Wanted. Small car.

[Mick F]

Yes.
A big fiddle.
I'll never forgive them.
VW diesel scandal for emissions, and Fiat 500 Twinairs for exceeding them too.
Goodness knows which other manufacturers fiddled the books as well.

[francovendee]

Yes.
A big fiddle.
I'll never forgive them.
VW diesel scandal for emissions, and Fiat 500 Twinairs for exceeding them too.
Goodness knows which other manufacturers fiddled the books as well.

Last November I finally got rid of my 2000 Focus 1.8 diesel. It had been a thoroughly economical car to own and regularly returned 55 mpg.
I bought a VW Golf 1.6 diesel Blue Motion. Having driven it for over 9000 miles guess what, 54 mpg
All manufacturers lie about their products, well not exactly lie but give you figures that are very difficult to achieve.

[Bonefishblues]

Yes.
A big fiddle.
I'll never forgive them.
VW diesel scandal for emissions, and Fiat 500 Twinairs for exceeding them too.
Goodness knows which other manufacturers fiddled the books as well.

Last November I finally got rid of my 2000 Focus 1.8 diesel. It had been a thoroughly economical car to own and regularly returned 55 mpg.
I bought a VW Golf 1.6 diesel Blue Motion. Having driven it for over 9000 miles guess what, 54 mpg
All manufacturers lie about their products, well not exactly lie but give you figures that are very difficult to achieve.

I think the fault is in the test tbh - you can't blame manufacturers, unless it's a VW scenario.

[Mick F]

The fault isn't the test per se, it's the manufacturers testing these cars artificially to pass the tests, not real people doing the same tests in real conditions.

[Vorpal]

The mpg is calculated using very specific sets of parameters on a rolling road. Things like reductions for aerodynamic drag are done fomulaically. They are not intended to describe *your* mileage, only to make it possible to compare vehicles. Previously, car manufacturers did it howeve they liked, so not only were they wildly innaccurrate, they were wildly innaccurrate in differnt ways.

The vehicles that are subjected to testing are set-up for it. Not to lie, but get as close to ideal performance conditions with factory production parts, as possible. They test a vehicle that is probably 85th or 90th percentile.

Only someone who gets a good performing vehicle and drives in mainly ideal, or nearly ideal conditions can hope to achieve anything near the stated mileage. Someone who uses a car professionally or at least driving mostly with a warmed up car (typical journeys of several hours), and not too much in congested traffic can likely come close.

If the tests were done in real conditions, they would not be comparable.

[Bonefishblues]

The mpg is calculated using very specific sets of parameters on a rolling road. Things like reductions for aerodynamic drag are done fomulaically. They are not intended to describe *your* mileage, only to make it possible to compare vehicles. Previously, car manufacturers did it howeve they liked, so not only were they wildly innaccurrate, they were wildly innaccurrate in differnt ways.

The vehicles that are subjected to testing are set-up for it. Not to lie, but get as close to ideal performance conditions with factory production parts, as possible. They test a vehicle that is probably 85th or 90th percentile.

Only someone who get a good perfroming vehicle and drives in mainly ideal, or nearly ideal conditions can hope to achieve anything near the stated mileage. Someone who uses a car professionally or at least driving mostly with a warmed up car (typical journeys of several hours), and not too much in congested traffic can likely come close.

If the tests were done in real conditions, they would not be comparable.

Indeed. What is required is something less formulaic which gives the consumer a better indication of real-world consumption. Until then, they will be a comparator only, with some odd anomalies like the twin air, which is notably bad in terms of test vs reality.

[Flinders]

Yes, a big leap in consumption.

I kept records for the 500 - I did it because I thought we'd get good figures, then kept on with it to show that they were terrible considering the zero VED.

Here's a screenshot of my spreadsheet latterly.Screen Shot 2017-09-06 at 15.49.17.png
Columns:
Date, Mileage, Fuel cost, Filled up yes or no, Litres, Imperial Gallons, MPG figure since last top-up.

I think we get 45mpg (about) with a fairly ancient focus estate -petrol. Genuinely mixed driving. Some short commuting, some long journeys on M ways, some medium journeys on A roads.

[Vorpal]

Indeed. What is required is something less formulaic which gives the consumer a better indication of real-world consumption. Until then, they will be a comparator only, with some odd anomalies like the twin air, which is notably bad in terms of test vs reality.

