Key takeaways
U.S. electricity demand is projected to increase nearly 40 percent from 2026 levels by 2035, but domestic manufacturing capacity for critical grid components has not kept pace.
Large power transformers are the backbone of the grid but remain a severe bottleneck, with 80 percent import reliance, a 30 percent supply deficit, and lead times of up to two and a half to five years.
High-voltage direct current transmission and power electronics manufacturing capacity remain largely undeveloped domestically.
This report recommends a five-part federal industrial policy to increase domestic manufacturing capacity and friend-shore key components and materials that are creating bottlenecks for grid expansion.
Introduction and summary
Read the fact sheet
Over the coming decade, the United States will undergo an expansion of the electric grid at an unprecedented rate. Meeting the expected demand for electric grid components will require the United States to either continue its overreliance on imports or launch a strategic buildout of the domestic industrial base.1 Given the importance of the electric grid to the nation’s economic and national security, the former risks putting the nation’s prosperity at risk to global bottlenecks and cybersecurity vulnerabilities, while also forestalling the economic benefits of self-sufficiency, including for the American workforce.
For decades, electricity consumption in the United States remained flat. Now, the explosive growth of data centers, coupled with the shift toward electrification across the economy, is causing electricity demand to skyrocket.2 Over the past five years, electricity demand has increased 2.1 percent per year on average, yet it is expected to increase nearly 40 percent from 2026 levels by 2035.3 This spike arrives as the nation’s grid infrastructure is already under strain and in need of widespread repair, with components across the grid having exceeded their designed lifespans, making many parts of the grid susceptible to damage from increasingly frequent extreme weather events. A resilient energy grid directly sustains critical infrastructure, from health care facilities to clean water systems, telecommunications networks, manufacturing, and transit. It is the foundational backbone of the U.S. economy, Americans’ basic safety, and national security.
Today, the United States relies heavily on imports of essential grid components as well as the materials and critical minerals required to make them. These components and materials are in high demand globally and can have wait times of several years. As a result, the utilities that operate the grid are often forced to pay high prices and pass the costs down to ratepayers. The cost of transmission and distribution components—including wire and cables, transformers, wood poles and switchgear—has increased 50 percent to 152 percent since 2019 and are expected to be a significant driver of higher electricity prices over the next five years.4 Long lead times in the domestic supply chain are also an energy security risk, as utilities can struggle to acquire essential components quickly as the grid grows or is in need of repair.
Investing in new domestic production of grid components offers an opportunity to lower energy costs, create and retain high-quality construction and manufacturing jobs, and increase energy security for Americans. It would diversify and onshore critical grid component manufacturing, directly buttressing national security. An updated and more reliable grid would also help ensure that domestic producers of all kinds of products such as iron, steel, and aluminum have the reliable, affordable energy they need to operate in an increasingly competitive global market. Additionally, by producing components domestically, the United States would be able to ensure that components were manufactured in accordance with high labor and environmental standards and do not contribute to exploitation or environmental degradation.
The Center for American Progress proposes a grid component industrial policy that prioritizes onshoring and increasing domestic production of components and materials that are currently creating supply chain bottlenecks. This includes the production of large power transformers (LPTs), grain-oriented electrical steel (GOES), and high-voltage direct current infrastructure including cables and power electronics, as well as the sourcing of critical minerals required to create them. In those instances where onshoring is not a viable short-term option, friend-shoring should be considered preferable to relying on suppliers from adversaries such as China. Limited domestic production capacity for these components and materials has long been identified as a problem by Republican and Democratic administrations alike, with the U.S. Department of Energy (DOE) noting as far back as 2012 that about 80 percent of all LPTs were imported and faced a lead time about one year—a problem that, despite various government efforts, has not improved since.5
The federal government, alongside states and utilities, should coordinate on a comprehensive and long-term grid component manufacturing strategy for the identified priority components, a list that may be expanded as needs are identified. The authors suggest establishing a five-part industrial policy for grid component manufacturing:
- Initiating supply increase
- Backstopping demand
- Addressing material shortage
- Supplying and supporting a high-quality workforce
- Comprehensively planning for the future
How the grid works
As electricity travels from a power plant, a solar panel, or a wind turbine to a home or business, it needs to be stepped up or down in voltage several times, which is done by transformers housed in substations. The first type of transformer is a generation step-up transformer, a type of large power transformer that increases (steps up) the voltage so that power can be moved over long distances on the transmission grid. There are also small and medium power transformers throughout the grid, with handling capacities from 5,000 kilovolt-amperes to 100,000 kilovolt-amperes (kVA), while LPTs exceed 100,000 kVA.6 Most of the United States currently uses high-voltage alternating current (HVAC) lines for transmission, but there are also several high-voltage direct current (HVDC) lines that can carry large amounts of power over long distances more efficiently.7
Most of the grid uses alternating current (AC), but HVDC lines use direct current (DC), so HVDC converter stations are on either side to convert the electricity from AC to DC and back again. These converter stations use power electronics, including converters, inverters, and rectifiers.
