/ Motoring

‘I went to court when my car didn’t meet its mileage claims’

Filling petrol tank

Our post on Convo last month about differences between claimed and actual car mileage prompted you to tell us your own stories. It’s a problem many of you have clearly faced, but how far would you go for a solution?

For Doug Clement, from Donegal, the answer was all the way to court.

Doug was so annoyed when his car failed to achieve the fuel economy that he’d been promised that he took legal action against the dealer – and won.

Our recent research found that 98% of the 200 cars we put through the same test couldn’t match or beat their miles per gallon (mpg) claim.

New Hyundai failed to meet MPG claim

Doug came across the problem when he bought a new Hyundai i10 in late 2010, after a salesman in the showroom assured him that he would be able to get the 57.6mpg, which it had been claimed for it on what is called a ‘combined cycle’ – this includes driving in towns and cites, and the countryside.

But Doug found there was no way he could squeeze that much out of the tank and so he told the dealer he was rejecting the car. The dealer refused him a refund, which was when Doug decided he’d have to go to court.

It took two years, but the court did eventually find in Doug’s favour – and interestingly the reason given was that he had specifically asked about fuel efficiency when he bought the car.

Because he couldn’t match the official mpg figures quoted by the salesman, the car was deemed not fit for purpose under the Sale of Goods Act and the dealer was ordered to pay a full refund.

The case is not considered to have set a precedent, but the decision does suggest that courts will take similar cases seriously – so make sure you ask in the showroom about fuel economy. Even better, try to get the dealer to commit to it in writing.

Fuel efficiency – drivers tell Which? Car Survey it’s a major issue

Being unable to match fuel efficiency is one of the big issues that drivers tell us about each year in the Which? Car Survey.

For example, the sixth-worst car for missing fuel economy claims in our most recent research was the Toyota Yaris Hybrid. Of the 63 owners who gave feedback on their car last year, 12 complained of poor fuel economy.

Have you ever asked a dealer about fuel efficiency? Would you ask them to write this figure down?


Good result.

When fuel efficiency is a critical factor in the choice of a car, put the seller on the spot, effectively making it ‘the essence of the contract’. Other features that are essential to your own requirements can be treated similarly. The same applies to any purchase where as a buyer you are relying on a particular perfomance characteristic that (a) you have raised with the seller and (b) the seller has assured you will be met. If you can’t get it in writing, write to confirm your understanding of the position.

Using a credit card to pay a minimum of £100 of the total purchase price also gives consumer protection in the event of a dispute.

It’s high time retailers woke up to the implications of the Sale of Goods Act and its successor the Consumer Rights Act coming into force in October this year. The more consumers exercise these rights all the way to the Court the better it will be. I wonder if, in this case, the car dealer had words with the manufacturer, and what will happen next.

The only consumption figures manufacturers are allowed to publish are, I understand, those derived from the EU NEDC tests. These come with a warning that they do not represent real life driving. They are for comparative use only. Don’t rely on these for a claim; you will need a statement, presumably in writing, from the dealer as to what consumption you will get. In view of the variation between drivers, journey types, terrain etc I doubt anyone with any common sense would give such a statement.

I have noticed on one website that the manufacturer gives a single fuel consumption figure without, as far as I can see, any qualification. I have asked them to explain this and wait to hear. It would be interesting whether this could give rise to a claim.

Congratulations to Doug Clement for doing what many of us could and should have done in the interests of justice.

My understanding is that car manufacturers prepare their cars for the fuel economy/emissions tests in various ways that will return better figures than a motorist might achieve. Examples of preparation include taping-up doors, using special tyres, over-inflating the tyres, removing parts to decrease the weight of the vehicle, using special oil, pushing pistons back into brake callipers, disconnecting the alternator, and so on. Though there may be no cheating in the tests themselves, manufacturers that play these games should hang their heads in shame.

There may be valid reasons why the outdated tests are not updated promptly but there is nothing to stop tests being carried out on unmodified vehicles.

It has been pointed out that the fuel economy figures published by manufacturers are not intended to reflect what the motorist can expect to achieve, but these figures are still used in marketing. When I replaced my car in 2012 I was looking at two models – one with a petrol and one with a diesel engine. I casually asked what mpg ‘I could expect to achieve’ and the experienced salesman reeled off the test figures, as I had expected. He received a pre-prepared lecture about being honest to customers.

John is spot on about making your requirements clear. That applies when making any purchase, but is particularly important when buying something as expensive as a car.

“these figures are still used in marketing”. As I understand it these figures, but only these figures, must be published by the manufacturer according to EU regulation. Anyone know differently?

The salesman lied to me, Malcolm. I gave him the opportunity to let him say that the figures he had quoted were just the test figures and not what I had asked for.

Why do you think it is acceptable for a manufacturer to specially prepare a car so that it returns higher fuel economy figures than most motorists could achieve?

As I said above if you are given false information about fuel economy by the retailer they should be held liable. The fact that the EU test cycle allows manufacturers to prepare vehicles is the fault of those who published the standard, not the manufacturers. You need to remember that these tests are designed to produce comparative figures, not real life, and they will remain comparative if they are carried out on the same basis. Out of interest what preparations in particular have a significant effect on the results that will give any one manufacturer an advantage?

Criticism might be levelled at the EU don’t you think? Unless we find out facts to the contrary.

Malcolm – As far as I am aware, it is standard practice to carry out tests on actual production samples when independent test laboratories are involved, so I don’t see why car testing should be any different. As John has pointed out below, you cannot have comparative figures if different manufacturers modify their cars in different ways. I read that one manufacturer even reprograms the engine management computer before tests are carried out. I don’t know which particular preparations will push up the official mpg to the greatest extent.

I have not paid much attention to criticism of the current EC testing, but even if there are fair criticisms these in no way justify the manufacturers preparing their vehicles so that they perform better during tests. I believe that most people would regard it as dishonest to prepare a vehicle to enable it to produce better fuel economy figures.

What if I – as a customer – was to prepare a vehicle while the salesperson was out of the showroom. I could switch the price label, put in a spare wheel and a set of car mats, and then innocently say that I would take that car. 🙂

The car fuel economy article in the May 2015 edition of Which Magazine is very apt in my case. I purchased a new Audi A4 2.0 tdi Allroad back in November ’14 – I have had the vehicle 6 months.

Initially I gave the vehicle the benefit of the doubt re fuel consumption; I had winter tyres fitted and fuel consumption is usually slightly more in Winter. Having recently changed to Summer wheels/tyres the fuel consumption did not alter. I contacted the Audi dealership where I bought the vehicle and they promptly offered to have a look at the car, software, etc. (Aberdeen Audi have been very helpful and I have no issue with them).

