Description

On 17 January 2023, an ATR 72-200 (G-NPTF) operated by West Atlantic UK on a domestic cargo flight from Belfast International to East Midlands was on a CAT 2 ILS approach at destination in night instrument conditions (IMC) due to reduced visibility in freezing fog when it experienced an electrical malfunction. This resulted in multiple system function and display failures, and intermittently compromised radio communications. The approach could not be continued and after going around, a successful MAYDAY diversion to Birmingham was completed.

Investigation

A Field Investigation was carried by the UK Air Accident Investigation Branch (AAIB). Relevant recorded data were recovered from the aircraft cockpit voice recorder (CVR), flight data recorder (FDR), and enhanced ground proximity warning system (EGPWS). Relevant recoded ATC radar information was also available.

It was noted that the 37-year-old captain had of a total of 5,264 hours flying experience, which included 583 hours on type. Similar experience and age information for the first officer was not recorded.

What Happened 

The captain as pilot flying (PF) briefed for a CAT 2 ILS approach at East Midlands, where freezing fog was delaying approaches. A ten-minute hold preceded clearance to begin the approach. The aircraft then proceeded normally, with the required dual autopilot indications seen as 1,000 feet agl was passed. The aircraft entered cloud at approximately 500 feet agl and at 400 feet agl, an ‘ELEC’ caution appeared on the Centralised Crew Alerting System (CCAS), alerting the crew to a DC GEN2 fault indication on the electrical system panel which was flashing in time with the first officer’s EADI and EHSI. The autopilot (AP) then disconnected and the flight director (FD) modes on both EADIs disappeared, although the FD guidance remained visible and appeared to the crew to be giving sensible indications. It was established during the investigation that at 400 feet agl all secondary radar returns from the aircraft had ceased.

A go-around was commenced and as power was increased, “numerous audio and visual warnings” including EGPWS ‘TERRAIN AHEAD’ and ‘TOO LOW GEAR’ occurred with 29 more such alerts and warnings subsequently following intermittently. The first officer’s EADI and EHSI continued flashing and his airspeed indicator (ASI) began cycling between maximum and minimum speed. A positive rate of climb was achieved and landing gear retraction followed. The captain’s EADI and EHSI both went blank for a few seconds, then reappeared before blanking again for a few seconds approximately one minute later. The standby instruments continued to work normally. As the climb continued, both pilots heard what they suspected were electrical relays cycling. The Autopilot Display Unit (ADU) continued to flash, and the AP could not be re-engaged. The aircraft levelled at 3,000 feet as per the published go-around procedure and having made the required right toward the EME nondirectional beacon (NDB), a ‘PAN’ call was then made to the control tower but no response followed. When the same call made on 121.5 also did not elicit a response, they changed back to tower frequency, set 7600 on the transponder and declared a ‘MAYDAY’. Despite what was described as having been “severe static”, they heard a faint response telling them to contact RADAR and on doing so, normal radio communications were restored. A diversion to Birmingham was requested and approved.

The Quick Reference Handbook (QRH) action for a DC GEN2 fault - to switch it off - was then completed, and this was followed by all audio and visual warnings, including false EGPWS activations, ceasing and the first officer’s EADI and EHSI going blank. Various non-flashing cautions remained illuminated on the overhead panel. The captain recalled that although he would have expected the Bus Tie Contactor (BTC) to close, this did not happen, so DC Bus 2 remained unpowered. He also reported that  he consulted the QRH DC BUS2 ‘Lost Equipment List.' In addition to the items on that list, the aircraft also had unlisted electrical failures, including loss of the Environmental Control System (ECS) recirculation fan and the Aviation Performance Measuring System (APMS).

The crew noted that the main battery was discharging and that the green hydraulic system - used to lower the landing gear - was unpowered and the blue hydraulic system was used to pressurise the green hydraulic system. The captain was concerned that further electrical failures may occur if the aircraft were to remain in the air for longer than the published 30-minute main battery life. The left NAV receiver was selected to the runway 33 ILS at Birmingham but “all automatic flight functions remained unavailable” so the captain flew a manual raw data approach and the flight was completed without further event.

Why It Happened

It was found that two ‘doll’s eye’ faults were indicated on magnetic latching indicators; one on the Bus Power Control Unit (BPCU) indicator (a contactor fault) and one on the Multi Function Computer (MFC) fault indicator on the right-hand side maintenance panel. However, it was not possible to establish whether these related to the flight under investigation or not because it was not known when these indicators were last checked.

A series of engine ground runs eventually established that the origin of the problems experienced was DC Gen 2 (see the illustration below). These showed that the fault symptoms experienced could be replicated even after the installed Generator Control Unit (GCU) had been substituted with the left engine GCU, so the DC Gen 2 was removed from the aircraft for a detailed investigation.