After long-distance transmission, the electricity reaches another LPT, which reduces (steps down) the voltage so the power can safely enter a local distribution grid. Finally, the electricity passes through a distribution transformer—either mounted on a utility pole or in a green ground-level box—which drops the voltage to a safe, usable level for homes and businesses.8 Distribution transformers handle capacities of up to 5,000 kVA.9
Prioritizing grid components for increased domestic manufacturing capacity
The U.S. power grid is massive, consisting of nearly 28,000 generators at power plants,10 several thousand large power transformers,11 and 60 million to 80 million distribution transformers,12 connected by 642,000 miles of high-voltage lines and 6.3 million miles of distribution lines.13 The length of the high-voltage transmission network is 13 times that of the interstate highway system, and there are 1.5 miles of distribution lines for every mile of roadway in the United States.14
Simply maintaining the grid at its current size would be challenging enough, given that many of these components are operating past their expected lifespan and should be replaced to maintain reliability. But of course, the United States needs to dramatically expand the grid to meet anticipated electricity demand. In 2024, the DOE estimated the transmission grid would need to expand 2.4 times to 3.5 times its 2020 size by 2050 under its central decarbonization scenario—an estimate that predates the recent surge in data center-driven demand forecasts.15
This is an unprecedented challenge that will not be met with a business-as-usual approach. Indeed, the United States is not currently on track to meet this growth, and insufficient domestic manufacturing capacity for key transmission components will make it even harder to achieve.16 Failing to meet this growth can lead to disruptive and even dangerous power outages, and recent analyses from Resources for the Future found that delays in building new transmission infrastructure lead to higher electricity bills.17
The United States has the opportunity to yield significant economic benefits by expanding production of grid components. Over the past decade, investor-owned utility investment in the grid doubled, and from 2025 to 2029, these utilities are projected to invest another $1.1 trillion in capital expenditures to upgrade and expand the grid, roughly half of it in transmission and distribution.18 The recent proliferation of large data centers has further increased demand in the sector.
While the United States does manufacture grid components and their materials, there are varying levels of existing capacity and production. And demand signals are sometimes cyclical and, in the case of recent demand driven by data centers, potentially unstable.19 Government policy is needed to turn these imbalanced and unreliable demand signals into long-term manufacturing capacity. As described in further detail below, for example, there is strong domestic production of distribution transformers but limited production of LPTs and minimal to no capacity of HVDC converters. Recent federal studies have identified LPTs, GOES, and HVDC infrastructure including power electronics among the most important components to onshore and for which domestic production should be increased.20
Transformers
Transformers are the connection points of the grid, responsible for increasing and decreasing voltage as electricity moves from generation to end users. Transformers are classified by power-handling capacity: distribution transformers (up to 5,000 kVA), which provide the final voltage step-down to end users, followed by small power transformers (5,000 kVA to 10,000 kVA), medium power transformers (10,000 kVA to 100,000 kVA), and large power transformers (above 100,000 kVA).21 LPTs step up voltage from generation sources onto the long-distance transmission lines that carry electricity across the country.22 LPTs are particularly important, since more than 90 percent of the grid flows through at least one of them.23
The United States has faced a critical shortage of transformers that has been growing since the COVID-19 pandemic due to rising electricity demand and worsened by a fleet that is largely at or past its designed lifespan.24 Demand has increased 116% for power transformers, and LPTs are currently in a supply deficit of 30 percent, while distribution faces a milder shortage with a 6 percent deficit.25 By 2027, transformer demand is expected to more than double its 2020 level.26
Figure 2 breaks down types of power grid transformers and some of their key components, showing how much of a priority each is for federal industrial policy, based on reliance on imports, supply deficits relative to demand growth, and importance to national security. It includes some of the key upstream components that create bottlenecks for transformer production: bushings, which provide insulated pathways for high-voltage current; on-load tap changers (OLTC), which regulate voltage under varying load conditions; continuously transposed conductors (CTCs) made from copper or aluminum, which are used in windings; and transformer cores, structures of thin laminated sheets of ferrous metal, typically grain-oriented electrical steel, wound or stacked together.27
Grain-oriented electrical steel creates an upstream bottleneck
One of the most severe bottlenecks to growing transformer manufacturing is the limited domestic production of GOES.28 The United States imports 90 percent of its GOES and currently has only one domestic producer: a Cleveland-Cliffs subsidiary named AK Steel with two facilities, one in Pennsylvania and one in Ohio.29 In 2022, the DOE noted that the United States couldn’t meet LPT demand with only one GOES manufacturer.30 Since then, U.S. Steel, which was recently acquired by Japan-based Nippon Steel, announced a new facility to be built in Arkansas, but production won’t begin until 2028.31
The United States needs to reduce its import reliance on GOES, and more than one additional facility will be necessary to do so as demand for transformers grows. This is particularly important as the GOES shortage is not only a U.S. problem but a global one. Global capacity for GOES production is 3.8 million metric tons per year, but demand is expected to exceed 5.5 million metric tons per year by the 2030s.32 It is essential that the United States be able to meet its own demand if other countries reduce exports to meet their own needs. Already, imports are becoming increasingly expensive, with global prices doubling between 2021 and mid-2023 followed by a slight decline into 2025, and tariffs under the Trump administration are likely to have further increased the cost post-2025.33 GOES is a prime target for effective industrial policy that relies not just on tariff policies but also on consistent and comprehensive trade and investment policy, as recommended below. In addition to GOES—which, according to industry analyses, is used in approximately 95 percent of all transformers—amorphous metal, an alternate material, should be considered a prime target for industrial capacity.34
Transformer manufacturing is expanding, but not fast enough
Several companies are already working to capitalize on the expected growth of the power grid, despite the uncertainties around demand growth due to changing clean energy policy at the federal level, tariffs, and unpredictable data center expansion.35 In September 2026, Hitachi Energy announced a new $528 million transformer manufacturing facility in Mississippi only a year after it announced a $1 billion investment in expanding grid component manufacturing, citing the demand pull from artificial intelligence (AI) data centers.36 Its previous investments included $457 million for a new LPT facility in Virginia, $106 million for transformer component manufacturing in Tennessee, $70 million to expand production of high-voltage components in Pennsylvania, and $22.5 million to expand transformer manufacturing in Virginia.37 Siemens Energy is following suit, announcing in February its own $1 billion investment to ramp up manufacturing for grid and gas turbine equipment.38 This investment will include the construction of a new high-voltage switchgear facility in Mississippi and an expansion of its new LPT manufacturing facility in Charlotte, North Carolina, which was announced two years earlier.39
This growth in manufacturing capacity is a step in the right direction, but according to a recent study by the National Laboratory of the Rockies, it hasn’t been fast enough. From 2011 to 2023, apparent consumption of LPTs increased 28 percent, but domestic manufacturing capacity only increased by 11 percent.40 And with electricity demand expected to increase over the next decade, the current planned increase in capacity is not enough, which is why additional federal intervention is still necessary. In addition, there are not enough workers to fill the skilled roles in transformer manufacturing, and with the current workforce aging out, meeting the labor demand for growing this industry will be a challenge.41
The United States has been meeting growing demand by increasing reliance on imports
Over the past five years, the United States has met this skyrocketing demand largely with imports, as detailed by component in Figure 2. For several components, including LPTs and GOES, import reliance is extremely high. According to data from the U.S. International Trade Commission (see Figure 3), imports for transformers stayed relatively flat from 2016 to 2021—in the range of $1.8 billion to $2.4 billion—before surging dramatically from 2022 ($3.3 billion) to 2025 ($9.1 billion).42 Meanwhile, exports remained far lower, in the $340,000–$600,000 range.