My average fuel consumption over the 6 month period, on a Combined Cycle, is 38.44 mpg. This equates to the Audi published Combined Cycle figure of 47.1 mpg – in other words Audis ‘published’ figure is 18.39% overstated. If the fuel consumption figures were the odd % out I could live with it. It is just so far out that the difference jumps at you; and I find it hard to comprehend that an Audi 2.0 tdi engine cannot return over 40 mpg (I have yet to see this – even on long haul motorway driving and keeping within the speed limit. Prior to the A4 Allroad I ran a 2.0 diesel Subaru Legacy Boxer and returned mid 40’s MPG on a long haul run, i.e., my driving pattern has not changed.)

My annual mileage is approx. 15K therefore I will need to purchase 323 ltrs more diesel than anticipated. Cost wise at £1.20 per ltr (estimated average price going forward), this works out to £387.60 per annum. I have eMailed my Audi dealership and copied Audi Customer Services to ascertain how they are going to compensate me. I have yet to hear back from Audi directly and the dealership have asked me to keep them informed as to how I get on with Audi!

It was interesting to read that Doug Clement, from Donegal, received a court ruling in his favour for his Hyundai i10 fuel consumption being overstated, with the Court seemingly ruling in his favour because he asked about fuel consumption at the point of or prior to purchase. Surely it has to be irrelevant whether you ask about fuel consumption, performance, etc, prior to making a purchase. Either published material is correct or it is not. If a supplier (of anything) is knowingly publishing inaccurate data they are technically in breach of contract which, in the case of us ‘mug motorists’, entitles us to return something which is not fit for purpose as bought or request compensation to offset a definite loss.

I will see what Audi come back with, but if there is not a willingness to make redress I shall be contacting Trading Standards.

I can’t understand why people blame the vehicle manufacturer for (a) complying with the EU regulations and (b) aiming for the best result under those rules. Doug did have the right idea when he asked the salesman to confirm he would be able to achieve the published figure. Don’t expect one manufacturer to quote “realistic” figures – after all we all use our vehicles in different circumstances. And who would go to the manufacturer who says my equivalent model does 15% less miles to the gallon than anybody else quotes? Be sensible, just use the published figures for comparison and ask the seller to confirm what economy you are likely to achieve in practice. I think we would all be amazed how quickly the vehicle retailers in Europe would force a change in the EU rules!

If you modify a vehicle so that it performs better in a set of tests you are using ‘unfair means’ to obtain an advantage over other companies. I have borrowed that term from higher education, where it is used as a euphemism for cheating.

Would you trust Which? testing if manufacturers supplied products? I would not because of the risk that they might be carefully checked or even modified, so that they were better than what you or I could buy in the shops.

What is wrong with honesty? It could be a powerful marketing tool.

If a manufacturer or retailerquotes you a fuel consumption without any qualification then you would have a case. If you ignore the caveat applied to the figures the EU requires the manufacturer to publish then I see no case. The deficiencies in the outdated NEDC test cycle, the laxity in the standard, dictated by the EU have been well-aired on another conversation. It is no good ignoring these issues as if, despite all its shortcomings, the existing EU test should somehow still deliver real life figures – something it was never claimed to do. It simply depends upon how the fuel consumption of your car is presented to you by the manufacturer or retailer.

Best place is to look at Honest John or other collections of figures that motorists have provided for their “real life” performance – like spritmonitor.de. The latter shows the large variation that occurs in practice.

Testing of products that feature in Which? reports is, I believe, done on products purchased in the same way that you or I would buy them. I can see various benefits of this and I trust Which? testing, even though I would prefer to have more information in many cases.

Why should car manufacturers be allowed to prepare their cars before testing? Why not just test cars exactly as they are sold to the public?

In practice it would be very difficult for a manufacturer to guarantee that any given customer will get the same fuel consumption as any given test result because driving conditions can vary so much.

However, it ought to be possible for a retailer’s service team to check that a given car is performing within expectations for the make and model of car.

In my day job, I sometimes get to do long motorway trips on using newish (hired or leased) examples of medium sized and large cars. If diesel powered, I can usually get about 60-65 mpg on such trips. (This does involve driving at sensible speeds within the prevailing speed limits.)

However I doubt I would get anything like as good fuel economy if I was only using these cars for a series of short “social domestic and pleasure” trips across town or even on the motorway, if I were running late and had a plane to catch.

When shopping around for cars that I might actually buy, I do hope that published mpg figures will provide a fair basis of comparison between models but I do not necessarily expect to be able to exactly match the official test figures in my own use.

Because we don’t know [presumably – trade secrets] what specific action the different manufacturers take to prepare each model for testing it is impossible even to rank the various vehicles in any order of fuel efficiency because of the inherent inconsistency. Some might do all the things that Wavechange has listed above, others might do only some of them. I can’t really see how anybody can rely on the test reults for anything, least of all for making a purchase decision. As Malcolm says, it’s best to check against other drivers’ experience of the same model and do a sort of mental averaging of the figures to get a more realistic guide to what you might expect in practice. Even then, no two sets of driving conditions and behaviours are the same in real life so you have to exercise a degree of tolerance up or down. All the more reason to bring forward a reliable and consistent independent test standard to which all manufacturers must submit unmodified vehicles. Most car buyers would probably be happy with a trustworthy ranking table of fuel effciency for the models in each bracket, whether that be price or engine capacity or some other criteria.

John, in the previous conversation I did give some % figures in response to particular items – of the order 1 or 2 %. If you put NEDC into a search engine you will find some critical analysis of the test with the sort of “loophole deviations” that might be expected. I do not see that these have a major impact on the comparative results to significantly effect ranking. It is important also to remember the tests seem largely aimed at establishing comparative emissions – principally used in the UK for VED, company car benefit-in-kind tax, congestion charging and, I believe car parking permit charges in some boroughs. It could be argued that if these are on the optimistic side they are of benefit to motorists – but I won’t! However, it needs to be emphasised these tests are not claimed to produce real-life figures.

I can see what you’re saying Malcolm but I think you are underestimating the reliance that many people place on the m.p.g. figures. In this Conversation we know that they are not real-life figures, or really any more than a rough indication, but I expect some people take them as gospel along the lines that “they must be meaningful if every car advert has to include them” and “if they’re not meaningful why publish them”. And I don’t think we should overestimate what most drivers think about emissions data unless they have a direct interest in their money-saving properties. I agree with you that the EU, not the manufacturers, are at the root of the problem; nevertheless manufacturers should not game the system even to the slightest extent. I would guess that leaving out the spare wheel and the tool kit, and firming up the tyres above the recommended pressures, would make a significant difference to the performance figures [otherwise they wouldn’t do it]. Today I’ve been de-cluttering our car and taking out the Winter kit; it’s surprising how much dead weight we’ve been hauling.

