Monitoring and measuring fuel use

This page gives some guidance to fuel monitoring and measurement, including tips for starting out and key performance indicators (KPIs).

What monitoring tells you

Without monitoring it’s impossible to know whether:

  • fuel is being wasted (or even stolen)
  • you've chosen the best vehicle for the job
  • the vehicle is being driven aggressively
  • the engine is left idling for long periods of time
  • drivers are using the skills they were taught during driver training
  • fuel saving claims made by salespeople for devices are true

What to measure

Measurements should give you useful information to monitor KPIs, which must be relevant, reliable and easy to measure.

Example KPIs

  • average fuel efficiency (km/litre)
  • energy Intensity = fuel consumed / (tonnes carried x distance travelled)
  • unnecessary idling
  • speeding events and time spent above the speed limit
  • fuel use per month in litres and $
  • incidents and accidents

Measurement methods

Measurement options range from a simple pen and paper approach to on-board GPS-based systems. Sophisticated systems measure more information, but it's important to understand what is being measured and how, and account for any inaccuracies in each system.

  • Hubodometers and odometers - have a measurement uncertainty of ±2% to ±4%. This is partly because they measure tyre rotations and as tyres wear, they become smaller.
  • In-service monitoring of fuel use - has a margin of error typically ±3 to ±5% if averaged over five or more fills, as long as the truck's tank is filled to the same level each time.
  • Fuel weighing - the most accurate way of measuring fuel with a margin of error of less than ±0.5%.
  • Specialist flow meters - have a margin of error better than ±1%, but are best for evaluating fuel saving devices and driver training because they are relatively expensive.

It’s unreliable to measure fuel use by filling the tank from the pump before and after a short trip (for example during driver training sessions) as the errors can be as high as 50%.

Errors can also happen during data entry and analysis. Be careful not to use averages of averages - this is a common mistake, which gives inaccurate results.