The United States imports transformers from dozens of countries—mostly from those with which the United States maintains strong trade relationships, such as Mexico, South Korea, and Canada. (see Figure 2 for breakdown by component) China is the largest global producer and exporter of transformers, with about 60 percent of global production capacity, and shipping $9.3 billion worth of transformers last year—36 percent higher than the year before. However, U.S. imports of transformers and most of their components from China remain low.43 In 2025, only 1.8 percent of LPT imports came from China, as well as 3.3 percent of medium power transformers and 5 percent of small power and distribution transformers.44 Even though the United States currently imports LPTs from close allies, it should maintain a baseline of production capacity for these components given their essential role on the grid.
Large power transformer manufacturing is a matter of national security
LPTs’ reliance on imports creates an energy security concern and, therefore, a national security risk. When LPTs fail, whether from old age, damage from extreme weather, or a cyberattack, they are extremely difficult to repair or replace in a timely manner. Utilities are now waiting up to five years to purchase LPTs, and some data center hyperscalers have begun paying to cut the line, a luxury that the average small utility does not have.45 LPTs are extremely large and expensive, costing up to $10 million per unit. Utilities typically cannot afford to purchase a spare and need to get permission from their state regulatory board in order to raise rates for the purpose of purchasing one.46 LPTs also cost hundreds of thousands of dollars to move, and they are often customized, making it difficult to acquire replacement parts.47 Therefore, having LPTs produced domestically, and ideally locally to where they are placed, so that they can be easily transported and repaired is both economically smart and essential to energy security.
More than a decade ago, a Federal Energy Regulatory Commission (FERC) analysis found that if only 9 of the country’s 55,000 transmission substations—likely the critical high-voltage substations with very large transformers—were compromised, the entire country would suffer a blackout from which it could take a year and a half to recover.48 While this estimate may no longer be entirely accurate, as grid operators have had many years to harden the grid and enhance cybersecurity measures, it illustrates the degree to which just one LPT failure may affect grid resilience.
Maintaining a skilled workforce has been a challenge for the transformer industry, particularly LPTs, since the precision and complexity of the manufacturing necessitates that it be done by hand and requires years to master.49 The current workforce is aging out and lacks the vocational pipeline needed to backfill skilled positions. By some estimates, meeting projected demand for transformers would require almost tripling the workforce.50 Overcoming this barrier will be critical to achieving stable, long-term grid component production. Similar to the lack of component manufacturing capacity, workforce difficulties have been recognized at the federal and state levels.51 Existing proposals to address workforce shortages include state-level workforce training programs, the co-location of LPT manufacturing with related industries such as insulation manufacturing, and increased internship and apprentice opportunities.52
While the United States should continue to expand production of all types of transformers, LPTs have reached a level that requires federal intervention. In contrast, distribution transformers are simpler to manufacture and smaller, so they are easier to import and are produced by more than 30 domestic manufacturers.53 However, the fact that distribution transformer imports have more than quadrupled over just the past five years does indicate that further domestic capacity would be beneficial to meet growing demand.54
Cleaning up grid component production
On either side of a transformer are circuit breakers, which are critical components of the transmission system that interrupt the electrical current during a fault to prevent short-circuiting or overloads from causing damage to grid infrastructure.
Circuit breakers and other transmission components commonly use sulfur hexafluoride (SF6), the most potent known greenhouse gas with a global warming potential (GWP) 23,500 times that of CO2.55 An estimated two-thirds of all SF6 emissions in the United States are attributed to the electrical transmission and distribution sector. While SF6-containing equipment is not meant to emit any gas into the atmosphere, leaks are common throughout the components’ life cycle, including in manufacturing, usage, and decommissioning.56
There are several alternatives to SF6 that have lower global warming potentials, as well as technologies that better detect and prevent leaks.57 For example, one promising innovation is Hitachi Energy’s EconiQ technology, which allows the replacement of SF6 gas in existing switchgear with a lower-GWP alternative gas mixture.58 This prevents the need to entirely replace otherwise functioning equipment, which not only can be costly but also creates waste and emissions associated with disposal.59
A phaseout in the European Union of SF6 in electrical switchgear, along with other fluorinated gases commonly used in industrial applications, is underway and is already helping push manufacturers to develop SF6-free components.60 In the United States, a few states, including California, Colorado, and Massachusetts, have enacted regulations to mitigate SF6 production and leaks.61 For the United States, prioritizing the domestic production of SF6-free grid components and working toward eliminating SF6 from grid components is a critical climate imperative, particularly as the grid expands over the coming decades. It also offers an opportunity for domestic manufacturing leadership, as the world hopefully moves toward SF6-free products over the coming years.
