Fuel Economy Figures and Actuals - Sale of Goods Act

Joe_Horner

Interesting and fairly in-depth assessment of the growing difference between "test" and real life economy figures here:

http://www.theicct.org/sites/default/files/publications/ICCT_EU_fuelconsumption2_workingpaper_2012.pdf

It also gives in-depth information about the test cycle, including the fact that 25% of it is spent at idle. So cars with stop-start technology will actually have their engines switched off and be getting infinine economy for about 1/4 of the test. Needless to say, that's very unlikely to translate to similar savings on the road!

Ultrasonic

^^^^^
Thanks very much for that link. I've bookmarked it to read later but it looks really interesting.

k12479

Ultrasonic wrote: »

Do they really do this?

No they do not.

Changing engine maps affects emissions as well as fuel consumption, try and cheat on this and the consequences are severe. In the US it can be unlimited fines and the whole truck industry was punished by fines and bringing the next legislation in a few years earlier. Things like external fan cooling (because the vehicle is not actually moving) are done, as well as other stuff, but the software and maps are the same.

The problem with fuel consumption figures is that the NEDC test cycle is not representative of anyone's driving, but any new cycle will have the same problem, lots of people have looked into it including the manufacturers. They are only allowed to state cycle figures, so obviously develop things targeting that although work has been going on for a long time to improve the 'real world' fuel consumption.

ILW

Joe_Horner wrote: »

The engines are "optimised" for economy - in other words, re-mapped (the only way to optimise an ECU controlled engine) for maximum economy. Those maps used may not be suitable for long term reliability but the test only lasts about 20 minutes, so no worry there.

Your evidence for that claim is?

GolfBravo

k12479 wrote: »

No they do not.

Changing engine maps affects emissions as well as fuel consumption, try and cheat on this and the consequences are severe. In the US it can be unlimited fines and the whole truck industry was punished by fines and bringing the next legislation in a few years earlier. Things like external fan cooling (because the vehicle is not actually moving) are done, as well as other stuff, but the software and maps are the same.

We're not talking about the US, we're talking about Great Britain and Europe. There are no EU penalties, and nobody monitors the accuracy of fuel consumption tests. Same with EU tyre labels - no penalties for misleading tyre label info.

Each new car model in EU must be officially certified and it is car manufacturers that actually pay for this certification - so they can select where their cars are certified.

However, since the "most manufacturer friendly test results" are achieved at IDIADA in Spain there is a 4-month waiting list there. Their facilities in Santa Oliva offer the most ideal conditions to achieve good fuel consumption results, and also their engineers are not "too demanding".

Firstly they start with a car rolling resistance test - the results are then used for the chassis dynamometer fuel consumption test, and good result there gives you the most "fuel savings": they get a car up to 120kph, put it into neutral, and measure how far the car can travel. Their track is on a 1.5º slope (maximum allowed), so cars actually roll downhill. Their track is perfectly smooth. There is a wind cone next to the track, so they can wait for the most favourable wind. For this test cars are usually fitted with barely legal tyres (low rolling resistance) and are pumped up to maximum pressure -10% - nothing wrong there, this is all perfectly legal.

And then you have the actual chassis dynamometer fuel consumption test - since most tested cars are not yet released they use ECU software that slightly differs from a standard/final ECU software.

Should a customer be disappointed with the fuel consumption IDIADA can help with the ECU too. Nothing too drastic, just a slightly less conservative engine mapping, but offering better performance/lower fuel consumption.

Also the engine oil differs from a standard spec car - the most widely used for test purposes is Castrol Edge 0W30 (IDIADA have a big 2000l oil tank available, in case any manufacturer forgets to use the "right" oil). Plus the engines are well run in, at least 10K km.

BTW, even the EU see that manufacturers are taking the !!!! (test/real life differences up to 38%), and they want to replace the current fuel consumption test with UN's World Harmonised Light Vehicle Test Procedure, and the current plan is to introduce it by 2017. The only problem is that Germany and Italy are opposing it, they want it well after 2020 - I guess their car manufacturers would be forced to revise their official fuel consumption figures up.

k12479

GolfBravo wrote: »

And then you have the actual chassis dynamometer fuel consumption test - since most tested cars are not yet released they use ECU software that slightly differs from a standard/final ECU software.

Yes, the calibration may not be complete, and some revisions to the software may occur later too, but there is not a 'certification' software and calibration and 'production' one solely to get round emissions and fuel consumption as has been alluded too.

Joe_Horner

k12479 wrote: »

Yes, the calibration may not be complete, and some revisions to the software may occur later too, but there is not a 'certification' software and calibration and 'production' one solely to get round emissions and fuel consumption as has been alluded too.

Not sure anyone was saying there were specific "certification" and "production" maps. Simply that they don't necessarily use the same ones for the tests and for production.

For the (pre-production) tests they map for maximum economy under the test conditions then, as they develop it for final production, they may find they need to change those maps for driveability and reliability.

How the manufacturers justify that (as you're trying to) is immaterial, the fact remains that the ECU on your production car is unlikely to be running the same software / maps as were tested, so is unlikely to have any chance of meeting the same consumption figures even if they kept everything else the same (which they don't).

Hence the documented fact that the difference between published and real life economy (and CO2 emissions) has been growing rapidly as manufacturers learn what works best "on the test".

Joe_Horner

ILW wrote: »

Your evidence for that claim is?

As I asked gloomanddoom earlier, which part? That they optimise settings, that the only way to optimise settings is by mapping on an ECU controlled engine, or that those optimum economy settings may not be suitable for rl road conditions?

GolfBravo

k12479 wrote: »

Yes, the calibration may not be complete, and some revisions to the software may occur later too, but there is not a 'certification' software and calibration and 'production' one solely to get round emissions and fuel consumption as has been alluded too.

During the whole fuel economy certification test nobody checks what ECU software is being used, it's not part of the certification process - and car manufacturers use this fact to their advantage.

There is no single test, they can keep testing as long as they like (IDIADA charge a flat daily fee), then simply pick the best result. Manufacturers go there with a number of ECU maps, their engineers are present during every test to make adjustments, IDIADA even offer advise re: ECU map tweaks.

Most manufacturers revise their ECU software on a regular basis (to improve reliability, driveability, etc.) and there is no requirement to retest to update the fuel consumption figures.

steve-L

k12479 wrote: »

No they do not.

Changing engine maps affects emissions as well as fuel consumption, try and cheat on this and the consequences are severe. In the US it can be unlimited fines and the whole truck industry was punished by fines and bringing the next legislation in a few years earlier. Things like external fan cooling (because the vehicle is not actually moving) are done, as well as other stuff, but the software and maps are the same.

The problem with fuel consumption figures is that the NEDC test cycle is not representative of anyone's driving, but any new cycle will have the same problem, lots of people have looked into it including the manufacturers. They are only allowed to state cycle figures, so obviously develop things targeting that although work has been going on for a long time to improve the 'real world' fuel consumption.

The problem is the manufacturers develop a car to pass artificial tests not real world.