zevOS

01Business

The unit economics of a public charger in India

What a DC fast charger actually earns, what it actually costs, and the three variables that decide whether a site pays back in three years or never.

ZOzevOS Editorial · Platform team
5 August 2026 · 4 min read

Charging businesses fail for one of two reasons: they built where the cars are not, or they priced without knowing their landed cost of energy. Both are arithmetic problems, and both are solvable before the civil work starts. This is the model we walk new operators through.

The revenue side has exactly three variables

Revenue per charger per month is utilisation × capacity × price. Everything else is detail.

  • Utilisation — the share of hours the connector is actually delivering energy. On a new public site this is 3–6%. A mature urban DC site reaches 10–15%. Above 20% you have a queueing problem, not a utilisation problem.
  • Capacity — the rated power, derated for reality. A 60kW charger rarely averages 60kW across a session; vehicle taper and state of charge mean 35–45kW average is more honest.
  • Price — your ₹/kWh to the driver, net of GST and net of promotions.

A 60kW DC charger at 8% utilisation, averaging 40kW when active, delivers roughly 60 × 0.08 × 24 × 30 ≈ 3,456 kWh a month if you use rated power — but closer to 2,300 kWh once you derate for taper. At ₹20/kWh net, that is about ₹46,000 of monthly revenue per charger. The two-thirds difference between those two calculations is where most business plans go wrong.

The cost side is where the surprises live

CostTypical shapeNotes
Energy₹8–11/kWh commercialThe single largest variable cost; check whether your slab moves at higher consumption
Demand charges₹150–400/kVA/monthCharged on sanctioned or peak demand, not on units consumed — this is the cost that kills low-utilisation sites
Site rent / revenue share8–20% of session valueOr a fixed monthly rent, which is far riskier before utilisation is proven
Payment gateway1.5–2.2% of collectionsUPI is cheaper than cards; the mix matters
Platform feePer session or per kWhShould scale with revenue, not sit as a fixed licence
Maintenance & field ops₹3,000–8,000/charger/monthDominated by dispatch frequency, not by parts
Connectivity₹200–400/charger/monthPlus the cost of the outage when the SIM lapses

Demand charges are the trap

Energy cost scales with usage; demand charges do not. A 60kW connection attracts its demand charge whether you sell 500 units or 5,000. At ₹250/kVA on a 75kVA sanction, that is around ₹18,750 a month before a single electron moves — roughly 40% of the revenue in our example above.

This is the arithmetic reason why an under-utilised DC site loses money even at a healthy margin per unit, and why the single highest-leverage decision in the whole model is site selection rather than pricing. It is also why smart charging matters commercially and not just technically: if charging profiles let you serve the same demand from a smaller sanction, the saving is immediate and permanent.

Working the model backwards

Rather than projecting revenue, solve for the break-even utilisation. Take your fixed monthly costs — demand charge, connectivity, fixed rent, allocated field ops — and divide by the contribution per kWh (price minus energy cost minus variable share minus gateway minus platform fee). That gives the kWh per month the site must move to break even; convert to utilisation and ask honestly whether the location supports it.

In our illustration, contribution is roughly ₹20 − ₹9.50 energy − ₹3 site share − ₹0.40 gateway − platform fee ≈ ₹6.50/kWh. Against ₹22,000 of fixed cost, break-even is about 3,400 kWh a month — which at a derated 40kW average means around 85 charging hours, or 12% utilisation. If the site cannot plausibly reach 12%, no pricing change will save it.

The three levers that actually move payback

  1. 01Location. Utilisation is a property of the site, and no operational excellence compensates for the wrong one. Traffic that stops for 30+ minutes anyway — food, retail, transit interchanges — beats traffic that merely passes.
  2. 02Sanctioned load per connector. Every kVA you do not need is a permanent monthly saving. Load sharing across connectors is the cheapest capex reduction available.
  3. 03Uptime. A charger that is unavailable 8% of the time loses 8% of a revenue line whose costs do not fall correspondingly. Reliability is a margin lever, not a customer-service one.

Price is deliberately not on that list. It matters, but it is the lever operators reach for first and it is the one with the least room — you are bounded below by your energy cost and above by the driver’s alternative, which in India is frequently a home socket at ₹8.

Where AC sites differ

The whole model inverts for AC. Capex is a tenth, demand charges are trivial, and utilisation can be much higher because dwell time is long. But revenue per connector is small, so AC networks are volume businesses: they work at fifty sites and fail at five. The operational implication is that AC economics live or die on how little attention each site needs.

unit economicsROIpricing

Get the next one by email

One email a month. Operations notes, unit economics and policy changes that affect your tariff.