HVDC transmission
The most efficient and resilient way to deliver large amounts of energy over long distances is high-voltage direct current transmission.62 Compared with high-voltage alternating current lines, which have dominated in the United States so far, HVDC is more resilient; reduces energy losses; and is better suited to integrating remote renewables, such as offshore wind, onto the grid. In addition, HVDC will be essential to handle the high-density load of emerging data center clusters.63 The United States, which currently has only eight HVDC lines, has already begun investing in initiatives to reduce the cost of HVDC technology and increase domestic manufacturing capacity for its components.64 However, more federal policy is needed to accelerate the growth of the United States’ HVDC manufacturing base and ensure the country is able to expand the HVDC network at the rate needed to meet growing demand.
Direct current is not a new invention: Developed in the 1880s by Thomas Edison, it was the standard in the United States in the early years of electricity.65 However, the difficulty with converting DC between high and low voltages at the time meant that alternating current—developed around the same time—won out, and it has dominated for the past century. Until recently, the conversion problem still posed an expensive challenge, but technological inventions, including solid-state devices called power electronics, have made it easier.66
Building out HVDC will be essential for U.S. national and energy security. HVDC can connect distant parts of the country where electricity is produced to where it is needed, such as bringing energy from massive wind farms in rural areas to cities with high energy demand. This increased connectivity can reduce concerns around energy scarcity, and unlike conventional transmission systems, grid operators can control HVDC systems to direct energy to where it is needed most, balancing supply and demand in real time.67
HVDC lines are emerging around the country
There are several HVDC projects planned for the coming years or recently completed across the United States, many of which are specifically designed to bring clean energy to areas where it is needed most.68 For example, the following three HVDC lines came online in 2026:
- The New England Clean Energy Connect transmission line, a 145-mile HVDC line bringing hydropower from Canada to Massachusetts to power nearly 20 percent of the state’s overall electricity, was completed in January.69
- The Champlain Hudson Power Express, a 1,250-megawatt HVDC transmission line delivering clean energy from Canada to New York City to power 20 percent of the city’s electricity needs, was completed in June.70
- The SunZia project, a 550-mile HVDC transmission line delivering wind energy from New Mexico to Arizona and Southern California, was also completed in June.71
While these projects represent significant progress, the Trump administration has also recently undertaken efforts to slow transmission build-out and grid decarbonization. In August 2026, the DOE announced it was halting work to advance three National Interest Electric Transmission Corridors first proposed in 2024 to speed transmission construction.72 Policy reversals such as these slow the growth of domestic manufacturing by creating uncertainty in the demand landscape, making it hard for manufacturers to plan long-term investments.
The United States needs to lead in manufacturing power electronics
HVDC transmission—and to a lesser extent, AC transmission—relies on power electronics devices such as converters, inverters, and rectifiers.73 These devices manage power flows and convert between alternating and direct current, with more modern devices also having key grid-stabilizing capabilities such as the ability to restore energy after a blackout. Power electronics are also critical for enabling the integration of solar and wind resources onto the grid.
At the core of power electronics are power semiconductors, devices similar to the semiconductors found in personal electronics, except that they process vastly more energy. Semiconductor manufacturing is a highly complex process with hundreds of material inputs going into finished semiconductor products.74 In 2021, the Biden administration noted that some semiconductor components cross international borders as many as 70 times before being integrated into a final product.75
While the United States is a leading holder of semiconductor and power electronics intellectual property, manufacturing capacity for these components is uneven, largely due to technological diversity within the sector.76 Current capacity is insufficient to support the build-out of a grid that fully supports renewables integration and deploys large-scale HVDC transmission. For instance, the capacity for domestic manufacture of solar inverters—small devices that enable the connection of solar panels to the grid—has grown significantly, from 19 gigawatts to 28 gigawatts between 2024 and 2025.77 Manufacturing capacity for HVDC converters—building-sized electrical installations—remains much further behind.78 Recent awards under the CHIPS and Science Act have been successful in onshoring significant semiconductor manufacturing, including power semiconductors, a positive sign for future U.S. power electronics manufacturing.79
Whether investment in domestic power semiconductor manufacturing translates into increased manufacturing capacity for HVDC and advanced AC power electronics is not guaranteed. A lack of domestic demand for HVDC components, driven by a commensurate lack of investment in HVDC projects, has to date prevented manufacturers from making long-term investments in domestic HVDC component manufacturing capacity, forcing existing HVDC projects to rely on suppliers based in Europe and Asia.80 This, in turn, delays project completion and drives up costs. Current delivery dates for new orders of HVDC converters and cables are in the early 2030s.81
Cable manufacturing is creating a bottleneck
Another component of HVDC transmission that is experiencing long backlogs and bottlenecks is direct current cables, which globally take more than five years to procure, significantly longer than AC cables.82 This is particularly problematic for the submarine cable industry, which produces HVDC cables for offshore wind, and has become one of the main constraints for offshore wind projects in the United States and in Europe.83
While there has been some recent progress to build out domestic cable manufacturing plants in recent years, it takes an estimated three to five years to build factories and develop the skilled labor necessary to manufacture cables, so while the investments of today will prevent bottlenecks in the early 2030s, they cannot alleviate those of today.84
HVDC manufacturing is a matter of national security
Sourcing certain digitally connected components domestically is also important for reducing the risk of cyberattacks. Some transmission components are particularly at risk of being compromised, such as inverters, semiconductors, and other power electronics produced in China.85 Some Chinese-manufactured inverters have been found to have undisclosed communications devices embedded, and in July, the Federal Communications Commission (FCC) banned the import of certain new power inverters86 produced in several foreign countries, including China.87 While this risk is not unique to HVDC infrastructure and applies to most digitally enabled technologies across the economy, it is particularly important for the United States to consider as it builds out a nationally connected HVDC network. The United States should incorporate a cybersecurity prioritization framework as it develops industrial policy to bolster HVDC manufacturing capacity.88
