How to measure your own fuel economy and actually change it
Where the official consumption figure comes from, why your own number is different, and a repeatable way to measure what your car really does on your roads.
Every driver notices it eventually. The brochure said one thing, the trip computer says another, and the receipts say something else again. The gap is real, it is normal, and it is mostly not the car's fault.
An official consumption figure exists so that two cars can be compared with each other. It is a measurement taken under identical, repeatable conditions, which is exactly what makes it useful for comparison and useless as a prediction. Nobody commutes through a laboratory.
What the official number measures
The tests are standardized driving cycles: a defined pattern of accelerations, steady speeds and stops, run on rolling road equipment in a temperature-controlled building, with the vehicle prepared in a specified state. Europe replaced its older cycle with a longer, more demanding one some years ago; the United States uses several cycles, including cold-start and air-conditioning runs, and applies correction factors before publishing. The details differ by region, and figures from different regimes are not comparable with each other.
What every version has in common is the removal of everything variable. No weather, no traffic, no hills, no crosswind, no roof box, no four passengers and a dog, and no you. That is the point. It is also the entire explanation for the gap.
Where your miles differ
| Cold starts | The first miles are the thirstiest, and a short trip can be all of them |
|---|---|
| Speed | Aerodynamic drag rises with the square of speed, and the power to push through it rises faster still |
| Traffic | Every acceleration buys energy that the next brake application throws away as heat |
| Terrain and wind | Climbing and headwinds are genuine work; a route can cost more than a driver does |
| Load and shape | Passengers, cargo, roof boxes and bike racks charge you continuously |
| Temperature | Cold air, thick fluids, seasonal fuel blends and everything electrically heated |
Cold starts deserve the top of that list because they dominate for so many people. Until the engine reaches operating temperature it runs a richer mixture, the oil is thick, and the catalyst is not yet doing its job. A car that only ever does four-mile trips may never once complete its warm-up. Two five-mile journeys cost noticeably more fuel than one ten-mile journey covering the same ground.
Speed is the other heavyweight, and it is not linear. Drag force grows roughly with the square of speed, so the power needed to overcome it grows roughly with the cube. Going faster costs more than most people expect, and the penalty accelerates. This is why a steady highway cruise a few miles per hour slower shows up on the receipt in a way that careful town driving rarely does.
The quiet leaks
- Tires below the pressure on the door placard, which is the most common and the easiest to fix
- A roof box or roof bars left on for the eleven months of the year you are not going skiing
- A trunk being used as long-term storage for things that live in a garage
- A neglected air filter, which matters more on older engines than on newer ones
- A brake that drags — suspect it if one wheel is much hotter than the others after a drive
- Wheel alignment out of specification, which scrubs energy away and eats tires at the same time
- Long idling warm-ups, which burn fuel while warming the engine more slowly than gentle driving does
Individually most of these are small. Together they are the difference between a car that roughly meets expectations and one that never does. Some are free and take five minutes; a dragging brake and an alignment problem are shop jobs, and the brake in particular is worth attending to promptly for reasons beyond fuel.
Measure your own, properly
Trip computers are convenient and generally optimistic. If you want a number you can act on, do it with the pump and a calculator.
- Fill the tank at a pump you can return to, and stop at the first automatic click rather than topping off
- Reset the trip meter to zero at the pump, not when you remember later
- Drive normally for a full tank — at least a couple of hundred miles, and no special effort
- Refill at the same pump, in the same way, to the first click
- Divide the distance on the trip meter by the fuel dispensed to refill it
- Repeat for three tanks and average them, because a single tank is mostly noise
Now you have a baseline, and a baseline is what makes experiments possible. Change one thing at a time — the highway speed you settle at, the route to work, the pressure in the tires, the roof bars — and measure another full tank. One variable, one tank, one comparison. Anything less controlled tells you nothing you can trust.
The habits that actually move the number
- Look far enough ahead that you lift off instead of braking; energy you do not scrub off is energy you do not have to buy back
- Get to your cruising speed without drama, then hold it steady rather than surging and lifting
- Use the highest gear the engine is comfortable in, without letting it labor
- Chain errands into one warm trip instead of three cold ones
- Take the roof box off the day you get home
- On the highway, use the air conditioning rather than open windows; around town it hardly matters
What is not on that list, deliberately: coasting in neutral, switching the engine off on the move, and tucking in behind trucks. These circulate as economy tricks, they compromise your control of the car for a trivial saving, and some of them are illegal where you live. The economical way to drive is the smooth way, which happens to be the safe way too.
When the gap is a symptom
A gradual drift downward as a car ages is ordinary. A sudden change is worth investigating. Consumption that jumps by a noticeable margin over a couple of tanks, with no change in your driving, points at something specific: a thermostat stuck open, so the engine never reaches temperature and the heater stays lukewarm; a sensor giving the engine bad information; a dragging brake; a fuel leak, which usually announces itself by smell.
If a warning light has appeared, have the stored fault codes read before guessing. An engine running on a substitute value because one sensor stopped reporting will happily use extra fuel for months without any other symptom.
Run three tanks brim to brim before you conclude anything at all. That number is the one spending your money, and it is the only one worth arguing with.