Advertised “Official Fuel Consumption Figures” are accompanied by the words “The MPG and CO2 figures quoted are sourced from official EU regulated test results. These are provided for comparability purposes and may not reflect your actual driving experience.” I presume this is officially dictated phraseology and, like the figures, the only information the manufacturer is allowed to publish. Perhaps the caveat should be made much stronger and direct. We are notoriously bad at reading the instructions until all else fails!

As regards the tests themselves I imagine the manufacturers are less concerned with fuel consumption but with emissions. These are used for taxation and other revenue – earning purposes by states and are not only artificially derived – by the driving cycle that does not represent real life – but artificial in the sense they end up as bands with different driver-charges moving from one band to another, rather than a sliding scale. It is therefore an incentive for the manufacturer to do all it can to just get into one band, rather than the higher one. This is always the difficulty with banded charging regimes – like making your income tax affairs avoid as much in the higher rate band as you can.

Modifying a car so that it will produce lower emission figures during testing is cheating.

Testing should be done independently, by people that can be trusted.

wavechange, you miss the point. The NEDC is a defined driving cycle (speed / time / stop start) that manufacturers must use to get type approval for their vehicles. One requirement is that manufacturer’s fleet must overall meet an emission target (currently I think 135 g/km). Therefore the vehicles will be developed to meet this objective. If you know that vehicles tested fall outside the requirements of the test specified by the EU then that would be cheating. Perhaps you would supply examples.

The NEDC test is somewhat lax and allows – inadvertently – some variations in testing but nothing I have seen is likley to give a subtantial advantage. The new test regime learns from these “loopholes” and as and when implemented will produce a driving cycle much more representative of EU countries driving. Although you wonder whether that will be different in 5 years!

It is important to understand the difference between “claimed mpg” and the NEDC issue. The latter was introduced long ago when car technology and driving conditions were very different. It appears that a result of this is it can give fuel consumptions that are further from reality than when originally introduced. This is a consequence of the test cycle, not cheating by manufacturers (who do exploit loopholes, which is not in the spirit of the tests, but does not seem to cause the large disparity)

Standards need to be carefully thought through and written to ensure consistent and sensible results are obtained. The new one seems much better. It will probably be 2020 before it is used as the sole test – NEDC will be used until then. However CO2 targets are being reduced, irrespective of the test cycle, to an objective 95 g/km so it does produce results.

I don’t think it’s me that’s missing points, Malcolm. You can call it ‘exploiting loopholes’ if you want, but I will continue to call it cheating.

We can agree on cars producing lower emissions. They are safer, more comfortable, quieter, more pleasant to drive and much more economical.

Within reasonable limits, fuel consumption figures depend much more on the driver and traffic conditions than the car! I really can’t see how any manufacturer or dealer can be expected to guarantee fuel consumption figures when the values achieved are so driver dependent. When you see the way that some drivers perform, seeming to be using either full throttle or heavy braking, it would be a surprise if they achieved even half the published figures. In my own case I have sometimes achieved worse figures than those published, sometimes equalled them and, on a few occasions, have even bettered them. The best that can be said about published figures is that they provide a basis of comparison if it understood that they must be treated with a degree of suspicion.

Jean says:
5 May 2015

When I told Toyota that my new Prius hybrid was not giving me the advertised mileage, they told me that was because I wasn’t driving it properly! Stupidly I believed them. So grateful for this article and all the comments indicating that I am not alone, and probably can drive ok.

You’ve probably been a naughty motorist and carried passengers!

Well done Mr. Clements.

An interesting story as I am curious as to wheher this an example of cross-border consumer resolution. Donegal, county and town, being in Eire and SoGA being UK law. Or is it an Irish law of similar type?

It was resolved in 2012 so that is pretty early stuff and it is a shame it was not widely promoted as by now every salesman and every buyer would be aware that the advertised figures are bogus for most real life use.

Here’s an example of consumers’ organisation that has taken action against a couple of European companies.

Can anyone think of a good consumers’ association that might be able to take this on in the UK?

Hi wavechange – the web address you’ve added links to subscriber-only content on the Financial Time’s website, just to bear in mind. 🙂

Sorry about that, Andrew. I don’t know why I was able to read the article earlier today. I don’t normally post links to the Daily Mail but this is the same story: http://www.dailymail.co.uk/news/article-2766915/Revealed-The-car-makers-tricks-boost-MPG-claims-leave-drivers-forking-400-year-petrol.html

Not a problem wavechange – I embedded the weblink into your text to avoid some greater confusion.

There seems to be confusion in these newspaper articles. The manufacturers are, I understand, required to publish their emission and fuel consumption results derived from tests specified in an EU standard. Those figures have a caveat that they may well not be met in real life driving. I also understand they can publish no other figures. The standard is well out of date and has loopholes that allow manufacturers to perform the test in ways that were not in the spirit of the standard. However, nothing I have seen suggests these loopholes make a substantial difference to the results.

If this is true it is quite incorrect to accuse the manufacturers of misleading customers to the extent indicated – they are simply doing what the EU instruct. Perhaps the EU should be blamed.

That is my reading of the situation. However, if any one has evidence that manufacturers are grossly exaggerating the results contrary to the NEDC tests or portraying the fuel consumption incorrectly then please provide it. What this conversation and the reports in the press need to be judged against are the facts. At the moment it seems to have the makings of a witch hunt.

There’s no shortage of press reports on cheating by car manufacturers: telegraph.co.uk/motoring/news/10803684/Car-buyers-lose-out-in-fuel-test-cheating.html#disqus_thread

It makes no difference if a car is specially prepared to produce better fuel economy figures or if the cheating goes on in the test itself. Either way, it is dishonest.

Which? Perhaps you could inject some facts into this conversation. As far as I can see, the loopholes in the NEDC test regime can produce just a few percent change in emissions and consumption if all are used. However the differences between the EU test cycle results and “real life” are substantial. I am looking at some individual tests – so a flavour not really a proper sample – of new cars which gives the EU combined results, their achieved result, and the % of the EU actually achieved.

Peugot 308: EU 60.1 mpg / “real” 40 / = 67%. Ford Fiesta 65.7 / 39 / 60%. Nissan X-Trail 53.3 / 41.4/ 78%. Renault Twingo 67.3 / 36.9 / 55%. Honda Civic Tourer 44.1 / 38.3 / 87%. BMW 4 Series 50.4 / 38 / 75%.

The question I ask is, do you have evidence that the difference between the two is largely down to the manufacturers “cheating” as one conversationalist puts it, or is it principally down to the deficiencies of the EU NEDC test?

I can accept newspapers sensationalising issues without proper scrutiny, but presumably if reputable consumer associations like BEUC and Which? follow the same tack they must have facts to support the allegations. I would like these to be declared so we can continue with a rational debate.