In August 2026, following the FCC ban on inverters, President Donald Trump issued an executive order (EO) to prohibit the importation of certain foreign-produced power system equipment deemed to pose a cybersecurity threat.89 Nearly all grid components—including transformers, circuit breakers, and inverters—may be covered by the final rule, which will be drafted and decided on over the next several months. While more than 20 countries are covered by the EO, China is the only one that is a significant manufacturer of grid components.90 The importation of some of these systems may pose a risk, but the wholesale prohibition on sourcing from China may be infeasible in the near term. The creation of a robust inspection, monitoring, and verification program—inclusive of the requirements of the North American Electric Reliability Corporation’s Critical Infrastructure Protection (CIP) standards—could address security concerns in the near term while domestic production capacity is expanded.91
Prior and existing grid component policies
Tariffs on grain-oriented electrical steel
In 2018, following a Section 232 investigation regarding steel imports, the United States imposed a 25 percent tariff on most imported steel products, including GOES.92 A subsequent Section 232 investigation of transformers and their components found that the tariffs on GOES may have contributed to the decline of demand for domestically produced GOES, as domestic transformer core manufacturers shifted operations overseas to avoid having to pay higher input costs for the share of their GOES inputs they could not source domestically.93
Infrastructure Investment and Jobs Act grid resilience programs
The Infrastructure Investment and Jobs Act (IIJA) of 2021 authorized several programs to promote resilience for the grid, including the Grid Resilience and Innovation Partnerships (GRIP) Program and the Grid Resilience State and Tribal Formula Grants program.94 Many of the awards under the original GRIP Program were canceled by the Trump administration, and some were reinstated, and the administration eventually rebranded the program as the Speed to Power through Accelerated Reconductoring and other Key Advanced Transmission Technology Upgrades (SPARK) program.95 The DOE also recently announced $967 million in funding for the Grid Resilience State and Tribal Formula Grants program, which was created with the goal of modernizing critical energy infrastructure.96
CHIPS and Science Act
While often thought of as a bill to onshore traditional semiconductor manufacturing, the domestic manufacturing of power semiconductors—critical to power electronics and discussed above—has also benefited from CHIPS and Science Act funding, including support for advanced silicon-carbide power semiconductors and power semiconductor packaging.97 It is notable that supported projects, to date, have mostly catered to electric vehicle and defense applications, rather than power electronics for grid applications.
Inflation Reduction Act’s Industrial Demonstrations Program
Funded by the Inflation Reduction Act and administered by the DOE, the Industrial Demonstrations Program is providing $75 million to Cleveland-Cliffs to electrify certain processes at its Butler, Pennsylvania, facility, which is a major producer of domestic GOES.98 This will expand the yield and reduce emissions at the site.99
DOE Grid Supply Chain Program
In 2026, the DOE announced plans to enhance the domestic supply chain for manufacturing of key grid components, including distribution and power transformers and the materials needed. This effort will include promoting greater refurbishment and reuse of existing materials as well as promoting greater standardization of grid equipment.100
Raw material inputs
Grid component manufacturing requires significant mineral inputs, namely aluminum, copper, nickel, chromium, tungsten, germanium, and polysilicon.101 Additionally, certain components make use of naturally and artificially created minerals or other compounds, including mica, ceramic, and various polymers—all largely used for electrical and thermal insulation.102 Domestic supply chains for these inputs vary significantly in development. The United States is a major primary producer, processor, and recycler of copper, but it has only one primary nickel mine and essentially no processing capacity.103 Where excessive supply chain concentrations exist, the federal government should make efforts to diversify and onshore production, especially processing and recycling capacity, areas where the United States is often almost completely dependent on foreign actors.104 Recent efforts by both the Biden and Trump administrations have made some progress in developing more diversified raw material supply chains. And while more progress is needed in addressing mineral processing bottlenecks, those bottlenecks are more pronounced in other supply chains than those for grid components.105
Regardless of where they are sourced from, changes to how the United States acquires raw materials for grid components must address the potential harms of mineral extraction and processing. Historically, mineral mining and processing operations have led to environmental degradation as well as human rights and labor abuses.106 Decades of lax regulation and harmful outcomes have eroded public trust in the mining and processing industries, and mines continue to be proposed in environmentally sensitive areas.107 Efforts to onshore or otherwise secure raw material supply chains must ensure that these harms are not repeated and address past harms. Specifically, the following should be prioritized:
- Reuse and recycling. The reuse and recycling of already-in-circulation materials should be prioritized to avoid the need for more primary mining. Recycling is already a well-established process for materials such as aluminum and copper, where significant portions of the domestic supply are already met by recycling.108 However, for some materials, recycling may require new processes and related policy support, including to mitigate adverse environmental and labor outcomes that have plagued some recycling operations and techniques.109
- Human, labor, and environmental rights. Materials sourced for grid components should come from operations that meet the highest human, labor, and environmental standards.
- Responsible revenue collection. Affected communities should benefit from the mining industry through direct financial support in the form of payments for the extraction of publicly owned resources. This includes the responsible collection and stewardship of such payments in a manner that does not create a resource dependency.110
The supply chains for various raw mineral inputs and their associated finished products vary significantly in length, diversity, and development. A policy intervention that works for one supply chain will not necessarily work for another. Similarly, broad-based support for only one section of the supply chain, such as mining, is likely insufficient to achieve desired outcomes such as onshoring. Mining is not the bottleneck in many American mineral supply chains—a lack of domestic processing capacity and expertise is a much more significant factor.111 Reforms to mineral supply chains must be appropriately targeted and consider the unique needs of the specific finished product.
Recommendations to increase the domestic manufacturing capacity of key grid components
The United States does not currently produce enough materials and components to support its electric grid, holding the nation back from building an advanced, clean grid that supplies reliable and affordable electricity. This can be alleviated through a multifaceted industrial policy—a five-part plan that attempts to mitigate the bottleneck of available components, put Americans to work, and get electrons flowing.
