How to Use This Tool
Record the distance since the previous fill and the fuel it took to fill up again, one line per tank. Several tanks is not a nicety — it is the whole method.
How to measure it properly
- Fill to the same point every time. First auto-shutoff click, no topping up. Topping up is what makes the fill level inconsistent, and consistency matters far more than fullness.
- Reset the trip meter at the pump, or note the odometer.
- Drive normally and fill up again the same way.
- Record distance and litres — that pair is one measurement.
- Repeat for several tanks and divide total distance by total fuel.
The fuel figure is always for the previous tank's driving. Getting that off by one is the usual mistake, and it produces a figure that looks plausible and is wrong.
Why one tank tells you nothing
The pump's auto-shutoff triggers on backpressure from the filler neck, which depends on the nozzle angle, how fast the fuel is flowing, how the car is parked and the temperature. A litre either way is entirely normal.
On a 44.8 L fill that is 2.2% of the measurement, and it lands directly on the result: the same 612 km comes out at 7.16, 7.32 or 7.48 L/100km depending only on where the nozzle clicked. Two tanks that differ by 3% may well be identical driving.
Averaging across tanks fixes it, because the errors are random in sign. Total distance divided by total fuel is the right calculation — not the average of the per-tank figures, which weights a short tank the same as a long one.
Why the trip computer disagrees
It does not measure fuel. It calculates it from how long the injectors were open, multiplied by an assumed flow rate, and divides by distance from the wheel speed sensor. Both inputs carry error, and the errors do not cancel.
In practice these displays read optimistically more often than not, typically by a few per cent. A computer showing 6.9 L/100km against a measured 7.32 is 6.1% out.
It matters less than it sounds for day-to-day driving, because a consistent bias still shows you whether today was better than yesterday. It matters a lot for anything absolute: comparing two cars, checking a manufacturer's claim, or working out a genuine cost per kilometre.
The pump is the better reference, and not a perfect one — pumps are periodically verified for accuracy, and fuel volume changes slightly with temperature, so a cold morning fill contains marginally more fuel per litre than a hot afternoon one.
Reading the spread
Once you have several tanks, the variation between them is the interesting part. Perhaps 2% of it is fill-level noise. The rest is real, and it is usually the mix of journeys: short cold trips are dramatically worse than long warm ones, and motorway speed costs more than most drivers expect because drag rises with the square of speed.
If one tank is far off the others, look at what was different about that fortnight before concluding something is wrong with the car.
