Engine-driven pump case-drain leak at the turnaround
At a regional operator's base, a twin turboprop is being turned around. During the walk-around you find a slow drip under the left engine nacelle at the hydraulic drain mast. With the left engine running and the system pressurised to 3,000 psi, you time it at about one drop every 8 seconds. It comes from the engine-driven pump case-drain line, not from a line fitting, and the fluid is clean and of the correct mineral-based type. The left system reservoir is 0.6 litre below the minimum mark; you top it up to the max mark before taking the measurement. The aircraft maintenance manual gives an external leak limit of one drop per 10 seconds for this system, and the operator's MEL allows dispatch with a known external leak of up to that limit, provided the reservoir is serviced and the level is checked at each transit. The sector is 1 h 20 min to a remote airfield with no engineering support; the alternate is a further 35 minutes away.
1. Interpret the measurement. Is it within limits, and what does the leak path tell you about the defect?
Read the worked answer
Start with the arithmetic, because the decision turns on it.
- One drop every 8 seconds is 10/8 = 1.25 drops per 10 seconds, so the leak is about 25 per cent above the AMM limit of one drop per 10 seconds. It is outside limits.
- Re-measure before acting on a single sample: stabilised fluid temperature, engine at ground idle and again at a higher power setting, the drip caught on a clean surface and timed over at least 60 seconds, and the reservoir level recorded before and after. Seal leakage often increases with temperature and pump output, so a reading taken cold can understate it.
What the path tells you:
- The fluid is coming from the case-drain line at the drain mast, not from a B-nut, hose or fitting. On an engine-driven pump, a case drain carries normal internal leakage back to the reservoir; fluid appearing there in excess means the shaft seal is passing, with the fluid travelling along the drive pad and out of the drain.
- This is an external leak, not an internal one. An internal leak (a worn pump or a leaking check valve) would show as low system pressure, high fluid temperature and a reservoir that drops without any visible fluid, and the AMM treats the two differently.
- The 0.6 litre shortfall is consistent with a seal that has been passing for a while, and it also tells you the pump has run with a low reservoir at some point, which is not good for it.
So: external, from the pump shaft seal, and above the published limit.
2. Make the dispatch decision, with the reasoning and the conditions you would attach.
Read the worked answer
The MEL permits dispatch with an external leak up to the AMM limit. It does not permit dispatch above it. At 1.25 drops per 10 seconds the aircraft does not meet the MEL entry, so a normal release is not available.
Options, in order of preference:
- Change the pump. If a serviceable engine-driven pump and the seals are available, this is the correct answer and the aircraft goes at the next available slot.
- Isolate the left system under the MEL if the operator's approved MEL has a system-inoperative entry that covers it. That requires: the affected services confirmed available on the remaining systems; the affected pump depressurised and the system capped or configured as the AMM requires; the reservoir at the correct level for an isolated system; a performance penalty check (landing distance, asymmetric braking, gear extension time); and any operational restriction in the MEL entry, such as flight in known icing or dispatch to an aerodrome without support.
- Special flight permit only if the operator can justify the flight and the CAA agrees, which for a deteriorating seal on a remote-sector route is unlikely to be attractive.
Why not just 'top it up and go':
- A shaft seal that is already passing above limit can fail suddenly, and then the system loses fluid rather than weeping it. Fluid sprayed into a hot nacelle is a fire risk.
- The reservoir has a finite volume. Even a slow loss over 1 h 20 min plus a 35-minute alternate, with no engineering support at the destination, leaves no recovery path.
Whichever route is taken, the defect goes in the technical log with the measured rate, the time, the power setting and the reservoir quantity, and the release references the MEL item or the permit.
3. What would change your decision, in either direction?
Read the worked answer
Things that would make it worse and stop the aircraft outright:
- A rate that increases with temperature or power, or a second measurement well above 1.25 drops per 10 seconds.
- Fluid tracking onto the exhaust or hot section, or any smell of burnt fluid in the nacelle - fire risk dominates everything else.
- A reservoir that keeps falling after servicing, or a system pressure that fluctuates or will not hold.
- Contamination: water, metal particles, or the wrong fluid type in the system. That would require flushing and a much larger investigation, and mixing mineral and phosphate-ester fluids produces a sludge that destroys seals.
- Any abnormal noise from the pump, or a case drain that is blocked, which can force fluid past the seal.
Things that would make a limited dispatch defensible:
- A stable rate comfortably below one drop per 10 seconds at all power settings, a reservoir that holds its level, and a short sector to a base with spares.
- An MEL entry that genuinely covers the configuration, with all the performance and operational conditions met.
- A clear monitoring plan: reservoir checked at each transit by a qualified person, a written limit above which the aircraft stops, and a nominated person to make that call at the outstation.
The thing I would not accept is 'it has always dripped a bit'. The limit exists because the seal has a failure mode, and the consequence of that failure in flight is fluid loss plus a fire hazard on a remote sector with no support.