Hitachi puts $528 million on transformers as the grid's constraint

Hitachi Energy is spending $528 million on a transformer factory in Gallman, Mississippi, that will more than double the company's US production capacity for small and medium-sized power transformers — the 10 MVA to 160 MVA units, rated for voltages up to 230 kV, that move power out of substations and into load. The company announced the project Tuesday and expects to begin production in 2029.

At its core the plant answers a shortage that has been documented for years. Backlogs for transformers and other grid equipment are slowing expansion of the power grid at the same moment data centers, manufacturing, and electrification are pulling demand upward, and Hitachi is funding the capacity on its own balance sheet rather than waiting for a utility's rate case to pay for it. That is where the capital stack has moved: the scarce input in the power buildout now sits upstream of the project, and the component maker is the one carrying the fixed cost.

Gallman is Hitachi Energy's largest single US investment, lifting its American manufacturing expansion commitment to roughly $1.5 billion, and it sits alongside capacity additions at plants in Pennsylvania, Virginia, and Tennessee. The new building will be more than twice the size of the company's Crystal Springs, Mississippi, facility, and production will transition from Crystal Springs to Gallman once the new lines are complete. Greg Callahan, the senior vice president who heads Hitachi's North American transformer business, told Utility Dive the company does not disclose production volumes by facility; the size of the investment, he said, reflects the scale of the expansion.

That leaves a capital cost as the only hard number on the table, since the report names no customers, order commitments, or state incentives behind the $528 million. Customer lead times, Callahan notes, vary with transformer specifications, project complexity, and market conditions — an honest answer that also concedes the shortage is not one uniform queue. For anyone trying to underwrite an interconnection date, a check size says how much Hitachi believes in the shortage, not how many megavolt-amperes the plant clears from it.

Which brings the vintage problem. Capacity that begins producing in 2029 cannot shorten a lead time quoted in 2026, and by the company's own framing the new supply eases industry-wide constraints only as it comes online. Hitachi is building for the second wave of load growth while holding an option that appreciates through the first. As this publication has argued, the grid's bottleneck is a price on new supply; Gallman attaches a delivery date to one node of it. On this beat the gap between the announcement and the first megavolt-ampere is now measured in years, and that gap is the thing developers and data-center sponsors are actually financing. Hitachi's answer arrives in 2029.