How Many Chargers Does an Electric Bus Depot Need?

Most transit agencies start with one depot charger per bus, and most of them later find that the number they should have planned was not chargers at all. A depot has three counts that matter: plugs, power units and megawatts at the grid connection. Each is set by a different constraint, and getting one of them right does not fix the other two.

This article is for planners who already know what a CCS plug and a pantograph are and need to defend a number: to a board, a utility or a charger vendor. We use published agency data from the US Alternative Fuels Data Center and the Federal Transit Administration [1, 2], field consumption data from Dutch e-bus fleets [3], and one worked example for a 100-bus depot.

The short answer: 1:1 is the default, not the rule

The one-to-one ratio is what agencies actually buy first. In a 2023 FTA survey of battery-electric bus operators, 13 of 25 respondents reported a one-to-one ratio of buses to chargers in their initial procurement [2]. The AFDC guide on deploying charging for transit buses confirms that most agencies plan one depot charger per bus, and that Santa Clara VTA, for example, started at 1:1 [1].

The same guide is explicit that there is no general rule: the right ratio depends on routes, schedules, labour and available power [1]. TriMet in Portland plans three buses per charger per night. Several agencies combine a large number of slow chargers with a few fast ones, and the AFDC notes that seven fast and nine slow chargers can do the work of 83 slow chargers in one modelled case [1]. A ratio of 1:1 is safe for a pilot of five or ten buses. At 100 buses it starts to cost real money in equipment, space and grid capacity that the night never uses.

Three numbers hiding behind "how many chargers"

Part of the confusion is vocabulary. A modern depot charger is often a power cabinet feeding several dispensers - plug-in posts or pantographs - so one "charger" can mean one cabinet or one parking position. The AFDC guide describes both setups: multiple dispensers per power cabinet, and TriMet's one dispenser per cabinet [1]. Separating the counts makes the planning tractable.

Table 1. The three counts in depot charging design

Count	What it is	What sets it
Dispensers	Plugs or pantographs, one per parking position that must charge	Number of buses parked overnight, depot layout, whether staff can move buses or plugs at night
Power units	Cabinets that convert grid power and deliver it to dispensers	Energy each bus needs, charging time per bus, length of the usable night window
Grid connection	Megawatts the utility supplies to the whole site	Total fleet energy per night divided by the window, plus headroom; tariff and demand charges

In practice, dispensers tend to stay close to one per bus, because a bus that has to be moved to another plug at 2 a.m. needs a person to move it. Power units are where the ratio can drop well below one. The grid connection is usually the binding constraint, and it is the one with the longest lead time.

Worked example: a 100-bus depot

Take a depot of 100 twelve-metre buses averaging 250 km per day. Field data from 79 twelve-metre e-buses in seven Dutch cities gives an average consumption of 1.15 kWh/km, rising by about 14% in cold months [3]. That is 288 kWh per bus per day on average and roughly 330 kWh in winter. Assuming about 90% efficiency from grid to battery, each bus draws about 365 kWh from the grid on a winter night, and the fleet needs about 36.5 MWh.

With 150 kW units, a typical depot power class (the AFDC guide places slow depot charging at 65-150 kW [1]), a session takes about 2.4 hours at full power. The constant-voltage phase at the top of the battery adds time, so call it three hours. If the usable window per bus is about six hours, after arrival, washing and inspection and before pull-out, one power unit can charge two buses in sequence.

Table 2. Three designs for the same 100-bus depot (winter night)

Design	Dispensers	Power units	Peak power	Trade-off
1:1 plug-in	100	100 x 150 kW	15 MW	Simplest operation; most equipment and the largest grid connection
Shared units, sequential	100	50 x 150 kW	7.5 MW	Half the power electronics; needs automatic sequencing between dispensers
Shared units, managed to a 6 MW cap	100	50 x 150 kW	6 MW	36.5 MWh takes about 6.1 hours at the cap; little slack for late returns or a failed unit

The average power the fleet needs over an eight-hour night is about 4.6 MW. The 15 MW connection of the 1:1 design is more than three times that, and it is paid for in equipment, transformer capacity and, in many tariffs, demand charges. The real-world equivalent is LA Metro, where managed charging reduced the required capacity from 20 MW to 10-12 MW [1].

The last row shows the other side. Once the cap gets close to the average, the night has no margin. A bus that returns at 01:30, a unit that faults at 03:00 or an unexpectedly cold week moves some departures from ready to not ready. That is the point at which the question stops being "how many chargers" and becomes "which bus gets energy first".

THE RULE IN ONE LINE

Buses per power unit = usable window per bus ÷ charging time per bus. Grid connection ≥ fleet energy per night ÷ the window, plus headroom for failures and late returns.

Resilience: the spare charger is a design decision

Sharing a power unit between buses magnifies the impact of an outage: if one cabinet fails, every bus behind it waits [1]. The FTA report found the same pattern in operations, where a single on-route charger going down noticeably reduced fleet availability [2]. Agencies plan around this explicitly. The Chicago Transit Authority plans at least one fast charger per garage for resiliency, and Metro Transit in Minnesota at least two fast chargers for recovery [1].

For our 100-bus example, two or three spare 150 kW units, or one 450-600 kW recovery charger, cost far less than holding the whole depot at 1:1. The condition is that someone, or something, decides which buses go to the recovery charger and in what order when it is needed.

Fewer chargers only work when the night is managed. Unmanaged, the cheapest design is also the most fragile.