This is perhaps the wrong conversation to have this discussion because the intro is about salesman misleading customers, not manufacturers. If so perhaps we should be discussing this in the previous conversation “How many miles can your car really do on a full tank”.

There are indeed very big differences between what a motorist will achieve and the EC NEDC tests. As has been said before, the official tests are obsolete. Every driver must be aware that how they drive affects their fuel economy, and with a digital display of mpg on modern cars, the fact is inescapable.

I have seen figures quoted for the effect of various preparations used by manufacturers to inflate their fuel economy figures but there is no way of being sure how these modifications will affect a particular make and model in a test. I suspect that the combined effect of the small percentage improvements works rather like compound interest.

Why not tell all manufacturers to stop all modification of cars prior to the EC NEDC tests?

If it is motor cars Honest John is the place to look.

I note that there are sixty thousand real life stats on MPG. And the derived figures from all the reader loaded data are compared to the Test figures showing how far out they are ! But then it tells you why.

” Why the EC figures do not represent true MPG
The prescribed EC test is a lab test carried out to simulate a mix of different types of driving and arrive at ‘combined’ CO2 emission and fuel consumption figures.
Because vehicle taxation in Europe is now based on CO2 emissions, manufacturers naturally optimise their engines to achieve the lowest possible CO2 in the tests – this skews the true MPG.
This gives a correspondingly low fuel consumption figure. Unfortunately the relevant EC Directives prescribe that this figure and only this figure can be publicised by manufacturers, even though it is unlikely to be achieved by the average driver in real life conditions.”

I have no problem with manufacturers seeking to achieve the lowest possible emissions figures. That is to be commended.

It is unacceptable that they prepare their vehicles before the tests so as to achieve better figures than they would do if the tests were carried out on unmodified vehicles, as sold to the general public. Is that too much to ask?

The point is, when the official fuel consumption figures become the received wisdom of the motor traderes, those are the figures they will blurt out when a customer asks the question. That’s what happened in Mr Clement’s case and it has probably happened every day for years. Car salesmen are on a bonus for every sale and having met a number of them [and they are all men – the women’s role in the dealerships seems to be to make the coffee and rearrange the magazines] I wouldn’t expect them to do any more than parrot the tests results and say nothing about their imperfections as an indicator of real performance. The buyer has do their homework and use websites like Honest John and motoring magazines to get closer to the actuality. Unfortunately, on this issue, the EU burns a lot of fuel and takes a long time to get anywhere. Like every schoolboy I used to cook my conkers before a contest; it’s natural that car makers will do the same if they can get away with it. The important thing now is to make sure car buyers realise that the official figures are faulty and not fit for the purpose for which they wish to use them. Disappointingly, according to what Diesel has posted above, they don’t seem to be much fitter for the other purpose for which they are produced, namely emisions. Malcolm suggests that the differences in performance caused by modifying the vehicles tested aren’t all that significant and don’t distort the overall rankings; I am not convinced, myself, because there is no way of knowing.

John – What is the point in publishing fuel consumption figures if manufacturers use various ways to artificially improve them? Since manufacturers pick different ways to prepare their cars, the value of the tests for comparative purposes is debased. It seems clear that the difference between the published figures and what the motorist achieves is growing.

No-one has given me a reason why tests should not be carried out on unmodified vehicles.

I completely agree with you. The official fuel efficiency figures are a “trade misdescription”. These Conversations prove that – we are debating whether it’s this or whether it’s that. The problem is it’s not possible to just stop publishing the figures; they exist, and car owners and buyers want to know them. If I understand it correctly, the NEDC results are the only figures the manufacturers are permitted to state in any advertising and in technical specifications and data that they produce. This is makes no sense. What I am not at all clear about is what a car dealer can say to a prospective purchaser about fuel economy. I suspect they are not so shackled and could offer an “illustration” of the typical performance likely to be achieved under the kind of driving conditions that correspond with the purchaser’s experience according to reputable research. Still quite vague perhaps but probably more useful. It would not be a daunting task to compile comparative tables and make them available across the industry; most dealerships only handle one or two marques so they would not have to spend a fortune acquiring the data for the specific models they sell. I suspect that the relationship between the manufacturers and the traders makes any of this practically impossible.

It would be interesting to know what the Society of Motor Manufacturers & Traders thinks about all this. In the interests of consumer knowledge, which is power, it is important that the discrepancies are publicised and that car buyers are pointed in the right direction to find useful information.

The reason that no-one has given a reason why tests should not be carried out on unmodified vehicles is no doubt because there isn’t one. Why such a massive element of the overall European economy [with significant implications in the case of vehicle imports], and such an important part of EU citizens’ lives, does not have a reputable and independent car testing establishment, with standardised inputs and outputs, beats me.

I don’t know how we will move on, John. I suspect that until the new tests are implemented we will continue to be fed worthless figures. We need consumer organisations across Europe to unite and ensure that preparation of cars to improve fuel economy figures ceases forthwith.

John, “The reason that no-one has given a reason why tests should not be carried out on unmodified vehicles is no doubt because there isn’t one.” As discussed elsewhere as far as I can see the modifications allowed by the laxity in the EU standard do not give substantial reductions in mpg or emissions – nothing like the large differences that appear between the artificial NEDC test cycle – outdated – and “real life”. The culprit seems overwhelmingly to be the test regime. If someone has evidence to the contrary then it would help this conversation move on – and perhaps manufacturer-bashing could be done on a sounder base.

I don’t know how much closer to reality the new standard will be when it eventually yields figures. But in practice single values will never represent reality; a range of values for mpg within which say 90% of drivers should fall is to my mind the only sensible way to present useful data. The new test will not do this, so will always be for comparative purposes.

It might be more constructive to have all those who collect real-life data across Europe – from drivers records, not from test regimes – to pool and publicise the results and produce an on-line data base that purchasers can access. That should satisfy those who want to establish likely costs, while the EU tests can be recognised as producing data for comparative purposes; the two objectives should not be confused.

I have assumed that the motor manufacturers, who virtually control the dealerships, would not be very happy to see the unofficial and realistic figures presented to car buyers. That is why I advocate publicising the alternative sources of data. After all, not all car buyers are canny Which? readers and dealers are trained and incentivised to steer them into a purchase with an element of economy in the information given.

We will never get perfect comparability in fuel efficiency statistics but we can at least try to stop the weighted results caused by modifying vehicles before test. Just because manufacturers can do it doesn’t mean they should. I am mindful that the CBI recently called for companies to earn more trust from the public. Well, here’s a starting point for one of our biggest industrial sectors.

Many people are indifferent to fuel economy figures but for just as many it is possibly critical. The consumer organisations must continue to press the EU authorities to get on with implementing a more sensible test regime with minimum delay, urge the EU to put an end to anything, however apparently insignificant, that compromises the trustworthiness and value of the NEDC tests, and tell the public what the data really means and how in its present form it is unhelpful.

