Comparison of Weather Sensitivity of Fuel Consumption, Prius vs. Conventional

This is a comparison between a 2007 Toyota Prius and a 1998 Mitsubishi Carisma GDI, with predominantly low speed highway driving in a temperate climate.

Prius data is from September 2007 to December 2010, over 110,000 km, and 88 fill-ups.
Carisma data is from August 1998 to August 2007, over 250,00km, and 252 fill-ups.
The Carisma has about the same size and performance as the Prius.
The Prius has a 1.5 liter gasoline VVT-i Atkinson engine with 78 hp, integrated with two synchronous electric motor/generators in the Hybrid Synergy Drive, for 113 total hp.
The Carisma has a 1.8 liter 125 hp engine with stratified charge lean burn capable gasoline direct injection, NOx cat, and five speed manual transmission.

The Carisma always had only summer tires mounted year-round, or all-season tires mounted year-round.
The Prius had winter tires mounted in winter that had about the same rolling resistance as the summer tires, and since September 2010, all-season tires.
Both cars had conservative grill blocking, and mostly 0W-30 motor oil. The Carisma also had additional mild intake air preheat.
Both cars were driven by the same drivers on the same routes with almost* the same speeds.
*The Prius is possibly driven slightly faster on average because of the recently added passing lane on the daily commute.

Routes:
About 50% of the distance was the 2 x 23 km daily commute rural highway with 3 km city + a few km errands & activities.
About 40% of the distance was 50 to 120 km rural highway and Autobahn trips with a few km city driving. The Prius is currently tending to be driven more of these trips than the Carisma was.
About 10% of the distance was over 120 km rural highway and Autobahn trips with a few km city driving.
Highway and Autobahn driving is mostly going with the flow of traffic in the right lane, which is usually about 88 km/h actual (speedometer shows higher, odometer is very accurate).
The city driving is mostly at the end of a trip, with a warm engine, and is about 10% of the total.

On average, conditions for both cars were about as comparable as practically possible.

The raw data can be found at Spritmonitor.de (Fred_H).
Three mpg data outliers, which were known to be caused by unusually high speed trips, were excluded from the calculations.

Average climate data from a nearby city is included for illustrative purposes only. Actual temperatures while driving were not recorded, and tend to be more extreme than the averages.
The mileage data lags behind the temperature data, because fill-ups often include trips from the month before.

The monthly mileage averages are not distance weighted, but are a simple average of the tank averages, and the seasonal averages are simple averages of the monthly averages. This gives equal weight to every month, regardless of how many miles were driven in each month.

The calculations do not give exact absolute values for the exact time frames, but they are calculated in exactly the same manner for both cars, giving a fairly exact relative comparison on average.
There are small variations due to the smaller sample size for the Prius combined with random weather and tank fullness variations, but on average, the mpg decrease of the Prius in winter is about the same as the Carisma. In time, with additional data, the Prius mileage curve is expected to smooth out.


Plot of average monthly average mpg of eight years for the Carisma, and three years and three months for the Prius, in predominantly low speed highway driving.



Graph of average seasonal average mpg of eight years for the Carisma, and three years and three months for the Prius, in predominantly low speed highway driving.



Graph of average seasonal average of eight years for the Carisma, and three years and three months for the Prius, in predominantly low speed highway driving, in liters per hundred kilometers.
Although the seasonal difference appears higher for the Prius when calculated as a ratio to the Prius's lower average consumption (mi./gal.), this graph shows that the absolute difference in the amount of fuel consumed in a given distance was actually lower for the Prius.


These results are for predominantly low speed highway driving in a temperate climate, and extrapolation to other driving patterns, cars, or climates may or may not prove to be accurate.

I consider this to be an interim report, as I intend to update it occasionally with additional data.

Fred_H

 

Thanks, and you're welcome everyone!

Answering the question about oxygenated gasoline in winter, as far as I know, there is little if any difference at all in the formulation in winter here.

Until very recently I think there was less than five percent ethanol in the gasoline, but now there is a new regulation that allows up to ten percent. I haven't noticed anything yet though. In fact my last tank-full was fairly good, considering the unusually cold and snowy weather recently.

Anyway, at least till now it was the same for both cars, so this comparison is not affected.


I also suspect that the seasonal difference might be greater for the Prius with more city driving, but my impression from my admittedly limited city driving experience is that the heater setting has a bigger impact than the warm-up cycle itself. But that is speculation on my part, and even though my driving cycles contain warm-ups and city driving, it doesn't appear to have a significant impact compared to the Carisma. However that may be in part due to the city routes typically driven, which tended to be low speed (50km/h) with relatively few stops.

Also, expressed in terms that better reflect the absolute amount of fuel consumed, the Prius actually has a slightly smaller difference in fuel consumption between winter and summer of .49 liters per hundred kilometers vs. .55 for the Carisma.

Not at all. Check back here again in twelve years.