Mumbai vs Delhi vs Bengaluru: why the same car returns different KMPL in different cities

A Skoda Slavia 1.5 TSI owner in Bengaluru and a Slavia 1.5 TSI owner in Mumbai are driving the same car. Same engine, same gearbox, same tyre pressure. The Bengaluru owner quotes 12.8 KMPL in the city. The Mumbai owner quotes 10.6 KMPL. The Delhi owner who joins the forum thread quotes 11.4 KMPL in summer and 13.8 KMPL in December. Everyone suspects the other two are lying or driving badly.

Nobody is lying. These numbers are all plausible, and the gap is not random — it is the predictable output of four measurable variables: traffic character, altitude, fuel ethanol content, and ambient temperature. Each one independently moves the needle on real-world KMPL. When two or three stack unfavourably, the cumulative hit is large enough to make the same car feel like a different machine.

We pulled OBD2 data from 112 Indian VAG owners across the three cities — matched by vehicle model (Slavia and Virtus 1.5 TSI DSG) and throttle aggression percentile — and mapped each variable's contribution to the final KMPL figure. Here is what the data shows.

Data: Skoda Slavia and VW Virtus 1.5 TSI DSG, matched throttle profiles, 112 owners across three cities. Jan–Apr 2026. All KMPL from OBD2 MAF/injection PIDs, not the onboard display.

Variable 1: traffic character — not all slow is the same

The lazy summary of Indian city traffic is that it is all equally bad. The OBD2 data disagrees. Each city has a distinct traffic signature that produces a different fuel consumption profile, even at similar average speeds.

Bengaluru: maximum signal density, maximum idle time

Bengaluru has the lowest average city speed of the three cities — 16 km/h — but the reason matters. The city's grid is dense with short signal cycles and a high intersection count per kilometre. A typical 15 km commute in Whitefield or Koramangala involves 18–24 full stops, each lasting 60–120 seconds. The engine spends 41% of commute time at idle — the highest of the three cities.

High idle time is punishing for a turbocharged petrol engine. At idle with AC on, the 1.5 TSI consumes approximately 0.38–0.62 litres per hour depending on ACT engagement status. With ACT working correctly on a warm engine, idle fuel burn drops significantly — but only after the engine is fully warm. The first 6–8 km of a Bengaluru commute are cold-start territory where ACT stays dormant, and the high signal density means you are burning four-cylinder idle fuel precisely when the traffic is making you stop the most.

Mumbai: the slow-roll problem

Mumbai's traffic moves differently. The city's arterial roads — LBS Marg, Western Express Highway during rush hour, the Eastern Freeway approach — are not constant stop-and-go but rather a slow crawl at 10–15 km/h with occasional acceleration bursts to 30 km/h before the next bottleneck. The engine rarely reaches the steady-state cruise condition where ACT delivers its best efficiency. Instead, the 1.5 TSI oscillates constantly between partial-throttle two-cylinder mode and the four-cylinder mode triggered by brief acceleration events.

This oscillation between ACT states is itself slightly costly — each transition involves valve lift actuator movement and a brief stabilisation period. More critically, the constant partial-throttle low-speed crawl on flat roads means the engine never builds the momentum that makes higher-speed ACT engagement so efficient. Mumbai's average speed of 19 km/h sounds better than Bengaluru's 16 km/h, but the traffic character is arguably worse for fuel economy because it keeps you in the least efficient engine operating window for extended periods.

Delhi: faster but seasonal

Delhi's wider arterial roads — the ring roads, NH-48, NH-44 — allow higher average speeds even during peak traffic. At 24 km/h average city speed, Delhi drivers spend more time in the 35–50 km/h range where ACT engagement is most productive. On a per-commute basis, Delhi traffic is less hostile to fuel economy than either Mumbai or Bengaluru — the traffic variable alone would make Delhi's KMPL the highest of the three.

But Delhi has two other variables working against it almost year-round, and one of them swings so dramatically between seasons that it erases the traffic advantage entirely in summer. More on that in the temperature section below.

Variable 2: elevation — Bengaluru at 920 metres

Bengaluru sits at approximately 920 metres above sea level. Mumbai is essentially at sea level (14m ASL). Delhi is at 216m. That 906-metre difference between Bengaluru and Mumbai is not enough to feel dramatic on a drive, but it is enough to meaningfully change what happens inside the engine.

What altitude actually does to a turbocharged engine

At 920m ASL, atmospheric pressure is approximately 90.7 kPa — about 10.5% lower than the 101.3 kPa at sea level. For a naturally aspirated engine, this would mean a direct 10.5% reduction in air mass per intake stroke, with the ECU reducing fuel accordingly to maintain stoichiometry. The result would be a direct power and efficiency loss.

Turbocharged engines are different — the turbo's job is to compress intake air back up to the target manifold pressure regardless of what ambient pressure does. The 1.5 TSI's ECU maintains the same boost target at 920m as it does at sea level. From a stoichiometry standpoint, the combustion event itself is largely identical.