Hi Malcolm, I’m sorry it’s taken a while for us to respond to this post.

Our experts have stated that changes in technology can partially account for the growing gap between official and real-world figures. However, Altroconsumo, the largest consumer association in Italy, recently conducted the official (NEDC) fuel consumption tests on a Volkswagen Golf 1.6 TDI and a Fiat Panda 1.2 and found significant differences between the declared results and their own re-testing.

They’ve also said that the organisation hired a certified car testing laboratory that performed the same tests that car manufacturers are obliged to perform. In order to investigate the effect of the flexibilities allowed under the NEDC, Altroconsumo identified a number of different parameters that were tested under different settings. The cars were tested under optimal conditions for both low fuel consumption and high fuel consumption.

The declared fuel consumption and CO2 emissions of the VW Golf were more than 50% lower than the test results obtained by Altroconsumo. The declared fuel consumption and CO2 emissions of the Fiat Panda were more than 18% lower than the test results obtained by Altroconsumo.

This has resulted in Altroconsumo launching legal action against these two manufacturers for overstating fuel economy performance. The organisation is seeking compensation for owners of these cars

You can find out more about Altroconsumo’s investigation here.

Andrew, thanks for providing this information. If the implication is that the manufacturers did not follow the test procedure and gave results, as a consequence, that were effectively falsified then they should be prosecuted. Is this what was alleged to have happened?

The NEDC test does, because it seems carelessly constructed, give leeway in the way cars are prepared for testing. One report I have seen suggests that between 2002 and 2010 the reduction in CO2 emissions overall was around 30%. Of this it was estimated 20% was due to technological improvements – so a real gain under the test conditions – whilst 10% on average could be accounted for by using the imprecision in the tests specified if manufacturers chose to do so. Under the proposed new tests many of these have been removed so hopefully more realistic and comparable results will be reported – in around 5 years!

Given these “loopholes” that give rise to so much criticism you have to ask why the EU has not acted to close them by modifying the standard.

I do not think we could ever use an artificial test to predict sensibly what real drivers will achieve, as they will inevitably cover a fairly wide range of fuel consumption for a given vehicle, depending upon driving style, load, journey type and terrain for example. Will Which? consider supporting a European database from driver reports – extending e.g. spritmonitor – that we are advised to use to check “real life” consumptions?

The BEUC report kindly provided by Andrew relates to the same cases as the newspaper articles I referred to earlier. It seems unlikely that the companies will be prosecuted or the unfortunate owners compensated.

Every motorist needs to know how manufacturers are preparing their cars for the EC NEDC tests, thereby producing false data that will push up the costs of running a car. I am surprised that ‘Honest John’ makes no reference the cheating that has become endemic in the motor industry.

This BEUC report more or less states what has been said elsewhere – that the EU NEDC test is well out of date with respect to current vehicle technology and driving habits and is lax in certain of its test requirements. You have to put much of the blame for this at the EU’s door and ask why they have not dealt with it.

If Altroconsumo has evidence that Fiat and VW have produced consumption figures by testing other than in accordance with the standard then they should be punished. As the EU relies on these results to meet its CO2 emissions targets I presume they would impose hefty fines.

If however Altroconsumo is simply accusing VW and Fiat of producing misleading consumptions by publishing the only figures allowed by the EU, and if these have been produced in accordance with the standard, then I do not see what their case is. It will be interesting to see how this develops.

I still want to know if the new standard is being delayed and, if it is, why. Meantime I still suggest we need a reliable database from drivers reports of real life consumptions that we can all access to see what we might achieve. A standardised test procedure is no substitute. I hope Which? might comment on supporting this.

There is no reason why any car should be modified prior to testing. Like cheating in the subsequent tests, it is falsifying the results. Perhaps those who devised the NEDC tests should have specified that vehicles must not be prepared in a way that would improve fuel economy or emissions figures.

Since vehicle testing is not carried out in independent laboratories, we cannot be sure how long the manufacturers have been modifying their vehicles prior to test, or which manufacturers are using which tricks. I do not know why the testing protocols are so out of date, but my understanding is that it is the manufacturers that are keen to delay implementation of tests that are more relevant to the 21st century.

Maybe Rob Hull could tell us how vehicles are prepared before testing for Which? I expect that the tyre pressures are set as recommended in the makers’ handbooks and the engine warmed up where specified. I very much doubt that the doors will be taped up or the tyres over-inflated.

Thanks Rob. That’s what I would expect from Which?. I would expect manufacturers to do the same.

Rob, your link states “All manufacturers follow the same test procedure, but can select any accredited lab to use for the test. It’s very hard to get truly repeatable and comparable results when using multiple labs.” I am a little surprised at the inference here – that your lab results may not be replicated by another lab. This means your lab may not give representative results then?

My experience of testing to EN and worldwide standards is that they tell you the tests you must undertake, the way the tests should be conducted and specify the key characteristics of the test equipment used – precisely to ensure that results are as repeatable and comparable as possible.

For example when carrying out a test for water ingress, the size of the water nozzle, the precise shape of the cavity behind it, the water flow rate and its temperature, the distance the nozzle is held from the test item, the temperature and preparation of the test item….etc…are all specified.

Are you saying that the NEDC test does not specify the test equipment and its operation?

If you look on the SMMT website they have a “Myth busting statement” about the NEDC tests. I have challenged their comments on some of the points raised and wait to hear back. If I do i will pass the response on.

I have been in industry that has had its own well-equipped test labs – also development labs of course – and my myth busting comment would be, for those who have not had that experience, do not be so cynical about their integrity. In my experience they worked honestly and constructively. As I have commented elsewhere it is quite impractical to have only independent test labs for every product manufactured so not a route worth pursuing (in my opinion). Existing labs are quite well regulated through their approvals processes. However, if someone has factual evidence that this is not the case this is the forum to present it.

I cannot comment about the practicality of having independent test labs. Nevertheless, independent testing is highly regarded and we would not have independent labs if their value was not recognised. My view is that if products are going to be tested, why not have the testing done in independent labs?

Which? no longer runs its own test labs but testing is still carried out by independent labs. Rather than asking manufacturers or retailers to supply goods for testing, they are purchased. Would you want a magazine compiled from data supplied by product manufacturers? I would cancel my subscription if that happened.

Oops – My post above was a response to Malcolm’s post below.

wavechange, I have had experience of independent test labs – these are often profit making organisations. Your total faith in them is misplaced – they depend on the quality of their personnel, and their investment in equipment. It is not the ideal world you might think.

I have been told this by one of my former postdocs, who went on to run a government lab in what used to be called county analysts. I don’t have total faith in independent labs any more than I have total faith in assessment in higher education, something I have had direct experience with.

Independent testing remains the best respected approach and if the gold standard is tarnished then action is needed to rectify the problem.

