How to calculate daily energy demand for an electric fleet

Manufacturer range does the opposite job. It tells you the distance a vehicle can cover before it stops, under a stated test or a stated load. Depot sizing does not care about that. It cares about how much energy has to go back into the battery between the last return and the first dispatch, which is a function of miles driven and nothing else.Using range also imports an error you cannot see. A 250-mile van running 90-mile days does not consume 250 miles of energy overnight. It consumes 90. Size a depot on rated range and you will buy roughly two and a half times the service capacity the operation needs, plus the transformer, the switchgear and the make-ready that come with it.If the electric vehicles are not on the road yet, use the diesel or gasoline vehicles they are replacing. Those odometers are running the same routes.

Light-duty vehicles are easier, because they carry an EPA efficiency rating on the window sticker and on fueleconomy.gov, expressed as kilowatt-hours per 100 miles. Divide by 100 and you have kWh per mile directly.Above Class 3 there is no EPA rating at all, and consumption moves enormously with gross weight, terrain and duty cycle. That is not a gap in the published data you can close by finding a better source. It is a genuine property of the vehicles, and it is why the number has to be replaced with your own within a few weeks of the first units arriving.

Cold has three separate mechanisms and they add up. Cabin heat is resistive or heat-pump load that comes straight off the pack. Battery thermal conditioning draws power to hold the cells in their operating window, including while parked and plugged. Regenerative braking is limited on a cold pack, so energy that would have come back does not.Payload is more linear and more predictable. A box truck at rated gross weight consumes materially more per mile than the same truck empty, and a route that runs full outbound and empty back does not average out the way people assume, because the loaded leg does the climbing on more days than not.The right way to handle both is a sensitivity table. Pick the multiplier you can defend, see what it does to the port hours, and check whether the design still fits the window.

Meter kWh is larger, because AC charging loses energy in the onboard charger, the cabling and the thermal management running during the session. The gap is small enough to ignore on one van and large enough to matter across a forty-vehicle yard.We use a ten percent allowance in the roll-up below as a stated planning assumption. It is not a measured figure and it should not survive contact with your first month of billing data — compare the meter reading against the telematics energy-added figure and use your own ratio.

The distribution matters more than the mean, because a depot that cannot fully charge the fleet on a heavy day starts the next day with a deficit, and deficits compound across a week.Pull the 90th-percentile day per vehicle rather than the mean, and separately look at the worst week in the data. If the 90th percentile is 40 percent above the mean, that is the design case, and it is also the argument for the ports and the capacity that the finance conversation will otherwise trim.Look for the vehicles that are not like the others while you are in the data. Most fleets have a handful of units running double the miles of the rest. Those may justify a higher-power port or a mid-shift top-up rather than dragging the whole depot's sizing upward.

Energy per vehicle per night divided by delivered kilowatts per port gives port hours, which with dwell time gives your port count. Fleet energy per night divided by the charging window gives the managed peak in kilowatts, which is the service capacity you actually need rather than the sum of the nameplates.It also drives the tariff conversation. Total kWh per month and the shape of the load across the day are what a utility account representative needs in order to say which rate schedule applies and whether an EV-specific schedule exists.Get this number wrong upward and you buy a transformer you did not need, on a lead time you did not need to wait for. Get it wrong downward and vehicles do not go out. The first error is expensive; the second stops the business, which is why the rounding on this calculation goes up.