Low power and P0299 after a replacement turbo: leak, restriction or sensor?
A 2018 2.0 L diesel van, 148,000 km, arrives with intermittent power loss, a stored P0299 (turbo underboost) and a pending P2002 (DPF efficiency below threshold). The turbocharger was replaced 8,000 km ago at another workshop and the customer refuses to pay for a second one. A logged full-throttle run in third gear at 2,000 rpm gives: desired MAP 1,780 mbar absolute, actual MAP 1,320 mbar absolute, MAF 46 g/s; the calculated reference at the desired boost is 62 g/s. VGT actuator is commanded 78% with feedback tracking within 2%; a smoke test at 0.3 bar shows no external leak; there is a light oil film in the intercooler outlet hose. The air filter is new, DPF soot mass 9 g, ash 42 g, and differential pressure 14 mbar at idle. Ambient 28 °C, altitude 40 m. The workshop allows 1.5 hours of diagnosis before any parts are authorised.
1. Use the MAF and MAP readings together to decide whether the fault is a charge-air leak. Justify your conclusion with numbers.
Read the worked answer
Estimate the airflow the engine should swallow at the boost it actually achieved. A 2.0 L four-stroke at 2,000 rpm pumps 0.0333 m³/s. At 1,320 mbar absolute and about 40 °C after the intercooler, ρ = 132,000 / (287 × 313) = 1.47 kg/m³; mass flow = 1.47 × 0.0333 = 49 g/s; at a volumetric efficiency near 0.95 that is about 46 g/s - exactly what the MAF reports.
Repeat at the desired 1,780 mbar: ρ = 1.98 kg/m³, mass flow 66 g/s, and with the same efficiency about 62 g/s, which is the reference value.
What that proves. If air escaped after the MAF, the MAF would report more air than the manifold receives - a high MAF for the boost seen. Here MAF and MAP are mutually consistent, so the engine is genuinely under-boosted and under-breathing, not leaking. The 0.3 bar smoke test proves little either way: a split hose or intercooler seam often only opens at 1.3 bar, which is why it finds nothing.
Conclusion: suspect a real boost deficit - exhaust restriction, pre-MAF intake restriction, sensor bias or a turbo not being driven - not a charge-air leak.
2. List the next measurements in priority order, with the acceptance criteria and what each rules out.
Read the worked answer
- Exhaust backpressure at the pre-DPF port under full load, against the manufacturer's limit (typically well under 200 mbar). Idle Δp of 14 mbar and an ash load of 42 g look acceptable, but only a loaded test reveals a restriction. A blocked DPF or collapsed catalyst produces low boost with low MAF - exactly this signature.
- Intake restriction - vacuum at the turbo inlet or pressure drop across the airbox, usually limited to about 50 mbar. This catches a delaminated hose, a collapsed intake or a foreign object upstream of the MAF, which gives the same low-flow, low-boost pattern.
- Verify the MAP sensor against a known-good gauge at idle and at full load. A biased sensor makes a healthy turbo look weak, and it is the cheapest possibility to eliminate.
- VGT vane travel - drive the actuator end to end and confirm the mechanical stops. Feedback tracking within 2% proves the actuator only; it does not prove the vane ring inside a rebuilt unit is free or correctly calibrated.
Only after these should the turbocharger come off. Each test above eliminates a whole group of causes, which is what a 1.5-hour budget demands.
3. The customer demands you back a warranty claim against the previous workshop. What do you record and tell him?
Read the worked answer
On the job card: the raw measurements, including the calculated airflow figures, the smoke-test pressure used and its limitation, the backpressure, intake and sensor results, and the conclusion - the turbo is not yet proven defective, and the under-boost has several possible causes including exhaust restriction and a pre-MAF intake fault.
To the customer: do not promise a claim against the other workshop. Explain that the previous repair has not been shown to be faulty; a correct replacement unit can still be mis-calibrated, fitted with a restricted oil feed, or damaged by debris from the intake, and 8,000 km is too short to blame the driver either. Give him the measurement sheet so he can raise it with the other shop if he chooses.
Commercially: sell the diagnosis as a priced, authorised sequence - backpressure and intake first, sensor verification second, turbo removal last. Get written authorisation before parts are ordered. Also flag now that if the compressor is damaged by a foreign object, the intake path must be repaired as well, or the next unit will fail the same way and the argument will repeat.