/ Motoring

How many miles can your car really do on a full tank?

Car fuel dial

Like many motorists you probably have a good idea of how many miles your car can do on a full tank of petrol. But do you know if it is anywhere near what the manufacturer claims?

When I bought my car, I wanted a reasonably priced vehicle, that was small and easy to park, and not too expensive to run. I chose a Peugeot 107, with an advertised 508 miles on a full tank of petrol. But after a few months I noticed I never managed to get near that – the closest I ever got was 330 miles.

Obviously I didn’t drive until the tank was completely empty so it could have done a few more miles, but that’s still quite a difference between the real and claimed figures.

New Which? research put 200 cars through the same test and found that 98% couldn’t match or beat their miles per gallon (mpg) claim. The car that performed worst compared to its official mpg figure was the Mitsubishi Outlander PHEV (a plug-in hybrid), which overstated its mpg by 120%, costing £459 a year in unexpected fuel costs.

Other studies have shown that the gap between official test results and how we drive in the real world has increased over the past years, so the problem could be getting worse. But why is this happening?

Why fuel claims aren’t accurate

The problem lies with the Europe-wide test used to measure fuel use, the New European Driving Cycle (NEDC).

The test is called ‘new’ but it was introduced in 1970 and last updated in 2007. It’s outdated and doesn’t properly measure the kind of driving we do today, for example on motorways.

It also contains a number of loopholes. Manufacturers can, for example, opt to switch off air con and lights when testing, or increase tyre pressure above recommended levels. And as there is no standard way of testing, this means that you can’t make a proper comparison between different cars.

Many of you have already told us you knew the claims were misleading. Malcom R told us in 2013: ‘EU fuel consumption figures do not represent real-life, and are not intended to.’ Vynor Hill said: ‘Manufacturers’ claims are of no use to anyone.’

It seems unfair that a product that isn’t meeting its claimed performance doesn’t suffer any consequences and consumers just have to bear the brunt of it, or do their own research.

Steering in the right direction

A new and improved fuel efficiency test that would address many of the current one’s weaknesses is due to be introduced in 2017 by the European Commission (EC). This actually closely mirrors our own test and more accurately estimates fuel use.

But some factions are trying to stall the introduction of the new test to delay it to 2020. That would be another five years of useless fuel claims.

We don’t think this is right, and are urging the EC to stick to the planned timing, so that you won’t be taken for a ride any longer.

Do you think official miles per gallon figures reflect reality? Do you agree the EC should bring in the new test by 2017?

Lee Williams says:
23 April 2015

Vw Golf said it can do 70mpg I get 48 at best Honda said it can drive from Edinburgh to Brussels on a full tank. I can not get to Cornwall on a full tank.
It’s all done in test conditions with no wind reactions, pay load, hill graduate, start stop, emergency stops, bad Servicing of car. Travel boxes. Bike racks. Even tryes make a difference to MPG.
NO DRIVER EVER GETS RETESTED. unless your a Hgv driver or take additional lessons.your driving habpits make the biggest difference of all. So No test can say it true or ackreate.

I would just be happy if the tests were done on cars from the factory without all the modifications they seem to be allowed to make to improve the figures.

The problem I have with this introduction is it appears to characterise the standardised fuel consumption tests as a “claimed performance”. The EU test results do not pretend to give claimed performance; it is a way of producing comparativer data for vehicles tested on a standard basis, in approved facilities. To suggest otherwise is misleading the consumer.

Like the intro to the May edition of Which? magazine this approach seems designed to capture cheap media attention. Let’s leave that to the politicians. Which should adopt an independent objective appraoch otherwise its standing as a Consumer Organisation will be compromised.

I am quoted in the intro as saying the figures do not represent real life and are not intended to. Exactly the point. I do not infer they are misleading.

They also produce emission data under standardised conditions, so vehicles can be ranked – but again they state they will not necessarily be achieved in practice. Comparative data again, but used in the UK for such things as VED and company benefit in kind calculation.

The problem with standards is they lag behind progress and innovation. The many parties involved take time to assemble facts, agree procedures, validate proposed methods and so on; anyone involved in the preparation of standards will know the problems.

I hope that the next standard is more representative of current vehicles and driving conditions – acceleration capability has increased for example, and is introduced when it is properly prepared, but no standard test will represent what individuals will achieve on the open road – the wide range of road topography, journey types, driving style, tyre inflation and so on will lead to a range of fuel consumptions within which most people will experience. And, inevitably, the standard will be slightly out of date when it is published.

