop new decarbonized heating solutions for large buildings. Among the finalists announced last year, six companies were developing heat pumps that could generate higher-temperature hot water and/or steam. One of them is now installing its first demonstration system in an apartment building in Harlem, and two others have similar demonstrations in the works.
The ACEEE report also mentions a few other promising new heat pump formats, such as an all-in-one wall-mounted heat pump from Italian company Ephoca. It’s similar to the window heat pump in that it’s contained in a single device rather than split into an indoor and outdoor unit, so it doesn’t require mounting anything to the outside of the building or worrying about refrigerant lines, although it does require drilling two six-inch holes in the wall for vents. These may be a good option for those whose windows won’t accommodate a window heat pump or who don’t like the aesthetics. New York State is also funding product development for better packaged terminal heat pumps that could slot into wall cavities occupied by less-efficient packaged terminal air conditioners and heat pumps today.
Gradient and Midea are not yet selling their cold-climate window heat pumps directly to consumers. Gradient brought a version of its technology for more moderate climates to market in 2023, which was only suitable for heating at outdoor temperatures of 40 degrees and higher. But the company has discontinued that model and is focusing on an “all-weather” version designed for cold climates, which is the one that has been installed in the New York City apartments. Gradient told me it is currently selling that model in bulk to multi-family building owners, utilities, and schools. Midea did not respond to my inquiry.
One big takeaway is that even the new school heat pumps designed to be easier and cheaper to install have higher capital costs than buying a boiler and air conditioners — a stubborn facet of many climate solutions, even when they save money in the long run. Canary Media previously reported that the Gradient product would start at $3,800 per unit and the Midea at $3,000. Experts expect the cost to come down as adoption and demand pick up, but the ACEEE report recommends that states develop incentives and financing to help with up-front costs.
“These are not just going to happen on their own. We do need some policy support for them,” Nadel said. In addition to incentives and building decarbonization standards, Nadel raised the idea of discounted electric rates for heat pump users, an idea that has started to gain traction among climate advocates that a few utilities have piloted.
“To oversimplify,” Nadel said, “in many jurisdictions, heat pumps subsidize other customers, and that probably needs to change if this is going to be viable.”
Editor’s note: This story has been updated to include comment from Gradient.
Log in
To continue reading, log in to your account.
Create a Free Account
To unlock more free articles, please create a free account.
The public-private project aims to help realize the president’s goal of building 10 new reactors by 2030.
The Department of Energy and the Westinghouse Electric Company have begun meeting with utilities and nuclear developers as part of a new project aimed at spurring the country’s largest buildout of new nuclear power plants in more than 30 years, according to two people who have been briefed on the plans.
The discussions suggest that the Trump administration’s ambitious plans to build a fleet of new nuclear reactors are moving forward at least in part through the Energy Department. President Trump set a goal last year of placing 10 new reactors under construction nationwide by 2030.
The project aims to purchase the parts for 8 gigawatts to 10 gigawatts of new nuclear reactors, the people said. The reactors would almost certainly be AP1000s, a third-generation reactor produced by Westinghouse capable of producing up to 1.1 gigawatts of electricity per unit.
The AP1000 is the only third-generation reactor successfully deployed in the United States. Two AP1000 reactors were completed — and powered on — at Plant Vogtle in eastern Georgia earlier this decade. Fifteen other units are operating or under construction worldwide.
Representatives from Westinghouse and the Energy Department did not respond to requests for comment.
The project would use government and private financing to buy advanced reactor equipment that requires particularly long lead times, the people said. It would seek to lower the cost of the reactors by placing what would essentially be a single bulk order for some of their parts, allowing Westinghouse to invest in and scale its production efforts. It could also speed up construction timelines for the plants themselves.
The department is in talks with four to five potential partners, including utilities, independent power producers, and nuclear development companies, about joining the project. Under the plan, these utilities or developers would agree to purchase parts for two new reactors each. The program would be handled in part by the department’s in-house bank, the Loan Programs Office, which the Trump administration has dubbed the Office of Energy Dominance Financing.
This fleet-based approach to nuclear construction has succeeded in the past. After the oil crisis struck France in the 1970s, the national government responded by planning more than three-dozen reactors in roughly a decade, allowing the country to build them quickly and at low cost. France still has some of the world’s lowest-carbon electricity.
By comparison, the United States has built three new nuclear reactors, totaling roughly 3.5 gigawatts of capacity, since the year 2000, and it has not significantly expanded its nuclear fleet since 1990. The Trump administration set a goal in May to quadruple total nuclear energy production — which stands at roughly 100 gigawatts today — to more than 400 gigawatts by the middle of the century.
