km/kg, Not km/l: Tracking a CNG or Bi-Fuel Car Without Mixing Up the Fuels

Published Aug 26, 2026 · Bharometer Blog · 7 min read

CNG has its own physics. Compressed natural gas is sold by the kilogram, not the litre, because gas density varies with temperature and pressure. When you fill up at a CNG station, the dispenser shows kilograms dispensed and rupees charged. Your mileage from that tank is therefore expressed as kilometres per kilogram: km/kg. It is not comparable to km/l without a conversion that almost nobody applies correctly.

For a dedicated CNG vehicle, this is straightforward once you understand it. The problem compounds on a bi-fuel car: those run CNG for the bulk of city driving, then switch to petrol when the CNG tank empties or when you're on a highway where CNG stations are sparse. You now have two fuels, two units, two efficiency numbers, and two ₹/km costs. Log them in a single bucket and the number you get is meaningless.

Why km/kg Is the Right Unit and What It Actually Tells You

One kilogram of CNG contains roughly 52–55 MJ of energy. One litre of petrol contains roughly 32–34 MJ. CNG is more energy-dense per kilogram, which is why a car doing 15 km/l on petrol might do 22–25 km/kg on CNG. Those two numbers are not the same quantity; you cannot compare them directly or average them together.

What you can compare is cost per kilometre: ₹/km. That is the single number that tells you which fuel is actually cheaper for your specific vehicle on your specific routes. To get ₹/km for CNG, you need price per kg (which varies by city) and your km/kg. To get ₹/km for petrol, you need price per litre and your km/l. Only when you compute both separately do you have the numbers to make decisions.

A worked example with real-ish numbers for Delhi in August 2026:

• CNG: say ₹76/kg, vehicle doing 22 km/kg. Cost per km: ₹76 ÷ 22 = ₹3.45/km.
• Petrol (E20): say ₹95/litre, vehicle doing 13 km/l in city. Cost per km: ₹95 ÷ 13 = ₹7.31/km.

That gap is why people retrofit or buy CNG. But the only way to know if your car is actually delivering 22 km/kg or has drifted to 18 km/kg is to log it tank-to-tank, every fill-up, separately for each fuel.

The Bi-Fuel Tracking Problem in Practice

Here's what happens without a system. You drive mostly on CNG. One evening the CNG runs out mid-highway so you switch to petrol, put in 5 litres to reach the next CNG station, then top up CNG. You now have a petrol fill-up sitting inside what you're mentally calling "a CNG period." If you log a single km/l or km/kg across that mixed stretch, neither number is right.

The discipline that fixes this is simple but requires consistency: log every CNG fill-up as a CNG entry and every petrol fill-up as a petrol entry, against the current odometer reading at the time of that fill-up. Don't wait until both tanks are empty to record anything. Each fuel has its own fill-up history; each produces its own tank-to-tank efficiency calculation. The odometer keeps climbing through both, and the software does the attribution.

This is not how most generic mileage apps work. Apps that assume a single fuel type per vehicle, or that only track litres, break immediately when you add CNG in kilograms. You end up on spreadsheets, or you give up and just watch the fuel gauge.

CNG Tank-to-Tank: The One Gotcha That Kills Your Numbers

CNG tanks don't have a truly reliable "empty" indicator the way a petrol gauge shows a quarter-tank. The pressure gauge drops, but temperature affects it. In practice, many drivers refuel CNG before the tank reads empty: they stop at a CNG station of opportunity rather than running the tank to a consistent level.

This matters because tank-to-tank mileage assumes you refill to the same level each time. If you fill to 4 kg one visit and 6 kg the next (because the queue was short), the km between those fills divided by the kg added can be misleading if your actual consumption crossed an inconsistent boundary.

The most reliable method mirrors what you'd do with petrol: fill to full (or as close as the station will go, typically until the dispenser auto-stops on high pressure), log the kg and the odometer, drive, fill again to the same full point, log again. The km driven divided by the kg added for the second fill is your actual km/kg for that stretch. The same method explained in detail for petrol in the tank-to-tank method guide applies here with kg substituted for litres.

For bi-fuel cars, the additional step is tracking which kilometres were driven on which fuel. If you switched mid-tank, the petrol odometer segment and the CNG odometer segment need separate accounting. In practice, most bi-fuel drivers start a journey on CNG and switch to petrol only when CNG depletes; logging the odometer at each fuel switch builds a clean record.