There are two problems with estimating real world consumption. One is the usage of the vehicles varies considerably from one range to another, and it is not a statistically normal distribution. It can be modelled by a log-normal distribution, but only once you have lots of data about the usage. So, if they download all the data from a few thousand Model ABC cars and analyze journey length, mileage, etc. They have a good idea what the usage distribution will be for the next generation of Model ABC, and can probably do a good job of estimating 'real world' mileage. But what group of users do they publish it for? Average? The average is not the mean, and this could easily be as misleading as the current information. 50th percentile? What if few users fall onto the 50th percentile?

And how do we define that? By annual mileage? Number of trips? Some combination?

Then, do you do a separate prediction for model ABCS, the estate version, which is heavily used by company car fleets for their travelling technicians, sales people, and other support staff?

[old_windbag]

only to make it possible to compare vehicles.

This is exactly so. As with so many other products by creating a test with specific criteria that all manufacturers use then it allows comparison by consumers( i.e energy ratings of appliances ). They cannot account for big dave with his leaden foot driving or miss daisy's featherlight throttle and hypermiling. They are for comparitive purposes, honest johns site which I used when looking for current car gives a general idea of real world returns by many and under wide ranging profiles. It's typically around 80% of the best manufacturer test figure as a mpg value. Work on that and it gives a better expectation of what you'll get.

I have just tank to tank filled and calculated my 1.0 ecoboost. It returned 53.3mpg on around 75% a-roads/25% urban( with jams ). I'm more than happy with that. The concern I have with hybrids( not pollution reduction in urban areas ) is that say mickf gets 4mpg average more than myself and around 10-15gm/km CO2 lower. Then factor in the extra energy costs of all the inverter( all components semi's, caps etc )/motor/battery then will it be energy wise beneficial. Perhaps if run for a few hundred thousand miles perhaps. The total energy equation of a 1.0ltre petrol versus say yaris hybrid would be interesting to see if the end justifies the means. It gets down to oil extraction versus lithium mining etc, a complex comparison for the whole supply chain.

[Bonefishblues]

I understand it's rather difficult to get a better test, but the current one causes such angst and difficulty, not least for manufacturers themselves that it surely can't be beyond us to come up with something which is better, given smart people being engaged to do so?

[squeaker]

only to make it possible to compare vehicles.

This is exactly so.

Except that the 'real world' delta to the official consumption has been getting worse, due to the technologies used to 'improve' fuel consumption ie engine downsizing combined with turbocharging (IMO).

[old_windbag]

The current test method may not be up to snuff but when out on the bike I see a lot of aggressive driving. Not in a person to person way( we know that happens ) but the way many drivers seem to attack the throttle from junctions and off roundabouts. There is a lot of impatience. When in the car yesterday I saw the usual accelerate towards hazard, then hit brakes technique. Perhaps they were hoping for some regen energy, the yaris method .

Perhaps people simply don't understand the aspects of driving that cause the biggest reductions in mpg...... or simply they don't care as they have more than enough money to pay for more fuel.

[Vorpal]

Except that the 'real world' delta to the official consumption has been getting worse, due to the technologies used to 'improve' fuel consumption ie engine downsizing combined with turbocharging (IMO).

I think it is likely a combination of things. More cars having turbochargers (which aren't used during mpg testing), is definitely a contributor. I also expect that the emission reduction equipment contributes. More parts in the system means more stuff to vary and less reliability. It also takes longer to warm up and achieve peak efficiency.

[Mick F]

I understand it's rather difficult to get a better test, but the current one causes such angst and difficulty, not least for manufacturers themselves that it surely can't be beyond us to come up with something which is better, given smart people being engaged to do so?

Exactly.

[Vorpal]

I understand it's rather difficult to get a better test, but the current one causes such angst and difficulty, not least for manufacturers themselves that it surely can't be beyond us to come up with something which is better, given smart people being engaged to do so?

I'm sure that it is, but these things take *years* in the automotive industry. It's all done by committee, and it's got to work in many places in the world, as well in Europe, and not put undue burden on manufacturers. they also have to consider where test equipment is going to come from, if that industry has the capacity to supply everything for implementation, etc.