I’d like some advice on the “real world” expected fuel consumption of semi automatic cars, when operating as automatics.

I need to give a bit of background first, so please bear with me.

As I understand it (from reading stuff on the internet) “official” fuel consumptions, as quoted by manufacturers, is determined in a very specific way. In summary, first they run the car up to speed on a track, and then freewheel to a stop. The time/distance to come to a stop gives them information on the forces slowing the car down (a combination of aerodynamic drag and mechanical friction). They then put the car on a rolling road and “drive” a series of specified routines – accelerating, decelerating, cruising etc to match a specific requirement. Also during the test they measure the actual CO2 emitted. They then include the weight, the aerodynamic drag and the mechanical friction losses to determine how this would have affected the CO2 figures if actually moving and not just on a rolling road. From this, they work out the fuel consumption during the test, in terms of the chemical make up of the fuel and the quantity of CO2 produced. At least that’s my understanding of the basic principle.

Now, on to automatic cars. Conventional automatics, ie those with torque convertors, have always shown worse fuel consumption than equivalent manual cars. They also have higher CO2 emissions and therefore pay more for the licence. I have assumed that this is due to a combination of losses through the torque convertor, usually less available gears (therefore more of a compromise in choosing the best one for any particular operating condition) and the timing of when to make a change of gear not being as good as a human making the same decision.

The semi automatic type of gearbox, ie more like a manual gearbox but with servo operated clutch and gear change, has become more common. (There are cheapish ones eg like on the Toyota Aygo, or more expensive double clutch ones like the VW-Audi Group’s DSG box). They make great play of being very efficient. They typically have the same CO2 figures and fuel consumptions as the manual versions, in fact VW claim improvements with their DSG. This gives them a strong selling point over conventional auto boxes.

Now, this is where we come back to the official fuel consumption tests, as described above. When doing a rolling road test, all the driver of the semi-auto car would need to do is to drive it like a manual, changing gears at exactly the same speed as he does in the fully manual version. If he does this, then the results will be identical to the manual version. This is what I presume they do. They do not run the tests as fully automatic. What would happen if he left it in auto, and allowed the car to make its own changes? There wouldn’t be any torque converter losses, but what about the cars ability to choose when to change gear, for example.

We are used to understanding that real world fuel consumption figures can be worse than the official figures for manual cars, so we can factor it in when choosing cars. But what about semi automatics? Do we get an extra hit on fuel consumption if we drive in full automatic mode? If so, someone intending to drive their car as a full automatic would be wrong to compare the official figures from a conventional automatic with those from a semi-automatic, as the semi-automatic would have an unfair advantage if it had been tested when driven manually. Is there anyone out there who knows what effect there is on the typical real world fuel consumption when operating as an auto in comparison to when they drive the same car as a manual?

Thanks for your patience in reading this far.