Fuel economy calculator: mpg, L/100 km, km/L and cost per mile

Fuel economy is one measurement expressed in at least four incompatible units. The same car can be described as 30 mpg in the United States, 36 mpg in the United Kingdom, 7.84 liters per 100 km across most of Europe, and 12.75 kilometers per liter in parts of Asia and South America -- all without a drop of fuel changing hands.

This calculator takes a distance and the fuel used to cover it, in either unit system, and returns every one of those figures plus the cost per mile, the cost per 100 miles and an annual fuel cost at a reference 12,000 miles a year.

The arithmetic is a single division. The hard part is getting a trustworthy pair of numbers to divide, which is what most of this page is about.

#The formula

Formula: mpg = miles driven / US gallons used, and in metric L/100 km = (liters used / kilometers driven) x 100. The two are linked by L/100 km = 235.215 / mpg.

Every symbol:

• miles driven -- odometer or trip-meter distance between two fills.
• gallons or liters used -- the amount added at the second fill, which is exactly what the car burned since the first one.
• mpg -- miles per US gallon.
• L/100 km -- liters consumed per 100 kilometers.
• km/L -- kilometers per liter, which is 100 / (L per 100 km).
• imperial mpg -- miles per imperial gallon, which is US mpg x 1.20095.
• price -- what you pay per gallon or liter, used only for the cost lines.

Cost follows directly: cost per mile = price per gallon / mpg, and annual cost = 12,000 / mpg x price.

#A worked example, done by hand

You reset the trip meter at a fill, drive 312 miles, and the next fill takes 10.4 gallons. The pump price on that day is $3.40, used here as an example.

1. Economy: 312 / 10.4 = 30.0 mpg.
2. Imperial mpg: 30.0 x 1.20095 = 36.03 mpg (imperial).
3. Liters per 100 km: 235.215 / 30.0 = 7.84 L/100 km.
4. Kilometers per liter: 100 / 7.84 = 12.75 km/L.
5. Cost per mile: 3.40 / 30.0 = $0.113.
6. Cost per 100 miles: 0.113 x 100 = $11.33.
7. Annual cost at 12,000 miles: 12,000 / 30.0 x 3.40 = $1,360.

Every one of those describes the identical performance. Only the units differ.

#How to measure it: the fill-to-fill method

The method is simple and the discipline is everything.

1. Fill the tank and stop at the first automatic click of the nozzle. Do not add more by hand.
2. Reset the trip meter, or write down the odometer reading.
3. Drive normally for at least 200 miles, ideally a mix of your usual conditions.
4. Refill at the same pump if you can, and again stop at the first click.
5. Divide the distance by the volume added. That volume is precisely what the engine consumed.

The nozzle rule matters more than anything else on the list. Fuel tanks have irregular shapes and long filler necks, so "full" depends entirely on how you stop filling. Topping off by hand at one fill and clicking off at the next injects an error straight into the numerator of your division.

Distance is the smaller worry: odometers are usually accurate within a couple of percent, and non-standard tire sizes are the main cause of drift.

Why one short tank is unreliable. Suppose the true economy is 30 mpg and your fill-level discipline is off by half a gallon in either direction. Here is the range of answers that produces:

A half-gallon slip over 150 miles moves the answer by six mpg -- larger than most of the differences people are trying to detect. The same slip over 400 miles moves it by a third as much, because the error is fixed while the denominator grows.

Averaging three or four consecutive tanks does the same job. The fill errors partly cancel, and you get a figure stable enough to plan with, which is the input the fuel cost calculator needs.

#Where each input comes from

Distance traveled. A trip meter reset at the pump is the cleanest source. If you forgot, subtract odometer readings. Do not use a mapping app's route distance -- it does not include your detours.

Fuel used. The pump total on the receipt at the second fill. Not the tank capacity, not an estimate, not what the fuel gauge suggests.

Fuel price. The price on that receipt, for the grade you buy. It only affects the cost lines, never the economy figures.

Unit system. Match it to how you measured. Kilometers with gallons produces a meaningless result.

The output is far more sensitive to the fuel figure than to the distance, because the fuel number is small and a fixed measurement error is a large fraction of it. Half a gallon out of 10 is a 5 percent error; two miles out of 312 is 0.6 percent.

#US gallons versus imperial gallons

A US gallon is 3.785411784 liters. An imperial gallon is 4.54609 liters -- about 20 percent larger. Both are called a gallon, and neither is a rounding of the other.

The consequence is that the same car "gets better mpg" in the United Kingdom without any change whatsoever. Miles per gallon means miles per larger container, so the number rises:

imperial mpg = US mpg x (4.54609 / 3.785411784) = US mpg x 1.20095

Thirty US mpg is 36.03 imperial mpg. Running that backward, a British review quoting 45 mpg is describing 37.5 US mpg. Comparing a US figure with a UK figure without converting overstates the British car by about a fifth, which is the single most common error in cross-market vehicle comparisons.