Labour and the night shift set the real window

Charging time on paper and charging time in a depot are different. The AFDC guide reports that Rochester tells floor managers not to plug buses in before 11 p.m., and that Santa Clara VTA has no labour on site between 1:00 and 3:45 a.m. [1]. If a power unit with two dispensers needs a person to swap plugs between the two buses, the six-hour window collapses to whatever the shift allows.

This is why sequential charging is a software problem as much as a hardware one. A cabinet with several dispensers, or a row of pantographs, can switch from one bus to the next without anyone on the floor, but only if the charging system knows the order. The FTA report notes that fleets with more buses than installed chargers rely on staff moving buses to ensure full charges [2]. Battery storage can widen the window: at VTA it adds three to four hours of off-peak charging time [1].

Space is the other night-time constraint. A Caltrans review of California's transition notes that depot charging is the dominant approach in the state, and that the space and storage problems are made worse because diesel buses are maintained at night, in the same yards [4]. Every dispenser competes for floor area with maintenance, washing and parking.

When managed charging pays off, and when it does not

Managed charging is not free. It needs a charge management system, OCPP smart charging profiles, integration with the block schedule and someone to own the rules. The AFDC guide says it is more likely to pay off for large fleets and time-of-use tariffs, and that for a small agency such as Mountain Line it would cost more than it saves [1]. The FTA report reaches a similar conclusion: savings from smart and managed charging grow with the size of the fleet [2]. Whichever system you choose, test its smart charging limits before the first real night: our OCPP Tester simulates hundreds of charging stations over OCPP 1.6J and 2.0.1, including dynamic load management.

A useful threshold for the decision is the gap between the nameplate power of the chargers and the average power the night needs. In our example that gap is 15 MW against 4.6 MW. When it is that wide, managed charging is the cheapest way to close it. The Optibus checklist for e-bus operators frames the same point for planning teams: set the bus-to-charger ratio together with a charging schedule that minimises peak demand within the depot's grid constraints [5]. For the software side of that schedule, see our article on depot charging planning and sharing power in an electric bus depot and on running the whole depot from one layer.

Want to see what happens to departures when the grid cap tightens or one charger misbehaves? Our Depot Night Simulator plays one night in a depot with 80 charging stations, 12 power units and one grid connection, minute by minute from 22:00 to 06:00, and shows which vehicles leave on time under each allocation policy.

FAQ

How many chargers per electric bus does a depot need?

Most agencies start with one depot charger per bus, and 13 of 25 operators in a 2023 FTA survey reported that ratio for their first procurement. Larger fleets often share power units between two or three buses per night, as TriMet does, and add one or two fast chargers for recovery. The right ratio depends on the usable night window, charging time per bus, labour and grid capacity.

What is the difference between a charger, a power unit and a dispenser?

A power unit, or power cabinet, converts grid power and delivers DC to the bus. A dispenser is the plug-in post or pantograph at the parking position. One power unit can feed several dispensers, charging buses one after another. When planners count chargers, they should say which they mean, because dispensers usually follow the number of buses and power units do not.

How long does it take to charge an electric bus overnight?

At a typical 150 kW depot charger, a twelve-metre bus that used about 330 kWh on a winter day needs roughly 2.5 to 3 hours, including the slower final phase. The AFDC guide states that a full charge on a slow depot charger takes five hours or less. With a six-hour window, one power unit can serve about two buses in sequence.

How much grid capacity does an electric bus depot need?

Start from energy, not charger count: total kWh the fleet draws per night divided by the charging window, plus headroom. A 100-bus depot of twelve-metre buses running 250 km a day needs about 36.5 MWh on a winter night, or about 4.6 MW on average over eight hours. Installing 150 kW per bus would mean 15 MW of nameplate capacity.

Does managed charging reduce the number of chargers needed?

It mainly reduces the grid connection and the number of power units, not the number of plugs. LA Metro cut its required capacity from 20 MW to 10-12 MW with managed charging. The AFDC guide notes it pays off mostly for large fleets and time-of-use tariffs; for small agencies, the software and integration can cost more than the savings.

Sources and assumptions

Lepre, Burget & McKenzie, Alternative Fuels Data Center (2022). Deploying Charging Infrastructure for Electric Transit Buses. Charger-to-bus ratios (pp. 9-14), slow and fast charger power, resiliency (pp. 10-11), storage (p. 25), managed charging and LA Metro (pp. 32-35), labour constraints (p. 34).
Godfrey & Sipiora, CUTR / FTA Report No. 0253 (2023). Procuring and Maintaining Battery Electric Buses and Charging Systems - Best Practices. Survey of initial charger-to-bus ratios (p. 38), charger outages and availability (pp. 15-16), managed charging (p. 5).
Sustainable Bus (2020). Electric bus energy consumption in ViriCiti's spotlight. 1.15 kWh/km average for 79 twelve-metre buses in seven Dutch cities; +14% in cold months.
Caltrans DRISI (2021). Research Results Summary, Task 3299: battery electric bus charging infrastructure in California. Depot charging as the dominant approach in California; space and night-time maintenance constraints; charging scenarios.
Optibus (2023). EV checklists for public transport operators (EMEA). Bus-to-charger ratio, charging schedule and grid constraints as planning steps.

Assumptions in the worked example: 250 km per bus per day, about 90% grid-to-battery efficiency, 150 kW power units, roughly 30 minutes added for the constant-voltage phase, a six-hour usable window per bus and an eight-hour night. Real depots should replace each of these with their own block schedules, telematics and tariff data.