Count the megavolt-amperes before you count the trucks

Freight companies, mine included, have learnt to announce electrification as a number of vehicles. Sixty electric prime movers by a given date. A tenth of the metropolitan fleet by the end of the decade. The number is easy to publish, easy to count and easy for investors to compare with a competitor's. In February 2023 Morrowgate's board approved exactly that kind of commitment: 60 battery-electric prime movers on the Melbourne port shuttle by June 2025.

Four months later the local distribution network business told us it would take thirty months and A$9.6 million to give our depot the supply that fleet would need. Nobody had asked the question before the commitment was made. That is not a criticism of the people involved, me included; it is how vehicle-led electrification is usually planned. My argument here is that the order should be reversed. A board that wants to govern a fleet transition should begin with the electricity connection at each depot, then the duty cycles the trucks must perform, and only then the number of vehicles. The number of trucks is a result, not a plan.

A vehicle count hides the real constraint

A battery-electric prime mover on our shuttle uses, on average, 1.28 kilowatt-hours for every kilometre. Across two shifts and about 205 kilometres, that is roughly 260 kilowatt-hours a day for each truck. Sixty trucks need more than 15 megawatt-hours every day. If most of that energy must go in during an eight-hour overnight window, the depot needs about two megawatts of continuous charging. In practice it needs more, because trucks return in clusters at shift change and every charger wants full power at once. That is how a depot whose existing peak was about 600 kilovolt-amperes came to need a 4 MVA connection.

A connection of that size is not a matter of ordering a thicker cable. The network business has to assess what its local feeder and zone substation can carry, design the works, schedule them among other customers' requests and recover part of the cost from us. The process is regulated and, in my experience, reasonable. It is also slow, and a fleet operator cannot buy its way out of the slowness.

The vehicle number hides the work the trucks can do as well. A battery adds roughly four tonnes of tare mass to a prime mover, and that mass comes out of the payload the combination may legally carry. On our shuttle about 18 per cent of containers are too heavy for the electric combinations, and they will stay on diesel until vehicle design or mass rules change. A commitment stated as sixty trucks says nothing about which loads they cannot carry, which shifts they cannot cover and how long they will wait at a charger.

What a measured load profile buys

When the quote arrived, I recommended that we start with 24 trucks on the 1.5 MVA supply the depot already had. Software would schedule charging so the site never exceeded 1.4 MVA, a 1.6 MWh battery would absorb the shift-change peak, and 450 kilowatts of rooftop solar would supply part of a midday top-up. The board accepted the recommendation, and we revised the public commitment.

The 24 trucks have all been in service since April 2025. The site's peak demand has not exceeded 1.38 MVA. Availability is 93 per cent, against 95 per cent for diesel units on the same work; most of the gap came from charger faults in the first six months and a defect in the scheduling software, both since fixed. Energy costs 38 per cent less per kilometre than diesel did, and the fleet avoids about 820,000 litres of diesel a year.

The more valuable result is the data. We now know how much energy a loaded container trip uses on a hot afternoon compared with an empty run on a cool night. We know how long trucks really sit on chargers, how much flexibility the roster gives us and how the site battery behaves across a week. When we returned to the network business with eighteen months of measured load, the conversation changed. It offered a 2.5 MVA upgrade for A$5.4 million, due in late 2027, sufficient for 20 more trucks. Our measured profile showed that a smaller, managed connection could do work the earlier estimate had assumed would need a far larger one.

There were costs, and I would rather list them than let a reader assume there were none. Our direct emissions fell later than we had told investors they would. Our managing director had to explain the revision on a results call, and some investors read it as hesitation. The first midday charging schedule clashed with drivers' meal breaks and had to be redesigned with them. And a smaller first fleet carried a higher cost per truck for chargers, switchboards and the battery.

Commitments a board can actually govern

If I were drafting a fleet transition commitment for any freight board now, I would ask it to govern four things before it counts a single truck.

The first is contracted connection capacity, depot by depot, with the date the network business has committed to. That is the hard constraint and the one most likely to slip. A board should see it as it sees a major capital project: scope, cost, dependencies and date.

The second is the share of duty cycles that electric vehicles can perform under today's mass limits and charging windows. Not every route is a candidate, and the honest share will be lower than the enthusiastic one. On our shuttle it is about 82 per cent of container moves; on interstate linehaul it is, for now, close to none.

The third is cost and availability per kilometre against diesel on the same work, measured rather than modelled. Electric trucks will not win every comparison, and a board needs to know which ones they lose and why.

The fourth is the vehicle count itself, presented as the consequence of the first three. A number of trucks follows from the energy a depot can deliver, the work the trucks can legally do and the price at which they do it.

None of this argues for delay. Our first 24 trucks were working eight months before the date on which the original quote would have let us connect even one. Electrifying short, return-to-base freight is practical now, and freight boards should press for it. My argument concerns what they press on. A board that asks only how many trucks will usually hear an optimistic number and learn about the substation later. A board that asks for megavolt-amperes, duty cycles and measured cost will hear something less exciting and considerably more likely to happen.