Use the responsible motoring press (try Honest John) for indications of everyday consumptions.

DfT give further information at http://www.dft.gov.uk/vca/fcb/the-fuel-consumption-testing-scheme.asp

Malcolm wrote: “The problem with standards is they lag behind progress and innovation. The many parties involved take time to assemble facts, agree procedures, validate proposed methods and so on; anyone involved in the preparation of standards will know the problems.”

Unfortunately the progress and innovation has included irresponsible behaviour by car manufacturers in devising ways to cheat. Replacing the tyres and oil with alternatives that will produce better results, over-inflating the tyres, taping up the doors, removing parts to reduce weight, disconnecting the alternator and air conditioning, modifying the brakes, etc, etc. Perhaps we need a dishonesty test for car makers.

You call for Which? to have an independent objective approach. I cannot fault that, but the top priority here is for proper independent testing, following the example of Which?

Car manufacturers have achieved an amazing improvement in fuel economy over the years. My present car is currently averaging 58.5 mpg according to the gauge. If the truth is impressive, why do manufacturers need to give us figures that motorists are unlikely to be able to achieve?

wavechange, the “progress and innovation” I was referring to was the improvements made, and continue to ber made, in cars – in safety, durability, economy and so on. This made no reference to the separate issue of what happens in performing the test cycle. Please, if you do want to refer to one of my comments, keep it in context. 🙂

Jon Harris says:
25 July 2015

Without wishing to hijack this thread i would challenge that modern vehicles are more durable than years ago. Indeed the very measures being introduced to improve economy are inevitably making engines less robust in many cases.

I accept what you are saying about the figures being used as comparative data, but I can’t see much use for that if they can not be applied to the real world. Car adverts certainly don’t go out of their way to disabuse the public. There might be a small asterisked footnote somewhere, but the figures they use are usually the lab ones.
The lab figures can be used for ranking purposes, but has anyone done any testing to see whether this ranking is actually correct? The laboratory might suggest that a BMW is more economical than a similar V.W. as tested. Does this hold good when both are actually driven?
My last car averaged 43 MPG. I know this because over the course of a few thousand miles I brimmed the tank every time I filled up. I couldn’t claim that a similar car would do the same if the driver had a heavy right foot, or a lighter one than mine. If a few of us got together then we might reach a ball-park figure by averaging our consumptions. Now that would be really useful and if that’s what Honest John has done it is up to the car manufacturers to tell the truth and give the driver a real target to aim for.

Vynor, I understand your point. However only standardising testing can give values that can be used for comparative purposes – but we need improved tests. Remember the EU tests also give emissions, used for other oruposes, like taxing and congestion charging – again they are relative values but how else do you get them?

The “real life” results can only give an indication of a range of what is achieved by a number of people, but how are they derived? Short journeys, long journeys, stop start, in Norfolk or the Highlands, driven carefully or to their full potential? The only way to get meaningful results is to either take what is achieved by all users and give the range (impossible), or use a standard set of driving conditions and enough samples to get statistically significant results. But if manufacturers were to claim these, and you as an individual didn’t acheive them, then they may no doubt be liable for presenting misleading information.

The answer is to have a better set of standard conditions under which the EU tests are carried out, to get comparative results nearer (but only nearer) reality, and collate real life data from motorists to improve the reliability of so-called real life consumptions. Two separate things for separate uses.

Phil Ingle says:
23 April 2015

These test results confirm what we have thought for some time. I feel there is another area for improvement. Many new cars have trip computers which show current and average mpg. The current mpg is often irrelevant as it changes so frequently. But the average mpg on my Subaru Legacy computer over reads by 10-15%. Surely the technology is available to accurately measure average mpg, without having to do a brim to brim check? Or are the manufacturers relying on most people believing the computer, and with present lower fuel prices, being content with feeling their car is doing 51.9 mpg (as my car computer claims)?

Phil, I’ve for many years (obsessive?) always filled to the brim and recorded the miles and litres – a habit with a company car! My Renault Espace is surprisingly accurate when comparing the indicated mpg with the calculated – within 1 or 2%. I don’t know how well fuel used is measured but presume indicated miles travelled will be affected by tyre pressure and degree of wear – the wheel gradual reduces in diameter – could be 2 or 3% overestimating distance from new to just legal.

I use my mpg display to monitor my driving – I think I drive more smoothly now to achieve economy as you realise how much fuel is wasted accelerating and slowing whilst you actually don’t gain any real advantage on our congested roads.

