Highway 18, City 11: Why One Mileage Number for Your Car Is Meaningless

Published Aug 17, 2026 · Bharometer Blog · 8 min read

The ARAI figure on a car's specification sheet is not a lie, exactly. It is a measurement taken under controlled conditions: a standardised drive cycle run on a chassis dynamometer in a lab, at a fixed ambient temperature, with no AC, no fuel degradation from ethanol blends, and no driver who is stuck behind a BEST bus at Hindmata junction for twenty minutes. When your brochure says 18 km/l and your daily commute delivers 11, neither number is wrong. They are measuring two completely different things.

The problem is that most people treat the 18 as the target and the 11 as a failure. The 11 is actually your number. It is the number that should be driving every decision about running costs, resale timing, and whether that weekend Pune expressway run changes anything meaningful about your monthly fuel spend.

Why the Gap Is So Large

Fuel economy is almost entirely a function of engine load over time. An engine burning fuel to move a car at constant 80 km/h on a flat highway is in a comfortable operating band: combustion is efficient, the transmission is in a high gear, and the thermal losses are relatively low. The same engine inching through a signal-to-signal commute in Chennai's Anna Salai at 8 AM is doing something structurally different. It stops, idles, restarts under load, uses first and second gear almost exclusively, runs the AC compressor against high ambient heat, and runs the cooling fan harder because the airflow across the radiator at walking speed is negligible.

Each of those factors alone would knock a few percent off efficiency. Together, they can cut stated mileage by 30 to 50 percent. The 18 km/l highway figure dropping to 11 in city traffic is not unusual for a mass-market petrol hatchback. For a larger SUV with a heavier kerb weight and a bigger AC load, the city-to-highway ratio can be even more severe.

Here are the four biggest contributors, each with a rough sense of magnitude:

• Idling at signals: A petrol engine idling burns roughly 0.3 to 0.6 litres per hour depending on displacement and AC load, producing zero forward distance. A commute with 15 minutes of cumulative idle time can burn 0.1 to 0.15 litres doing nothing. Over 25 working days, that is 2.5 to 3.75 litres per month simply evaporated at red lights.
• Stop-start load cycling: Every time you accelerate from rest, the engine works against inertia. Fuel consumption spikes sharply during that phase. Frequent stops mean frequent spikes. On a highway, you accelerate once and coast.
• Air conditioning: AC compressor draw varies, but in a compact car in city traffic in May in Hyderabad, it can account for 1.5 to 2.5 km/l of effective mileage loss on its own. At highway speeds with the windows up, aerodynamic drag makes AC the more efficient choice. In slow city traffic, it just adds load.
• Gear range: Most petrol engines are most thermally efficient between 1800 and 2800 RPM at partial throttle in top gear. City driving keeps you in first and second, where mechanical losses in the drivetrain are also higher.

The Weekend Highway Run Does Not Balance the Week

A common belief: "My city mileage is terrible but I do a long drive every month, so it averages out." The math on this rarely works the way drivers expect.

Say you do a 250 km Bengaluru-to-Mysuru return trip once a month. At highway mileage of 17 km/l, that trip burns roughly 14.7 litres. Your monthly city commute is 800 km at 11 km/l, which burns 72.7 litres. The highway trip is about 17 percent of your total distance and 17 percent of your total litres. It moves your blended monthly average from 11.0 to roughly 11.8 km/l. Noticeable, but your real operating cost is dominated by the city portion, not the weekend run.

This matters when you are thinking about which vehicle to own. A car with strong city fuel economy and modest highway efficiency will almost always deliver a better cost-per-kilometre outcome for an urban driver than the reverse. The ARAI highway figure is largely irrelevant to that calculation.

The calculation that does matter is cost per kilometre across all your actual driving, which requires knowing your real blended number from real fill-ups, not a lab figure.

How to Separate Your City and Highway Numbers

Most drivers have a rough intuition that their mileage is better on trips. Very few have actually measured it in a way that separates the two. You need tank-to-tank fill-up data tagged by the type of driving in that tank.

The practical method: treat any fill-up that follows a predominantly highway drive as a highway measurement, and any fill-up that closes a city-only tank as a city measurement. You do not need GPS tagging or sophisticated route analysis. You need consistent notes on what the tank contained.

