How many EV charging ports does a fleet depot need?
Fewer than you have vehicles, in almost every depot. Ports are not a headcount, they are energy divided by time: take the kilowatt-hours each vehicle needs overnight, divide by the kilowatts a port actually delivers, and you get the hours that vehicle occupies a port. Ten vans that each need 60 kWh, on 9.6 kW ports, occupy a port for 6.25 hours each. In a twelve-hour dwell window that is six ports, not ten.
Updated 2026-08-20
The arithmetic, in one line
kWh per vehicle per night comes from your telematics, not from a brochure range figure. It is the input every other number on the project depends on, and it is the one most often replaced by a guess.kW delivered per port is the charger's continuous output, not the breaker size. NFPA 70 Article 625 treats EV supply equipment as a continuous load, so a 40-amp charger sits on a 50-amp circuit and delivers 40 A × 240 V = 9.6 kW. A 48-amp unit on a 60-amp circuit delivers 11.5 kW. Use the delivered number. People who use the breaker number end up 25 percent optimistic on every line of the table below.Dwell hours is the gap between the last vehicle in and the first vehicle out, minus plugging, minus repositioning, minus slack. It is the number operators overstate most reliably, and it is the divisor, so every hour you overstate it deletes ports you actually needed.
Dwell time is the divisor, which is why it halves your port count
Look at the 60 kWh row. At four hours the arithmetic breaks. At eight hours it is eight ports. At twelve hours it is six. The vehicles did not change, the chargers did not change, and the difference between the second case and the third is two ports of hardware plus two ports of trench, conduit and switchgear.So measure dwell before you price anything. Pull the last-in and first-out timestamps for every vehicle for a full month, including the bad weeks. The number you want is not the average — it is roughly the tenth percentile, because the depot has to work on the tight nights, not the typical ones.Then subtract the operational overhead honestly. If nobody is on the yard between the last return and the morning dispatch, the plug goes in when the driver walks away, and the cable has to reach where the driver actually parks. A twelve-hour gate-to-gate window with a ninety-minute plugging lag is a ten-and-a-half-hour charging window.
Ports, connectors and parking spaces are three different counts
If six ports serve ten vans, four vans have to take a turn on a port another van has finished with. Nobody is moving vehicles at two in the morning to make that happen.The mechanism that makes it work without labour is a dual-connector unit: one circuit, two cables, two adjacent spaces, with the unit sequencing between them — full power to the first vehicle, then automatically to the second when the first is done. Ten spaces, five dual units, five circuits.That is the shape most depots actually buy, and it is why an honest specification carries three numbers: spaces with a cable that reaches, connectors, and circuits. A proposal quoting only one of those has not been drawn against your yard.The distinction also changes the cost conversation. Circuits carry the civil and electrical cost. Connectors carry hardware cost. Going from ten single-connector units to five dual-connector units halves the expensive count and barely moves the cheap one.
When the arithmetic returns more than one port per vehicle
Raise the amperage per port. Going from 40 A to 48 A buys about 20 percent more energy in the same window for the cost of heavier conductors, which is cheap while the trench is still open.Lengthen the window. Staggering dispatch by ninety minutes across two waves is free and often decisive.Split the fleet into two charging waves and accept moving some vehicles, if you have yard staff overnight anyway.Add a small number of DC ports for the exception cases — the van that came back nearly empty, or the one reassigned to a second shift. That is a different decision with a very different price tag.What does not work is buying eleven Level 2 ports for ten vehicles. The eleventh port draws power the service has to carry and delivers energy nobody is parked long enough to take.
Size the passive infrastructure for the fleet you will have
Installing conduit and panel capacity for the full build up front is reported to cut per-port cost by 40 to 60 percent against retrofitting the same yard later. The reasoning is that the expensive parts of a port are the parts you cannot cheaply redo: the trench, the conduit, the pull boxes, the pads and the distribution equipment.The charger itself is the cheap and replaceable part, and it is also the part whose standards move.The honest risk is stranded capital. If the electrification plan slips three years, you paid for empty conduit. Empty conduit is also the cheapest thing in the trench, so the usual split is to size the civil work for the plan and the switchgear for the next two years.
What each port costs, so you can price the difference
Ports close to the existing service on an open trench sit at the bottom. Ports at the far end of a paved yard sit at the top.Which means the two ports you saved by measuring dwell properly are not a hardware saving. They are a saving on trench, conduit, conductors, a breaker position and a pad — plus the share of service capacity those ports would have obliged you to buy.There is no federal tax credit to net against any of it. Section 30C terminated for property placed in service after June 30, 2026 under Public Law 119-21, and there is no federal replacement. What remains is your utility's make-ready and fleet programs, which is where the recoverable money in this segment now sits.
Do I need one charger per vehicle?
Almost never. One port per vehicle is correct only when the energy each vehicle needs fills the entire dwell window, which happens when dwell is short or the vehicles are heavy. For an overnight van depot the ratio usually lands between one port per one-and-a-half vehicles and one port per two. Run the arithmetic on your own dwell data before accepting a one-to-one proposal.
What is a good charger-to-vehicle ratio for a fleet?
There is no good ratio in the abstract, because the ratio is an output rather than an input. It falls out of energy per vehicle, delivered kilowatts per port and dwell hours. Two depots with identical vehicle counts can correctly land on six ports and on twelve, and the difference will be dwell time.
Does load management change how many ports I need?
It changes how much service capacity the ports need, not how many ports you need. Managed charging spreads the same energy across the window at a lower peak. If the window is already full, management cannot create hours that do not exist, so it lowers your kilowatt number without lowering your port number.
Should I add spare ports for growth?
Add spare conduit, pads and breaker positions for growth. Add spare energised ports only when the vehicles are already ordered. Passive infrastructure does not go obsolete and is expensive to add later; charging hardware does go obsolete and is cheap to add later.
How do I count ports if some vehicles come back mid-shift?
Treat them as a separate population with their own dwell window and their own arithmetic, then add the two port counts together. Mixing a two-hour mid-shift population into a ten-hour overnight average produces a number that is wrong for both.
- NFPA 70, National Electrical Code, Article 625 — EV supply equipment as a continuous load, behind the delivered-kW figures
- SolarTech, EV fleet charging complete guide — depot per-port installed cost and the 40 to 60 percent saving from building make-ready once
- Recharged, How much does a commercial EV charging station cost — per-port installed cost and make-ready as a share of the bill
- IRS, Alternative Fuel Vehicle Refueling Property Credit — Section 30C termination for property placed in service after June 30, 2026