MPG calculations

[5wany]

This was on MSN today:

How are official mpg figures calculated?
Various standardised mpg tests have been employed over the years, with the latest being adopted across the EU in 2001. The current regime sees tests carried out by the car manufacturers themselves, under the watchful eye of the Department for Transport and its Vehicle Certification Agency. The tests, however, aren't done out on the road.
Mpg testing takes place in a laboratory with the car on a set of rollers known as a rolling road. A series of accelerating, braking and cruising drills are performed, designed to replicate different types of driving.
Equipment is attached to the car that precisely measures the exhaust gases that are released during the test. From this information, the official emissions ratings for the vehicle are calculated and fuel economy is determined very accurately at the same time.
There are two parts to the test. The first, the urban cycle, begins with a cold start of the engine and proceeds through a series of starts and stops designed to mimic town driving. The maximum speed is 30mph, the average is 12mph and the 'journey' lasts for two and a half miles.
Following that is an extra urban cycle which attempts to replicate out-of-town driving. This lasts for 4.3 miles at speeds of up to 75mph with a 39mph average.
Car manufacturers publish the results for both parts of the test but also a so-called combined figure which is an average of the two adjusted for the mileage. It's this combined mpg that's most regularly quoted as the car's official fuel economy.

Why can't I match the official mpg in my car?
The current mpg tests were designed to give accurate results but they also needed to be cheap and easily repeatable.
In the quest for repeatability and accuracy in the testing process, the huge range of variables which affect the fuel economy of different cars, driving on different roads in different locations were ironed out of the tests. When we drive our cars, those variables are reinstated so it shouldn't be a surprise that the laboratory results can be tricky to replicate.
There's so much that can impact on a car's fuel economy that isn't successfully reflected in the official mpg tests. Different aerodynamic effects from strong winds to an open window can have a huge bearing, as can the extra weight of passengers, luggage or optional extras fitted to the car.
The operating temperature of the engine is also important to how efficiently it performs with cold weather sharply reducing efficiency. Then there's the extra energy consumption from air-conditioning systems, headlights, heated windscreens and stereos.
The most important variable, though, is the driver. Different driving styles and journey types can produce vastly different fuel economy figures in the same car. A car's fuel consumption can vary by as much as 50% depending on when, where and how it is driven.

Despite all of this, the tests we all rely on to gauge fuel economy put an experienced technician, who's aiming to achieve the best possible economy, in a car, on a rolling road, in a laboratory. The tests are carried out at a balmy 25 degrees centigrade and the engine is warmed up beforehand.
It becomes clear why this is done when you consider the cost and technical difficulties that would be involved in testing every new car on the road. The weather, roads and traffic levels would need to be the same for each test and the tests themselves would have to account for different driving conditions in countries across the EU.
Do some cars get closer to their official mpg than others?
The Vehicle Certification Agency admits that: "fuel consumption achieved on the road will not necessarily be the same as the official test results." These days, most motorists are well aware of this and expect to fall below the official combined economy figures in their everyday driving.
There's also a widespread belief, however, that the official figures are useful as a comparison to help us when choosing a new car. They are, but only up to a point. Just because one type of car performs better than another on the tests, it doesn't necessarily mean you'll get better fuel economy from it on the road.

Hybrid cars perform well in town where the electric motor can assist the engine with all that accelerating from low speeds. They're less accomplished on the motorway where they rely on their internal combustion engines. Manual cars are usually more economical than automatics but the gap shrinks in larger cars, especially if they're being driven on the motorway.
Size doesn't matter as much as you might think either. A large car with a low-revving diesel engine will often be more economical on the motorway than a much smaller one with a high-revving petrol motor. There's a lot to think about but, unfortunately, things get even more complicated.
Car manufacturers are well aware of what it takes to excel on the official fuel economy tests. They'll test their new cars throughout the development process, continually making tweaks to extract the best results possible.
Who can blame them, a good official combined mpg figure is invaluable when promoting a new car to the buying public. Even more importantly, it means a low CO2 emissions rating which can save customers a packet in tax.
The upshot of this is that cars are designed and set-up to achieve the lowest possible fuel consumption in the tests. This can be a subtly different thing from achieving the best possible fuel economy out on the road.
Is there a better way that mpg could be calculated?
The mpg tests carried out at the moment are easy to perform and repeat for lots of different vehicles but the results only show us how good a car is on the tests. They're a rough guide to efficiency on the road but buyers should be careful about reading too much into the official economy figures when comparing different vehicles.
Peter De Nayer is one of the UK's leading authorities on fuel economy in cars. Formally a technical researcher and writer with the AA, he now conducts his own real world mpg tests on cars for many of the UK's leading motoring publications. He's in no doubt about the problems with the official testing.
"Fuel economy in the real world is very complex but the official tests were designed to be simple and repeatable. The results we get from them often don't indicate the sort of fuel economy motorists can expect or produce a level playing field for comparing different cars. Basically, if you ask a silly question, you get a silly answer."
Peter's own tests involve five test cycles carried out on quiet roads with stops and starts built in at predetermined points to replicate different journey types. An average is then calculated from these but, crucially, motorists can also use the raw data to work out a personalised average based on the particular mix of driving they do.

