You are here
Real-world fuel consumption is getting worse
Report: EQUA tests made public showing disparity between official figures and actual
Whatever the laughably discredited official tests might tell you, real-world fuel consumption of new cars is now getting worse. That’s according to the most comprehensive and accurately-measured set of on-road tests, which are today made public for the first time.
Emissions Analytics is the British company that does the tests, calling them EQUA. They’re taken from the same on-road tests as the NOX and CO tests EQUA launched earlier this year.
Simply put, EA straps a portable emissions lab onto the back of cars. This looks like a bodge-up but is actually extremely sophisticated and accurate. EA then connects up the tailpipe, and drives in a very controlled way around a fixed road route. The roads are in the South of England, and the speeds and style of driving are meant to mirror the average driver. Not an economy-obsessed engineer. Or a lead-footed Top Gear road-tester.
The system produces analysis of several exhaust gases, as well as the actual mpg. You can see the results by clicking on the website here.
As we’d all have predicted, the EQUA mpg measurements are a whole lot worse than the manufacturer’s quoted numbers.
Those official numbers come from a test called NEDC, or New European Driving Cycle. This was designed many years ago and was meant to give a valid basis for comparison across cars.
But the trouble with the official test is manufacturers have found more and more ways to game the system. They run the test with over-inflated tyres, special oils, mirrors removed and several other tweaks. They also design engines and gear ratios that work well in the test – which involves very gentle acceleration and slow speeds – rather than the real world.
A striking result of the NEDC test’s slow speeds is that in the real-world EQUA tests, some cars drink nearly twice as much fuel as their official figures. Small-capacity petrol engines are among the worst. See the results for the Fiat/Alfa 0.9 Twinair engine, Ford’s Ecoboost 1.0, the Mini One three-cylinder and the Smart triple. Between them they account for most of the top 10 cars that diverge furthest from their official ratings.
At the other end of the scale, cars that are relatively honest to their official ratings tend to be the ones where thirst isn’t regarded as a barrier to sales. EQUA actually found that the Aston V8 Vantage and Nissan 370Z were better than their official figures. The Mazda MX-5, Toyota GT86 and last year’s Jaguar XK, Porsche Cayman and 911 figure in that list too.
Those seven cars are all naturally aspirated. These numbers don’t of course tell us that NA engines are more economical than turbos. But they do heavily imply that turbo engines are engineered more to do well in the official test than they are on the road.
EA’s boss Nick Molden told TopGear.com that “engine downsizing has gone too far in Europe. It only exists because of the NEDC test.” Turbocharging with a small amount of downsizing can help in the real world, he says. But when the engine gets too small, the only time it’ll be running without thirsty turbo-boost is, you guessed it, in NEDC-test-type conditions.
Molden also says multi-speed gearboxes do help economy in larger cars.
Hybrids? He says the new Prius has very clever hybrid algorithms that make it work well on the EQUA drive. But other hybrids are less successful.
Molden is also sceptical about the benefit of stop-start systems. They are helpful in the nice warm NEDC lab, but they often don’t activate in cold short real-world journeys.
OK, so those things help explain why the consumption we notice – and the emissions we perhaps don’t notice but are responsible for – are so different from the NEDC ratings. But why is absolute real-word thirst for fuel getting worse? Over the past two years, EQUA mpg figures for all types of powertrain fell from 41mpg to 40, out of which the diesels dropped from 45mpg to 42.
Molden reckons engineers have taken their eye off the economy ball. Instead they are concentrating on NOx, especially since the VW scandal. In fact the recent VW diesels running the latest 1.6 and 2.0 engines have come out very clean for NOx in his test.
He also says that because from September 2017 cars will have to quote economy figures done on a new test. “We’re in a holding pattern until then.” This is the World harmonised Light vehicle Test Procedure (WLTP). WLTP is supposed to more closely approach real-world driving. But Molden says he still thinks there will be a gap. And also, WLTP is still a lab dyno test.
EQUA has now tested 800 cars. And it reckons it now has enough experience to calculate a figure for many cars it hasn’t actually tested. So you can look up most cars on the website.
For the ones not actually tested, EQUA applies a correction for weight and drag and other factors. So it can get a figure for a 4WD version, or an automatic, of a car it has tested. Or indeed a bigger car with the same engine as a small one it’s tested.
But the website is honest enough to distinguish these calculated figures from the actual measured ones. Molden says the tested cars have an error range of about 2-3 percent. The error is about double that for the cars with a calculated value.