EV Demand Charge Management: The Bill Forecourts And Fleet Depots Did Not Budget For

EV Demand Charge Management

Most EV charging business cases are built on energy: so many kWh a day, at so much margin per kWh. Then the first full bill arrives carrying a second number — a charge for the highest power the site drew in one short interval, however quietly it ran for the rest of the month. EV demand charge management is the work of controlling that number, and at the utilisation real sites see, it moves the P&L further than tariff design does.

Two segments read this differently, so each section says which one it addresses. Fuel retailers and forecourt operators are adding fast charging to a site that already has a metered load and an established peak. Enterprise and fleet operators charge their own vehicles behind their own meter — which, as the tariff categories below show, is not the same situation at all.

One clarification, because the terms blur constantly: a demand charge is a cost your distribution utility bills you for peak power drawn; demand response is a payment for cutting that draw when the grid asks — a cost line and a revenue line, and our guide to utility demand response participation covers the second. This post is about the first, and load management basics covers the underlying control if the idea is new to you.

What A Demand Charge Actually Is

An electricity bill has two halves. The energy half pays for what you consumed, metered in kWh or kVAh. The capacity half pays for the most you asked the network to stand ready to deliver — the highest average power over a short interval, usually 15 to 30 minutes, anywhere in the billing month. Draw 100 kW for twenty minutes on one Tuesday and that month’s capacity charge is set, whether the site then sits idle or runs flat out.

One mechanic, two sets of words

In the US and much of Europe it is a demand charge, billed per kW of monthly peak and often ratcheted, so one summer peak stays in the bill for months afterwards. In India it arrives as sanctioned load and contract demand — the capacity you contract for — plus maximum demand charges in rupees per kVA per month, with a billing-demand formula doing the ratchet’s job. Maharashtra’s current order is a clean example: billing demand is the higher of 65% of the maximum demand recorded between 06:00 and 22:00, or 40% of contract demand. That second term is a floor, payable on capacity you may never use.

Why the number gets so large

Demand charges scale with peak power while revenue scales with energy sold, and the two diverge exactly where charging networks live today: low utilisation. RMI’s analysis of US public fast charging found demand charges can be responsible for over 90% of a station’s electricity costs, with peak demand varying only about 16% month to month while energy sales swing by as much as 70%. The fixed half of the bill does not fall when traffic does.

Why One Fast Charger Resets A Forecourt’s Fixed Bill

Fuel retailers and c-store operators first. A typical highway site — dispensers, canopy lighting, a compressor, a c-store with chillers — might peak around 40 kW. Add one 60 kW DC charger and that ceiling more than doubles the moment a single vehicle takes a full-power session during billed hours. The session lasted ten minutes and sold perhaps 10 kWh; the capacity charge it created is monthly, and it repeats until the recorded peak falls.

The pass-through that cannot pass through

In India this collides with how a public charging fee is legally built. The Ministry of Power’s 2024 EV charging infrastructure guidelines define the driver’s fee as four components: the electricity tariff as a pass-through per kWh, a service charge, land cost and GST. Clause 10 caps that service charge until 31 March 2028 at ₹3 and ₹4 per unit for AC, and ₹11 and ₹13 for DC, in solar and non-solar hours. The operator’s margin per kWh is therefore a capped number — and a per-kW cost has no per-kWh line to sit in. Repricing sessions cannot fix it.

Depots do not get the EV tariff

For fleet operators the exposure is less widely understood. EV tariff categories are written for stations serving other people’s vehicles: Maharashtra’s LT VIII and HT IX both state that where a consumer charges their own vehicles at their premises, the tariff applicable is that of the premises. A captive depot is therefore billed as commercial or industrial — ₹520 to ₹555 per kVA per month in MSEDCL’s area — with the full maximum-demand machinery attached. A depot’s bill is a capacity bill first, which is why sequencing vehicles through depot chargers overnight pays for itself.

Four Levers For EV Demand Charge Management, Ranked By Cost

Ranked from cheapest to implement to most expensive. The order matters, because the first three exist mainly to postpone the fourth.

1. A real-time site kW ceiling (configuration, not capex)

The only lever that costs nothing but setup is a site-level power ceiling enforced while sessions run: the platform allocates headroom across active chargers and throttles them rather than letting aggregate draw reset the month’s peak. Dynamic load management is precisely this control loop. For a forecourt the ceiling must be set against measured site load, not charger nameplate, so the meter feeding the c-store has to be visible to the system dispatching the chargers — the same integration discipline behind forecourt POS connectivity, and usually the hard part.