Over a five-year period, only 20% of the cars Peter has tested showed a reasonable correlation to the official combined economy figures. Two out of three cars fell more than 10% below their official figure and some missed by as much as 30%.

Peter De Nayer says
The disparities aren't uniform across similar types of car either. A survey of 20 modern eco-special models with modifications to improve fuel efficiency showed they were on average 18% thirstier than the official figures suggested.
Within those figures however, one car's fuel consumption was only 5.1% greater than its official combined mpg but another recorded consumption that was 27% higher. This highlights the minefield consumers enter when choosing cars based on official economy figures.

Peter recognises that the time and cost required to adopt a similar testing regime to his own for every new car would be considerable but recommends a few changes to the current tests that wouldn't cost the earth.
"I'd like to see the addition of a steady speed test at around 70mph. It was used in the past and a lot of people do large amounts of motorway mileage. The current extra urban cycle doesn't really reflect this kind of driving."
"A lower start temperature on the urban cycle would also be beneficial. The 20 to 30°C range where current testing takes place isn't very indicative of everyday driving in the UK and cars use a lot more fuel while they're warming up."

Will things change?
Improvements to close the gap between official mpg and the reality as experienced by motorists would be desirable but don't hold your breath.

Aside from the cost issue, a key reason for this is that official mpg figures which are on the optimistic side suit car manufacturers. The major brands must all meet EU emissions targets or face fines and any reworking of the testing procedures to get mpg results closer to real life would push official CO2 emissions higher. This would make the targets far harder to hit and make most cars look less environmentally-friendly overnight.
It all adds up to a real headache for motorists who just want to know which cars will give them the best fuel economy. The official figures can be used as a guide but we should be wary about making direct car-to-car comparisons using them.
More and more motoring publications are doing their own real world mpg testing and the results they publish should be a better guide but not a perfect one.
No single mpg figure can give a totally accurate reflection of the fuel economy an individual can expect but by choosing a car that's broadly efficient and suited to the kind of driving you do, you shouldn't go far wrong.

[X5Sport]

Interesting article and confirms what I think most of us were thinking. The fact that no testing is done during the warming up phase shows just how badly thought out the common test standard is. Certainly all the BMW automatics I have owned have totally different gear change points when cold compared to hot, and I expect the engines themselves to do the same. After all, it's why we had Chokes on old cars.

To date the only car I've owned that was tested to the common standard that ever came close to its 'Official' figures was a Skoda Octavia. Noe of the BMWs have - albeit the X6 is getting closer now it's getting some miles on it.

2/3rds of cars being significantly below the figures shows just how bad the test is, and 25C as the temp is way too high for the UK, and most likely large areas of Europe too, particularly when you take Continental Winters into account as they last longer and the temperature ranges are even lower than experienced here in the UK. There ought to be a set of tests done at say 10C and 20C rather than just 25C.

But as the report says, don't hold your breath for changes because (a) Governments don't want to admit the test is flawed and under 'real world' use the CO2 emissions are higher, and (b) neither do the manufacturers!

I wonder how many Hybrid owners know how much of a shock they're in for if a battery pack fails? Some on even the most modest cars are about £19k!! And we think BMW bits are pricey!

[graemeX5]

A lot of it is down t driving style if like my father in law you coast down hill time roundabouts and don't go above 50 turn engine off in traffic then he gets about 53mpg out of his ford ka overall and a lot is gown driving.

Where for everyday work I bought a brabus smart car is 45 mpg but I sit in traffic or up motorway at 75 so not very light footed but it's a bit of fun and you can park it most anywhere.

But the x5 well about 18mpg approx but it's only used once a week so i don't worry

[IvanG]

I use an app called Road Trip on my iphone to keep an eye on economy. It's more accurate than the car's calculations. At my last fill I got 30.85mpg with mixed driving. Not bad for a 40D when you consider the performance!

[X5Sport]

There's an interesting article in this werk's Auto Express looking at the future of how these figures are obtained. The current standard is designed to all motor manufacturers to achieve low emissions and as we all have long since twigged mean that the car is driven with minimal fuel, no ancillaries on, no passengers, at a non-representative air temperature and using such gentle throttle changes that if we were to do it on real roads we would have a queue of angry drivers behind us! This means that of course in the real world none of us get close to the figures quoted by the manufactures in their glossy publications.

Finally it seems that the message from seriously p!ssed off motorists has got through. The UN and EU amongst others are discussing the new Worldwide Light-Duty Vehicle Test Cycle. In essence this will run a test twice, once with an unladen vehicle and everything turned off, and then again with a fully laden car with everything turned on. Also, a new agreed standard for how diesel particulates are measured will be adopted, as well as more representative data for electric and hybrid vehicles. All these tests will be valid in the UK, EU, Asia, China, US, India and Japan.

The UK Industry body has apparently welcomed the change from tests designed to meet the CO2 emissions needs and are useless for real world motoring.

At least it's a move to tell the real story with best and worst case data. However don't expect the new tests to be in place before 2017 though..