The five-point plan spans: 1) initiating supply increase; 2) backstopping demand; 3) addressing material shortage; 4) supplying and supporting a high-quality workforce; and 5) planning and evaluation. This plan builds off recommendations from years of research from sources both within and outside government. The current federal government has taken actions within this blueprint—such as the utilization of the Defense Production Act (DPA) and the instigation of tariffs—but the effort is haphazard and incomplete. This plan aims to coalesce the full slate of recommendations into an achievable blueprint for a willing federal government.
Spurring rapid supply increase
Ramping up domestic production is an immediate, near-term need given the multiyear backlog of critical equipment and the excessive reliance on imported goods. Speeding that ramp-up requires a large injection of direct federal investments to kick-start a manufacturing ecosystem of grid components. The investment portion should rely on direct grantmaking via the Department of Energy and production tax credits. These types of interventions proved to be successful at initiating manufacturing investments after the passage of the Inflation Reduction Act.112 Additionally, the U.S. Department of Defense has access to grantmaking specifically to keep the defense industrial base strong. Bolstering the domestic production of power grid components qualifies, according to a recent presidential memorandum.113 The federal government also should not assume that investments alone will spur a new, better ecosystem for grid component manufacturing. Nonmarket measures, such as those authorized under the Defense Production Act, may be necessary to move the private sector hastily. Finally, one of the difficulties facing manufacturers of grid equipment and their components, especially newer entrants into the supply chain, is that a lack of standardization in transformer design114 and HVDC equipment115 has made automation and mass production difficult.
In order to spur rapid supply increase of key components, the authors recommend the following:
Create a fund to facilitate large-scale, consistent investment
- Congress should establish a fund, administered by the DOE, to deliver large initial grants to manufacturers to subsidize capital costs associated with beginning or expanding production of critical grid components, notably prioritizing large power transformers, electrical steel, and HVDC transmission cables and converters.
Utilize the Defense Production Act
- The Defense Production Act empowers the executive branch with an array of authorities to shape national defense not only through programs for the military but also for energy production, emergency preparedness, stockpiling, and other activities in civilian markets.116 Title I notably empowers the president to compel private businesses to accept contracts. This authority should be strongly considered to hasten initial capital expenditures and get new and expanded grid component production up and running.
- DPA Title III allows the president to execute financial incentives to alleviate supply chain vulnerabilities. President Biden invoked DPA Title III in 2022 to address the high reliance on critical electric grid imports and support an increase in domestic production.117 President Trump followed suit with his invocation of Section 303, aiming to unlock funds to support grid infrastructure needs, including components.118 The One Big Beautiful Bill Act also allocated $1 billion for DPA use. Much more is needed. The Energy Bills Relief Act, introduced in March 2026 by Reps. Mike Levin (D-CA) and Sean Casten (D-IL), could begin to address this need by appropriating $2.1 billion to DPA Title III to expand domestic production of transformers and grid components, including GOES and amorphous steel.119
Expand the use of the Industrial Base Fund
- Established by the fiscal year 2011 National Defense Authorization Act (NDAA) and expanded by the fiscal year 2026 NDAA, the Industrial Base Fund exists to ensure robust domestic supply chains for needed materials.120 This includes advanced manufacturing and power sources. The language should be clarified to ensure the inclusion of power grid components, and then funds should be appropriated to support the production of components and materials.
Standardize transformer and HVDC design and equipment
- According to a report by Johns Hopkins University, standardizing HVDC transmission equipment can minimize supply chain challenges, maximize forward compatibility between systems, accelerate investment in domestic supply chains, and eventually support the adoption of global HVDC standards.121 This will require collaboration from the federal government, state governments, and the private sector. The Department of Energy should take charge of instigating and guiding this collaboration by convening key stakeholders from states, Tribes, and industry to develop recommendations for standardization. From there, the federal government should adopt these standards and apply them to further rulemaking and investment decisions.
- A similar process may be undertaken to standardize transformer design after further study. Given that utilities currently procure custom designs due to varying demands, additional study and research may be needed to identify parts of the transformer design and production process that can and should be standardized.122 The DOE should spearhead this research in collaboration with industry, and it should be done as the effort to convene states, Tribes, and industry is launched.
Backstopping demand
The lack of expanded grid component manufacturing capacity is in part due to a history of fluctuating demand, which has made investing in building out supply a risk for manufacturers. The uncertainty around data center growth and new electricity demand, while assuredly requiring an increase in domestic capacity, creates further confusion around the degree to which capacity should be expanded. Building and expanding an industry that has sustainable high-quality production will require consistent demand. The market may naturally provide this demand. However, for the industries identified in this report, there are challenges—supply chains dominated by China123 for existing production as well as supporting124 innovations to reach commercialization—that necessitate public intervention to spur demand, notably for SF6-free grid components. Additionally, in a globalized world, exposure to products made with another nation’s subsidies or under a different regulatory environment, nefariously or not, can force a manufacturer to compete with goods priced lower due to unfair trade policies. A holistic industrial policy requires utilizing trade enforcement to ensure that the investments are not undermined. Policies associated with this intervention include creating demand guarantees to ensure certain products are purchased and subsidizing production costs to help producers keep those products cost effective.
Create demand certainty
- To ensure the industry can persist through market and geoeconomic fluctuations, the federal government should provide long-term procurement guarantees. In addition to utilizing and applying the Build America, Buy America Act,125 one particular method for doing so comes from the president’s National Infrastructure Advisory Council, which recommends establishing a strategic virtual reserve of transformers.126 This is not a physical reserve but instead an economic tool that allows the federal government to step in as the buyer of last resort. It also recommends convening all parties who drive demand to achieve more accurate 10- to 15-year forecasting and encouraging long-term contracts between transformer manufacturers and utilities and other transformer purchasers. These last two recommendations should include the U.S. Department of Commerce as a convener to ensure a consistent and fair process and results.