Reading Your Real ₹/km Across Both Fuels

Once you have separate per-fill records, the useful number to watch monthly is weighted average ₹/km across both fuels combined. That accounts for the reality that you probably drove 1,200 km on CNG and 200 km on petrol in a given month, so petrol's higher ₹/km matters proportionally less.

Compute it like this:

1. Total CNG spent that month (₹) divided by CNG kilometres: CNG ₹/km.
2. Total petrol spent that month (₹) divided by petrol kilometres: petrol ₹/km.
3. Total fuel spend (₹) divided by total kilometres: blended ₹/km for the month.

The blended number is what to compare month-over-month. If it climbs, check whether it's the CNG efficiency dropping (km/kg fell), the petrol efficiency dropping (km/l fell), CNG price rising, or petrol price rising. You can isolate each variable only because you logged them separately.

A city like Mumbai where CNG queues in some areas run long during peak hours pushes more drivers toward petrol top-ups than the same driver would use in Pune. Your blended ₹/km in August (monsoon, heavier traffic, more AC) will typically be higher than March. That seasonal context matters when you're assessing whether a number is normal or a sign of something mechanical.

What Bharometer Does Differently for CNG Vehicles

Bharometer handles CNG as a first-class fuel type, with mileage tracked in km/kg and cost calculated in ₹/kg. When you add a fill-up for a CNG-capable vehicle, you enter kilograms dispensed (or scan the pump screen and let the app read the figure), the amount paid, and the current odometer. The app calculates km/kg for that tank and ₹/km for that fill-up, separately from any petrol entries on the same vehicle.

For bi-fuel cars, both fuel types sit on the same vehicle record. The log shows CNG entries and petrol entries in chronological order, each with their own efficiency figure. Monthly summaries break out CNG spend vs. petrol spend, CNG km/kg vs. petrol km/l, and the blended ₹/km for the month. You can see at a glance whether a month's higher fuel cost came from more petrol use or from a drop in CNG efficiency.

The pump-scan feature, covered in the fill-up scanning walkthrough, works for CNG dispensers that show the reading on a screen: photograph the display, and Bharometer reads the kg, amount, and rate fields automatically, the same way it handles petrol pumps. This matters at busy CNG stations where you're often rushing through the fill before the car behind you honks.

No account is required. Data stays on your device. CNG prices in your city are available in the app so you can cross-check the rate at the pump against the day's official price, which at CNG stations can sometimes vary in ways that petrol pricing doesn't.

When the km/kg Number Should Worry You

CNG efficiency drops are worth watching for a specific reason beyond just fuel cost: the CNG system itself, particularly the pressure regulator and injectors on a sequential CNG kit, can degrade without obvious drivability symptoms. A 10–15% fall in km/kg that isn't explained by heavier traffic or more AC use points toward a system that needs a service check.

The same early-warning logic that applies to petrol mileage, described in the mileage drop as a service signal post, applies here. The difference is that CNG system issues are often missed because the car still runs fine on petrol. If you never tracked CNG separately, you'd never see the CNG efficiency slide. You'd just keep switching to petrol earlier in the day than you used to and attribute it to "the queue was long."

A log catches this. Three consecutive CNG fill-ups where km/kg has dropped from 23 to 19 is a clear prompt to check the kit. No single fill-up tells you that; only the log does.

Setting Up Your Vehicle Correctly

In Bharometer, when you add a CNG or bi-fuel vehicle, select CNG as the fuel type (or petrol+CNG for a bi-fuel). From that point on, CNG fill-ups are logged in kg and petrol fill-ups in litres. The app keeps them in separate efficiency tracks but on one vehicle record so your monthly cost view combines both correctly.

If you've been logging a bi-fuel car as petrol-only and adding CNG fills as litres (a common workaround people use with generic apps), the historical data will be wrong. The cleaner fix is to start a fresh vehicle record set up correctly and begin a clean log from now. A month of correct data gives you a real baseline. Six months gives you seasonal comparison. That's the data that actually tells you what your car costs.

For a broader look at how ₹/km math works across different scenarios, the CNG vs. petrol cost-per-km breakdown covers the full comparison including how the gap changes with city-level CNG pricing.

Keep reading

• All blog posts
• Fuel & mileage guides