The Trump administration and congressional Republicans have periodically announced plans to expand the nuclear fleet over the past year, although details on its projects have been scant.
Senator Dave McCormick, a Republican of Pennsylvania, announced at an energy summit last July that Westinghouse was moving forward with plans to build 10 new reactors nationwide by 2030.
In October, Commerce Secretary Howard Lutnick announced a new deal between the U.S. government, the private equity firm Brookfield Asset Management, and the uranium company Cameco to deploy $80 billion in new Westinghouse reactors across the United States. (A Brookfield subsidiary and Cameco have jointly owned Westinghouse since it went bankrupt in 2017 due to construction cost overruns.) Reuters reported last month that this deal aimed to satisfy the Trump administration’s 2030 goal.
While there have been other Republican attempts to expand the nuclear fleet over the years, rising electricity demand and the boom in artificial intelligence data centers have brought new focus to the issue. This time, Democratic politicians have announced their own plans to boost nuclear power in their states.
In January, New York Governor Kathy Hochul set a goal of building 4 gigawatts of new nuclear power plants in the Empire State.
In his State of the State address, Governor JB Pritzker of Illinois told lawmakers last week that he hopes to see at least 2 gigawatts of new nuclear power capacity operating in his state by 2033.
Meeting Trump’s nuclear ambitions has been a source of contention between federal agencies. Politico reported on Thursday that the Energy Department had spent months negotiating a nuclear strategy with Westinghouse last year when Lutnick inserted himself directly into negotiations with the company. Soon after, the Commerce Department issued an announcement for the $80 billion megadeal, which was big on hype but short on details.
The announcement threw a wrench in the Energy Department’s plans, but the agency now seems to have returned to the table. According to Politico, it is now also “engaging” with GE Hitachi, another provider of advanced nuclear reactors.
On nuclear tax credits, BLM controversy, and a fusion maverick’s fundraise
Current conditions: A third storm could dust New York City and the surrounding area with more snow • Floods and landslides have killed at least 25 people in Brazil’s southeastern state of Minas Gerais • A heat dome in Western Europe is pushing up temperatures in parts of Portugal, Spain, and France as high as 15 degrees Celsius above average.

The Department of Energy’s in-house lender, the Loan Programs Office — dubbed the Office of Energy Dominance Financing by the Trump administration — just gave out the largest loan in its history to Southern Company. The nearly $27 billion loan will “build or upgrade over 16 gigawatts of firm reliable power,” including 5 gigawatts of new gas generation, 6 gigawatts of uprates and license renewals for six different reactors, and more than 1,300 miles of transmission and grid enhancement projects. In total, the package will “deliver $7 billion in electricity cost savings” to millions of ratepayers in Georgia and Alabama by reducing the utility giant’s interest expenses by over $300 million per year. “These loans will not only lower energy costs but also create thousands of jobs and increase grid reliability for the people of Georgia and Alabama,” Secretary of Energy Chris Wright said in a statement.
Over in Utah, meanwhile, the state government is seeking the authority to speed up its own deployment of nuclear reactors as electricity demand surges in the desert state. In a letter to the Nuclear Regulatory Commission dated November 10 — but which E&E News published this week — Tim Davis, the executive director of Utah’s Department of Environmental Quality, requested that the federal agency consider granting the state the power to oversee uranium enrichment, microreactor licensing, fuel storage, and reprocessing on its own. All of those sectors fall under the NRC’s exclusive purview. At least one program at the NRC grants states limited regulatory primacy for some low-level radiological material. While there’s no precedent for a transfer of power as significant as what Utah is requesting, the current administration is upending norms at the NRC more than any other government since the agency’s founding in 1975.
Building a new nuclear plant on a previously undeveloped site is already a steep challenge in electricity markets such as New York, California, or the Midwest, which broke up monopoly utilities in the 1990s and created competitive auctions that make decade-long, multibillion-dollar reactors all but impossible to finance. A growing chorus argues, as Heatmap’s Matthew Zeitlin wrote, that these markets “are no longer working.” Even in markets with vertically-integrated power companies, the federal tax credits meant to spur construction of new reactors would make financing a greenfield plant is just as impossible, despite federal tax credits meant to spur construction of new reactors. That’s the conclusion of a new analysis by a trio of government finance researchers at the Center for Public Enterprise. The investment tax credit, “large as it is, cannot easily provide them with upfront construction-period support,” the report found. “The ITC is essential to nuclear project economics, but monetizing it during construction poses distinct challenges for nuclear developers that do not arise for renewable energy projects. Absent a public agency’s ability to leverage access to the elective payment of tax credits, it is challenging to see a path forward for attracting sufficient risk capital for a new nuclear project under the current circumstances.”