Quite right!
Regarding emission testing, I would have thought that the lab would be accurate. It simply has to run the engine at various revs and loading and then measure what comes out of the back!
I would also have thought that the various car pools used by company drivers could provide good sample data for real life driving. Same car and several drivers, while the manager keeps a check on the fuel and mileage. “What Car” does this, so it is possible to get a reasonable idea of how a car will perform without relying on manufacturers’ hype.

Tortoise says:
23 April 2015

So, does the current (and future) test really not make any allowance for the aerodynamic properties of the car? From the fact that they are rolling road tests, it seems that must be the case. Surely that is one sure way of making sure that the test figures will differ from real world performance.

Tortoise, as far as I know with the test being carried out on a rolling road a roller resistance is included to represent the effect of drag.

In her introduction, Emmanuelle has written: “And as there is no standard way of testing, this means that you can’t make a proper comparison between different cars.” I knew that official fuel economy figures were of little use as an indicator of what a driver might achieve, but that since the tests are carried out under standard conditions, at least they would be useful for comparing different vehicles.

I am really disappointed that manufacturers are allowed to manipulate tests by over-inflating tyres, etc. but at least some of these games are well publicised.

I wish Which? every success in forcing the introduction of a new test regime by 2017. The less say that manufacturers are allowed to make an input the greater the chance we might see more honest figures.

“Official fuel consumption test procedures have been in use since the 1970s. EU Directive 80/1268/EEC as amended or, for Euro 5 vehicles onwards, Regulation 692/2008 describe the tests which all new cars on sale after 1 January 2001 are required to take.” Perhaps Emmanuelle would explain what she means.

Many standards are written by groups of people with expertise, and does and should include manufacturers. To continually demonise manufacturers is neither helpful nor accurate. No more than blaming the poor standard of education in this country just on the teachers and lecturers.

What is the point of having official fuel consumption figures if they are generally substantially above what most motorists could achieve? Many drivers understand what factors affect fuel consumption, so why not design tests that are fit for the purpose.

Optimistic fuel consumption figures have been quoted since before I started driving and it’s about time that the consumer was rewarded with more honesty.

I think these questions have already been covered. Honesty is not the issue – we should stop characterising it in this way. The reasons for official tests, their drawbacks, their future is being discussed together with the difficulty in giving other than general data on what an individual might achieve in practice.

So it’s not dishonest to blow up the tyres to achieve higher fuel economy figures? 🙂 🙂

Honesty in this case means testing according to the standard. If that is not done correctly then it is dishonest, but if it is within the provisions of the standard then it may be the standard that needs changing – which is apparently underway. The mistake still being made is to regard the EU tests as being pressented as representing real life. They do not and statements are made to that effect. Use the real life data that is published – but don’t rely on that applying to you either for the reasons said before.

Jan H says:
23 April 2015

Have methods of measuring mpg changed in the last 10 years?

I just looked up my 2004 diesel Zafira on Parkers (I don’t know where else to look?) – they quote 44mpg, & I generally get 42-45 depending on the journeys I’ve done. 44 seems to be a reasonable estimate.

From reading Which? magazine, it appears that it is the car manufacturers that are pushing for a delay in introducing new tests that should provide more realistic figures for fuel economy.

Product testing for Which? is carried out completely independently of manufacturers, for very good reasons. I don’t believe that the manufacturers should have any involvement in how tests are designed and implemented.

Jon Harris says:
25 July 2015

Which? may or may not be independent [I’ve no knowledge either way] but some of their reviews and conclusions do make me wonder why they come to the conclusions they do. Their apparent love affair with the Honda Jazz for example whilst NEVER seemingly mentioning the very real issues Honda had for years with gearboxes escapes me if they are really there to lobby for a better deal for their customers/subscribers. My wives Jaxx g’box went after less than 35k miles and this is not uncommon if you have a quick google.

For those interested in the detail of the proposed new test conditions this paper gives an analysis: http://www.theicct.org/sites/default/files/publications/ICCT_WLTP_EffectEU_20141029.pdf

CO2 emissions form a key part of this because of legislation. What the report explains is the background to the NEDC, its deficiencies and the proposals industry, government and other experts are making to minimise deficiences and make the testing more representative of current car technologies and driving conditions.

We must still distinguish between the need to do type testing under strictly controlled and standardised conditions for a particular purpose, and performance in the real world; the new tests aim to narrow the gap but we must be careful to only use the data for its intended purpose and not misuse or misrepresent it. Perhaps it would be better if only emission values were quoted in the motoring literature, and fuel consumption was left to the individual to discover. How helpful might that be?