2. Scheduling and simultaneity limits (depot operations)

At a depot the binding constraint is rarely energy: 40 vehicles at 60 kWh each is 2.4 MWh, which an eight-hour window absorbs comfortably. The cost comes from how many chargers run at once. In Maharashtra only demand registered between 06:00 and 22:00 counts toward billing demand, and energy drawn between midnight and 06:00 carries a 10% time-of-day discount — but the 40%-of-contract-demand floor means the simultaneity you design for is billed whether you use it or not. Twelve 30 kW chargers together imply roughly 380 kVA of contract demand and a floor near 152 kVA, about ₹84,000 a month at HT commercial rates before a vehicle plugs in. Six at a time roughly halves it.

3. On-site storage or solar buffering (capex)

A battery lets the charger draw 120 kW while the meter sees 60 kW — the only way to keep full-power sessions and a low ceiling at once. It is real capital, so it earns its place when the alternative is a network upgrade or a permanently throttled site. In India the timing question is sharper: MoP directs 0.7 times the average cost of supply during solar hours, 09:00 to 16:00, and 1.3 times outside them.

4. More sanctioned load — or a different meter

Raising contract demand is the expensive answer the other three exist to avoid — you pay the higher floor every month, permanently. But in India there is a cheaper structural move. MoP clause 5(3) requires the licensee to provide an LT connection up to 150 kW where a separate application is made for an EV charging station, and clause 9 directs that supply to EV charging stations be on a single-part tariff not exceeding the average cost of supply until 31 March 2028. Maharashtra implemented that literally: in FY 2025-26 neither the LT VIII nor the HT IX EV category carries a demand charge. A separate EV connection can therefore remove the charge rather than manage it — while the depot next door, charging its own fleet, cannot use that category. Read your own state commission’s order before assuming either outcome.

Worked Example: One 60 kW Charger At A Maharashtra Forecourt

A highway petrol pump with a c-store sits in MSEDCL’s LT II (C) commercial category — the category that explicitly names petrol pumps and service stations — at ₹520 per kVA per month for FY 2025-26. Contract demand is 60 kVA and the site records a maximum demand of about 42 kVA. The operator adds one 60 kW DC charger and considers two ways to connect it.

The headline is the ₹20,280 a month — about ₹2.4 lakh a year — that one charger adds behind the existing meter. Spread across 200 sessions a month, roughly 3,000 kWh at 15 kWh a session, that is ₹101 per session in fixed cost, against a DC service charge capped at ₹13 per unit, or about ₹195 of gross margin on the same session. Over half the capped margin is gone before hardware, land, staff or payment costs.

The third row is the one that surprises operators: in a month where nobody charges during billed hours, the contract-demand floor still bills 48 kVA — ₹10,764 more than the site paid before the charger existed, for capacity requested and untouched. The fourth shows why the connection decision outranks every optimisation downstream of it: on a separate EV connection the same charger buys energy at ₹7.21 per kVAh plus ₹1.24 wheeling, against ₹12.47 plus ₹1.17 on the commercial category, with no capacity charge on top. The strategy work behind turning a petrol station into an EV energy hub starts here, not at the charger spec.

What To Ask Your Charging Platform

Broad platform energy-savings claims are covered elsewhere; these five questions are specific to the kW ceiling. Can it enforce a site-level ceiling in real time, rather than alert after the interval has been recorded? How does it allocate capacity across chargers when the ceiling binds? Can schedules be written against a tariff calendar, including the hours your state counts toward billing demand? What does the driver see when a session is throttled mid-charge? And can site consumption outside the chargers feed the same ceiling? A charging management platform like YoCharge should answer all five with configuration, not a development ticket.

Demand charges are a design constraint, not a tariff to be negotiated away: decide which meter the charger sits behind, size the contract demand you are willing to pay for every month, and hold the ceiling in software. Done in that order, the fixed half of the bill stays proportionate to the business the site actually does.

Frequently Asked Questions

Sources: Ministry of Power — Guidelines for Installation and Operation of Electric Vehicle Charging Infrastructure, 2024 | MERC Order, Case No. 217 of 2024 — MSEDCL tariff for FY 2025-26 to FY 2029-30 | RMI — Rate-Design Best Practices for Public EV Chargers | RMI — DCFC Rate Design Study