Networked or standalone chargers for a depot?

Most writing about networked charging is aimed at offices, retail and multifamily, where the chargers serve people who are not you. There, networking pays for itself through access control, per-session billing, reimbursement and usage reporting to a landlord or an incentive program. Bring that reasoning into a fleet yard and it collapses: there is no access control problem worth solving with software when the gate is locked, and there is no billing relationship when every kilowatt-hour goes into a vehicle you own and pay for anyway. So set the workplace argument aside completely. The depot case for networking rests on different ground.

Four things, in descending order of how often they justify the cost.

Load management and cloud networking are frequently sold together and they are not the same thing. A local energy-management controller measures the site's actual current and throttles or sequences the charging load to hold the site under a set limit, and it does that whether or not the internet is up. The National Electrical Code addresses these systems directly — Article 625 permits an energy management system to set the load for EV charging, with the requirements for those systems in Article 750 — and jurisdictions commonly accept them as an alternative to upsizing the service. Cloud networking adds visibility, reporting, remote management and vehicle-aware scheduling on top of that. Those are real benefits and they are worth paying for on many sites. But if the whole reason you are networking is 'we need load management', ask your bidder whether a local controller achieves it, because the answer is often yes and the answer has no recurring cost.

A per-port subscription looks small next to a capital budget and large next to nothing, and the mistake is to evaluate it over year one. Chargers get operated for a decade. Do the arithmetic over that decade and compare it against the thing networking is claimed to avoid. We do not publish market subscription rates — they vary by vendor and are usually negotiated inside a bigger deal, and the numbers that circulate are not sourced well enough to print. Get yours quoted, including the renewal price after the initial term, and then read the row that matches it.

Set the ten-year subscription total against the cost of the service upgrade that managed charging avoids. Only your utility can give you the second number, because it depends on the existing service, the transformer, the distance, the utility's cost-allocation rules and its construction charges — which is exactly the kind of fact we refuse to publish as a national average. Ask for it early, in writing, as part of the load-letter conversation. What the published data does tell you is where the money sits in a commercial charging project generally: make-ready electrical and civil work commonly accounts for 50 to 70 percent of a commercial charging bill, and electrical infrastructure runs 40 to 60 percent of a DC fast project. Service work is the big number, subscriptions are the small recurring one, and if managed charging genuinely removes a service upgrade from the project then networking has usually paid for itself before you get to the data benefits.

A small depot — say up to six or eight ports — with comfortable spare service capacity, a long dwell window, one duty cycle and telematics you already trust for energy data is a genuine standalone candidate. Buy good non-networked units, add a local load-management controller if the load calculation needs one, and you have no subscription, no software vendor relationship, no renewal negotiation and nothing that stops working when a contract lapses. The costs you accept in exchange are real: a truck roll for every fault diagnosis, no remote reset, manual firmware handling, no automatic session records, and a much harder time proving usage to a utility program that wants metered data. Those costs scale with port count, which is why the standalone case weakens quickly as a yard grows.

A network subscription is a dependency, and dependencies deserve terms. Three questions decide how bad the dependency is. What happens to the chargers if you stop paying — do they keep charging on a local schedule, do they lose load management, or do they stop. Can the hardware be moved to a different network operator, and what does that require — this is where the open protocol question does real work, and it is worth reading up on before you sign rather than after. And who is the contracting party: if your installer signed the network agreement on your behalf, the software vendor's customer of record is a company that finished its work three years ago. Require assignment of the agreement to you at acceptance, with the renewal price stated or capped.

Network a subset. On a phased depot there is a defensible design where the load-management controller and the metering are site-wide and cloud-connected, while the individual charging units are inexpensive and dumb. You get the site-level visibility, the demand control and the tariff scheduling — which is where nearly all the money is — without paying a per-port subscription on every position in the yard. Not every product line supports it and not every vendor will propose it, since per-port recurring revenue is the business model. Ask anyway. On a twenty-four port yard the difference over ten years is the size of several ports, and the questions in the RFP checklist are the ones that surface whether the option exists.