Whatever the shortcomings of an independent lab, I hope they would not prepare cars for testing by taping up the doors, swapping and over-inflating the tyres, and so on.

Given the volume and economic value of the European car industry [including imports] I should have thought it would be cost-effective to establish an institution that could undertake the testing on behalf of manufacturers. At present manufacturers are put on trust to test their vehicles in accordance with the terms and conditions of the New European Driving Cycle [NEDC] but, other than the company’s own quality assurance system, there is no external validation of the test conditions and the results. Introducing that might be a worthwhile interim measure to boost confidence in the tests. I am sure that the personnel in the companies’ test workshops will strive to ensure that the highest scientific standards of compliance with the NEDC rules will be applied but, as we all know, commercial pressures can lead to contrivance and there would be much merit in applying a degree of external rigour to the process.

One of my personal desires would be to ensure that motor cars can last longer before being scrapped. I feel that every five years we should achieve the addition of one extra year to average longevity of the overall vehicle fleet as a result of design and product development. Anything that would harness the advances in performance and safety acrosss the industry so they can be applied to existing vehicles, and that enables cars to be retro-fitted to avoid obsolescence, would progress towards that objective. A respected pan-European testing house and research institute could make a major contribution. It could bolster the European manufacturing industry and lead to fewer imports [I am not in favour of protectionism and there is a lot to learn from technological developments elsewhere but we need to keep the playing field level]. This might all be too idealistic, but as Chairman Mao is reputed to have said, “the longest journey starts with a single step”.

John, as I understand it the tests are subject to independent scrutiny and the laboratories are accredited. Accreditation with an international standard against which there activities are assessed is standard for many industrial laboratories.

Cars do last much longer than they ever used to – largely through lack of corrosion of bodywork as well as better manufacturing of the engineered components. My two cars are 11 and 21 years old respectively (we sold the 30 year old car last year) – all in good working order and just MoTd. The main problem is with the new technology modern vehicles incorporate – when that goes wrong is it worth the cost of repair? Partly high component cost, but also high labour costs. The penalty of making cars safer and more energy-efficient.

Thanks Malcolm. Wikipedia quotes Which? [May 2015] as saying that “no official body polices the tests and the car companies can arbitrarily reduce their results by 4% at the end of the cycle”. I have left my May magazine in someone else’s house so I am not able to check right now whether that quotation has been taken out of context.

John, I have seen statements that the tests are independently looked at. I have asked SMMT, one organisation that states this, whether this is true.

The 4% is part of the requirements of the EU test. Not something the car manufacturers arbitrarily do, seemingly. This is in one of the documents –

“UNECE regulation 101 allows extending the type approval to all vehicles within a 4% interval on their respective CO2 emissions. Such provisions were not discussed in the WLPT process” (WLTP is the proposed new test regime).

In another “In addition, the NEDC regulation provides tolerances for some of the parameters that also have impacts on CO2 emissions, for example, tolerances around the speed schedule of the driving cycle (±2 km/h and ±1 second), tolerances of dynamometer control coast-down times (±5% and ±10%), temperatures, differences between measured and official CO2 values (±4%), tolerances of measurement devices, etc.”

The concerns I have, with no links to the industry, are that when Which? looks at topic such as this it gets the facts in front of the readers it is seeking to inform and presents a balanced and objective case. At present the facts do not seem to have been either well researched or fairly presented and I do not want to be misled into a conclusion that may not be fair. So my aim has been to try to find out facts and put them into this conversation. If my “facts” are incorrect then I’d welcome them being challenged – I’m just a layman digging arount the internet and may be reaching the wrong conclusions. But facts are the key.

John wrote: “At present manufacturers are put on trust to test their vehicles in accordance with the terms and conditions of the New European Driving Cycle [NEDC] but, other than the company’s own quality assurance system, there is no external validation of the test conditions and the results. Introducing that might be a worthwhile interim measure to boost confidence in the tests. I am sure that the personnel in the companies’ test workshops will strive to ensure that the highest scientific standards of compliance with the NEDC rules will be applied but, as we all know, commercial pressures can lead to contrivance and there would be much merit in applying a degree of external rigour to the process.”

I totally support this as an interim measure, but in the longer term we need a robust testing procedure that any informed person can understand and will respect as fair.

If, for the sake of argument, more realistic tests were required and carried out, how do you all think potential customers might react to seeing test figures like:

Toyota Yaris 1.0 petrol 48 mpg +/- 8 mpg uncertainty

Ford Fiesta 1.4 petrol 42 mpg +/- 12 mpg uncertainty

Nissan Note 1.4 petrol 45 mpg +/- 10 mpg uncertainty.

So you would be able to reject the Ford Fiesta as being sold under false pretences if your style of driving led it to return less than 30 mpg.

DerekP, I think the only sensible way of giving drivers a good idea of their likely consumption is a range, as you suggest, within which 90%, say, would fall. These can only be derived, I think, from real life experience – the sort of data collected by e.g. spritmonitor. This data cannot be provided I suggest by the tests set for type approval by the EU – whilst the new test regime is likely to be more like the “average” European driver it is still an artificial test under laboratory conditions. I don’t know how it could be used to give the spread of consumption that a range of drivers would get in practice. I suggest it is left to do its main job – to reduce CO2 emmissions to specified levels and let’s have a separate publicised and independent data base against which we can estimate our own likely consumption.

Your Ford Fiesta might be run on badly inflated tyres, continually fully loaded, in hilly country with a heavy right foot. I don’t see how a manufacturer can deal with a fuel guarantee under such circumstances.

Seems to me that we are looking for spurious accuracy in some respects. Rather like the 1.8 children and on average humans have one testicle. Figures can be accurate but not very useful.

Perhaps the simplest and most useful method would be to test the engine sans the car and just report on what the engine potentially can do in best case scenarios. Then nobody can possible mistake the figures for what a car might achieve in real life.

Average km per litre at 30, 120 and at 80kph for all engines will be a fair representation of the engine efficiency.
After that I imagine wieght of car , weather, driving style and other traffic can mess it up as much as it likes but we will have one set of unambiguous figures.

It is well known that manufacturers removing parts of vehicles to reduce weight and return better figures in the official tests, but removing everything bar the engine is truly innovative. However, I’m not sure how you plan to measure mpg if there are no wheels.

There are test cycles that just use the engine removed from the vehicle, aimed at commercial vehicles. However in most cases it is simpler to test the whole vehicle. Remember the emphasis is on emissions, to which l/100km (or mpg if you are an imperialist) is linked.

I see. I thought we were discussing the fuel economy figures. Apart from the VED being a reflection of carbon dioxide emission, I doubt that the majority of motorists car about emission figures. What they want is an honest indication of the fuel consumption of vehicles.

