Fuel consumption is up: telling normal variation from a real fault in ten minutes

“Consumption has gone up” is the hardest complaint to verify, because it has the most innocent explanations. Before hunting a fault, rule out the things that are the explanation.

What explains the difference with nothing broken

Factor	Consumption increase
Winter versus summer	+15…30 %
City versus highway	+30…50 %
Air conditioning	+5…15 %
Roof rack	+10…20 % on the highway
Tyres 7 psi under-inflated	+3…5 %
Driving style	up to +30 %
Trips under 3 miles	+20…40 %

Stack two or three of those and you have a third more consumption with no faulty component at all.

So the first question is not “what broke” but “what am I comparing against”. Comparing winter city driving with summer highway driving is meaningless.

How to measure properly

The trip computer shows a calculated figure, and it tends to be optimistic. There is one reliable method:

1. Fill to the click at a particular pump.
2. Reset the trip meter.
3. Drive at least 200–250 miles in your normal pattern.
4. Fill to the click again at the same pump.
5. Divide.

A single measurement means nothing — you need at least two or three consecutive ones.

Three parameters that answer the question

If consumption is still high after filtering out normal factors, look at these — in this order.

1. Coolant temperature

The cheapest and most common cause. If the sensor lies and reports a cold engine, the module enriches continuously “for warm-up”. Consumption rises 20–30 % and the lamp may never light.

Normal warm reading is 185–205 °F in live data. Covered in the temperature sensor guide.

Same category: a thermostat holding 160 °F, P0128 — the engine never reaches its designed operating regime.

2. Fuel trims

Long-term trim shows how far the module permanently deviates from its base calculation.

• Steady positive — something is adding air or fuel is short: a vacuum leak, a dirty airflow meter, a weak pump. Covered in P0171.
• Steady negative — too much fuel: an injector, pressure, the temperature sensor. Covered in P0172.

Reading the numbers themselves is in STFT and LTFT.

3. Oxygen sensors

An aged sensor responds slowly and the module constantly overshoots. There may be no codes at all — P0420 arrives later, once the catalyst has suffered.

Watch the upstream sensor signal in live data: a healthy one oscillates several times per second, a lazy one noticeably less. Covered in its own article.

What else raises consumption with no codes

These produce no faults at all, which is why they get checked last — wrongly:

• A clogged air filter. The engine is choked and the module compensates.
• Spark plugs past their service life. Misfires are not logged yet, but combustion is already incomplete.
• A seized caliper. The car drives permanently part-braked. The sign is one hot wheel after a drive — see brake noise.
• A dead knock sensor. The module permanently retards timing for safety — plus 5–10 % consumption. Covered in P0325.
• Tyre pressure.

A ten-minute sequence

1. Compare against the right period — same season, same route.
2. Coolant temperature in live data when warm.
3. Long-term trim at idle and at 2500 RPM.
4. Upstream oxygen sensor signal.
5. Knock correction under load.
6. Read pending codes — they may hold what has not surfaced yet.
7. Check tyre pressure and air filter by hand.

Steps 2 and 3 close most genuine cases.

In short

• Season, route, and style explain up to a third of the difference — filter those first.
• Measure only by “full tank, 250 miles, full tank”, twice in a row.
• A lying temperature sensor is the cheapest and most common real cause.
• Steady positive trim means air; steady negative means fuel.
• A dead knock sensor and a seized caliper produce no codes at all.

Download the app — temperature, trims, oxygen sensor signal, and instant consumption are read with an ELM327 adapter over Bluetooth SPP on Android.