From the above document: “NEDC imprecision allows for modifications of a vehicle that is used for official coast-down experiments, for example, replacement of normal road tyres by conditioned low resistance tyrses, atypically high tyre pressures, manual adjustment of brakes, etc. These artificial measures do not appear during real-world driving of normal production vehicles, were never intended by the NEDC regulator and should therefore not be included in official WLTP-NEDC CO2 conversion factors.”

This looks as if we are heading in the right direction. Perhaps we could include a driver or equivalent weight in the new test schedules. Leave the spare wheel in too. In fact, go as far as possible to carry tests under realistic conditions.

The test is “artificial” in the sense of standardised conditions and, if you look into the report, it deals with more realistic payloads and otherconditions. But understand the test’s purpose.

In the cynical world of manufacturer-bashing, it might be worth remembering that these same villains have developed cars that are safer, more durable and more economical for us to buy – the downside being the motive for doing so has been to maintain their share of the market and be profitable :-).

It is these self-same companies that provide employment outside the public sector and help fund your and my pension via investments. Those who have worked in industry might suggest the real-world situation is different from the perception of some others.

Malcolm – I just want to see useful figures for fuel economy. It might be the fault of the current test allowing manufacturers to manipulate test vehicles to produce better fuel consumption figures but that’s dishonesty in my book. Look at how frequently the term is used by others. I have not noticed that the manufacturers working together to push the EC to bring in the new tests.

Apart from criticising manufacturers for removing spare wheels, I have generally been complimentary about the car manufacturers, particularly in respect of improvements in warranties.

My interest is to promote fairness and honesty.

Tortoise says:
24 April 2015

Thanks Malcolm. So, do you know if the rolling road resistance is adjusted to correspond to the measured drag of each vehicle tested? That would make more sense, but would still give only a rough approximation of the actual drag at various speeds. At motorway speeds, where Which? says a large percentage of real world fuel consumption takes place, aerodynamic performance makes a significant difference, which is why roof bars and boxes, or having your windows open, can significantly reduce fuel efficiency.

Tortoise, that is what I understand. I believe that the resistance in the test is dependent upon the speed.

” I just want to see useful figures for fuel economy.” Then use the specialist press figures, or Which? if they are representative. The current test is deficient, as those who operate it agree, but testing in accordance with an EU standard is not dishonest. As I have said before, it is not a test designed to produce real life figures, nor does it claim to be. So we should not be branding something as dishonest when it does not purport to do what some accuse it of.

I see many of these conversations brand companies as dishonest, rip off, shocking, deceitful, and indeed some practices are and should be stopped. But let’s reserve such serious accusations for those who merit it, otherwise their shock value simply gets devalued.

I thought we’d have exhausted this line of attack by now? 🙂

@malcolm the figures taken for what they are supposed to represent by may valid, but I would suggest the shameless use of those same figures in the numerous TV ads by car manufacturers with little or no nod to what they actually stand for IS dishonest, rip off, shocking, deceitful.

Thank you William.

Michael H says:
24 April 2015

Having the Mitsubishi PHEV on the top of that list (and other plug-in hybrids on the list) is potentially a little unfair.

Modern ‘plug-in hybrid cars’ are designed for a certain type of motorist – the short commuter, and are designed to be plugged in overnight. If you have the wrong tool for the job (such as a PHEV for long motorway journeys) you are never going to get this sort of MPG. Unless the test accounts for this “right tool for the job” approach any mpg figures for such cars are always going to be under or over stated – depending on the test.

A “a lot of people do a lot of motorway miles” approach to the test is biased against such vehicles. That is not what they are for.

I drive a Mitsubishi PHEV and can get 30 miles range on the battiers if I drive in a sensible “battery focused” way (i.e. relatively mild acceleration and braking within the regeneration tolerance). My commute is less than 20 miles return, and I regularly plug it in – so I get amazing mileage from a tank of fuel. On a 100 mile motorway run however this drops to <50mpg. Averaged out, though, I reckon I get 100+ mpg.

In their defence Mitsubishi have (in my expereince at least) been telling their customers that if you plan to do a lot of motorway miles then the diesel is a better option – which is fair play in my book.

I agree with Michael H. The consumption figure given for plug-in cars gives some indication for comparison, but in practice the actual consumption can vary wildly either higher or lower and depends heavily on the individual driver.

