Measuring fuel economy

The EPA measures the fuel economy of most cars on the road in the US at fueleconomy.gov These estimates are computed using a fairly detailed procedure:

• Drive the vehicle on a dynamometer, whose resistance is adjusted to account for air resistance and vehicle weight.
• For city estimates: drive the vehicle for about 30 minutes at an average speed of 21 mph, covering a distance of 11 miles.
• For highway estimates: drive the vehicle for about 10 minutes, at an average speed of 48 mpg, covering a distance of 10 miles.
• Additional specialized tests are done for high-speed driving, air conditioning, and so on.

For the city estimate, note that the 21 mph average speed can be considerably higher than the speed you might experience in heavy stop-and-go traffic. If you live in an area where there are persistent traffic delays, you can easily fall far short of this 21 mph average, so your gas mileage will likely be worse than the EPA's estimate.

Let's look at a concrete example: the 2013 VW Golf, 2.5 L gasoline engine:

The EPA estimates that this car will achieve 24 MPG in the city, 31 MPG on the highway, and a "blended" rate of 26 MPG. This blended rate depends on your personal mix of city/highway driving: the EPA default is 45% highway, 55% city, but you can personalize it to reflect your circumstances.

Note that, although the default city/highway mix is pretty close to 50%, the blended MPG is a lot closer to the city estimate than the highway estimate. I discuss this in a separate blog post, but the basic idea is that, for any car, the city MPG matters a lot more than the highway MPG towards the final blended MPG.

Are these EPA estimates "accurate", that is will they match what you, as the driver of the car, will actually see in terms of how much fuel you are using? The EPA website allows users to report their own MPG numbers, using the "Your MPG" feature. For the car above, you can see that 1 driver of this particular car reports an actual MPG of 27, which is not far from the official EPA estimate.

Using data from only 1 driver is not very accurate, however, so there are websites which aggregate far more data, from many more drivers, to provide more accurate fuel consumption estimates:

• fuelly (World-wide, although seems to be heavily skewed towards the US)
• honestjohn (UK): uses imperial gallon in its MPG definition, which is different from a US gallon
• spritmonitor (Germany): uses liters/100km for fuel efficiency

In the case of the above vehicle, fuelly reports a comparable average of 26.2 MPG, from 11 different cars:

When evaluating a new car, be sure to look at both the EPA estimates as well as user-reported estimates, to get a more accurate picture of the vehicle's fuel efficiency. Better yet, report your own fuel efficiency estimates at the appropriate site above, so that everyone else can benefit from your information.

Diesel and CO2

Diesel engines are more fuel efficient than gasoline equivalents: it takes less Diesel fuel to drive any distance vs. a a comparable gasoline engine. However, Diesel engines produce more CO2 from burning Diesel fuel vs. a comparable gasoline engine. If you simply compare fuel consumption, Diesel engines appear to be much better than gasoline equivalents. However, if you compare CO2 emissions, Diesel engines, although still better than gasoline equivalents, are a lot closer.

All numbers in this post were obtained from fueleconomy.gov I try to compare equivalent engines as much as possible (i.e. similar engine volume and horse-power, on the same model car).

• 2014 Audi Q5
• 3.0 TDI: 27 MPG, 385 CO2 grams/mile
• 3.0 Gasoline: 21 MPG, 429 CO2 grams/mile
• If you drive 10000 miles in one year:
• Diesel will burn 22% (=21/27) fewer gallons of gas overall
• But Diesel will only produce 11% (=385/429) less CO2 overall
• 2014 VW Golf
• 2.0 TDI: 34 MPG, 297 CO2 grams/mile
• 2.5 Gasoline: 26 MPG, 342 CO2 grams/mile
• Diesel will burn 23% fewer gallons of gas, but produce 13% less CO2 overall
• 2014 VW Jetta SportWagen
• 2.0 TDI: 33 MPG, 310 CO2 grams/mile
• 2.5 Gasoline: 26 MPG, 342 CO2 grams/mile
• Diesel will burn 21% fewer gallons of gas, but produce 9% less CO2 overall
• 2014 VW Passat
• 2.0 TDI: 34 MPG, 290 CO2 grams/mile
• 2.5 Gasoline: 25 MPG, 350 CO2 grams/mile
• Diesel will burn 26% fewer gallons of gas, but produce 17% less CO2 overall
• 2014 Chevrolet Cruze
• 2.0 Diesel = 33 MPG, 307 CO2 grams/mile
• 1.8 Gasoline = 27 MPG, 333 CO2 grams/mile
• Diesel will burn 18% fewer gallons of gas, but produce 8% less CO2 overall

Diesel engines do reduce overall CO2 emissions. However, the gains are more modest that it appears from fuel consumption alone. Additional gains can be obtained from reducing the size of the engine, and the weight of the car.