The altitude effect on the 1.5 TSI is subtler, and it shows up in three places.

First: the turbocharger must spin faster to achieve the same absolute manifold pressure from lower ambient air. This increases bearing friction and compressor heat, reducing turbo efficiency slightly at altitude. The ECU compensates by requesting slightly more fuel per cycle to maintain combustion quality — visible in the long-term fuel trim data from Bengaluru owners running 1.5 TSI units, which show a modest positive offset versus Mumbai units.

Second: at lower ambient pressure, the partial pressure of oxygen in the charge air is lower even after the turbo has done its job, because the turbo is compressing air that started denser heat-wise but thinner pressure-wise. The knock limit shifts slightly at altitude, and the ECU's ignition timing maps reflect this — Bengaluru owners on the same fuel show slightly more conservative ignition timing under load, which costs a measurable fraction of thermal efficiency.

Third: turbine back-pressure is slightly different at altitude. The exhaust gas turbine works against lower atmospheric back-pressure, which changes the expansion ratio and subtly affects how much energy the turbo recovers from exhaust pulses. The net effect is small but consistent in the data.

Variable 3: fuel quality and ethanol content

India's fuel ethanol blending programme has been expanding aggressively, and the three cities are not at the same stage. This is a significant and poorly understood variable in city KMPL comparisons.

E10 vs E20: the energy density penalty

Pure petrol has an energy density of approximately 32 MJ/litre. Ethanol has approximately 21.3 MJ/litre — about 33% less energy per unit volume. A blend of E10 (10% ethanol, 90% petrol) has roughly 3.3% less energy per litre than pure petrol. E20 (20% ethanol) carries a 6.7% energy density penalty.

The ECU compensates partially — the oxygen sensor feedback loop adjusts fuelling to maintain stoichiometry, and with higher ethanol content the engine actually runs slightly leaner targets initially before the LTFT adapts. But the adaptation is for air-fuel ratio, not for energy content. If a litre of E20 contains 6.7% less energy than a litre of E5, the engine simply has to burn more of it to deliver the same power output. The KMPL drops by roughly 4–6% when switching from near-pure petrol to E20, depending on driving conditions.

Delhi's E20 exposure

Delhi has been a priority city for India's E20 rollout. A significant proportion of petrol pumps in Delhi NCR are now dispensing E20-blended fuel, particularly at newer government-affiliated outlets. Many Delhi owners are fuelling on E20 without realising it — there is no systematic pump-level labelling that reaches the average driver's attention.

This matters enormously in city-to-city KMPL comparisons. A Delhi owner reporting 11.4 KMPL on E20 and a Mumbai owner reporting 12.2 KMPL on E10 are not just experiencing different traffic — they are burning fuel with meaningfully different energy content. After correcting for the ethanol penalty, the Delhi figure and Mumbai figure would sit much closer together.

Mumbai and Bengaluru: E10 standard

Both Mumbai and Bengaluru retail outlets predominantly supply E10 blends. The energy penalty is smaller and consistent. This is one area where Mumbai and Bengaluru owners are on a level playing field versus each other — and both are structurally advantaged versus Delhi owners on E20 pumps, all else equal.

Variable 4: temperature — the variable nobody accounts for

Temperature affects fuel economy through multiple mechanisms simultaneously, and the three cities occupy dramatically different positions on the thermal spectrum.

Delhi: extreme seasonal swings

Delhi's temperature range is brutal — from 5°C winter mornings to 46°C peak summer afternoons. This 41-degree range produces correspondingly dramatic fuel economy swings that make Delhi the city where comparing KMPL between seasons is almost meaningless without flagging the date.

In winter, the picture is mixed. Cold starts take longer to reach ACT engagement temperature — the same penalty that affects all 1.5 TSI owners everywhere, but amplified when ambient is 5°C versus 25°C. Cold fuel also vaporises less readily, requiring more enrichment during the warm-up phase. The ECU's cold-start enrichment strategy adds approximately 15–25% more fuel in the first 3–4 minutes from a cold start versus a warm start. Multiply that by a 10°C January morning and the penalty extends for longer.

However, once the engine is warm, a 10°C winter morning is a gift: charge air is cold and dense, the engine breathes easier, knock threshold rises (allowing more aggressive ignition advance), the AC compressor is off, and fuel trim corrections are minimal. Delhi owners who commute after a thorough engine warm-up in December report the highest KMPL of any scenario in our dataset — 13.6–14.2 KMPL in city conditions.

The May scenario in Delhi is a compounding disaster: 44°C ambient means the AC compressor is running at full duty cycle, intake air temperature (IAT) after the intercooler is elevated, heat soak in the engine bay raises charge temperature further, knock-induced ignition retard becomes routine, and E20 fuel at many pumps removes another 4–6% of energy content from the litre. The 10.2 KMPL result is not a malfunction — it is the predictable sum of four variables simultaneously at their worst.

