How to measure real fuel consumption with a fill-up log

Fuel is usually the largest variable cost in a fleet, and the budget is usually built on the vehicle's catalogue consumption. The problem is that the catalogue figure is not even an estimate — it is the result of a standard test cycle, and that cycle contains no site roads, no loaded body and no air conditioning running while the vehicle waits at a junction.

Measuring real consumption needs no GPS or CAN bus hardware. A simple method that has been in use for decades gives a reliable L/100 km per vehicle as long as it is kept consistently: the fill-up log.

Why does catalogue consumption not match reality?

Because it is measured on a standard test cycle — a set temperature, an unladen vehicle and a fixed driving profile. Most of what determines consumption in service sits outside that cycle.

The main factors that push the real figure up:

• Load — the gap between a full and an empty body is the single biggest factor in commercial vehicles.
• Idling — an engine running during loading, unloading and waiting produces no kilometres but burns fuel, quietly raising cost per kilometre.
• Share of urban driving — the more stop-start traffic in the mix, the higher the consumption.
• Season — warm-up time in the cold, air conditioning in the heat; both push the same way.
• Tyre pressure and maintenance state — small omissions create a permanent percentage gap.
• Driving style — acceleration and cruising speed choices produce a measurable difference between drivers in the same vehicle.

None of these are exceptions; they are normal operating conditions. Real consumption coming out above the catalogue figure is therefore not a sign of a fault. What carries meaning is the size of the deviation and where it sits relative to other vehicles working in the same conditions.

What is the fill-up log method?

Fill the tank completely and record the odometer, then fill completely at every subsequent stop, recording litres and kilometres. The fuel burned over the interval equals the litres added afterwards — excluding the first fill-up.

The logic is straightforward. If the tank is at the same level at the start and at the end — full — then the fuel burned over the distance equals the fuel put in over that distance. The litres of the first fill-up do not count, because that fuel was burned before the measurement started; it only starts the odometer clock.

Worked example: four fill-ups

Records for a diesel van with a declared consumption of 7.5 L/100 km:

The gap between the declared 7.5 L/100 km and the measured 8.22 L/100 km is 9.6%. For a vehicle covering 2,400 km a month at the July 2026 diesel price of roughly ₺73.50 per litre, that small-looking percentage is about ₺1,270 a month, or ₺15,200 a year. The money is already being spent; measuring only brings it into view.

How many fill-ups before the figure is reliable?

Two are mathematically enough, but a single interval is very exposed to chance. Before acting on the number, wait for at least three intervals and roughly 1,500 km.

A single interval misleads because its content is not homogeneous. Six hundred kilometres largely on the motorway and six hundred kilometres in town describe the same vehicle very differently. As intervals accumulate, those differences balance out.

• 1 interval: indicative, not decision-grade.
• 3 intervals / ~1,500 km: a figure you can discuss per vehicle.
• 6+ intervals / a full season: seasonal effects even out and a trend becomes readable.

What should the deviation be compared against?

The average of vehicles doing the same job, not the catalogue figure. The catalogue value is an optimistic baseline; the fleet average already contains real operating conditions.

The practical consequence: if most vehicles sit around 10% above the catalogue figure, the likely explanation is not that ten drivers have simultaneously become wasteful, but that the baseline is too low for this kind of work. Talking to drivers one by one would be the wrong address.

When one vehicle sits clearly above its peers on the same route, however, the place to look is obvious: the vehicle's technical condition, its loading profile, or its driver. Comparison is what makes a deviation meaningful.

That last row matters. The only way to tell whether you are discussing the vehicle or the driver is to see the same vehicle with different drivers and the same driver in different vehicles. A single average cannot make that separation.

Six errors that ruin the measurement

1. 1Not filling the tank completely. The method rests on one assumption: the same level at the start and the end. A partial fill invalidates that interval.
2. 2Not recording the odometer at the pump. A reading remembered in the evening makes the distance wrong, and the error passes straight into the result.
3. 3Counting the first fill-up. The most common mistake, and it makes consumption look higher than it is.
4. 4Mixing fill-ups between vehicles. A receipt without a plate or vehicle id is a record you cannot trace when a figure goes wrong.
5. 5Not stripping out one-off events. Theft, leaks or fuel bought in cans will send consumption through the roof if they enter the same table. Those records need to be flagged separately.
6. 6Comparing across seasons. Setting January against July tells you about the weather, not the driver. Comparisons belong between like periods.

What do you do with the result?

Measured consumption is not a performance score on its own; it is an input to several calculations. A fuel budget built on it lands close to reality. The fuel line in the cost per kilometre calculation fills correctly. Trip pricing is based on field data rather than a spec sheet. In a replacement decision, a vehicle whose consumption is drifting upward makes itself visible early.

The hard part is not the arithmetic, it is continuity: unless the record is entered in the field, at the pump, against the right vehicle, the table has to be rebuilt a month later. Once the logging discipline holds, the rest is four operations.