Single-Cylinder Rich Misfire on a GDI Four
A 2018 sedan with a 2.4 L gasoline direct-injection four-cylinder, 96,000 miles, arrives with the MIL on, a rough idle, and a strong fuel smell after shutdown. Codes stored: P0302 and P0172. Long-term fuel trim is minus 24% at idle and minus 9% at 2,500 rpm; short-term trim swings between minus 3% and minus 10%. The cylinder 2 spark plug is wet and sooty black, while cylinders 1, 3, and 4 are light tan. Relative compression is within 3% across all four cylinders. A cylinder cut test produces no change in idle quality when cylinder 2 is disabled. High-pressure rail pressure drops only 6 bar in the 10 minutes after key-off. The EVAP purge valve holds vacuum and does not flow when commanded closed. The customer authorizes one repair visit and asks whether the catalytic converter should be replaced at the same time.
1. Which single fault best accounts for the trim, plug, and cut-test data, and which competing faults does the evidence eliminate? Show the reasoning chain.
Read the worked answer
Best fit: a cylinder 2 injector that is stuck partly open / leaking.
- The trims are negative (rich), and they are much worse at idle (minus 24%) than at 2,500 rpm (minus 9%). A leaking injector passes a roughly constant amount of fuel per unit time, so its effect on mixture is largest when the injector pulse width is shortest. That matches the idle-dominant pattern.
- Only cylinder 2's plug is black and wet. A fault common to the whole intake tract would color all four plugs.
- The cut test is not contradictory. Cylinder 2 is already misfiring from an over-rich mixture, so disabling its ignition changes little. A cut test is only informative on a cylinder that was contributing normally.
Eliminated by the data:
- EVAP purge valve stuck open - would enrich all four cylinders and blacken all four plugs, and the valve was verified closed.
- High-pressure pump internal leak - would bleed into the cam housing and affect all cylinders; rail pressure hold-down here is nearly spec.
- Intake vacuum leak - would drive trims positive, not negative.
- Mechanical fault (valve, piston, ring) - relative compression is within 3%.
2. The technician proposes ordering a new injector immediately. What one additional confirmation would you require, and how would a failed result change the diagnosis?
Read the worked answer
Confirm delivery, not just the symptom.
Run an injector balance / delivery test on cylinder 2 (command each injector and watch rail-pressure decay or use the manufacturer's balance routine). If cylinder 2 produces a clearly larger pressure drop than its siblings, the injector is condemned on evidence rather than on a black plug.
If the balance test is normal, the diagnosis must change. Re-open the following possibilities, in this order:
- An injector that seals but has a bad spray pattern or a sticking nozzle - it would still foul a plug without over-delivering at the test pressure. A hot soak and restart check, plus a second plug reading after a road test, separates this from a leaking seat.
- A cylinder-specific ignition or compression issue that is masking itself - re-run a running compression and check secondary ignition if accessible.
- A leaking injector seal or a rail-side leak pooling at cylinder 2 only - inspect the rail and the injector bore for fuel.
A swap of injector 2 to another cylinder with new seals, followed by a fault-follows check, is a valid fallback only if the manufacturer permits re-use and coding is available.
3. The customer asks whether the converter should be replaced at the same time. Give a decision rule and the evidence you would need before answering yes.
Read the worked answer
Default answer: no, not yet.
A converter is only condemned on evidence. There is no P0420, and a rich misfire is very often recoverable once the fuel problem is corrected and the unit is driven through a complete monitor cycle.
Decision rule:
- Repair the injector, clear the codes, verify trims return to within roughly plus or minus 10%, then run a full drive cycle and re-check.
- If P0420 returns, or if rear oxygen sensor activity is switching at the same rate as the front sensor, or if the converter's oxygen-storage check falls out of spec, then replacement is justified.
- A converter damaged by raw fuel may take several hundred miles of normal operation to clean up; condemning it on the day of repair is a guess.
Practical points:
- Document the condition and the reasoning in writing so the customer understands why the converter was not replaced today and what would trigger it later.
- If the converter fails the monitor after repair, check that the root cause is truly gone before installing a new unit - a second converter ruined by an uncorrected rich condition is a comeback you will own.