For most road freight operators, fuel is one of the largest costs on every trip, and one of the least questioned. The fuel card statement arrives, the total looks roughly right, and it gets paid. The problem is that a total tells you nothing about whether each truck, on each trip, used what it should have.
Fuel leakage rarely looks like one big event. It is a few litres here, a mis-recorded odometer there, a truck that has quietly started burning more than it did six months ago. Individually these are invisible. Over a fleet and a year, they add up to real money.
Start with a baseline, not a guess
You cannot spot a bad number without a good one to compare it against. For each vehicle, you need an expected consumption figure, usually litres per 100 km, that reflects how that truck actually runs. A few things should shape it:
- Vehicle and configuration: a truck-tractor with a tautliner behaves differently from a rigid, and a unit pulling a superlink differs again.
- Load condition: loaded, part-loaded and empty running will give different results. A single average hides this.
- Route type: flat highway, mountain passes and congested urban delivery do not burn the same.
- Vehicle age and condition: a baseline set two years ago may no longer be fair.
Build your baseline from your own history on clean data, not from the manufacturer's brochure. Brochure figures are a useful sanity check, but your roads, loads and drivers are what count.
Capture the right data at every fill
Consumption analysis is only as good as what is recorded when fuel goes into the tank. At minimum, every fill should record:
- The vehicle (registration), confirmed against the card or bowser entry
- The driver
- The odometer reading at the time of the fill
- Litres dispensed and the price per litre
- Location and date and time
- Whether it was a full-tank fill
- The trip or load it belongs to
The odometer and the full-tank flag are the two that most often go missing, and they are the two that matter most. Consumption is calculated between full fills: distance travelled between two full tanks, divided by the litres needed to refill. Without a reliable odometer, the calculation is meaningless. Without the full-tank flag, you are dividing by the wrong number.
Look at variance, not just totals
Once you have a baseline and clean fill data, the useful question for each trip or period becomes: how far was actual consumption from expected, and is that gap normal?
Set a tolerance band rather than expecting exact matches. Real consumption moves with wind, traffic, border delays and load weight. A small variance is noise. A consistent or large one is a signal. Many operators use a simple rule: anything outside the band gets a quick review, and anything outside it repeatedly gets a proper investigation.
Patterns worth watching
- One vehicle consistently high: points toward a mechanical issue, tyre pressure or alignment problems, an injector or turbo fault, or a driver using it differently.
- One driver high across different vehicles: usually driving style, excessive idling, speeding or route choice. Occasionally something less innocent.
- Litres that exceed tank capacity or do not fit the distance: a fill that could not physically have gone into that truck is a red flag for fuel going into another vehicle or containers.
- Fills in unusual places or at odd times: off-route, late at night, or two fills close together.
- Sudden step changes: consumption that jumps overnight often has a cause you can identify, such as a repair, a new trailer, or a change in routes.
- Odometer jumps or going backwards: a data entry problem, or someone manipulating the reading.
Separate causes before you assign blame
The temptation, when numbers look bad, is to assume theft. Often it is something much more ordinary. Before confronting a driver, rule out:
- Data entry errors on the odometer or litres
- A wrong vehicle registration captured against the fill
- A partial fill that was recorded as a full tank
- A genuine route deviation caused by a road closure or a customer instruction
- Long waiting time with the engine running, particularly at busy collection or delivery points
- A real mechanical or tyre fault
Handled this way, fuel reviews become a maintenance and driver-coaching tool first, and a control tool second. Drivers are far more likely to accept the process if it is clearly fair and based on evidence.
Cross-border and price differences
On cross-border runs, fuel prices and availability differ, and operators often plan fills to buy where it is cheapest. That is sensible, but it complicates consumption analysis. A tank filled in one country and burned in another means the fill-to-fill method still works, but only if every fill is recorded with its own odometer and location. Keep the cost per litre as a separate analysis from the consumption one, so a cheap fill is not mistaken for an efficient trip.
Feed it back into your costing
Fuel consumption should not live in a spreadsheet nobody opens. It should feed your cost per kilometre, your quoting and your trip profitability. If you quote on a fuel figure that is a year out of date, you are either losing loads by overquoting or losing margin by underquoting. Recalculating baselines every quarter or after any major change keeps those numbers honest.
Make the process routine
The operators who control fuel well tend to do the same few things consistently:
- Record the full set of data at every fill, with no exceptions.
- Calculate consumption per vehicle and per trip as part of closing the trip, not months later.
- Review exceptions weekly, while the trip is still fresh in everyone's memory.
- Share results with drivers, including the good ones.
- Adjust baselines when vehicles, trailers or routes change.
This is much easier when fuel entries are tied directly to the vehicle and the trip they belong to, rather than reconciled by hand from card statements. A transport management system such as TitanFleet can hold that link, so the variance is visible when the trip closes instead of surfacing at month-end.
The takeaway
You do not need to catch every missing litre. You need a baseline you trust, clean data at the point of fill, and a habit of looking at variance while it can still be explained. Do that, and fuel moves from a line you simply pay to a cost you actually manage.