Beware of MPG bringing gifts

Background

The standard measure for a vehicle's fuel efficiency in the US is MPG (Miles per Gallon). It represents how many miles you can drive this vehicle on one gallon of gas. Vehicles have two important MPG numbers: one for city driving and another for highway driving. It is also possible to compute a blended MPG number, as a function of how much you drive in the city vs. the highway. Here's an example from fueleconomy.gov for the 2013 Honda Civic:

In contrast, the standard measure for fuel efficiency in many European countries is "liters per 100 kilometers". It represents how many liters of fuel you need to drive this vehicle for 100km. For example, the 2013 VW Golf has an advertised fuel efficiency of 3.2 liters (of Diesel fuel) per 100 kilometers. 

MPG and liters/100km are interchangeable. Knowing that:

• 1 mile = 1.6 kilometers
• 1 gallon = 3.78 liters

The 2013 Honda Civic has a fuel efficiency of:

• 32 MPG = (32 * 1.6) / 3.78 = 13.54 km / liter
• To drive 100 km, you need 100 / 13.54 = 7.38 liters / 100 km

Conversely, the 2013 VW Golf has a fuel efficiency of:

• 3.2 liters / 100 km = (3.2 / 3.78) / (100 / 1.6) = 0.84 gallons / 62.5 miles
• This car will drive 74.4 MPG on one gallon of gas

Is there a reason to prefer MPG vs. liters/100km?

Suppose you want to upgrade your car to a newer model that burns less fuel. Which upgrade will result in greater fuel savings:

• (A) Going from a car with 18 MPG to a car with 20 MPG?
• (B) Going from a car with 30 MPG to a car with 32 MPG?

Choice (A) results in more than twice as much fuel saved than choice (B), even though the absolute MPG difference is the same (2 MPG), because the MPG varies inversely with gallons (and therefore cost).

 

Say you drive 10,000 miles in one year, and gas costs $4/gallon.

• Choice (A)
• At 18 MPG you spend: 10000 / 18 * 4 = $2222 on gas
• At 20 MPG you spend: 10000 / 20 * 4 = $2000 on gas
• Net savings: $222 (= 55 fewer gallons of fuel)
• Choice (B)
• At 30 MPG you spend: 10000 / 30 * 4 = $1333 on gas
• At 32 MPG you spend: 10000 / 32 * 4 = $1250 on gas
• Net savings: $83 (= 20 fewer gallons of fuel)

You save more than twice as much fuel by going from 18 to 20 MPG than by going from 30 to 32 MPG.

Now let's look at the same calculation in liters/100km:

• Choice (A)
• 18 MPG = 13.06 liters/100km
• 20 MPG = 11.76 liters/100km
• Net savings: 1.3 fewer liters of fuel/100km
• Choice (B)
• 30 MPG = 7.84 liters/100km
• 32 MPG = 7.35 liters/100km
• Net savings: 0.5 fewer liters of fuel/100km

In liters/100km, Choice (A) saves more than twice as much fuel than Choice (B), because liters/100km varies proportionally with gallons (and therefore cost).

Conclusion

Equal improvements in MPG do not translate to equal savings in fuel. Upgrading an 18 MPG car to a 20 MPG car makes a lot more sense than upgrading a 30 MPG car to a 32 MPG car, even though the difference is 2 MPG in both cases. The same goes for upgrading an 18 MPG car to a 19.8 MPG car vs. a 30 MPG car to a 33 MPG car, even though the difference is 10% in both cases.

Liters/100km is unambiguous about fuel savings. Upgrading a car from 8 to 7 liters/100km will result in the same fuel savings as upgrading a car from 15 to 14 liters/100km.

By the same reasoning, the city MPG number for a car matters more than the highway MPG number. A 2013 Honda Civic driven for 5000 miles in the city (at 28 MPG) and 5000 miles on the highway (at 39 MPG) will have an overall fuel economy of 32, which is a lot closer to the city number than to the highway number.

If your goal is to reduce fuel consumption by purchasing a more efficient car, be sure to convert from MPG to liters/100km in order to get a more accurate picture!