Every presentation about Indian EV adoption eventually shows the same map: fourteen thousand fast chargers by 2027, strung along the golden quadrilateral and the major freight corridors like beads. The maps are not wrong about where the chargers will go. They are wrong about when, because nearly every bead on that string depends on a piece of equipment that almost nobody in the EV industry orders, stocks or thinks about: a distribution transformer.
Between March and December 2025 we surveyed 1,180 proposed highway charging sites across nine states, working backward from tender documents, discom sanction letters and the commissioning data the charge-point operators file. The finding is simple. The charger hardware is not the bottleneck. Site acquisition is not the bottleneck. The sanctioned load is.
Where the queue forms
A typical four-gun highway hub needs a dedicated 250 to 400 kVA transformer. In the states we surveyed, transformer procurement lead times have lengthened from fourteen months in 2023 to twenty-two months today, driven by the same distribution-reform schemes that are finally fixing discom balance sheets — every discom in the country is upgrading its network at once, and the winding shops are full.
The consequence shows up in the commissioning data. Of the sites in our sample that had hardware installed by September 2025, 41 percent were still waiting on their sanctioned connection four months later. The industry calls these dead sites, and the drivers' apps, which mostly do not list them, disguise the scale of the problem.
India does not have a charger shortage. It has a transformer queue, and the queue runs through the same discoms that are simultaneously rebuilding half the grid.
There is a second ceiling hiding inside the first. Many corridor sites receive their connection at 400 volts rather than the 11 kV feed the site design assumed. A 400-volt connection caps practical hub output around 180 kVA. The operator installs four guns anyway — the tender demands it — but only two can run at full rated power simultaneously. Ask around the industry and everyone has a story about a 240-kilowatt hub that behaves, at peak, like a 90-kilowatt one.
What breaks the ceiling
Three fixes matter and only one is expensive. The first is administrative: Maharashtra and Gujarat now process dedicated charging connections through a single-window cell, cutting sanction time by a third. The second is inventory: two of the largest charge-point operators have started pre-ordering transformers against their tender pipeline, a dreary working-capital decision that will do more for corridor coverage than any software feature.
The third fix is the expensive one. High-capacity corridors eventually need 33 kV feeds and that is grid capex, with grid timelines. Our modelling suggests the realistic 2027 figure is not fourteen thousand corridor fast chargers but about nine thousand fully powered ones, plus a long tail of partially powered sites that will frustrate drivers for years.
None of this is an argument against the rollout. The queue is proof the rollout is real. But the next time a deck shows beads on a string, ask about the copper.
