UK investigators have determined that a wire defect, which led to a cascade of electrical problems on an ATR 72-200 freighter, had resulted from incorrect use of wire-stripping tools by a third-party maintenance provider.
The West Atlantic aircraft – arriving from Belfast on 17 January last year – had been conducting the final stage of a Category II approach in low visibility, at night, to East Midlands airport’s runway 27.
As it descended into cloud at 500ft, the crew received indications of a DC generator fault. The autopilot disconnect horn sounded and, as the crew opted to execute a go-around, they received multiple audio and visual alerts including spurious terrain, bank angle, and height warnings.
Although standby instruments continued to function normally, both pilots experienced periods of blank displays and the crew was unable to reselect the autopilot.
Attempts to contact air traffic control to request a diversion to Birmingham – on both the regular radio frequency, and the distress frequency – were initially unsuccessful.
“With the prospect of a manually-flown diversion, using degraded flight instruments and without radio communications, crew workload was high,” says the UK Air Accidents Investigation Branch.
But the pilots managed to reach East Midlands radar controllers and request vectors to Birmingham.
The crew, carrying out quick-reference checks, turned off DC generator 2 about 7min after the incident began, and power was also lost to the corresponding DC bus.
Audio and visual warnings ceased, and the first officer’s display screens – which had been flashing – went blank.
The crew sought to determine which systems had been lost, recognising that the main battery – which had a 30min life – was discharging, and the green hydraulic system, which extends the landing-gear, was unpowered.
All automatic flight functions were unavailable, says the inquiry, so the captain flew a manual raw-data ILS approach to Birmingham’s runway 33, and the crew used the blue hydraulic circuit to pressurise the green circuit and lower the undercarriage.
The aircraft (G-NPTF) landed safely and neither pilot was injured.
Investigators inspected the aircraft’s right-hand DC starter-generator and focused on a speed sensor which measures the high-pressure compressor’s rotation speed. Its output connector is connected to the sensor unit by two wires.
One of these wires was found to be broken, but held in place by its heat-shrink sleeve, close enough for intermittent contact. The inquiry says the manufacturer’s component maintenance manual did not specify the tool to be used for stripping the wire, and the maintenance organisation used a mechanical – rather than a thermal – tool, resulting in the fracture.
While the inquiry could not determine why the wire failed when it did, the break occurred during a critical phase of flight. The partial connection of the wire resulted in rapid, intermittent speed signals being sent to the generator control unit, and triggering power-distribution anomalies and various electrical problems.
“Due to the rapidly changing and unusual power distribution configuration of the aircraft during this event, it has not been possible to fully explain the behaviour of some of the flight instruments that was observed by the crew and that of other aircraft systems,” the inquiry says.
Investigators state that the starter-generator manufacturer and the overhaul organisation have identified a number of safety actions they intend to take to prevent a recurrence.