- As mentioned above, the president should fully utilize the authorities given under Title I of the Defense Production Act. This could include compelling electric utilities to purchase a growing percentage of domestically produced transformers as the industry ramps up.
- In addition to providing long-term procurement guarantees, the federal government should consider the creation of a new federal authority to build, own, and operate a national HVDC network. This naturally would involve the procurement of large numbers of transformers and other grid components, provide the necessary demand signal for long-term HVDC supply chain investment, and provide significant grid stability benefits when completed. National HVDC networks have been studied before, notably in the DOE’s 2020 Interconnections Seam Study.127 Similarly, states have established state transmission authorities to help get transmission projects built.128 These authorities should prioritize the procurement of domestically produced grid components.
Subsidize production costs
- Section 45X of the federal tax code—the advanced manufacturing production credit—provides a tax credit for producing specific goods. It was passed in 2022 as part of the Inflation Reduction Act and has proved successful in boosting production of numerous goods.129 The credit is already applied to inverters, as well as primary aluminum, and consideration should be given to increasing the value of those credits. In addition, the credit should be applied to the production of LPTs, GOES, and HVDC transmission components.
- Grid component manufacturers also should be given an investment tax credit to cover the upfront costs of equipment and building a manufacturing facility. This credit should support the build-out of facilities for the production of LPTs, GOES, and HVDC transmission components as well as the production of SF6-free components.
Utilize the trade enforcement toolkit
- It will be critical to utilize Section 232 of the Trade Expansion Act to initiate an investigation by the Department of Commerce to determine that key grid components and associated inputs are critical for national security and thus eligible for tariffs to be imposed. Initiating an investigation under Section 301 of the Trade Act of 1974 will be equally critical as well as considering anti-dumping and countervailing duties to overcome unfair foreign trade practices by specific nations.130
What action should states take?
Building out the manufacturing capacity for grid components does not need to be solely undertaken by the federal government. States can utilize planning, procurement, and investment tools as well. The scale will not be the same as for the federal government, but these efforts can support growing a domestic industry.
One example currently stands out. The California Grid Manufacturing Initiative is an innovative effort considered by the California legislature as A.B. 2516 in 2026. It initiates a statewide assessment of grid equipment needs and demand, authorizes the state to pursue public-private joint ventures, and leverages lower-cost public financing tools such as structuring procurement agreements and preferred supplier arrangements. Effectively, it combines California’s procurement apparatus with strategic investments to help fill gaps and, ideally, to ignite grid component manufacturing in the state. For more, see the report by the Climate and Community Institute, United Auto Workers Region 6, BlueGreen Alliance, and the Center for Public Enterprise.131
Addressing material shortage
Addressing the needs for building key grid component products can be undercut if a similar effort is not made to ensure that the raw materials that make up those products are sufficiently available. As such, a grid component industrial strategy must identify and support the production of key raw materials domestically and build out an effective trade strategy for those not produced in the United States. In particular, the production of GOES and aluminum should be expanded domestically utilizing direct financing as well as incentives, and the federal government should undergo a significantly deeper commitment to circularity to extract raw materials from goods and components at the end of their lifespans and reduce the environmental impacts of primary material production. Attention should also be paid to onshoring polysilicon production. While polysilicon is the beneficiary of a recent price floor established following a Section 232 investigation, future action may be necessary to fully onshore it and support advanced power semiconductor manufacturing.132
Develop raw material strategy
- In addition to continuing and expanding support for key raw materials under Section 45X, the federal government should provide additional direct investments—through DOE-administered grants—toward primary aluminum and GOES production, with the assurance that a necessary percentage is utilized for transformer production or otherwise purchased by the federal government to be resold for transformer production.
Create domestic circular supply chains
- As mentioned above, expanding grid component manufacturing will require significant raw material inputs, the extraction and processing of which will raise environmental and human rights concerns. Wherever feasible, demand for raw materials should be minimized through the creation of domestic circular supply chains—with a critical emphasis on mitigating labor and environmental hazards—that recycle end-of-life components and production scrap for reuse in new components. An estimated 99 percent of transformer materials can be recycled, and researchers have already developed at least one process to recycle transformer core materials for reuse in new transformers.133 Federal support for recycling research should be significantly expanded, and incentives for transformer production should be expanded for manufacturers who utilize recycled components.134 A recent funding opportunity by the DOE for projects that promote the refurbishment and reuse of grid components and materials is a good start toward this goal.135
- Alongside recycling, raw material demand reduction should entail ensuring that components are well-maintained and refurbished to extend the lifespans of transformers and other components—particularly LPTs, due to how expensive and difficult they are to replace. According to Hitachi Energy, refurbishment and repair of transformers can extend lifetimes by 20 to 30 years.136 The existing transformer supply backlog has already pushed some transmission developers to adopt refurbishment practices.137 The standardization of transformers—particularly LPTs, which are typically customized—would make refurbishment easier, as spare parts would be more available and local skilled workers would be more familiar with replacement processes.
- In order to make refurbishment and remanufacturing more efficient, transformers—particularly LPTs—should be designed with modularity in mind, with standardized subunits that are easy to replace. Therefore, when one part breaks or needs repair, that subunit can be replaced without needing to replace the entire unit. Further, when the transformer does reach the end of its life, a modular system is easier to disassemble and recycle so that valuable materials and minerals can be reused in new units.
Supplying and supporting a high-quality workforce
As recently as June 2024, the National Infrastructure Advisory Council publicly noted: “the domestic transformer manufacturing industry has had difficulties attracting and retaining qualified workers. The previous cyclical nature of the industry has made manufacturers wary of increasing capacity despite rising demand.138 Domestic LPT manufacturers report difficulty in hiring qualified workers, and the workforce is aging due to underinvestment in the specialized skill set required for these jobs, which have gone offshore for years.139 LPT manufacturing requires specially trained engineers, and few universities teach the required specializations.140 Many engineers are choosing to work in fields other than power electronics that may be more lucrative, such as software engineering. LPT manufacturing also requires workers with vocational training, which is also declining, and who would be willing to live in relatively remote locations.141 Even if there were not an existing workforce shortage, a consistent pipeline of people moving into grid component manufacturing over the next generation would be needed to sustain the domestic workforce required to support the coming grid expansion.