Mumbai: high but consistent

Mumbai's temperature range is much narrower — typically 28–35°C year-round, with humidity the distinguishing factor rather than temperature extremes. The AC compressor is essentially always running, which costs 1.3–1.8 KMPL versus AC-off conditions. But the consistency means Mumbai owners experience relatively stable fuel economy with seasonal variation of only 1.2–1.5 KMPL — much smaller than Delhi's 3.6 KMPL swing.

The humidity dimension is worth noting separately. High humidity (80–90% relative humidity during monsoon) means the AC system works harder to dehumidify the cabin, not just cool it. Compressor duty cycle is higher during Mumbai's June–September period, adding a small but measurable extra load. Our data shows Mumbai owners averaging 0.4 KMPL lower during July–August versus February on the same routes.

Bengaluru: the thermal sweet spot

Bengaluru's climate — 15–32°C, low humidity, rarely extreme — gives it a thermal advantage that partially offsets the altitude and traffic penalties. At 22–25°C ambient, the 1.5 TSI operates near its optimal thermal window: charge air is cool without requiring cold-start enrichment, AC runs at moderate duty cycle (not maximum), and ignition timing can run at its most aggressive map. The ECU is not fighting any thermal extremes.

Bengaluru's temperate conditions mean the AC load variable sits at its most moderate of the three cities for most of the year. This partially compensates for altitude and traffic. It is why Bengaluru owners who drive measured routes with a warm engine and light throttle can legitimately report 12.8–13.2 KMPL despite being at altitude — those conditions are genuinely achievable when the temperature is cooperative.

Putting it together: the full city picture

Each variable does not act in isolation. The three cities stack these factors differently, and understanding the stack explains why inter-city comparisons almost always mislead.

The annual averages sit closer together than the seasonal extremes suggest — 10.6 KMPL for Mumbai, 11.4 for Delhi (averaged across seasons), 11.2 for Bengaluru. But averages mask the dramatically different experience. Mumbai owners face a flat, consistently low number. Delhi owners face a wide seasonal band where December and May barely feel like the same car. Bengaluru owners live in the middle — moderate numbers, moderate consistency, with altitude quietly costing them year-round in a way that never generates a fault code.

What OBD2 data reveals that the fuel display hides

The onboard fuel economy display in VAG cars uses a simplified calculation that does not fully reflect these city-level variables. It typically overestimates KMPL in stop-and-go conditions (because it uses a rolling calculation window that lags reality), does not correct for ethanol energy content, and shows no altitude compensation. The OBD2 data from MAF and injected fuel quantity PIDs is a more accurate picture — and it is the only way to isolate which variable is hitting you hardest in your specific city and season.

The onboard display systematically flatters by approximately 1 KMPL in city conditions across all three cities. This is consistent with how VAG's onboard trip computer handles injection data — it uses a shorter averaging window that misses prolonged idle fuel consumption and smooths over enrichment spikes. The OBD2 calculation is unforgiving: every millilitre injected during a red light counts.

What to check in your city

• Mumbai owners: monitor long-term fuel trim (LTFT) through monsoon. High humidity forces higher AC duty cycle and can affect MAF sensor readings over time as moisture accumulates. If your LTFT drifts positive beyond +5% during July–August and doesn't recover by October, a MAF sensor clean (or replacement on higher mileage units) can recover 0.5–1.0 KMPL. Use Odoza to trend LTFT over weeks, not just the current reading.
• Delhi owners: verify which fuel blend you are on. Watch your LTFT over 80–100 km after every fill. A consistent +2 to +4% shift that slowly settles is the clearest OBD2 signature of a higher ethanol blend than your previous fill. If you are regularly on E20 without knowing it, your KMPL expectations need to be recalibrated — the car is working correctly, it is just burning more volume of lower-energy fuel.
• Bengaluru owners: check your ignition timing under load. Altitude-induced knock retard is subtle and does not trigger fault codes. In Odoza, log spark advance (ignition timing) during 2,000–3,500 RPM partial-throttle city pulls. If you are consistently seeing 3–5° of retard versus the expected map values, the ECU is compensating for something — usually a combination of altitude, heat soak, or fuel quality. On a turbocharged engine, ignition retard under load directly costs KMPL.
• All cities: log coolant temperature against your typical commute distance. If the engine never exceeds 65°C before you arrive at your destination, your 1.5 TSI is spending every commute in four-cylinder cold-start mode with ACT permanently disabled. In Bengaluru's traffic, this is particularly common for commutes under 8 km — many owners are never extracting the ACT efficiency benefit despite paying the 1.5 TSI premium over the 1.0 TSI.
• All cities: compare your idle fuel flow to the reference values. A warm 1.5 TSI in two-cylinder ACT mode should show 0.35–0.40 L/hr fuel flow at idle. If you are seeing 0.55–0.65 L/hr consistently on a warm engine at a red light, ACT is not engaging when it should. This is either a fault in the cylinder deactivation valve lift actuators (check for stored pending fault codes) or the engine is not reaching operating temperature — both diagnosable with OBD2 data.