I drive a Chevrolet Volt which has an official consumption of 235.4mpg (1.2l/100km). In practice during the first two years I owned the car it recorded a lifetime average about 220mpg. This was because my daily commute was entirely within the electric range so I only used petrol on longer trips (and a few trips to Scotland and to Cornwall pulled the overall figure down). Without the long trips the calculated consumption would have been close to 0l/100km but of course none of those figures account anything at all for the electricity from the wall socket.

Last year I changed jobs to a location outside the battery range. Depending on the weather my daily commute can range from about 60mpg to 200mpg. Am I upset that suddenly I’m getting half the ‘official’ figures or even less? No, those figures make assumptions about my pattern of use that are no longer remotely accurate.

I don’t think there’s an easy solution to the problem. If you’re trying to choose between two plug-in hybrids then your normal distance between charges makes a massive difference to the relative efficiency; there is no single figure that will work for everyone.

As you both say, the figures are rather meaningless because there will be a much greater variation in fuel consumption according to use than seen with petrol and diesel vehicles.

I am not sure why Which? chose to include hybrid vehicles in its report.

Chris Sims says:
24 April 2015

If you are interested in finding out realistic mpg. visit ‘spritmonitor.de’. This German web site tracks feedback from hundreds of thousands of motorists and is the largest data bank of real life mpg. that I have found. It offers English translation and can be configured in mpg. instead of km/l. after registering.

Chris, an interesting site. So far I find it gives consumptions as reported by individuals listed – haven’t found accumulated data. For my car with around 11 entries the spread of consumption about the mean is +/- 11% – 25% overall ignoring the two extremes. It makes the point that consumption depends very much on the individual and to attempt to gave a simple answer is whistling in the wind. An expected range is the best, I suggest, that can be offered.

It would be constructive of Which? to point to sites such as this if they want motorists to get helpful information.

Incidentally the mean was about what I average in my car on a mix of short (mostly), medium and occasional longer journeys. My range from a long journey to all short is about 33% difference in consumption.

Damian McDevitt says:
24 April 2015

Your conclusion that it would take three years to recoup the price difference between the Outlander PHEV and its diesel brother is completely misleading. The PHEV has a much higher base specification than the diesel version you selected. If you compare the PHEV with the GX3 Auto diesel with identical equipment specs you will find that the prices are the same to the penny.

This detail would normally be of very little interest, but in this case it is innovatory and important. Mitsubishi have introduced an explicit policy of taking the price factor out of the choice between plug-in hybrid and diesel versions of the same vehicle. Widely adopted by manufacturers this policy could make for a big step forward in the acceptance of plug-in hybrid vehicles. It would be interesting to see Which? comment on the policy as well as the vehicles.

NB – I do not own and never have owned a Mitsubishi, and I have no connection with Company or its distributors.

Tim in Suffolk says:
25 April 2015

My vote would be for a policy that sees a manufacturer have to release average mpg figures for urban and extra urban. If after a year of each models availability the complaints show a lower figure on average them the manufacture should be forced to compensate every buyer yearly whilst owning the vehicle, an amount based on a (provable) mpg figure over average mileage and fuel price for that year.

As you may know I am a great fan of seeing what BEUC and the other European consumer groups do :
This 47 page report makes excellent reading

Much more ballsy and most recent is this quote from the BEUC site:
“On 24 February 2015 Altroconsumo launched their class action in court in Italy to help owners of these two cars get their money back. Check Altroconsumo’s press release and video (illustrating a car confession about the real fuel consumption and call for consumers to find out if they are entitled to a refund). BEUC issued a press statement backing this class action and a position paper on EU testing overhaul.
Italian consumers who want to join the class action to reclaim the money can register here and find out how much they overspent.”

Thanks for that, Dieseltaylor. From the Transport & Environment report you have cited:

“By producing artificially low fuel economy figures in tests and using technologies that work far more effectively in tests than in the real world carmakers are cheating their customers of anticipated improvements in fuel economy. Customers become disillusioned, no longer trust official figures and are more reluctant to buy more fuel efficient cars. This in turn makes it more difficult to sell these cars that manufacturers need to do to achieve their targets. Carmakers are creating a vicious circle in which the only way they can compete and achieve their targets is to cheat.”

One of the recommendations of the report is: “By 2020, a new system for type approval of vehicles should be introduced to ensure certifying and testing bodies are entirely independent of carmakers.”

This backs up my allegation of dishonesty and the need to carry out proper independent testing.