The removed DC Gen 2. [Reproduced from the Official Report]

It was noted that routine scheduled maintenance completed on the aircraft two days before the flight under investigation had found that the DC Gen 2 brushes were worn to their service limit and it had been replaced with an overhauled unit. This unit had been overhauled and bench tested in accordance with the Component Maintenance Manual (CMM) six months previously and had been supplied with a valid EASA Form 1 Authorised Release Certificate.

The DC Gen 2 was removed from the aircraft and examined by the manufacturer, first by means of a visual inspection and then using the manufacturer’s acceptance test procedure. Further component-level examination was then conducted at the AAIB’s laboratories. The speed sensor unit transport cover was removed, and the speed sensor unit and its connector were removed. The speed sensor connector was in good condition and its pins were undamaged. Externally, the cable between the connector and the sensor unit was in good condition and noted to be correctly sealed and both the speed disc and sensor unit were in good condition.

Acceptance tests from the unit CMM were performed. These included a resistance test of the speed sensor which failed. The unit was then mounted in a test rig for dynamic testing which it failed because of malfunctioning sensor signal. No signal was measured at the beginning of the five-minute test, and the signal was intermittent for one minute before recovering and remaining stable until the end of the test. The speed sensor peak voltage was also tested and found to be within limits.

The speed sensor output connector was removed from its housing for examination of wiring condition and further resistance checks. The speed sensor output connector is connected to the sensor unit by two seven strand wires. Activating an open circuit in this wiring was found to be repeatable by gentle pressure. These wires were soldered to pins at the rear face of the connector and the soldered joint covered with a heat-shrink sleeve. When this sleeve was removed, one wire was found to be completely broken but whilst held in place by the sleeve was close enough to make intermittent contact. Removing the sleeve on the other wire found damage to the wire core. Both wires showed evidence of mechanical insulation removal.

Speed sensor connector with heat-shrink sleeve removed from the two wires covered by it. [Reproduced from the Official Report]

Because the manufacturer’s generator overhaul requires removal of the speed sensor connector from the wires to facilitate cleaning of the main generator unit, additional wire is provided at manufacture so this process can be repeated multiple times during the unit’s service life. To reconnect the wires during each overhaul, a small part of the wire outer sleeve is removed to expose the wire core so that it can be soldered onto the connector pins.

Since no wire stripping method was specified in the CMM, the third-party organisation which had carried out the overhaul of the removed generator used a standard mechanical stripping tool. With this method, misalignment of the wire with the correct wire gauge on the tool will result in either the insulation not being stripped, or the conducting wires being part-severed by the stripping tool cutters as they close on the wire. The generator original equipment manufacturer (OEM) advised that it had changed to using thermal insulation stripping tools after observing the potential for wire damage when mechanical stripping tools were used.

Finally, it was possible to attribute some data downloaded from the MFCs to the electrical malfunction, but this built-in test equipment (BITE) data did not indicate when, or for how long a condition persisted, or whether there were multiple instances detected during a flight. However, system or sensor failures were identified which were consistent with interruption to all the affected busbars.

The narrative Conclusion of the Investigation was as follows:

A wiring defect on the DC Gen 2 speed sensor resulted in rapidly changing erroneous signals being sent to the GCU. This resulted in the rapid opening and closing of contactor 23PA in response to these inputs and, due to the rapidly fluctuating conditions, the BTC entered a self-protection mode and remained open for the remainder of the flight.

As a consequence of the above, the crew lost a significant number of instruments and systems during the final phases of a CAT II approach in reduced visibility at night. The power distribution anomalies also resulted in a number of spurious and potentially distracting EGPWS aural alerts. The crew conducted a go-around and, following a temporary loss of communications and permanent loss of autoflight capability, manually flew the aircraft to Birmingham Airport where it landed without further incident.

The wiring defect was probably caused by incorrect use of wire stripping tools at the third-party organisation that had overhauled the starter-generator.

The starter-generator manufacturer and the overhaul organisation have identified a number of safety actions they intend to take to prevent a reoccurrence.

Safety Action

• The Overhaul Facility involved stated that it has:
  • Updated the method of wire stripping used to bring it in line with the standard practices.
  • Clarified that tool choice is based on reference to the CMM, Standard Practice and any applicable Technical Instruction in that order.
  • Its Method Department technicians have been informed that for any task where a wire stripping method is not specified, they must assist in the assessment of the best way to strip a wire.
  • Wire stripping has been classified as an industrial process and specific training on it is therefore now required.
  • Technicians have been informed that the preferred method of wire stripping is to use thermal wire strippers.
• The OEM has stated that it intends to:
  • Modify the CMM procedure for this unit to highly recommend the use of thermal wire strippers.
  • Extend the use of thermal wire strippers to all other applications.

The Final Report was published on 16 May 2024. No Safety Recommendations were made.

Related Articles

• Aircraft Electrical Systems
• Electrical Problems: Guidance for Controllers
• Aircraft Maintenance
• Component Maintenance Manual
• Automatic Flight
• Electronic Flight Instrument System (EFIS)
• Loss of Communication