Excerpt: ArcBest’s Tesla Semi pilot averaged 1.55 kWh per mile, and fleet managers are calling it a breakthrough. But that number doesn’t mean what most people think it means.
The Number Everyone Is Repeating
A logistics executive at a trade show tells you the Tesla Semi gets 1.55 kWh per mile. That’s the number ArcBest reported after running two units in their ABF Freight pilot program. Another fleet manager overhears and nods knowingly. “That’s incredible efficiency,” he says. “Way better than I expected for something that heavy.”
The claim spreads. Forum posts compare it to passenger EVs. A Reddit thread calculates what it would cost to run cross-country routes at that rate. The consensus forms: Tesla Semi efficiency is shockingly good, maybe even suspiciously good, and it proves electric trucks can replace diesels without compromise.
Except the 1.55 kWh/mile figure doesn’t tell you what people think it tells you. It’s a real number from a real pilot program, but the way it’s being interpreted misses three critical constraints that determine whether that efficiency translates to viable operations.
Where the Confusion Started
The 1.55 kWh/mile number comes from ArcBest’s pilot program. They ran two Tesla Semis in their fleet and tracked energy consumption over several thousand miles of real-world driving. That’s not a lab test. It’s actual freight routes with actual cargo.
The problem is that people hear “1.55 kWh per mile” and immediately start doing back-of-the-envelope math. They know a diesel Class 8 truck gets around 6.5 miles per gallon. They know diesel costs about $3.50 per gallon. They assume electricity costs $0.12 per kWh at industrial rates. They multiply and divide and conclude the operating cost advantage is massive.
What they’re missing is that kWh per mile is a partial metric. It measures one thing: energy going into the wheels relative to distance traveled. It does not account for payload capacity utilized, route topology, charging infrastructure availability, or battery degradation over the duty cycle. Those constraints determine whether 1.55 kWh/mile represents a working business model or an optimized demonstration run.
What That Efficiency Number Actually Measures
Start with the physics. A Tesla Semi has a battery pack estimated at 900-1000 kWh based on Tesla’s claimed 500-mile range. At 1.55 kWh per mile, that gives you 580-645 miles of theoretical range. That’s not 500 miles of operational range. It’s 580-645 miles if you run the battery to zero, which no fleet does because battery longevity depends on avoiding deep discharge cycles.
Real-world practice keeps the state of charge between 20% and 80%. That gives you 540-600 kWh of usable energy. At 1.55 kWh per mile, your operational range is 348-387 miles before you need to find a charger capable of delivering 750 kW, which is Tesla’s target charge rate. Those chargers do not yet exist in meaningful density outside of Tesla’s planned network.
Now add payload. The 1.55 kWh/mile figure came from ArcBest’s pilot routes. We don’t know the average payload weight they carried. A Class 8 tractor-trailer combination can legally gross up to 80,000 pounds. If ArcBest was running partial loads or prioritizing efficiency to validate the technology, the real-world number could be higher. Aerodynamic drag is constant, but rolling resistance and the energy required for acceleration scale with mass. Add 10,000 pounds of cargo and you might see 1.65 or 1.70 kWh per mile. That cuts your operational range to 318-353 miles.
Route matters too. The miles ArcBest logged could have been optimized for efficiency testing. Flat highways in temperate weather yield better numbers than mountain passes in January. Regenerative braking helps on downhill grades, but it doesn’t fully recover the energy you spent climbing. The Tesla Semi efficiency you get in the Midwest is not the same as what you get crossing the Rockies.
The Constraint Everyone Ignores
Here’s the part that doesn’t show up in press releases: a diesel Class 8 truck can refuel in 15 minutes and has an operational range of 1,000 to 1,500 miles per tank, depending on tank size and driving conditions. The Tesla Semi, even with 750 kW charging at Tesla’s claimed rate, needs at least 30 minutes to add 200 miles of range. That assumes the charger is available, functional, and located where your route needs it.
The constraint is not energy efficiency. The constraint is duty cycle compatibility. A 350-mile operational range works for regional freight. It does not work for long-haul operations unless you rebuild your entire logistics network around charging infrastructure that doesn’t exist yet. ArcBest buying additional units tells you they found routes where the math works. It doesn’t tell you the math works for the majority of freight miles driven in North America.
Fleet operators care about ton-miles per day. If your truck spends 30 minutes charging every 300 miles, you’re losing productive time. If you run two shifts, you can charge during driver breaks and overnight. If you run three shifts or do irregular long hauls, the efficiency number stops mattering because you can’t keep the truck moving.
Why This Keeps Happening
The market wants Tesla Semi efficiency to validate the entire electric truck thesis. Battery costs are falling. Policy is pushing electrification. Fleet managers are under pressure to decarbonize. A good efficiency number feels like proof that the technology is ready.
But readiness is not binary. The technology works for specific duty cycles. ArcBest’s pilot proved that. The mistake is extrapolating from “this works for regional routes with predictable distances and access to charging” to “diesel trucks are obsolete.” The latter requires infrastructure deployment at a scale that won’t exist for another decade, even with aggressive investment.
Manufacturers also have an incentive to emphasize efficiency because it shifts the conversation away from operational constraints. If people are arguing about kWh per mile, they’re not asking about charger availability or battery replacement costs after 500,000 miles.
What the Actual Reality Looks Like
Tesla Semi efficiency at 1.55 kWh per mile is achievable under real-world conditions. That part is true. The question is whether it’s achievable at the payload weights, route profiles, and duty cycles that define most freight operations.
For regional hauls under 300 miles with depot charging, the math probably works. For long-haul freight, it doesn’t work yet because the charging infrastructure isn’t there and the operational range isn’t sufficient. The efficiency is real. The constraint is not energy consumption per mile. The constraint is total system throughput when you account for charging time, route planning, and infrastructure availability.
ArcBest is buying additional units because they identified routes where those constraints don’t bind. That’s smart fleet management. It’s not evidence that the average freight operation can replace its diesel fleet tomorrow. The efficiency number is impressive. The deployment envelope is still narrow.