Electric Vehicles
What the EV transition costs the grid
Total electricity demand rises modestly. Peak demand is the problem, and it is almost entirely a question of when people plug in.
The claim that electrifying transport will overwhelm the grid is made confidently in both directions. The arithmetic is less dramatic than either version.
The energy question is modest
A typical car driven fifteen thousand kilometres a year at around 18 kilowatt-hours per hundred kilometres consumes roughly 2,700 kilowatt-hours annually.
That is comparable to a household's water heating, or somewhat more than a fridge and freezer combined over a year.
Full electrification of a national car fleet typically implies an increase in total electricity demand in the region of fifteen to thirty percent, phased over two decades.
Grids have absorbed growth at that pace before. It is a substantial build programme and not an unprecedented one.
The peak question is the real one
Networks are sized for peak, not for total energy.
If a large number of vehicles begin charging at seven in the evening — the moment people arrive home, which is already the daily peak — the coincident load is severe, and it lands on local distribution networks that were sized decades ago.
The same vehicles charging spread across the overnight hours add almost nothing to peak, because demand at three in the morning is far below capacity.
Which means the entire infrastructure question turns on scheduling rather than on generation.
Why the local network matters more than the national one
National transmission and generation adequacy is a manageable planning problem.
The distribution network — the low-voltage cables and transformers serving a street — is where the constraint bites first.
A residential transformer serving a few dozen homes was sized for lighting, appliances and perhaps electric showers. Several simultaneous 7 kW charging loads on the same transformer is a different proposition.
Distribution operators are addressing this through monitoring, targeted reinforcement, and increasingly through flexibility contracts that pay for load reduction instead of building copper.
The tools available
Time-of-use tariffs. The simplest and most effective. Price the overnight period low and most drivers shift without any technology beyond a timer.
Evidence from markets with EV-specific tariffs shows very high compliance, because the saving is large and the behaviour change is trivial.
Managed charging. The supplier or network operator schedules charging within a window the driver specifies.
More effective than static tariffs, because it avoids the secondary problem: a low-price window starting at a fixed time creates a new synchronised spike at that moment.
Randomising start times within a window solves this and requires the charger to be controllable.
Smart charger mandates. Several jurisdictions now require domestic charge points sold to be capable of scheduling and of receiving signals, with default off-peak settings.
Workplace and depot charging, which shifts a share of demand to daytime when solar output is high.
Bidirectional charging, discussed elsewhere, which turns the fleet from a problem into a resource — and which is further off.
The commercial fleet case
Depot charging for buses and delivery vehicles is a much larger local load than domestic charging and is far easier to manage.
A depot has a single operator, known schedules, and a strong incentive to minimise demand charges. Sequencing chargers so that the depot never exceeds a target draw is straightforward software.
Which is why fleet electrification has generally proceeded with fewer grid problems than expected, despite the loads being large.
What the evidence from high-adoption areas shows
Regions with EV penetration well above the average — parts of Norway, California and China — provide the useful data.
The consistent finding is that grid problems have been local and manageable, concentrated in specific distribution assets, and addressed through targeted reinforcement rather than wholesale rebuilding.
Predicted system-level crises have not materialised, largely because adoption is gradual enough for networks to respond and because charging behaviour has proved responsive to price.
What would actually cause trouble
Rapid adoption concentrated geographically, on a network with no monitoring, with flat tariffs giving no reason to shift, and with unmanaged high-power home charging.
Every element of that is avoidable and the fixes are cheap relative to reinforcement.
Which makes this a policy and metering problem rather than a physical one — a recurring theme in electrification generally.