Testing under standard conditions is obviously not going to show what any individual driver will achieve but if the manufactures stop cheating and preparing their vehicles, we might see slightly more useful figures when we are in the market for a new car.

This conversation is about salesman or dealers who deliberately or otherwise give fuel consumption figures to customers that are wrong. Perhaps it should revert to that theme as we have deviated away from it. The questions about EU testing etc might be better continued in the parallel conversation “How many miles can your car really do on a full tank?”. Just a suggestion to keep the themes from being confused.

I think the threads are already hopelessly intertwined and most will go to the “live” thread.!!

The point about testing the engine was to divorce its testing from any confusion in users minds from what a car might get in actual life. Testing as to what a car might do in real life could be run by consumer magazines or auto magazines on their own courses and consumers can take their choice on how far any consensus exists … or not : ) And whose results and reviews they trust.

Given the automated driven car any test track can be run precisely by an AI and for considerable periods so achieving reproducible figures accurately.

Re; The query on running engines without wheels that is actually something that has been going on for decades as manufacturers calibrate for ECU and gear-shifting and measuring torque. To simply run it in top gear or any series of gears to measure fuel usage AND emissions is actually trivial.

I have already made a very on-topic point about how I was lied to by the dealer who I bought my car from. The farce of manufacturers preparing cars to perform better in tests is relevant to the discussion, in my view, since it is the reason that cars consume more fuel when tested by Which? and driven by the general public.

I’m not sure how many miles my car will do on a tank because of the inconvenience and possible danger of running out of fuel. I would estimate 650 miles.

I’m sure our moderators will give us a gentle nudge back on-topic if necessary. 🙂

The official fuel consumption figures are suppose to provide us with figures that we can use to compare vehicles, but that’s no use because some companies are bigger cheats than others: http://www.transportenvironment.org/press/makers-gas-guzzling-cars-cheat-emissions-tests-most

Here is an article courtesy of the Munich University, entitled ‘100 tips to beat the New European Driving Cycle (NEDC)’: http://www.researchgate.net/publication/260811509_100_tips_to_beat_the_New_European_Driving_Cycle_(NEDC)

Interestingly it says that taping up seams may be illegal.

Which? has submitted a super-complaint about misleading and opaque pricing practices in the grocery market. I believe that international action should be taken against every car manufacturer that has engaged in preparing vehicles to produce improved fuel economy figures.

Honesy John publishes real life figures provided by motorists on his website, together with the “Official” (EU) published fuel consumptions. He also gives a brief critique of the NEDC tests and its relation to “real life”.

According to his data, on average, manufacturers meet 86% of the Official Test Results in real life.

A quick look at major manufacturers shows quite a wide range of achieved figures across their car fleet. For example:
Honda average 91%, range 76-117%. VW 89%, range 71-104%. Fiat 86%, range 74-97%, Ford 83%, range 71-92%.

I suspect the range will partly be explained by deficiencies in the Official test in not representing real life EU driving or car technical advances.

I would like to know how honest the Honest John website is. To quote from the Real MPG section:

“Because vehicle taxation in Europe is now based on CO2 emissions, manufacturers naturally optimise their engines to achieve the lowest possible CO2 in the tests – this skews the true MPG.”

While it is widely recognised that manufacturers are modifying cars to make them perform better in the official tests, I am surprised by the fact that all we have is a positive comment about manufacturers aiming to achieve low carbon dioxide emissions. Having read the motoring section of the Telegraph for years I have assumed that it is generally supportive of the motor industry.

Drivers’ figures put on this or any other website are no substitute for good quality independent testing under controlled conditions.

wavechange, HJ says on his website:
“Overstated official mpg figures cost UK drivers a colossal £4.45 billion per year in fuel and can result in misery for millions of cash strapped motorists. That’s why we created Real MPG, which allows readers to show how many miles their cars actually do to the gallon.

To date, we have collated more than 46,000 entries, covering all the major manufacturers and models. On average, 87 per cent of cars in the UK achieve their official MPG figures, but some fare better than others.”

Thats more like the hype we like to see, isn’t it!

Figures from “the general public” if collected in sufficient quantity to be a significant sample are the best way I can think of to give other drivers a good idea of the range of consumption they might achieve. It is data from the “general public” that gives the basis on which driving cycles are constructed. I’d rely on such data, just as if consumers associations were to collect data on the trouble free lives of domestic appliances I would use them to assess how long mine might last – better than an artificail test.

If facts presented get in the way of an argument it is sensible to challenge the facts. Perhaps you should ask HJ but simply casting aspersions is not the way to support a stance. I am simply trying to look for facts and infotrmation rather than to pursue a witch hunt that may, or may not, have foundation.

Which?, maybe you would like to comment on Honest John’s data and whether it is honest or not. The site gives the number of driver reports upon which the data is based and any of us can add our own data, like we can on spritmonitor.

I could make the same comment to you, Malcolm. 🙂 It is common knowledge that manufacturers are cheating the public by modifying their cars so that they perform better in the tests. I would like facts about the extent of this cheating, but it seems very unlikely that we will be given them.

If you are happy with numbers put into databases by the general public, that’s fine. I want to see figures generated under standard conditions using unmodified vehicles. People use the official figures. We have both given examples of this.

I would like to know why we cannot ban manufacturers from modifying their vehicles prior to testing.

I see that Parkers guide reckons that a drivers style can affect the mpg between 2% to 20% so I think nitpicking on the margins is probably not that useful given the effects a driver has. Let alone the types of journey, weather, traffic conditions etc.

On Saurday I bettered the manufacturers figures by the odd mile per gallon because I can drive smoothly and by anticipation in built up areas try to avoid any stopping. All of this done safely. Admittedly it was around 11 pm but a mixture of roads and only 18 miles.

The journey down I thrashed the revs a bit and was near 8mpg under for journey. But then at 5 pm with three stops at lights and junctions also. SO overall my average would be near the bottom end of the range. However taking the individual journeys it proves without any doubt that style is perhaps a much bigger factor than removing a wing mirror etc.

Therefore I submit that more accurate figures from the manufacturers derived from a no chicanery tests would still be pretty darn useless . Of course we may all be paying a little more in Road Tax …. ..!

Of course driving style has a big influence, as does terrain, speed and holdups. Every driver will be aware of this. But when the large majority of motorists are getting lower fuel economy than the test figures then it is hardly surprising that so many people are fed-up with the fact. I’m not too concerned about the figures that I achieve, which are nearly 10 mpg better than my previous car, but cheating by the manufacturers to achieve artificially high figures must be stopped.

If all manufacturers test unmodified vehicles then at least there will be a chance that a significant percentage of drivers will exceed the official fuel consumption figures. Many of us are paying little or no VED these days.

