Generator loss in flight: battery endurance and dispatch under the MEL
Ten minutes into the climb out of Auckland, the number 2 starter-generator on a 28 V DC twin turboprop drops off line. The crew's reset attempt fails, and the generator control unit shows no overvoltage or overcurrent trip; the loadmeter reads zero. You are the AME on the ground taking the call. The aircraft is 55 minutes from its destination with a 25-minute alternate. The number 1 generator is carrying the full load. The battery is a 24 V, 34 Ah lead-acid unit, capacity-checked three months ago at 85 per cent of rated capacity. With the failed generator off line and normal load shedding applied, the essential DC load is about 20 A. The approved MEL permits dispatch with one generator inoperative if the remaining generator and the battery can supply the essential loads for the planned flight plus reserve, and requires the battery to be at least 80 per cent capacity. It also prohibits night operations with one generator inoperative unless standby power is verified.
1. Run the endurance calculation and state whether the planned flight is acceptable.
Read the worked answer
The battery only carries the load if the remaining generator also fails, but that is exactly the case the MEL is asking about, so the calculation has to be done properly.
Usable capacity: 34 Ah rated × 0.85 measured capacity = 28.9 Ah. The measured figure is three months old, so it is an upper bound; capacity falls with age and with cold soak.
Time required: 55 min to destination + 25 min to alternate = 80 min = 1.33 h.
Endurance at 20 A: 28.9 / 20 = 1.44 h = about 86 min.
Reserve: 86 − 80 = 6 minutes. That is not a reserve. It is inside the noise of the measurement. In addition, the 20 A figure is a nominal load: pitot heat, landing light, de-ice and a fuel pump cycling on will push it up, and higher loads shorten endurance directly. Cold soak at altitude reduces available capacity further.
Conclusion: as planned, at 20 A, the flight does not have a defensible reserve. It fails the intent of the MEL, even though the numbers superficially 'fit'. It could be made acceptable in one of three ways:
- Shed load to about 12–15 A (single comm, transponder, essential instruments, pitot heat only as required), which gives 1.9–2.4 h and a real reserve.
- Shorten the route, for example to a destination without the 25-minute alternate requirement, in daylight VFR.
- Fix the generator before dispatch.
If it is night or icing is forecast, the flight does not go at all under this MEL.
2. What ground troubleshooting would you do before condemning the generator?
Read the worked answer
A 'generator off line' squawk has a chain of possible causes, and the expensive components are not the most likely.
- Confirm the indication. Was there a genuine loss of output, or a sensing/indication problem? Check the loadmeter, the ammeter shunt and the GCU indications against a known load.
- Field circuit. Check the field/flash supply and the generator control unit's field output. A generator with no field voltage produces nothing, and the fault may be in the GCU, the wiring or a fuse, not the generator.
- Wiring and connectors. Look for chafed loom, corroded pins, a loose cannon plug at the generator or the GCU, and a bad ground. A pin pushed back in a connector can produce exactly this intermittent failure.
- Drive. Belt condition and tension, or the pad coupling and shear section on a pad-mounted starter-generator.
- The machine itself. Brushes and their springs, commutator or slip-ring condition, brush length against limits, and bearing condition. A worn brush gives a gradual or intermittent loss rather than a clean trip.
- Bench or on-aircraft output test. Bring it up and measure voltage and current at the generator terminal, then compare with the other side. If the generator produces correctly on the bench but not in the aircraft, the fault is in the aircraft wiring or the GCU.
The professional failure mode here is changing the GCU first because it is easy to reach, then the generator, and never finding the chafed wire that caused the original trip.
3. What would you record, and what operational restrictions would you impose for the flight?
Read the worked answer
Recording:
- A defect entry in the technical log stating the failure, the time, the phase of flight, the reset attempt and the indications observed.
- The MEL reference and the deferral category, with the expiry date. Most MELs give a defined rectification interval, and the deferral must be closed within it.
- The battery make, part and serial number, the date of the last capacity check and the measured capacity (85 per cent).
- The load survey: the essential load figure used (20 A, or the reduced figure), and the endurance calculation showing the time available against the time required.
- The engineer's certification and the crew's acceptance of the restrictions.
Restrictions I would impose on this flight:
- Daylight VFR only, per the MEL, and no flight into known icing, because de-ice and propeller heat are large DC loads.
- No night operations unless standby power is verified as the MEL requires.
- Route planning with a reduced alternate requirement, or an en-route diversion aerodrome within the computed endurance at the reduced load.
- Load shedding confirmed before departure and a written load schedule in the cockpit, so the crew does not restore the load in flight.
- Battery fully charged and verified before dispatch, and a re-check of the battery capacity at the next opportunity: a battery at 85 per cent and three months older is not a dependable reserve.
The deferral stays open in the technical log, and the next shift should know that the aircraft is running on one generator with a marginal battery, not treat it as a routine item.