BEV Ready
Product Case studies Blog Book an engineer call
← All posts

You Probably Don't Need the Grid Upgrade You Think You Need

By BEV Ready · May 11, 2026

Table of Contents

Somewhere in nearly every depot electrification plan, there's a number that looks like this: fleet size, multiplied by charger power rating, equals required capacity. It's the number that gets quoted in a boardroom, the number that triggers a call to the utility, and — more often than anyone would guess — the number that's wrong.

Not wrong because the math is bad. Wrong because the assumption underneath it — that every vehicle plugs in at the same moment, at full power, for the whole session — almost never describes how a real depot actually operates.

The Assumption Hiding in Plain Sight

Worst-case simultaneous draw is the easiest capacity estimate to produce. It requires two inputs — how many vehicles, how many kW per charger — and no data about your actual operation. That simplicity is exactly why it shows up so often in early planning conversations, and exactly why it's so frequently misleading.

It's not a wrong number in the sense of being miscalculated. It's the correct answer to a question nobody's operation actually asks: what happens if every vehicle plugs in at the exact same second and stays at full power the entire time? For most depots, the honest answer is: that never happens, and planning as if it will overstates the capacity requirement substantially.

A Worked Example: 30 Vehicles, Two Estimates

The following uses illustrative numbers to demonstrate the concept, not a real customer result.

Consider a depot running 30 delivery vans, each drawing 19.2 kW at full charge rate, returning between 4pm and 7pm and departing again at 6am.

Worst-case simultaneous draw: 30 × 19.2 kW = 576 kW. Against a typical commercial service in the 200–400 kW range, this looks like an obvious, non-negotiable upgrade — and a large, expensive one.

Realistic staggered draw: The vans don't return in the same minute — they trickle in across a three-hour window based on route length. Charging doesn't need to run at full power continuously either; with a 12-hour overnight dwell window, most vehicles only need a fraction of that time at full rate to reach full charge. Modeled against actual return times and a managed charging schedule that sequences sessions across the dwell window, realistic peak draw for the same fleet often lands at a third to a half of the worst-case figure — sometimes lower.

That's the difference between a depot that needs a six-figure utility upgrade with a multi-year timeline, and one that needs a managed-charging system and no upgrade at all. Both numbers came from the same 30 vehicles. Only one of them describes what actually happens at the depot.

Why the "Conservative" Number Isn't Actually Safe

It's tempting to treat the worst-case number as the safe default — better to overestimate than underestimate, the thinking goes. But overestimating capacity needs has real costs, not just theoretical ones:

  • Unnecessary capital. A utility upgrade that isn't needed can run into six or seven figures, plus the multi-year timeline to secure it.
  • Delayed timelines. Applying for a service upgrade that turns out to be unnecessary still consumes the queue slot and the lead time, delaying a project that could have proceeded without it.
  • Wrong site decisions. A depot that looks infeasible under a worst-case estimate might be entirely workable under a realistic one — and get ruled out before anyone checks.

Underestimating capacity is a real risk too, which is exactly why the fix isn't to guess more optimistically — it's to model the realistic number properly instead of substituting a worst-case placeholder for either direction of guess.

What Actually Changes the Real Peak

Four variables separate the worst-case number from the realistic one, and all four are knowable before a single charger is installed:

  1. Return-time distribution. Do vehicles trickle back over hours, or cluster tightly around one shift change?
  2. Dwell window length. A 10-hour overnight window has far more slack to spread charging across than a 5-hour one.
  3. State of charge on return. Vehicles returning at 60% need less energy — and less time at full power — than ones returning at 15%.
  4. Whether charging is managed or unmanaged. Managed charging software that sequences sessions by departure time, rather than plug-in time, is usually the single biggest lever on realistic peak demand.

When You Genuinely Do Need the Upgrade

None of this means grid upgrades are never necessary — plenty of depots genuinely need one, particularly larger fleets on tight dwell windows or sites with limited baseline headroom to begin with. The point isn't to assume you don't need an upgrade. It's to find out based on a realistic model rather than a worst-case placeholder, so that when an upgrade is recommended, it's because the site genuinely needs it — not because nobody modeled the alternative.

The Other Peak Problem: Demand Charges

Even a fleet that fits within its site's electrical capacity can get a costly surprise on its first utility bill. Most commercial electricity tariffs include a demand charge — a fee based on the highest instantaneous power draw recorded in a billing period, measured in kW, separate from the per-kWh energy charge. One bad 15-minute window where too many vehicles charge simultaneously can set the demand charge for the entire month.

This is a distinct problem from the capacity constraint, and it doesn't go away once the capacity question is resolved. A depot that has plenty of headroom to charge its full fleet can still see demand charges spike if charging sessions cluster together — for example, when an entire shift returns and plugs in within the same window. The grid connection can handle it electrically; the tariff structure still penalizes it financially.

Managed charging addresses both problems with the same mechanism — spreading sessions across the available dwell window — but for different reasons. On the capacity side, it keeps peak draw below the site's service limit. On the cost side, it keeps the demand charge peak as flat as possible. The fleet that skips managed charging because its site has enough raw capacity often discovers that the demand charge alone justified the investment.

Model Your Real Peak Before You Call the Utility

If a worst-case estimate is the number currently driving your depot's electrification budget or timeline, it's worth a second look before that number turns into a utility application. The gap between worst-case and realistic peak demand is often the difference between a project that needs a multi-year upgrade and one that doesn't.

Try the BEV Ready feasibility calculator — free, no account required.

FAQ

Does every vehicle in a fleet really charge at the same time? Rarely, in practice. Return times, shift patterns, and dispatch schedules stagger naturally, and managed charging software can stagger them further. But most back-of-envelope capacity estimates assume simultaneous full-power plug-in for every vehicle, which is the worst realistic case, not the typical one.

Why do capacity estimates default to worst-case simultaneous draw? It's the easiest number to calculate — multiply vehicle count by charger power rating — and it's conservative, which feels safer. The problem is that it's often so conservative it recommends a utility upgrade a site doesn't actually need, based on a charging pattern that will never occur operationally.

How much difference does staggered charging actually make to peak demand? It varies by fleet, but the effect is frequently large — a fleet with a wide return window and a long overnight charging period can see peak demand drop by half or more once charging is scheduled by departure time rather than arrival time. The exact figure depends on fleet size, dwell window, and charger power, which is why it needs to be modeled per site rather than assumed.

When does a fleet actually need a grid upgrade? When the realistic peak demand — modeled against actual dwell windows and staggered or managed charging, not worst-case simultaneous draw — still exceeds the site's available service capacity. That's a genuinely different number from the worst-case estimate, and it's the one that should drive the upgrade decision.