Official fuel economy figures: why are they so wrong?

Every car on the market comes with a fuel economy figure most commonly expressed in miles per gallon or, if you’re buying a grey import, sometimes litres per 100km. The tests have been around for over 30 years and no longer reflect modern vehicle performance. New tests called the Worldwide Harmonized Light Vehicle Test Procedures (WLTP) are being introduced soon, but this will have an impact on the tax breaks some manufacturers get for their cars being under 100g/km of CO2 because the tests are more like real driving.

What the tests are good for, though, is comparing one car against another, because in all the tests I have done (and bear in mind I have driven and reviewed over 400 cars), I have only once beat the posted fuel economy and that was in a Toyota Prius c.

The usual situation is that I am 20-30% less economical in daily driving than the official figures suggest. So, if you have a fuel budget that you are calculating based on your estimated weekly mileage, then remember to take into account a little bit extra.

Of course there are many ways to drive more efficiently – check out 11 tips for improving fuel economy here – but everyone will get different results.

When the tests are conducted the distances and measurements are:

• 0-780 seconds: simulated driving in an urban environment.
• 780 – 1180 seconds: simulated driving in an extra-urban environment, i.e. on open roads or a motorway
• The total output of the test is called the combined cycle or complete cycle.
• CO2 emissions are calculated on the complete cycle.

The tests are conducted either by a professional driver tester on a flat surface, or on a roller test bench with specific resistances.

Urban cycle

The urban cycle is designed to simulate city driving at no more than 50kph or 31mph:

1. Start the engine and wait 6 seconds in neutral
2. Change into first gear and wait 5 seconds before moving off
3. Accelerate to 15kph / 9mph in 4 seconds
4. Hold 15kph /9mph for 8 seconds
5. Brake to zero in 5 seconds
6. Stop for 21 seconds
7. At 49 seconds accelerate to 32kph / 20mph in 12 seconds
8. Cruise at 32kph / 20mph for 24s
9. Brake to a full stop in 11s
10. Pause for 21s
11. At 117s accelerate to 50kph / 31mph in 26 seconds
12. Cruise at 50kph /31mph for 12s
13. Decelerate to 35kph / 21.5mph in 8s
14. Cruise at 35kph /21.5 mph for 13s
15. Brake to a full stop in 12s
16. Pause for 7s

This is now 195s and a theoretical distance of 1017m. The cycle is repeated another three times giving a total theoretical distance of 4067m over 780s with an average speed of 18.77kph / 11.5mph.

If the tested car is manual then specific times are given for changing gear.

Extra-urban cycle

The extra urban cycle is designed to simulate driving at higher speed. The maximum speed is limited to 120kph / 74.5mph.

• Idle for 20s
• Accelerate to 70kph / 43.5mph in 41s
• Cruise for 50s
• Decelerate to 50kph / 31mph  in 8s
• Cruise for 69s at 50kph / 31mph
• Accelerate back to 70kph / 43.5mph in 13s
• Cruise at 70kph / 43.5mph for 50s
• Accelerate to 100kph / 62mph in 35s
• Cruise for 30s
• Accelerate to 120kph / 74.5mph in 20s
• Cruise for 10s
• Brake to a stop in 34s
• Idle for a final 20s.

Total duration is 400s and the theoretical distance is 6956m with an average speed of 62.6kph / 39mph.

As you can see from the measurements above, it doesn’t approximate real world driving. for example, in the extra urban cycle, taking 34 seconds to stop from 74.5mph is very unrealistic, and taking 41 seconds to get to 43.5mph when most cars will do it in well under 10 seconds is also unrealistic.

However, the figures are good to benchmark one car against another.