EV Charging Infrastructure: The Connection Is the Constraint, Not the Charger

Almost every electric-vehicle charging project we are asked about starts as a procurement question: which chargers, how many, what power rating. In practice the chargers are the easy part of the problem. The constraint that decides the whole design is the site's grid connection, and it is usually fixed long before anyone chooses a brand.

The connection is the budget

A charging installation does not get a power allowance of its own. It shares whatever headroom is left after the building's existing loads — HVAC, lifts, process equipment, kitchens. So the honest first exercise is not a charger count but a measurement: log the actual demand profile at the main incomer for at least a full week, ideally across a seasonal extreme, and find the real peak rather than the one implied by the connection rating on the paperwork. Nameplate capacity is what you are permitted to draw; the measured profile tells you what is genuinely free, and at what hours.

Diversity, not arithmetic

Adding up charger ratings and requesting that much power is the most expensive mistake available. Vehicles do not all charge at full power simultaneously: they taper as the battery fills, they arrive and leave at different times, and most are connected far longer than they are actually drawing current. A workplace car park where cars sit for eight hours has enormous natural diversity. A depot that must return a fleet to full state of charge inside a fixed overnight window has almost none. The two cases justify completely different connections, and the difference between them is a scheduling question before it is an electrical one.

Load management before reinforcement

Static load limiting — a fixed ceiling on the charging cluster — is simple and cheap, but it sizes for the worst case and therefore wastes headroom for most of the day. Dynamic load management meters the incomer in real time and shares whatever is spare among the active sessions, which generally allows a site to support considerably more charge points on the same connection. The trade-off is that charging duration becomes variable, so it suits an operational requirement of "ready by morning" far better than "ready in twenty minutes". Where the duty is genuinely time-critical, no control scheme substitutes for capacity, and it is better to say so early than to discover it at commissioning.

Where a battery buffer earns its place

A battery placed between the grid and the chargers does one genuinely useful thing: it decouples the power the vehicle sees from the power the connection can provide. It charges slowly from available headroom and discharges quickly into a session. That is worth doing in three situations — a connection too small for the peak but adequate for the daily energy, a tariff with demand charges where the peak itself is what is being billed, and a site where grid reinforcement is possible but slow or disproportionately costly.

Sizing follows from the duty, in a specific order. The energy delivered per busy period sets the usable capacity. The gap between charger peak power and available grid power sets the converter rating. The time between busy periods sets whether the buffer can recover at all. That last item is where buffered-charger proposals most often fail in review. A buffer does not create energy; it moves energy in time. If a site's busy periods follow one another with no recovery window, the battery is flat when the second peak arrives and the connection is exposed exactly as it was before.

Details that decide the commissioning

Three things routinely surface too late. First, DC chargers perform their rectification inside the unit, so harmonic performance and any associated filtering belong in the specification rather than in an afterthought — check the declared distortion against what the rest of the site's equipment can tolerate. Second, earth-fault protection for DC charging requires devices rated to detect DC residual current, and the interaction with the installation's existing protective arrangement has to be worked through rather than assumed compatible. Third, back-office integration — protocol version, authentication, metering data suitable for billing — determines whether the site can be operated as one system or only as a collection of independent boxes.

What to settle before you buy

A measured demand profile at the incomer. Realistic dwell times and the true deadline for each class of vehicle. Whether the requirement is energy per day or power per session — they lead to different systems. The tariff structure, including any demand charges. The available physical routes for cabling and civil works, which on many sites cost more than the charging equipment itself. Only after those are settled does the equipment list mean anything.

If you are planning a charging installation and would like the electrical side assessed before the procurement is locked in — connection headroom, load management strategy, and whether a battery buffer is justified in your case — we are glad to review it with you. Write to info@lmelectro.com and we respond within 24 hours.