A transformer market has more than one unit of growth.

Choose the market before choosing the forecast.

For a transformer manufacturer, “grid investment” is too broad a demand signal. Distribution and transmission equipment have different ratings, procurement paths and production requirements. This perspective concerns the US distribution fleet described in a November 2024 NREL fact sheet: its scope is not the global transformer market.

The study separates the fleet by phase, mounting arrangement and rating. Its estimated mix shows why a count-based forecast and a capacity-based forecast can produce very different priorities. [1] Neither measure establishes a manufacturer's obtainable revenue. That requires a further match to specifications, approvals, factory capability and competitive position.

Translate grid requirements into an order mix.

An additional megavolt-ampere can be supplied through different combinations of equipment. New connections create equipment requirements; higher loads at existing connections require a review of ratings and available capacity. A manufacturer therefore needs a bridge from the load location to the required product class before calculating equipment demand.

Three movements deserve separate treatment: expansion of the installed fleet, replacement of existing assets, and a change in the rating or specification of what is bought. Combining them into a single growth percentage hides both product mix and the timing of procurement.

Revenue then adds another layer: units by class multiplied by realised prices for that class. Equipment size, specification mix and like-for-like prices each contribute to the average invoice. Those explanations should be tested separately.

A bottleneck changes the meaning of demand.

DOE records a shift in reported distribution-transformer lead times from 3–6 months in 2019 to 12–30 months in 2023. It also describes extensive variation in utility specifications. These are dated observations, not a claim about today's delivery times. [2]

During a shortage, orders can contain precautionary purchasing and deferred projects as well as underlying expansion. Once supply improves, order growth may slow while deliveries increase. A useful historical series would distinguish requests for quotation, firm orders, backlog, shipments and energised installations, with cancellations and rescheduling recorded.

For capacity planning, this changes the question from “How large is the shortage?” to “Which constraints persist for the product classes our factory can qualify and deliver?”

What would change the investment judgement?

A case for additional capacity becomes stronger if demand remains after cancelled and duplicate orders are removed, customers accept the proposed specifications, and a credible qualification schedule connects the factory to deliverable orders.

It weakens if backlog is concentrated in projects unlikely to proceed, standardisation shifts demand away from a proposed product, or the factory's bottleneck lies outside the new equipment being purchased.

The next research deliverable should therefore be a product-class demand bridge, a dated backlog reconciliation and a constraint map. National capacity growth is the context for that work, not a substitute for it.