After four or five of each type, you will have enough data to see your actual city average and your actual highway average. For most Indian urban drivers, the highway number runs 40 to 60 percent higher than the city number. If your gap is smaller than that, either your city traffic is unusually light or your highway speed habits are aggressive enough to close the gap from the other direction.

High highway cruise speeds hit aerodynamic drag hard. Fuel consumption at 120 km/h is meaningfully higher than at 90 km/h on the same road. Drivers who push 130 on the expressway sometimes find their highway mileage is not much better than their city figure, because they have traded signal-stop losses for aerodynamic losses. The efficient highway speed for most Indian petrol cars is in the 80 to 90 km/h range, which is not how most people actually drive once a highway opens up.

E20 and the City-Highway Split

Since April 2026, all petrol at Indian retail pumps is E20. The blend's effect on mileage is not uniform across driving conditions. Ethanol has a lower energy density than pure petrol, and that gap shows up more acutely in low-efficiency operating modes. City driving, which is already thermally inefficient, will tend to show a slightly larger absolute drop from E20 than highway driving will.

Controlled tests published in July 2026 measured 2 to 6 percent higher fuel consumption on E20 versus E10 across vehicles in the 3-to-10-year age range. That range reflects vehicle age and calibration, but it also reflects usage pattern: a driver who does 90 percent city commuting is likely to sit toward the higher end of that range, while a frequent highway user may sit toward the lower end. You cannot know which end you are on from the published averages. You can only know from your own fill-up log, which is the point.

If you have been logging fill-ups in Bharometer through 2025 and into 2026, you already have a before-and-after record. Your petrol tracking in the app is E20 tracking by definition, because there is no other petrol at the pump. What the log shows you is what the blend actually does in your specific vehicle under your specific driving pattern, which is more useful than any fleet-average figure.

The Number Worth Optimising Is Cost Per Kilometre, Not km/l

Once you have separated your city and highway mileage, the next step is translating both into rupees-per-kilometre. Mileage in km/l is a rate; cost per kilometre is what you actually spend.

At a petrol price of ₹96/litre (roughly what several major cities were showing in mid-2026), a city average of 11 km/l translates to ₹8.73 per kilometre. A highway average of 17 km/l translates to ₹5.65 per kilometre. The highway kilometre costs 35 percent less. That gap is real money on a long trip but it does not change what your month costs, because the month is mostly city kilometres.

Knowing this lets you make accurate decisions: when evaluating a vehicle upgrade, you compare city cost-per-km, not highway ratings. When planning a road trip budget, you use highway cost-per-km, not the city average. When calculating monthly fuel spend, you weight by the kilometre split, not by the number that appears on the manufacturer's sticker.

A useful cross-reference for the underlying method: the tank-to-tank measurement guide covers why partial-tank readings and dashboard displays consistently mislead, and why the fill-up-to-fill-up method is the only baseline you can trust. City versus highway analysis only works if the raw numbers feeding it are solid, and that means proper tank-to-tank discipline.

Logging This Without Adding Friction

The barrier to actually building a city-versus-highway data set is almost never motivation. It is the friction of recording each fill-up consistently. Most drivers intend to track, then forget to note the odometer reading, then guess, then give up because the guessed data feels unreliable.

Bharometer's pump-scan feature removes most of that friction. You photograph the pump display after filling, and the app reads the litres dispensed, the amount paid, and the rate per litre directly from the screen. You add the current odometer reading (or let the app calculate from the previous entry), and the fill-up is logged. The whole sequence takes well under a minute at the pump.

From that log, the app computes your tank-to-tank mileage for each fill-up, your running monthly spend, and your cost per kilometre. What it gives back is a record you can actually use: not a single number, but a series of numbers that shows you whether the tank you just closed was a city tank or a highway tank, and what each type of driving is actually costing you over time.

Bharometer runs on iPhone (iOS 16 and above) and has a free version available on Android as well. There is no account, no server, and no data leaving your device.

The ARAI number on the brochure will not change. Your real numbers will, every month, depending on traffic, temperature, AC use, load, and road type. Measuring them is the only way to know what your car actually costs to run, and whether anything you do about it is working. The 18 is a ceiling built in a lab. The 11 is your floor. What you do between those two numbers is your actual driving life.

Keep reading

• All blog posts
• Fuel & mileage guides