Steve Pearce, Trump’s pick to lead the Department of the Interior’s Bureau of Land Management, wavered when asked about his record of pushing to sell off federal lands during his nomination hearing Wednesday. A former Republican lawmaker from New Mexico, Pearce has faced what the public lands news site Public Domain called “broad backlash from environmental, conservation, and hunting groups for his record of working to undermine public land protections and push land sales as a way to reduce the federal deficit.” Faced with questions from Democratic senators, Pearce said, “I’m not so sure that I’ve changed,” but insisted he didn’t “believe that we’re going to go out and wholesale land from the federal government.” That has, however, been the plan since the start of the administration. As Heatmap’s Jeva Lange wrote last year, Republicans looked poised to use their trifecta to sell off some of the approximately 640 million acres of land the federal government owns.
Sign up to receive Heatmap AM in your inbox every morning:
At Tuesday’s State of the Union address, as I told you yesterday, Trump vowed to force major data center companies to build, bring, or buy their own power plants to keep the artificial intelligence boom from driving up electricity prices. On Wednesday, Fox News reported that Amazon, Google, Meta, Microsoft, xAI, Oracle, and OpenAI planned to come to the White House to sign onto the deal. The meeting is set to take place sometime next month. Data centers are facing mounting backlash. Developers abandoned at least 25 data centers last year amid mounting pushback from local opponents, Heatmap's Robinson Meyer recently reported.
Shine Technologies is a rare fusion company that’s actually making money today. That’s because the Wisconsin-based firm uses its plasma beam fusion technology to produce isotopes for testing and medical therapies. Next, the company plans to start recycling nuclear waste for fresh reactor fuel. To get there, Shine Technologies has raised $240 million to fund its efforts for the next few years, as I reported this morning in an exclusive for Heatmap. Nearly 63% of the funding came from biotech billionaire Patrick Soon-Shiong, who will join the board. The capital will carry the company through the launch of the world’s largest medical isotope producer and lay the foundations of a new business recycling nuclear waste in the early 2030s that essentially just reorders its existing assembly line.
Vineyard Wind is nearly complete. As of Wednesday, 60 of the project’s 62 turbines have been installed off the coast of Massachusetts. Of those, E&E News reported, 52 have been cleared to start producing power. The developer Iberdrola said the final two turbines may be installed in the next few days. “For me, as an engineer, the farm is already completed,” Iberdrola’s executive chair, Ignacio Sánchez Galán, told analysts on an earnings call. “I think these numbers mean the level of availability is similar for other offshore wind farms we have in operation. So for me, that is completed.”
That doesn’t mean it plans to produce electricity anytime soon.
Greg Piefer thinks nearly all his rivals in the race to commercialize fusion are doing it backward.
Of the 59 companies tracked in the Fusion Industry Association’s latest annual survey, 48 are primarily focused on generating electricity, off-grid energy, or industrial heat by harnessing the power produced when two atoms fuse together in the same type of reaction that fuels the sun. Just four are following the path of Shine Technologies and using plasma beam energy to manufacture rare and extremely valuable radioisotopes for breakthrough cancer treatments — 10 if you count the startups with a secondary medical business.
“We’re a bit different from fusion companies trying to sell the single product of electricity,” Piefer, the chief executive of Wisconsin-based Shine Technologies, told me. “The basic premise of our business is fusion is expensive today, so we’re starting by selling it to the highest-paying customers first.”
Shine Technologies’ contrarian strategy is winning over investors. On Thursday, the company plans to announce a $240 million Series E round, Heatmap can report exclusively. The funding, nearly 63% of which came from biotech billionaire Patrick Soon-Shiong, will provide enough capital to carry the company to the launch of the world’s largest medical isotope producer and lay the foundations of a new business recycling nuclear waste.
For now, Piefer said, Shine’s business is blasting uranium with enough extremely hot plasma beam energy to generate medical isotopes such as molybdenum-99 for diagnostic imaging or lutetium-177 for targeted cancer therapies. In the next few years, however, Shine Technologies is looking to apply its methods to recycling and reducing radioactive waste from commercial fission reactors’ spent fuel. Only then, sometime a decade from now, will the company start working on power plants.
“I would essentially define electricity as the lowest-paying customer of significance for fusion today,” Piefer said.
Soon-Shiong contributed $150 million to the funding pool via NantWorks, the biotech company he founded. Other investors include the financial services giant Fidelity Investments, the American division of the Japanese industrial conglomerate Sumitomo Corporation, the Texas investment bank Pelican Energy Partners, the healthcare-focused investor Deerfield Management, and the global asset manager Oaktree Capital. As part of the deal, Soon-Shiong — known outside the medical industry as the owner of the Los Angeles Times — will join Shine Technologies’ board of directors.
