Last spring a charging operator called me from a half-finished site outside Manchester. Four 150 kW DC fast chargers standing in a gravel lot. The grid connection quote had just come back: fourteen months, plus a six-figure upgrade fee he hadn’t budgeted for. “Can a diesel generator hold us over?” he asked. I said yes. Then I told him what would probably kill the genset within a year if he did it the way most people do.
EV charging stations and diesel generators meet in exactly this spot: a site that needs megawatts of power the grid can’t deliver yet. It happens constantly now — fleet depots, highway corridors, retail parking lots. The diesel set becomes a bridge. Two years, maybe three, until the utility catches up.
Chargers are not like most loads I’ve fed with generators. They’re worse in three specific ways: they ramp from zero to full power in seconds, they shut off just as fast, and their rectifiers turn clean sine waves into harmonic soup. A genset sized like it’s feeding an office building will hate this duty cycle. So here are the five mistakes I keep seeing — the ones that turn a perfectly good diesel generator into the charging station’s worst enemy.
This is the mistake everybody makes first, and honestly it’s the most understandable one. Four chargers at 150 kW each means 600 kW, so buy an 800 kVA set with margin, right? Wrong. Here’s the thing: chargers almost never all run at full tilt at the same time. Cars taper their charge curves. One driver grabs coffee at 90% while another’s just plugging in.
A bridge genset sized at nameplate total will loaf along at 20–30% load most of the day. Diesel engines despise light loading. Wet stacking — unburned fuel and carbon glazing the cylinders — starts below about 30% load and it doesn’t forgive. I walked onto a fleet depot in Leeds where a 500 kVA set had been feeding six chargers for eight months at barely a quarter load. The exhaust was spitting black sludge. The fix cost more than a proper load study would have.
Size for the simultaneous load, not the nameplate load. Monitor a comparable site, log real demand curves, then pick a genset that sits at 60–75% of its prime rating during typical operation. That’s where diesels are happiest — hot enough to burn clean, with headroom for the spikes.
Every EV fast charger is basically a giant rectifier. It draws current in sharp pulses, not smooth sine waves. This is harmonics, and it matters more than most charging operators expect. A 400 kVA alternator that runs a motor load all day at 400 kVA might handle only 300 kVA of charger load before the neutral conductor and windings start cooking.
I’ve seen installers connect chargers to a standard genset with a straight face, watch the voltage waveform turn ragged, then blame “generator quality” when chargers start faulting. It wasn’t the generator. It was the harmonics. The alternator needed derating — typically 1.25 to 1.5 times the calculated load — or a 12-lead reconnection tuned for non-linear duty.
If your supplier doesn’t ask what percentage of the load is non-linear, find a different supplier. Ask specifically about alternator derating for rectifier loads. It costs nothing at quote stage and saves the machine later.
Here’s a scenario I watched play out in real time on a motorway service site: grid connection delayed, temporary genset humming along, and at 6:40 p.m. six cars plugged in within four minutes of each other. Each charger ramped to 120 kW in under ten seconds. The genset saw a 700 kW step load it never agreed to, frequency dipped, voltage followed, and two chargers threw over/under-voltage faults and dropped out. The drivers blamed the site. The site blamed the generator.
This is a control problem, not a machine problem. A bridge-power setup needs load management — a controller that staggers charger starts and caps each dispenser’s output so total demand stays inside the genset’s step-load capability. ISO 8528 sets the standard for how much load a genset must accept in one step (typically 60% of prime rating for the common G2 class). Respect it.
The chargers themselves usually support external current limiting through their back-end. Use it. A site running each charger at 80% of max but never faulting will always beat a site that promises full speed and delivers blackouts.
This one is psychological. The diesel arrives, the chargers work, revenue starts flowing, and the grid upgrade — still fourteen months away — quietly stops being a priority. I know a retail-park site that ran its “temporary” 800 kVA set for three years. The fuel bills alone would have paid for the transformer upgrade twice over.
Do the honest math before day one: diesel at bridge-site volumes typically costs three to four times grid electricity per kWh once you include maintenance, fuel logistics, and the eventual overhaul. That math only works as a countdown. Put the genset in with a contract that has an end date, and keep the utility upgrade moving even when the chargers are printing money. The machine also needs to be spec’d as temporary — containerized, skid-mounted, resale-ready — not buried in a concrete pit you’ll pay to demolish.
There’s also the permit side. Plenty of sites get temporary diesel running, then discover the local authority considers it a permanent installation after twelve months and applies a completely different emissions rulebook. Check before you plug in, not after.
When the grid finally arrives, someone has to switch between utility and generator. The bridge period often means months of running on diesel with the grid connection coming in stages — utility live for the site lighting one month, chargers still on the genset the next. A bargain transfer switch with slow changeover and no load monitoring will cost you in ways that show up at the worst moments.
Spend on the controls, not just the iron. A proper setup for this job has an automatic transfer switch rated for the actual step loads, generator controls that log every fault and load profile (you’ll want that data for the permanent design), and remote monitoring — because a bridge site in a gravel lot two hours away will develop problems at 2 a.m., and you’d rather know from your phone than from an angry driver on social media.
The operators who get this right treat the bridge period as a live rehearsal for the permanent site. Every month of generator data — real load curves, real peak simultaneity, real charger behavior — makes the final grid design better. That’s the silver lining most people miss.
Can a diesel generator hold an EV charging station over until the grid catches up? Yes — and I’ve seen it done beautifully. The sites that pull it off have three things in common: a genset sized to real simultaneous demand instead of nameplate totals, alternators derated for the harmonic reality of rectifiers, and a load controller that keeps charger chaos inside the machine’s step-load limits.
The ones that fail usually failed on paper before the first cable was pulled. If you’re planning a bridge-power site, start with the load study, not the generator quote. And keep chasing the utility — the diesel is your best friend for two years and your most expensive mistake in year three.
Email：[email protected]
Work：Monday to Friday,9:30-18:30
