Eighty Weeks for a Transformer

Lead times for large power transformers, the units rated above 100 MVA that connect generation to transmission and transmission to distribution, have doubled. In the United States, transformer prices have risen 79%. Some transmission-class units now require three to six years from order to delivery. The machines that make transformers take two to four years to procure themselves, a recursive bottleneck with no shortcut.

The constraint is physical. Factories that build grid equipment cannot keep pace with the volume of projects now competing for their output, and neither policy nor financing can compress a production timeline set by specialized manufacturing.

The demand side. The United States plans to add 86 GW of new generating capacity in 2026, roughly 99% of it solar, wind, or battery storage, according to EIA data visualized by Canary Media. Each of those projects needs grid interconnection equipment: transformers, switchgear, breakers, protection relays. So do the data centers that Nvidia CEO Jensen Huang described last week when he told investors that “everybody is power limited.” Nvidia posted $68.1 billion in quarterly revenue. The AI buildout it enables is competing with renewables and storage for the same pool of electrical hardware.

The supply side. Manufacturers have committed roughly $2 billion to new transformer capacity since 2023. Hitachi Energy accounts for $1 billion of that. Siemens Energy is building a new plant in Charlotte, North Carolina, expected to be operational in early 2027.

It is not enough. Transformer factories were underutilized before the current demand cycle, which has made manufacturers cautious about investing in capacity that could become stranded if demand proves cyclical. Manufacturers want guaranteed order books before they build factories for equipment that takes years to produce. The memory of overcapacity is recent enough to discipline capital allocation even in the face of clear demand signals.

The gap. Demand for transformers is projected to exceed 2024 levels by 21% in 2026. New manufacturing capacity takes years to build. The result is a queue that grows faster than factories can clear it.

Framework agreements between manufacturers and large buyers provide the demand visibility manufacturers need to justify expansion. Standardized substation designs can reduce custom engineering timelines. These are sensible incremental responses. They do not solve a doubling of lead times.

The legislative response. Senators Dave McCormick (R-Pa.) and Peter Welch (D-Vt.) introduced the REWIRE Act this week, targeting grid expansion through reconductoring: replacing aging transmission wires with advanced conductors that can double or triple transmission capacity on existing rights-of-way. Reconductoring reuses existing towers, existing easements, and existing environmental permits. It avoids the decade-long siting battles that have stalled new transmission corridors across the country. Whether the bill survives committee is uncertain, but the bipartisan sponsorship reflects a Washington consensus that new transmission lines are politically unviable on any timeline that matches demand growth.

Reconductoring addresses the wire. It does not address the transformer at either end of it.

What deploys anyway. The 58 GWh of energy storage the United States installed in 2025, a record, did not require 58 GWh worth of new transmission-class transformers. Utility-scale battery projects do need substation equipment and interconnection hardware, and many face the same queue as solar and wind farms. But behind-the-meter commercial storage operates within the existing electrical infrastructure of a building. It connects behind the meter, behind the existing service transformer, behind the existing switchgear. A 250 kW battery system in a commercial building does not require a new transformer order. It requires an electrician and a permit.

This distinction, invisible in normal markets, becomes material when transformer lead times exceed two years. A commercial building owner deciding between a grid upgrade and a behind-the-meter battery is no longer comparing just cost. The owner is comparing 80 weeks of waiting against 8 weeks of installation. The transformer shortage has turned time-to-deployment into a competitive advantage for distributed resources.

The same logic applies to distributed solar, demand response, and building electrification projects that operate within existing service capacity. Anything that avoids triggering a utility infrastructure upgrade avoids the queue.

The deeper problem. The energy transition is often discussed as a contest between political will and economic gravity. Policy incentivizes. Markets respond. Deployment follows. The transformer shortage reveals a third variable that neither policy nor markets can accelerate past a certain point: industrial production of specialized heavy equipment by a concentrated group of global manufacturers who remember the last time they overbuilt capacity.

A handful of companies, Hitachi Energy and Siemens Energy among them, control the bulk of large power transformer production worldwide. Their capital allocation decisions, shaped by memories of post-2008 overcapacity, now determine the pace at which grid investment can be physically executed. The bottleneck sits on the factory floor.

The projects that find a way around it will define the next three years of the transition. The ones that cannot will wait.

Sources