How Solar EV Charging Actually Works

The architecture in one paragraph

Solar panels feed string inverters; inverters feed the site's AC distribution; EV chargers draw from that same distribution; the grid covers any shortfall and absorbs any surplus. A metering layer watches generation and consumption in real time, and a load-management controller — talking OCPP to the chargers — decides how much power each socket gets and when. That is the entire trick: one electrical system, one set of measurements, one piece of software steering vehicle load into the hours when the roof is producing.

Solar-following charging: where the value is made

Dumb charging starts every vehicle at 100% rate the moment it plugs in — usually 8.30am, usually all at once, creating a demand spike precisely when the array is still warming up. Smart scheduling does three better things: it spreads charging across the parked hours (a van parked 9am–4pm needs only a fraction of its socket rating to leave full), it biases charging power towards measured surplus generation, and it caps total site demand below the supply limit. The result on real sites: vehicle energy lands 30–60% solar across the year, building self-consumption climbs past 90%, and the demand profile the DNO sees stays flat. The same controller logic is covered from the compliance angle in the load management guide.

AC, DC and what fleet operators actually need

Workplace and depot charging is overwhelmingly AC at 7–22kW — cheap sockets, gentle on the supply, perfectly matched to vehicles that sit parked for hours. DC rapid units (50–150kW+) earn their place for mid-shift turnarounds and HGV operations, but they are grid-led by nature: no rooftop array sustains 150kW into one vehicle on a winter afternoon. The honest design rule: AC for dwell time, DC for turnaround, solar offsetting the annual energy of both rather than the instantaneous power of either. Hardware selection is its own decision — the charger guide covers it.

Sizing: energy first, power second

Two budgets must close. The energy budget — annual kWh for building plus vehicles versus array yield (about 850–950kWh per kWp in the UK) — sets the array size and the savings. The power budget — worst-case simultaneous demand versus supply capacity — sets the load management requirement and the DNO conversation. Most "we need a bigger supply" verdicts dissolve under scheduling: forty vans needing 240,000kWh a year sounds enormous, but spread across nightly and daytime dwell windows it averages well under 100kW. Where the power budget genuinely cannot close, the options ladder runs: smarter scheduling, battery buffering, then supply upgrade — in that cost order.

The DNO treatment

Generation needs a G99 application; significant new charging load needs notification or assessment under the load procedures; a combined project bundles both into one engagement with the network operator. This is one of the quiet wins of designing together — one study, one set of protection settings, one witnessing visit, and a demand profile (flattened by load management) that is easier to approve than either project naïvely presented. Canopy projects add structural and metering specifics, covered in the canopy guide; whole-depot conversions get the full treatment in fleet depots.

Monitoring: the system that proves itself

A combined system should report three numbers monthly without anyone asking: solar share of vehicle energy, building self-consumption, and effective pence-per-mile by vehicle. Those are the numbers the business case promised, and the handover isn't complete until the dashboard shows them. Under-performance visible in week two is a warranty conversation; discovered at year-end, it's a write-off.