In short: Electrifying a bus depot is almost always framed as a grid-connection problem. But once the chargers are in, the binding constraint is connectivity — smart charge scheduling, OCPP telemetry, and the yard moves that get the right bus to the right plug all run on a depot network that most operators never had to build.
Key Takeaways
- The grid is only half the job — A depot's power upgrade gets the electrons in; software and connectivity decide whether 100 buses charge overnight without tripping the supply.
- Every charger is a data device — Smart charging, load balancing and demand-response all depend on continuous OCPP telemetry that has to move reliably across the whole yard, not just near the office.
- Plug WiFi was never built for a depot — A bus depot is a steel-clad, vehicle-dense, RF-hostile box; one managed private network covers chargers, telematics and yard movements where access-point WiFi drops out.
In a nutshell
Why does electric bus depot charging come down to the network?
Electric bus depot charging is sold as an electrical engineering problem, and at the front end it is one. The headline number on every depot electrification business case is the grid connection — the new substation, the reinforced supply, the £-per-megawatt cost of getting enough power onto the site to charge a hundred or more buses between the last run-in and the first run-out. The ZEBRA programme (Zero Emission Bus Regional Areas), through which the Department for Transport has put hundreds of millions of pounds toward English bus electrification, has funded exactly this kind of upgrade across operators including Stagecoach, First Bus and Go-Ahead, and the National Express depot in Coventry has gone fully electric on the same logic.
But the grid connection only gets the electrons to the yard. What decides whether a hundred buses are charged, in the right order, ready for the right diagrams by 5am — without exceeding the site's agreed supply capacity — is software, and software needs a network. Electric bus depot charging is, past the substation, a data-coordination problem: every charger has to report its state, every bus has to be matched to a charge plan and a departure time, and the depot management system has to throttle and stagger the load continuously through the night. None of that works if the charger on the far row of the yard cannot reliably talk to the system that is meant to be managing it.
This is the part the business cases routinely under-cost. The grid connection is visible, expensive and gets the scrutiny. The network that makes the chargers useful is assumed to be "just WiFi" — and that assumption is where depots come unstuck.
What does the charging software actually need to move?
It is worth being specific about the data, because "the chargers need connectivity" is too vague to plan around.
Each charge point speaks a protocol — in practice OCPP (the Open Charge Point Protocol) — back to a central management system. That conversation is continuous: state of charge, fault codes, available power, session start and stop, and the commands that throttle a charger up or down. Multiply that by a hundred or more chargers and add the depot management system's own layer — which bus is on which plug, what its next diagram is, how much energy it needs and by when — and you have a yard full of devices that must all stay in constant, reliable contact.
The reason this matters operationally is smart charging. A depot that simply plugs every bus in at once and lets them all draw full power would need a grid connection several times larger — and more expensive — than one that staggers and balances the load across the night. Smart charging is what keeps the connection cost down, and smart charging is entirely a function of the chargers and the management system being able to talk without interruption. Add demand-response and time-of-use tariffs — charging hardest when electricity is cheapest and greenest, easing off at peak — and the network becomes the thing that protects the depot's single largest operating cost. A dropped connection is not a minor inconvenience; it is a charger that has stopped following the plan, and potentially a bus that is not ready for service.
Why doesn't depot WiFi cope?
Here is the uncomfortable engineering reality: a bus depot is close to a worst-case environment for the wireless technology most operators reach for first.
A depot is a large, steel-framed, steel-clad shed, often with an outdoor yard attached, packed with metal vehicles that move. Every one of those properties is hostile to WiFi. Steel structure reflects and blocks the signal; the metal mass of the buses themselves creates a constantly shifting pattern of shadows and dead spots; and the sheer scale means a single access point covers only a fraction of the floor. The conventional fix is to hang access points throughout the shed and across the yard, cable each one back, and accept that handover between them — as a tablet, a telematics unit or a charger roams the coverage edges — will sometimes drop the connection. In an office, a momentary WiFi drop is invisible. On a charger mid-session, or a yard-management tablet directing a shunt, it is an operational fault.
Private 5G is built for exactly the conditions WiFi struggles with: range, density, and a cluttered metallic environment. A small number of cells can cover an entire depot — shed and yard together — with one continuous signal and proper handover, so a device does not drop as it crosses the site. Importantly, it carries everything on one network: the OCPP traffic to the chargers, the telematics off the buses, the yard-management terminals and the CCTV, with the capacity to keep them all live at once and the ability to slice the network so that safety-critical charge-management traffic is never crowded out by, say, a depot CCTV upload. The result is not a marginal improvement in coverage; it is the difference between a charging operation you can trust to run unattended overnight and one that needs a human walking the rows checking plugs.
How does this fit the wider depot, not just the chargers?
The chargers are the reason the network gets built, but they are rarely the only thing that ends up on it — and that is where the economics start to make sense.
A modern bus depot is already a connected operation, or wants to be. Telematics off every vehicle for predictive maintenance; pre-departure checks logged on tablets; cleaning and fuelling — now charging — sequenced through a yard-management system; CCTV across the yard for security and incident review; increasingly, automated or assisted shunting of buses into charge bays. Each of these has, historically, been solved on its own little network — a SIM here, a WiFi island there — and the result is the same fragmentation that bedevils every industrial site: nothing joins up, and the depot manager is left reconciling four systems that cannot see each other.
By way of example, the same network that carries OCPP telemetry to the chargers can carry the yard-management instruction that says "bus 142 needs 180kWh, send it to bay 30, which is free and on a cheap-rate charger". That is one decision, drawing on charger data, vehicle data and diagram data — and it is only possible if all three live on the same reliable network. The connectivity that electrification forces a depot to build is, conveniently, the connectivity the rest of the depot has wanted for years. The charging case pays for the network; the wider operation gets the benefit.
What are the trade-offs, honestly?
We are not going to pretend a private network is the answer to every depot, because it is not, and depot operators have a finely tuned ear for vendors who over-claim.
A small depot — a handful of electric buses, a single charging row within sight of the office router — does not need a private network, and we would not propose one. Conventional WiFi or a few cellular routers will cover it, and the cost of a managed network would not be justified by the scale. The case changes with size and with how hard the operation leans on smart charging: a large urban depot electrifying a hundred-plus vehicles, with a constrained grid connection that makes load management non-negotiable, is a different proposition entirely. There, the network is not an optional refinement — it is the thing standing between the depot and either a far larger grid connection or buses that miss their diagrams.
The honest framing is a trade-off between three costs: the grid connection, the network, and the risk of charging that does not reliably follow the plan. Spending more on the network is often what lets a depot spend less on the grid connection, because reliable smart charging is what keeps the peak demand down. That relationship — network spend buying down grid spend — is the calculation we think depot electrification business cases most often get wrong, because the two sit in different columns and are rarely weighed against each other.
Where this goes next
Britain is electrifying its bus fleets at pace, and the depots are being rebuilt around the chargers. We think the operators who come out of this well will be the ones who treat the depot network as core infrastructure — designed in alongside the grid connection, sized for the whole operation rather than just the chargers — rather than as an afterthought bolted on once the buses arrive and the WiFi turns out not to reach.
We aim to bring the same managed-network approach we have proven in other demanding outdoor and industrial settings to the depot: one network across the shed and the yard, carrying the chargers, the telematics and the yard management together, run as a service so the operator does not need a telecoms team to keep it alive. The grid connection gets the power in. The network is what turns a yard full of plugs into a charging operation that runs itself overnight — and that, increasingly, is the part of electrification worth getting right first.