Since its founding in 2005, Shine has brought down the cost per fusion reaction by a thousandsfold. Over a Zoom call, Piefer pointed out the window behind him in his office in Janesville, Wisconsin, nearly two hours southwest of Milwaukee. In the afternoon sun was a gray, nondescript-looking warehouse. Inside, construction was underway on the world’s largest facility for producing medical isotopes. Dubbed Chrysalis, the flagship plant is set to come online in 2028.
“We’ll make 20 million doses of medicine per year with it,” he said. “It’ll be the biggest beneficial use of fusion for humans ever, and we expect it to be the dominant technology for decades. This will be the way the United States produces neutron-based radioisotopes probably for the next 50 years.”
To make medicine, the company follows four steps. First, it dissolves uranium. Next, it irradiates the material with the plasma beam. Then comes the separation process to remove valuable isotopes from the other radioactive material. Finally leftover uranium gets recycled back into the process. Rinse and repeat.
“It’s the first closed loop ever used for producing medicine this way,” Piefer said.
To recycle spent nuclear fuel, the company just remixes those steps, he said.
“You dissolve uranium from the nuclear waste. You separate out valuable materials. You recycle the uranium and plutonium in a reactor,” Piefer said. Then fusion comes in with the plasma beam technology to transform highly radioactive material that stays dangerous for longer than Homo sapiens is known to have existed into something that decays in half-lives that take years, decades, or centuries rather than millennia, decamillennia, and centimillennia.
“There’s about half a percent of long-lived nuclear waste from fission that we don’t know what to do with. It lives basically forever. We don’t have a use for it. But if you hit it with fusion neutrons, it becomes short-lived,” Piefer said. “So it’s the same four steps. For medicine, it goes one, two, three, four. For recycling it goes one, three, four, two.”
Not only is the market for testing and medical isotopes already worth billions of dollars, it’s on track to more than double in the next decade. Currently, it’s largely served by what Piefer called “60-year-old fission reactors.”
“These are specialized research reactors that are very cold and very constrained from a capacity standpoint,” he said. “You can buy new ones, but it takes billions of dollars and probably two decades to bring a new reactor online.”
By contrast, Shine Technologies broke ground on Chrysalis in 2019, and is set to complete the project at what Piefer said would be an eighth the cost of building a new research reactor.
The U.S. government, meanwhile, is helping to fund the next phase of Shine Technologies’ business. Just a few weeks ago, the Department of Energy gave the company a share of $19 million split between five companies looking to commercialize reprocessing technology. Last year, the company inked a deal with the reactor fuel startup Standard Nuclear to sell the fuel-grade material it recovers from recycling.
In both the fusion and next-generation fission industries, companies often lure investors by promising to pull off several very challenging things at once, said Chris Gadomski, the lead nuclear analyst at the consultancy BloombergNEF.
Oklo, a stock market darling for its planned microreactor and power plant business, was also among the recipients of the federal funding for waste reprocessing. Amazon-backed microreactor developer X-energy just won approval to start manufacturing the rare and expensive form of reactor fuel known as TRISO. TAE Technologies, the fusion startup that merged in December with the parent company of President Donald Trump’s social media network TruthSocial in a bid to build the world’s first fusion power plant, also has a subsidiary producing medical isotopes.
“I usually look at it as a distressing sign when you have an energy company tackling four or five different things,” Gadomski said. “But Shine is really a medical device company that is focused on isotopes but whose technology can also reprocess spent fuel — and, by the way, it can be applied down the road to energy.”
So far, Shine’s technology has followed a similar Moore’s Law trajectory to semiconductors.
From roughly 1990 to 2000, microchips used in workstations increased their computation rate per dollar. Then came the gaming era from 2000 to 2015, when videogames drove demand for more and more efficient semiconductors, with upgrades on average every other year. From 2015 until roughly the debut of ChatGPT in 2022, the high-speed computing applications spurred on chip upgrades at a similar rate. Now the artificial intelligence era is upon us, transforming chipmakers such as Nvidia into goliaths seemingly overnight.
Piefer sees Shine Technologies on its own 35-year timeline. From 2010 to roughly 2023, testing dominated the business. From then until about 2028, medical isotopes are the new play. The recycling pilot plant set to come online after 2030 will kick off the reprocessing period. And finally, sometime in the 2040s, Piefer wants to get into energy production.
“It’s a different approach than most,” he said.
“Don’t get me wrong, moonshots have their place, too,” he added. “But I feel very confident in this path.”