If you run interjurisdictionally you already collect both halves of this number every quarter. The work is not measurement — it is attributing fuel and distance to a unit rather than to a fleet.
Key takeaways
Fuel economy per truck is litres of fuel issued to that unit, divided by the distance that unit travelled over the same period, multiplied by 100. Measure it over a period long enough to absorb partial tank fills — a month is usually the shortest honest window for a single unit — and rebuild it the same way every time.
The reason carriers do not have this number is never the formula. It is that fuel is attributed to a driver or a card rather than to a unit, and distance is attributed to a trip rather than to a unit. Fix the attribution and the metric appears.
If you run in more than one province or state, you file under the International Fuel Tax Agreement, which means you already do this work at fleet level. Ontario describes the obligation directly: under IFTA, interjurisdictional carriers report the amount of motor fuel consumed and the distance travelled in each jurisdiction, through a quarterly return filed with the base jurisdiction. The vehicles in scope are those with two axles and a gross or registered gross weight over 11,797 kg, three or more axles regardless of weight, or a combination total over 11,797 kg.
You are also already required to keep the underlying records: an Ontario licensee must keep all records supporting the information reported on the IFTA return for four years from the return due date or filing date, whichever is later. So the data exists, it is retained, and it is auditable. What it is not, in most small fleets, is broken down per unit — and that is the only change required.
One caution about scope. IFTA credentials are valid in the ten provinces and 48 states, but not in the Northwest Territories, Nunavut, Yukon, Alaska, Hawaii or the District of Columbia. If you run north, distance and fuel from those trips has to be captured some other way or your per-unit figure will quietly under-count.
Distance per unit can come from three places, and they do not agree. Odometer readings are the most defensible and the most annoying, because somebody has to read them at a consistent point in the month. Telematics distance is easy and continuous; Geotab, for instance, describes fleet dashboard reports that monitor fleet fuel consumption and idling, with total distance driven taken from vehicle odometer readings. Trip or dispatch distance is the least reliable, because planned kilometres are not driven kilometres.
Pick one source and keep it. A per-unit figure whose denominator changes source between months measures your reporting, not your trucks.
Three attribution problems, all fixable:
A fourth thing worth capturing even though it is not fuel: fuel transaction anomalies. Geotab describes visibility into fuel usage, spend and idling to identify inefficiencies and detect transaction anomalies. A litre discrepancy that repeats on the same card at the same time of day is not a fuel economy problem, and it will not be found by looking at the economy figure.
This is where most fuel programmes go wrong, and the failure is a management failure rather than a data one. Raw litres per 100 km across a mixed fleet compares vehicles doing different work, and the biggest variables are not driver behaviour:
One variable you can largely stop worrying about in Ontario: top speed. Most large trucks there are required to use electronic speed limiters capping them at 105 km/h — applying to commercial motor vehicles built after 31 December 1994, equipped with an electronic control module, with a manufacturer’s gross vehicle weight rating of 11,794 kg or more. Where limiters are set and enforced, the highway-speed component of the comparison is already flattened.
Three uses, in increasing order of value.
Coaching. Once figures are normalised, persistent outliers are worth a conversation about idling, progressive shifting and cruise use. Natural Resources Canada runs freight-sector training and tools for exactly this, including a Smart Driver challenge for driving a truck in a more energy-efficient way as part of its greening freight training, tools and resources to help Canadian fleets lower fuel consumption, operating costs and emissions.
Maintenance signal. A single unit whose figure degrades against its own history — not against the fleet — is the most useful signal fuel data produces. It is a maintenance ticket before it is a coaching conversation, and it belongs in the same work order flow as any other defect: see a work order workflow for a small shop.
Spec decisions. The highest-value use, and the slowest. Two years of per-unit data across different specs tells you what to buy next, which is a bigger number than anything coaching will recover.
Worked example: the unit that was not the problem
A carrier running nine tractors out of Ontario pulled its IFTA data apart by unit for the first time. One tractor came out clearly worst, and the driver assigned to it had been there eight months.
Before the conversation happened, the owner normalised three things. He separated the two units that pulled reefers, because their fuel purchases included trailer fuel. He split the fleet into highway and regional work. And he pulled idle hours from the telematics rather than assuming.
The worst unit stayed worst. But the gap shrank by more than half once reefer fuel was separated and idle hours were accounted for, and what remained tracked a single lane with a long urban leg that only that unit ran.
The driver conversation never happened. What happened instead was a shop inspection prompted by the unit’s decline against its own previous months, and a review of the lane. The lesson the owner drew was the useful one: the ranking was real, and the explanation for it had nothing to do with the person the ranking pointed at.
We will not quote one, and you should be suspicious of anyone who does without knowing your weights, lanes and specs. Published averages combine fleets doing incomparable work. The figure worth managing is each unit against its own history, where the comparison controls for everything automatically.
Litres per 100 km, because it is the Canadian convention, it scales linearly with fuel cost, and it makes small improvements legible. Kilometres per litre compresses exactly the range you care about.
No. Odometer readings taken at a fixed point each month plus fuel purchases attributed to a unit will produce the number. Telematics makes it continuous rather than monthly and adds idle time, which is the variable most worth having.
If they buy their own fuel you will not see the numerator, and per-unit economy is their business rather than yours. What you can still see is distance, and if you supply the trailer you can still measure the equipment. What belongs in the agreement is covered in what an owner-operator agreement covers.
Yes, and arguably before economy itself. Idle fuel is the only component of the numerator that produces no distance at all, so it is the one place where a change in behaviour maps directly onto a change in the number.
A short call is enough to work out where your fuel and distance attribution breaks and what it would take to fix it.