“If all manufacturers test unmodified vehicles then at least there will be a chance that a significant percentage of drivers will exceed the official fuel consumption figures.”. Unlikely; it is the test that is at fault as it neither represents modern car technology nor driving in real life. To its credit the Which? May article says “The culprit: an obsolete test”.

Official reports estimate the test is likely to give up to 30% higher figures than are realised “on the road”. The ways manufacturers can exploit weaknesses in the test are unlikely, an official report states, to contribute more than 10% to the CO2 reduction (and consequent mpg or l/100km).

The WLTP cycle proposed should be more realistic but I maintain the best source of fuel consumptions for drivers are data from other drivers collected on properly-managed databases.

The EU tests are primarily aimed at reducing emissions. The EU sets CO2 targets that manufacturers have to meet. If they don’t, the EU fines them.

Malcolm – In the tests carried out for Which? the cars are not modified and the tyres are at the specified pressures – see the recent post by Rob Hull.

I know the current test is outdated, and have done for years. Since we started to discuss this topic I have learned something about the new test, though I still have no idea if the manufacturers are pushing for its implementation.

As I have said before, it is cheating to exploit weaknesses.

wavechange, I am asking two organisations about the alleged delay to implementing the WRTL test. I will be interested to find out. Which? could contact the EU to ask them perhaps? Two possible explanations – but they are not official – are that the EU are setting CO2 targets alongside the new test regime that manufacturers regard as economically or otherwise unachievable in the proposed time scale, and/or that the CO2 targets are siginificantly less than set by other countries (e.g. USA, China, Japan) which will lead to a lack of competitiveness in those markets. Suppositions only but there are usually two sides to any story.

We are aware of the loopholes, or laxity, within the NEDC standard. What we do not know is how far individual manufacturers make use of these, and to what extent. Perhaps you, or Which?, have information about this?

What does seem clear from reports prepared for the EU is that the majority of the “discrepancy” between real life consumption and the artificial consumption provided by the test is due to the outdated test. The impact of what you describe as cheating is what matters in practice and it seems, in theory, to be about 10% of the reduction if manufacturers all took advantage of the test – but unless anyone knows how much advantage they take then this is conjecture. By all means berate them but let us stay focussed on the root problem – an outdated test. I’m sure we wouldn’t want to milslead people into thinking all the problem lies at the door of the manufacturers – unless we have supporting information.

Perhaps we should pause now to see if we get some facts?

I find it quite astonishing that anyone could envisage a test regime that reflects actual usage since there are such huge differences in the way that vehicles are driven and used. Any test regime merely gives a figure for the specific test conditions and these simply cannot cover all eventualities. As has been mentioned several times, test figures to a specific standard simply provide one method of comparing cars, anyone who thinks different is, frankly, living in cloud cuckoo land. The variations in actual usage are, probably, an order of magnitude or more greater than any difference in the test results from what wavechange refers to as cheating.

Has anyone has claimed that official tests could ever reflect any driver’s actual usage? I often get different figures when I make the same journey.

My concern is that manufacturers are modifying their cars to produce better figures than could be achieved if the tests were carried out by a good independent test lab, and that since different manufacturers are carrying out different fiddles, tests are NOT carried out under standard conditions.

tonyp, This conversation is headed ” I went to court when my car didn’t meet its mileage claims” and the previous one on a similar topic began “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?” The inference is that manufacturers are making claims that cannot be met. Whereas manufacturers are not making claims but simply publishing the only figures I believe they are allowed – those derived in a laboratory from standard tests done to an outdated and defective EU standard and driving cycle They are officially declared as not likely to be what you get in practice.

You are quite right – real consumption will depend. for an individual driver, on many factors that a standard test will not cover. We can only hope the proposed new driving cycle will get nearer an average for most drivers; the present one seems, from one source (Honest John), to be around 16% optimistic overall.

In practice it would only be sensible to publish a range of consumptions within which most drivers would fall The EU test will not do this; that must be left I believe to collated data supplied from experience by real drivers.

The theoretical impact on test results of manufacturers modifying their cars, because of the leeways allowed in the NEDC test, has so far been estimated at around 10% of the 30% savings in CO2 (and presumably fuel). To my knowledge no evidence has been produced of what individual manufacturers might have done or to what extent.

Test labs should be using the standard test regime and test equipment of specified performance if the standard is formulated as EN standards normally are to ensure reasonable repeatability and consistency from lab to lab – whether manufacturers’ in house facility or independent.

I’d like to see evidence before I draw firm conclusions.

Yes, Malcolm, I know how the conversation is headed, but the contributions have morphed into the accuracy of test figures. As it so happens, I do know the published information for my car and, since it is fitted with displays of instantaneous consumption, journey consumption and cumulative consumption, I am very aware of the actual values achieved as well as being able to see which driving techniques cause the highest consumption. As I mentioned in a previous post, I find that sometimes I achieve worse figures than those published, sometimes the published figures and even, admittedly rarely, I have achieved better than the quoted figures.

Recent activities have illustrated to me the effects of a series of short journeys in heavy traffic on fuel consumption. For the last 10 months I have been travelling, on average, twice a week to the local hospital for treatment. This is a journey of about 4 miles, but the journey time and fuel consumption can vary quite significantly depending on the time that I am making the journey – the worst case is school traffic in the afternoon – even in the best case, my consumption figures are significantly lower, in the region of 20% less than my more normal usage. With a mix of hospital visits and my more normal usage I find that my cumulative consumption is some 17% lower than it used to be. I think my experience shows the difficulty of providing ‘real life’ figures and the idiocy of expecting dealers to provide a guaranteed consumption performance.

tonyp, I’ve kept accurate records of my car also and achieve on average 37.1 mpg against an “official” 38.7. However I, like many, have mixed driving patterns and the range is from 31 to 42 depending on the type of journeys. So I am in agreement about the difficulty of providing “real life figures” – only a range would make sense.

I agree that a quoted range would be preferable to a single figure, but only to those who understand the concept of probabilities. How, for instance, would the minimum figure be defined? Would it be for a sensible driver who just happens to be spending most of the journey time in heavy traffic although driving sensibly or for the driver who has a load of heavy junk in the boot and a roof rack whose driving style is that the right foot should be either holding the accelerator pedal flat on the floor or flooring the brake pedal. If the range were to be such that all situations are covered then then it would be so wide as to be virtually useless.

If the range is based on sensible usage then there will be drivers who fail to achieve the minimum and claim that they have been cheated. It would be possible to assign a probability to the figures but my experience is that most people do not understand this concept, they fail to appreciate, for instance, that a 90% probability of something happening means that there is a 10% probability that it will not. However the figures are presented there will be those who will find cause to claim that they are wrong and that they have been cheated, it’s called human nature! To some extent the concept of a single figure with a caveat saying that it is merely representative and may not be achieved in practice is better than quoting a range with a similar caveat.