West Atlantic B734 at Brussels on Jun 4th 2019, partial electrical failure

Last Update: June 18, 2020 / 14:56:43 GMT/Zulu time

Bookmark this article

Photo of West Atlantic G-JMCR, Boeing 737-400 — West Atlantic G-JMCR, Boeing 737-400 (Photo credit: Alec BHX/KKC / Flickr / License: CC by-sa)

Incident Facts

Date of incident
Jun 4, 2019

Classification
Report

Airline
West Atlantic

Flight number
QY-3319

Departure
Oslo, Norway

Destination
Brussels, Belgium

Aircraft Registration
G-JMCR

Aircraft Type
Boeing 737-400

ICAO Type Designator
B734

A West Atlantic Boeing 737-400 freighter on behalf of EAT Leipzig, registration G-JMCR performing flight QY-3319 from Oslo (Norway) to Brussels (Belgium) with 2 crew, was descending towards Brussels with the commander (34, ATPL, 2,525 hours total, 2,325 hours on type), a line training captain in the right hand seat as pilot monitoring, and the co-pilot, who was completing his command upgrade training as pilot flying, in the left hand seat. Thunderstorms were in the area, many aircraft deviated around the thunderstorms. The crew prepared for an automated ILS landing runway 25R. During the descent an unusual noise, decribed as "large electrical clunk" occurred associated with the loss of primary EFIS screens on the left hand side of the cockpit and the disconnection of autopilot and autothrottle. The commander took control of the aircraft assuming the role as pilot flying, the copilot assumed the role as pilot monitoring for the remainder of the flight. ATC advised their transponder secondary returns had ceased. 2 minutes after the clunk, while descending through 8400 feet MSL, the crew declared PAN PAN and requested priority for the approach to runway 25R. The crew established that in addition to the loss of primary EFIS displays on the left side both CDUs were inoperative and various caution and advisory warnings illuminated including the #1 aft fuel pump low pressure, pressurization system AUTOFAIL and STANDBY, left side pitot static system, L ALPHA VANE and YAW DAMPER. The backlighting of the overhead panel was not working. No cautions had illumiated for the electrical system. The crew discussed the threats in relation to flying a manual ILS approach, as the approach could also be completed in VMC the crew decided to continue the approach. 2 minutes after the PAN the crew upgraded their call to Mayday and requested an immediate vector for an ILS approach to runway 25R. While about to intercept the localizer the aircraft was about 1500 feet above the target altitude, the crew advised they needed more track miles, ATC instructed to fly through the localizer. While nearing the localizer the crew reported they could see the runway and opted for a visual approach, ATC cleared the aircraft for a visual approach. Extending the flaps to 40 degrees the crew intercepted the glide path from above and were stable between 1500 and 1000 feet AGL. Tower cleared the aircraft to land reporting the winds were from 230 degrees at 5 to 8 knots. The commander noticed the EGPWS was not working. While the aircraft turned final the crew noticed a big cell near the end of the runway which was fairly active with a "wall of water" and lightning about every 20 seconds, the weather was clear to the south of runway though. The crew thus requested and was approved an immediate left turn in case of a missed approach. On short final, at about 300 feet AGL and 1nm before the runway threshold, the crew lost visual references in heavy rain, the commander therefore initiated a go around by estimating the engine power. The crew reported they felt a "sinking in the air" and the aircraft was initially slow to accelerate and establish a positive rate of climb before achieving a climb rate of 2,500 to 3,000 fpm. The commander followed the missed approach and visually orbited south-east. At about that time the copilot selected the transponder to ATC2, which restored the transponder returns enabing ATC to see the aircraft again and determine its position and altitude. The crew reviewed the effects of electrical failure and associated indications, the copilot noticed the normal circuit breaker for Transfer Bus 1 was open and identified the most appropriate checklist as "TRANSFER BUS OFF", however, as the required TRANSFER BUS OFF indication had not illuminated the checklist was not performed. The crew decided to not reset the circuit breaker reasoning they had sufficient systems to complete a safe landing. The crew assessed they had sufficient fuel to hold and wait for the weather to clear, after ATC reported the weather was clear they positioned for another approach and landed safely. On touchdown left intercom, VHF 1 radio and both engines' N2 and EGT gauges failed.

On Jun 18th 2020 the UK AAIB reported, Belgium's AAIU had delegated the investigation to the AAIB and released their final bulletin concluding the probably causes of the serious incident were:

The electrical failure was caused by a fault in the transfer relay which resulted in the loss of power to a number of electrical buses.

Following the electrical failure, the commander’s assessment was that the aircraft was in a stable condition so continued the approach to land at Brussels National Airport. This gave the pilots relatively little time to assess the situation and a number of non-normal checklists actions were not carried out; consequently, the aircraft was incorrectly configured for the approach and landing.

At a late stage of the approach the pilots lost visual references and executed a go-around.

The aircraft then orbited while the thunderstorms cleared the airfield and the pilots used the time to further analyse the failure. The second approach and landing were uneventful.