Ensuring that people are attracted to this industry now and in the future will require the industry to create good jobs to attract workers with many other choices. To ensure a steady supply of well-qualified workers and prevent labor disruptions, the government should mandate that companies that receive federal investments and incentives abide by neutrality or proactively sign memoranda of understanding that lead to collective bargaining agreements. Direct collaboration with unions ensures high-quality wages and benefits and a voice on the job. The federal government should explore collaborative efforts with unions and manufacturers to support registered apprenticeships, preapprenticeships, or other sorts of joint labor-management training programs. Inflation Reduction Act and CHIPS investments can serve as a model for these future legislative efforts.142 Similarly, public investments such as grants and tax credits that support the construction of new manufacturing facilities should require that construction jobs also include job quality standards, such as Davis-Bacon prevailing wage and apprenticeship utilization requirements.
Data collection, future planning, and coordination
Properly functioning industrial policy requires clear-eyed planning and coordination of industrial outcomes, activities which themselves require a full understanding of current capabilities and capacity. For infrastructure as indispensable as the electric grid, the lack of consistent federal direction setting for industry has led to deteriorating equipment and shrinking domestic production capacity. This has to change to ensure that the country does not waste investments and find itself in a similar or worse position a few years down the road. The federal government should:
- Collect adequate data from the nation’s utilities and manufacturers every few years. The government should have a clear, timely picture of the country’s industrial capacity and current output for certain grid components, alongside a full understanding of where those components are deployed and how they are operated.
- Evaluate future demand for grid components.
- Ensure that future demand can be met through the appropriate balance of domestic industry and imports.
Given that, this report proposes that any grid component industrial policy incorporate a multiagency study and reporting process, managed and held at the Department of Energy, that launches alongside the investments and programmatic changes, offering a significant overview in the year following passage. An office within the DOE should be appropriated funds not only to manage the process but also to make immediate initial disbursements to support conclusions without waiting for lengthy political debates. From there, this multiagency process should provide a consistent review and analysis every two years over a 10-year span. Below is the initial construct and framework that the Center for American Progress proposes:
Study and reporting
The departments of Energy and Commerce, with advice from the U.S. trade representative and departments of Treasury and Defense, should be jointly tasked with examining and then publicly reporting on:
- Domestic needs and production capacity, at present, and providing a five-year scope. Notably, this should include a comprehensive study of the number of LPTs and other essential transformers on the grid. These data should be updated regularly and be used to inform policy on increasing domestic manufacturing capacity of transformers.
- Global supply chain trade flows and any associated vulnerabilities or excessive dependencies.
- Analysis of existing federal authorities and how they should be utilized to address the availability of grid components in congruence with any additional investments and programmatic changes underway.
This should not delay initial investments and policymaking but instead should be incorporated into evaluations and updates to programs instituted by the rest of the five-point plan.
Conclusion
The United States needs a broad, long-term industrial policy that covers multiple industries over many years. It also needs significantly more grid components. Too heavy a reliance on foreign-made products will expose the economy to geopolitical leverage, give leeway to exploitative companies and governments, forgo positive job-creating economic effects for people and communities, and increase the nation’s cybersecurity risk. With a committed five-part industrial strategy, the United States can build out sufficient domestic production capacity of large power transformers and HVDC transmission, as well as key inputs such as grain-oriented electrical steel and power electronics. Doing so will be essential to build a next-generation electrical system that supports a highly functioning economy.
Acknowledgments
The authors would like to thank Ryan Mulholland, Karla Walter, Jonathan Fritz, Fiona Fitzpatrick, Cathleen Kelly, Anh Nguyen, Bill Rapp, Hannah Mast, Trevor Higgins, Mark Haggerty, Alia Hidayat, Devon Lespier, and Frederick Bell of the Center for American Progress, as well as Shannon Baker-Branstetter, formerly of CAP, Zoe Lipman at the International Brotherhood of Electrical Workers, and Tom Lewis of the BlueGreen Alliance, for their contributions to this analysis.
Methodology
For Figure 2, the authors relied on specifications and harmonized tariff schedule (HTS) codes from both Figure III-1, “Product Scope of the Investigation,” in the Department of Commerce Bureau of Industry and Security report on the “Effect of Imports of Transformers and Transformer Components on the National Security,”143 and Table 4, “HTS Code and Import Highlights by Component,” in the National Laboratory of the Rockies (NLR) report on “Large Power Transformer Supply Chain Gap Analysis and Domestic Content Strategies for Hydropower Rehabilitation.”144 Since these reports did not specify the HTS code for aluminum, the U.S. Geological Survey’s aluminum publication from February 2026 was used instead.145 Based on a combination of these tables, the following HTS codes were used to determine which countries the United States imports components and materials from, via the U.S. International Trade Commission DataWeb146 tool.
As stated in the NLR report, limitations exist on using HTS codes to evaluate component imports, as CTCs, OLTCs, and bushings do not have unique HTS codes and are tracked under broader product categories. When using DataWeb, the authors searched for 2025 import data and the general customs value. Classifications of importance in Figure 2 (low, medium, high, and very high) were based on authors’ judgement relative to the other materials included in the table. The table is not inclusive of all grid components or materials.
For Figure 3, the authors used a similar process via DataWeb by HTS code to determine both imports (General Customs Value) and exports (Exports: Total, FAS Value) from 2016 to 2025 by transformer type. Here, small, distribution, and dry-type/other transformers were broken out as shown in the HTS code table above.