The metric constant changes too. For imperial gallons it is L/100 km = 282.481 / imperial mpg, not 235.215.

#Where the constant 235.215 comes from

It is not arbitrary; it falls straight out of the unit definitions.

Start with 1 mpg, meaning one mile traveled per US gallon burned. Substitute the exact conversions: one mile is 1.609344 km, and one US gallon is 3.785411784 liters. So 1 mpg is 1.609344 km per 3.785411784 liters.

Invert to get consumption:

liters per km = 3.785411784 / 1.609344 = 2.3521458

Multiply by 100 to get liters per 100 km:

235.21458 liters per 100 km, at 1 mpg

Because consumption is inversely proportional to efficiency, any other value scales the same way:

L/100 km = 235.215 / mpg

The relationship is symmetric -- mpg = 235.215 / (L/100 km) -- which is why one constant converts in both directions. Kilometers per liter is a third form: km/L = 100 / (L/100 km), or equivalently mpg / 2.3521458.

#Trip computers and official ratings

Dashboard readouts usually read high. They estimate fuel flow from injector pulse width and other modeled quantities rather than measuring liquid volume, and the calibration tends to be optimistic. A readout two to five percent above the fill-to-fill figure is common. The instantaneous display is even less useful; it responds to throttle position more than to anything you can act on.

Trip computers are still worth watching for changes. A readout that is consistently optimistic by the same margin is a perfectly good relative instrument, and you can calibrate it once against a few measured tanks.

Official ratings are test results, not predictions. Ratings come from standardized laboratory cycles designed for comparability between vehicles, with defined speed traces, ambient temperature and no accessory load. Combined figures are a weighted blend of a city cycle and a highway cycle.

Real driving departs from the cycle in ways the cycle cannot capture: sustained speeds above roughly 55 mph, cold starts on short trips, hills, roof racks, heavy traffic and air conditioning. Use official ratings to compare one car with another -- that is what they are built for -- and your own measurement to predict your own costs.

#Turning economy into cost

Economy is only interesting once it becomes money. All figures below use an assumed $3.40 a gallon.

Read the L/100 km column rather than the mpg column when comparing. The step from 20 to 25 mpg cuts consumption by 2.35 L/100 km; the step from 45 to 50 cuts it by 0.53, less than a quarter as much, despite being the same five-mpg move. Fuel cost tracks consumption, not efficiency.

The annual column assumes 12,000 miles as a reference. Scale it to your own mileage: at 18,000 miles multiply by 1.5. For a car used mostly to commute, the commute cost calculator builds the same per-mile figure into a fuller picture, and the EV versus gas comparison restates it in cents per mile against electricity.

#Using economy tracking to spot a problem early

A logged economy figure is a cheap diagnostic. Fuel consumption responds to almost everything mechanical, often before anything else is noticeable.

A sustained drop of 10 percent or more, once weather and driving pattern are ruled out, is worth investigating. Common causes include underinflated tires, a dragging brake caliper, a clogged air filter, a failing oxygen sensor or thermostat, and simple added drag from a roof rack left on.

Three cautions when reading a drop:

• Seasonality is real. Winter economy falls for reasons that have nothing to do with the car: cold starts, denser air, winter fuel blends, heating and defrosting load, and more idling.
• Compare like with like. A month of highway driving followed by a month of city errands will show a large drop that means nothing.
• One tank is noise. Use a rolling average of three or more, for the reasons in the spread table above.

Logging tank by tank -- distance, volume and price -- takes a few seconds per fill and builds a baseline that makes a genuine change obvious.

#What this model leaves out

• Driving conditions. A single figure averages whatever mix of driving happened between the two fills.
• Load and passengers. Extra weight and roof loads are not separated out.
• Weather and season. No temperature adjustment.
• Fuel grade and blend. Ethanol content varies by region and season, and blends carry slightly different energy per gallon.
• Odometer error. Non-standard tire sizes shift the distance reading.
• Idling time. Fuel burned while stationary lowers the average with no distance to show for it.
• Everything that is not fuel. Tires, maintenance and depreciation are excluded; the cost per use calculator is one way to spread those.
• Price movement. The annual figure holds your entered price constant for a whole year.

#Common mistakes

Using tank capacity instead of fuel added. The tank is never empty at a fill. The pump total is the only figure that reflects what was actually burned.

Inconsistent filling. Clicking off at one fill and topping up at the next is the largest single source of error in the whole method.

Measuring over too short a distance. Under 200 miles, a small fill error swamps the result.

Comparing US and imperial mpg directly. They differ by a factor of 1.201, and the imperial figure always looks better.

Trusting the dashboard readout as an absolute. It is a good relative instrument and an optimistic absolute one.

Reading one bad tank as a fault. Weather, traffic and route mix explain most single-tank swings.

Mixing units. Kilometers divided by gallons is not a fuel economy figure in any system.

Every driving tool on the site is listed on the calculators index.