Both Banks Lean on a V6: Fuel Trim Pattern Versus the MAF Suggestion

Reject the MAF as the primary cause, based on the shape of the trim curve rather than any single number.

The key discriminator is where in the rpm band the trims are worst:

• +22 percent at idle, +8 percent at 2,500 rpm. This is the classic signature of unmetered air entering downstream of the throttle plate. A fixed vacuum leak is a large fraction of total airflow at idle and a small fraction at higher rpm, so trimming converges as rpm rises. The direction of the change is what matters, not the absolute magnitude.
• A genuinely under-reporting MAF produces lean trims that are roughly proportional to load at all speeds (often a bit worse at higher flow), because the same percentage error is applied to every airflow sample. You would expect something like +20 percent at idle and +18 to +20 percent at 2,500 rpm. That is not what this data shows.
• MAP corroborates a leak. 14.6 in-Hg absolute is only about 15 in-Hg of vacuum at idle - low. Manifold vacuum is lower when extra air bypasses the throttle. The 29.4 in-Hg key-on reading confirms the sensor and its reference are sound.
• MAF grams per second. 3.4 g/s at idle on a 3.5 L V6 is in the normal range, and so is roughly 14 g/s at 2,500 rpm no-load. If the sensor were skewed low, you would expect both trim and g/s to be visibly off.
• Fuel pressure holds at 58 psi, which removes a weak pump or leaking regulator as the lean cause.

Verdict: the data describes unmetered air, not a misreporting airflow sensor. Replacing the MAF would not change the trims.

Both locations are downstream of the MAF, so both should be able to pull in unmetered air - but only one of them responded to propane.

Why the discrepancy matters:

• Propane only changes fuel trim if it actually gets into the intake stream and is burned. A scuffed but unbroken plastic flex section may simply be cosmetic - no through-crack, so the propane outside it has no path in and trims do not move. That is a negative result for that location, not a contradiction.
• The throttle body base and manifold flanges dropped trims from +22 to +8, a 14-point correction. Propane is a rich fuel with no oxygen sensor feedback delay worth worrying about at idle, so a large, repeatable trim movement when you spray a joint is strong evidence that joint is drawing air.
• Note that propane enrichment will lower trims on any lean-running engine to some degree, so the test is only meaningful when you can bracket a specific location and reproduce a large, localized change. Repeat each spray three times and confirm the response is tied to one spot.

Next steps:

1. Narrow the leak to a single joint - throttle body base gasket, intake plenum-to-manifold seals, PCV hose and grommet, brake booster check valve and hose, EVAP purge line, and any vacuum-operated actuator.
2. Use a smoke machine on the intake with the throttle closed for a visual confirmation of the exact source, especially on a plastic manifold that can crack invisibly.
3. Check PCV flow separately - a stuck-open PCV valve is a calibrated vacuum leak that behaves exactly like this and does not show up as an external leak.
4. Repair, clear adaptives, and re-verify trims at both idle and 2,500 rpm before returning the truck. A correct repair should bring idle trims to single digits and leave cruise trims unchanged.

The trim curve has changed shape, and that invalidates the vacuum-leak diagnosis.

Compare the two datasets side by side:

Measurement	Leak scenario	New scenario
Idle LTFT	+22%	+22%
2,500 rpm LTFT	+8%	+21%
MAP vacuum at idle	~15 in-Hg	19 in-Hg (normal)

• A leak that is a fixed orifice cannot hold trims at +21 percent at 2,500 rpm, because the leak becomes a smaller share of total airflow as rpm climbs. Trims that stay flat across the rpm range point to a percentage error in the metered air measurement - i.e., the MAF under-reporting across its whole range - or a fuel volume shortfall that scales with demand.
• MAP vacuum now reading 19 in-Hg removes unmetered air as the explanation entirely; manifold vacuum is normal, so the engine is not ingesting air past the throttle.
• Rail pressure is static at 58 psi with the key on, but static pressure does not prove volume. A restricted filter sock or a pump that cannot keep up under flow can hold pressure at idle and starve at load.

Revised test plan:

1. Compare actual MAF g/s against calculated load and against a known-good vehicle of the same engine, or scope the MAF signal for a clean, linear rise with rpm.
2. Perform a fuel volume test (delivery over time into a graduated container), not just a pressure test.
3. Inspect the air filter and intake for restriction - a plugged filter skews the MAF's airflow sample low in exactly this pattern.
4. Only after those are excluded should the MAF be replaced, and then with a relearn and a trim re-check at idle and 2,500 rpm.