The AAIB analysed:

Failure of the 115V AC Transfer Bus 1

The failure of the 115V AC Transfer Bus 1 resulted from a fault in the transfer relay which caused circuit breaker C819 to open with the loss of electrical power from Gen Bus 1 to Transfer Bus 1. This resulted in the loss of electrical power to ADC 1, the primary EFIS displays and analogue instruments on the left side of the cockpit.

A loss of electrical power from a generator should result in the transfer relay automatically operating to allow the remaining generator to provide electrical power to the opposite electrical system through the transfer bus. Failure of electrical power to connect to the transfer bus is normally indicated by the illumination of the transfer bus off caption. However, the nature of the failure meant that the caption did not illuminate; this would have been contrary to the pilots’ expectation following the failure of the transfer bus. Consequently, the crew would not have recognised that the partial loss of electrical power was caused by the loss of power to the transfer bus.

Response by the flight crew

This partial electrical failure was a situation that the pilots would not have specifically trained for in the simulator, nor was it one for which their understanding of the electrical system would have provided a clear understanding of the cause and its implications. Consequently, they would have had to manage the situation by assessing which systems had failed and work through the implications using a decision-making tool and the QRH.

The pilots were aware of the thunderstorms in the vicinity of airfield and said they considered the options of continuing with the flight or delaying the approach while they investigated the problem. The commander was of the opinion that manually flying the aircraft while following radar vectors in a busy airspace environment without a serviceable transponder, while diagnosing the problem, would have significantly increased his workload. He assessed both visually and from the ATIS weather reports that he could complete the flight in VMC and had established that the aircraft was in a stable situation with sufficient systems to complete the approach. The pilots had already briefed and prepared the aircraft for a landing on Runway 25R and the commander was confident in manually flying the aircraft and conducting a go-around. The commander, therefore, decided that the safest option was to continue and land at Brussels.

Cockpit workload

It took approximately five minutes from when the electrical failure occurred until the aircraft intercepted the localiser. The PM requested priority to land, declaring a PAN, which would have reduced the time spent manually flying and allowed the aircraft to land before the thunderstorm reached the airfield. The PM requested extra distance from 22 nm to 32 nm; less than one minute later the PF requested immediate vectors. ATC advised they were 17 nm from touchdown and asked if they were ready for the base turn, which the PF accepted. Thirty seconds later the PF declared a MAYDAY and they were vectored tighter onto the localiser, thereby further reducing the distance and time available.

The misreporting of the distance as 6 nm, one minute after being informed it was 12 nm, would have upset the pilots’ mental picture and their decision to level off and ask for extra distance would have given them time to assess the situation and review their plan. The correct distance (DME) would have been displayed on the PF’s EHSI and on both pilots Radio Distance Magnetic Indicator; however, neither pilot questioned this discrepancy with ATC. Thirty seconds later the PF reported that he was visual with the runway and was cleared for a visual approach. The extra distance, and time, previously requested was not used and as a result of arresting the descent the aircraft was approximately 700 ft above the glideslope.

During this five-minute period the cockpit workload would have been high and the heavy static on the remaining VHF radio would have made communication more difficult.

The approach The aircraft was flying in twilight, in VMC, towards an active thunderstorm. The pilots reported that the aircraft was stable on the approach at 1,500 ft, with 40° of flap selected which was what was briefed, and the speed bugs set for. However, the QRA advises that if the yaw damper caption is illuminated, ‘Do not exceed flaps 30°’. During the approach, the PF realised that the EGPWS was not working, which meant there would be no reactive windshear warnings or automatic radio altimeter announcements during the approach.

At about 300 ft agl, a heavy rain shower obscured the end of the runway causing the PF to lose visual references and so he commenced a go-around during which he estimated the amount of thrust to set. With the FMC having failed, the N1 reference bugs should have been manually set, but this had not been actioned. Reports of the aircraft momentarily ‘sinking’, being slow to accelerate and achieve a positive rate of climb might have been due to the aircraft encountering windshear or insufficient thrust having been set. However, both pilots were of the opinion that they did not encounter windshear and felt that sufficient thrust had been applied.

Cumulative risk

Following the electrical failure, the commander followed a decision-making tool to help diagnose the problem and decide on the best course of action, which would be reviewed as new information became available and the situation developed.

The perception of the pilots was that there had been a significant electrical failure that coincided with a loud “electrical clunk”. They would not have known what caused the noise, or if the aircraft had been damaged, and would have needed to weigh the threat in orbiting to assess the problem against continuing with the landing. The pilots had already briefed and prepared for the landing and the commander’s assessment was that the best course of action would be to continue and land at Brussels. While the flight and go-around were flown safely, the crew did not complete a number of QRH procedures for systems that were not operating and, therefore, might not have identified and mitigated all the potential threats.

While the risk from each of these threats might be small, the cumulative effect can result in a reduction in the overall safety margin.

Time available

During a busy period of flight, the pilots had relatively little time to assess the situation, develop and review their plan as things changed. The time was further reduced by asking ATC for immediate vectors to the approach. They could have provided themselves with more time to assess the situation by being more specific and requesting a minimum distance to start the final approach.

The aircraft had plenty of fuel onboard and the probability of having to go-around could have been reduced by initially orbiting until the thunderstorms had cleared the area.