Gas mileage at specific speeds

I've been trying to measure gas mileage at specific speeds to get some idea of what the relationship between the two looks like (more or less quadratic I suppose, but the parameters would be interesting). However, it's hard for me to measure accurately due to wind and traffic conditions.

My estimate so far is based on the following conditions:

• Temperature around 20 C (68 F)
• Averaged between headwind and tailwind
• No AC
• Hybrid battery fully charged
• Cruise control at the indicated speed on the speedometer
• Actual speed measured using the trip meter (which is lower than the speedometer, and consistent with GPS speed measurements)
• Fuel consumption measured using the trip meter
• Measured over at least 2.0 km (1.2 mi) with almost no net height difference

My own results (again, probably inaccurate):

• 90 km/h (56 mph); real speed: 84 km/h (53 mph); consumption: 3.5 l/100km (68 mpg)
• 110 km/h (69 mph); real speed: 103 km/h (64 mph); consumption: 4.0 l/100km (59 mpg)
• 120 km/h (75 mph); real speed: 113 km/h (71 mph); consumption: 4.5 l/100km (53 mpg)
• 140 km/h (88 mph); real speed: 132 km/h (83 mph); consumption: 5.7 l/100km (42 mpg)

Has anyone else done this? If so, what are your results?

 

Yes, it's pretty normal here for the deviation the be about 7%, though supposedly it's getting smaller in newer cars. From what I've read, the manufacturer must guarantee that the displayed speed is never below the actual speed, regardless of conditions, so they add a large margin.

Note that the fact that the trip meter is accurate shows that the car does actually know how fast it's going.

 

Thanks Kevin_Denver! Exactly what I was looking for. Seems pretty consistent with my own measurements too.

Amazing fuel efficiency at low speeds BTW, I guess the serial hybrid design really pays off.