This is clearly an important factor in determining your vehicles fuel consumption, but it's pretty difficult to make quantitative estimates and without it comparisons are pretty meaningless.
My car's computer can tell the average speed since the last reset (i.e. re-fueling). And it's always suprisingly low! I use this to estimate the %City driving by assuming it to be a linear function of speed from 20km/h=100%City to 120km/h=100%Highway.
My wife gets an average speed of 40-50km/h on the school run (which includes motorway with heavy traffic). I got averages of 80km/h on my commute, which is mostly motorway with light traffic.
You could also argue that 120km/h is too high, as the efficiency has already dropped a lot from its peak by then. Maybe 100km/h could represent 100% highway and anything above that counts negative!
I notice the US urban fuel tests for an average of 32km/h, while in the EU it is 19km/h. And the US "highway" test is for an average of 77km/h (max 97km/h), while the EU "extra-urban" one is 63km/h (max 120km/h). That is why the latter figures are hard to beat.
My "highway" commute here in Oregon usually consists of steady sub-60mph cruising. Someone in Los Angeles or Phoenix might drive the 40-90-40mph sprint through traffic and call that highway too. The two are entirely subjective.
Tim, thanks for the stats on the US vs EU testing. That explains why all of the cars that come across the pond (Smart, 500, etc) always get terrible ratings in the EPA cycle.
Where I live, within the city the speed limit is 60 and the central area is 50, but as soon as you hit the perimeter it's up to 100. I consider 80km/h as my threshold for city-highway speed.
I use location and mileage as my primary criteria though. It's pretty obvious where which ends and the other starts though and I just Google post travel if I'm not sure of my highway travel and factor that from my total mileage for that fillup. My trips are pretty regular too so sometimes I don't even need to check for actual distance since I almost know it by heart.
BDC: European cars are increasingly being designed to give good test result figures, not real world figures. Guess the same is true in the US too. Must be time to change the tests again, especially with plug-in-hybrids.
Also, the EPA add 10%/22% to the urban/highway test result to give the official figures. And the Fiat 500 TwinAir is famously optimistic figures here.
Also found my average speed surprisingly low, particularly as the bulk of my time is motorway drving.
I've decided to tott up my motorway miles only & use that figure to calculate my city / motorway split, but it still won't reflect my stop / start motorway commute, maybe I should weight it down a bit if I encounter slow trips - food for thought...
During my research for a more economic car I looked into our (UK) mpg tests, this is what ours are tested too:
MPG Urban cycle
The urban mpg test cycle is carried out in a laboratory at an ambient temperature of 20°C to 30°C on a rolling road from a cold start where the engine has not run for several hours. It is intended to mimic the short car runs that are often performed, e.g. the school run or going to the supermarket. In practically all cases the urban cycle will give the lowest MPG figure.
The cycle consists of a series of accelerations, steady speeds, decelerations and idling. Maximum speed is 31 mph , average speed 12 mp and the distance covered is 2.5 miles.
MPG Extra-urban cycle
The extra-urban mpg cycle is conducted immediately following the urban cycle and consists of roughly half steady-speed driving and the remainder accelerations, decelerations, and some idling. Maximum speed is 75 mph , average speed is 39 mph and the distance covered is 4.3.
The extra-urban cycle is intended to mimic the longer journeys which typically include dual carriage way or motorway driving. As the car has been allowed to warm up and spends a good deal of time in the higher gears, the MPG obtained is higher than for the urban cycle.
MPG Combined Fuel Consumption Figure
The combined figure presented is for the urban and the extra-urban cycle together. It is therefore an average of the two parts of the test, weighted by the distances covered in each part.
No wonder why the manufacturers marketing is not realistic!!!!
For my values I blend the time and distance that I drive in town versus on the highway. Basically I use a weighted average for the time/mileage value between fill ups. Basically if I'm on a road trip where all my miles are freeway I use 100% highway, if I only drive around town I use 100% city. My commute is about 15 miles on city streets and 9 miles on freeways but the time breakdown is more than 70/30 because of the higher freeway speeds.
I do little inner city driving, probably average 15%. Suburban is probably 50-60% most tanks, average speed 60kmph with sporadic stops except winter. I plan to report about 40-50% city driving based on this, sort of a "correction" to reflect what I think true city driving would be.