83ft above the sea at night and no investigation?
I’m often asked about government cover-ups and the “real story” behind well-known commercial incidents. Generally, it’s easy to give an optimistic answer. Most investigations carried out by most of the accident investigation bureaus are top class: thorough and honest and with a detailed report which allows us, the outsiders, to check their work.
This case, however, makes me extremely uncomfortable. Although there is no direct evidence of a cover-up, it was certainly at best incompetently handled by the operator and the Norwegian CAA. In a perfect world, a situation like this should never arise. However, I’m pleased to say that with the investigation and reporting atmosphere around the world, it is at least possible to spot when it goes awry. In this instance, Aviation Herald is doing a brilliant job of pushing for answers about a serious incident which had been dismissed as not requiring an investigation at all.
Let’s start with the facts that everyone agrees on.
The date was 2nd December 2010. The aircraft was a de Havilland Dash 8-100 operated by Widerøe. It departed on a domestic Norwegian flight from Bodø to Svolvær, scheduled as flight WF-814 with 38 people on board. The weather was bad. The captain was the Pilot Flying and the first officer was the Pilot Monitoring.
Svolvær runway is 01/19. The aircraft was turning base for 19 for its approach to Svolvær when something went wrong. The airspeed dropped and the stick-shaker activated. The engines were accelerated to maximum power. The flight crew agreed to abort the approach and diverted to Leknes, where they landed without incident. From Leknes, they returned to Bodø.
The captain reported the occurrence to the operator. Both engines were checked for over-torque before the aircraft was returned to service, which should have been done before returning to Bodø. The chief pilot met with the captain and the first officer to collect information on the occurrence. As a part of this meeting, the first officer requested a copy of the print-out from the flight data recorder. Widerøe’s operating report stated that the aircraft went around at Svolvær because of turbulence and a downdraft, while on left base to runway 19. It also noted that the reports from the captain and the first officer did not agree.
Widerøe forwarded the captain’s report to the Norwegian Civil Aviation Authority. The Norwegian CAA filed the occurrence as an “incident” and did not refer it to Norway’s Accident Investigation Bureau (AIBN) for an investigation.
So the obvious initial issue is that the “incident” as reported by the captain did not match the events as related by the first officer. Of note is also that the first officer clearly felt strongly enough about this discrepancy to request a copy of the Flight Recorder data. It was two years later, when he left the operator, that he realised that no investigation had taken place. He also discovered that only the captain’s report had been forwarded to the CAA and based on the Captain’s version of events, the CAA did not believe it merited an investigation.
On the 17th December, 2012, the first officer phoned the AIBN and submitted his initial report directly to them, along with the print-out from the flight data recorder.
As a starting point, let’s look at what the print-out from the flight data recorder demonstrates. Aviation Herald posted a summary of the most important points:
- The stick shaker activated, which is a warning for the pilot that the aircraft is flying too slowly and is at risk of a stall
- The aircraft was at 400 feet above sea level when the stick shaker activated
- Eight seconds later, the aircraft stopped its descent. It was 83 feet above sea level
I need to stop there. The aircraft was not yet established on final approach. It was 83 feet (25 metres) above the sea. This alone seems to be a clear case of a serious incident.
To put it into perspective: their height above sea level was less than the wingspan of the Dash 8. It was less than the height of New York City’s Christmas tree. It was a hell of a lot less than the height they needed to be safe.
- The average rate of descent was 2,377 feet per minute; however, with an initial vertical acceleration of 0.75G and a final vertical deceleration of 2.7G, it’s likely that the maximum vertical rate of descent probably reached 3,500 feet per minute less than 150 feet above the sea.
- In the eight seconds while the aircraft was descending, the pitch angle of the aircraft increased by 10 degrees nose up, then nose down by 15 degrees and then back to 10 degrees nose up.
- While the aircraft was being pulled out of the descent, it experienced over 2.7Gs, which exceeds the structural limit of the Dash 8
- Both engines were over-torqued, reaching 118% and 120% of torque over a period of 35 seconds.
Without any other information, the information from the flight data recorder should have triggered an investigation.
Obviously, the next question is what happened in the cockpit? Both pilots agree that the captain was flying while the first officer was monitoring. The weather was bad. They were turning left onto base for their approach, which meant that the airport lights were behind them. It was completely dark and neither the horizon nor the sea below were visible.
The captain said he first noticed something wrong before he initiated the turn. There was a significant drop in airspeed and the aircraft started buffeting, a warning of an impending stall. He said that at the same time, the first officer called out “check speed”.
The captain said that he responded immediately: he pushed engine power to full and pitched up; however the aircraft continued to lose speed and to descend. He said that it felt like the aircraft was falling or being pushed down.
He pushed the control column forward to gain airspeed and prevent the stall. When he pulled the stick back again in order to climb out, the aircraft’s stick shaker triggered. He eased off to build up more speed and saw the red obstacle lights in front of him. He recalls looking at the altimeter and that it showed them as approximately 300 feet (90 metres) over the sea. He focused on one of the red lights and stayed low, in order to build up speed. It worked and he initiated the climb. He saw that they would pass over the light at a safe altitude. After he gained sufficient speed and the aircraft started to climb, the first officer unexpectedly took control. He decided not to oppose this but left the first officer to climb away.
The general facts are the same in the first officer’s version, however the actions taken by the captain are very different. The first officer said that he was monitoring the instruments when he called for the captain to “check speed” during downwind. Then again, as the captain had initiated the turn, he called out “check speed”. He felt that under the circumstances, the corrections made by the captain were too small. As they turned in for final approach, the stick shaker triggered. He remembers that he was startled by the shaking. He was prepared to perform corrective measures but the captain did not make the expected call out and reaction. Next, the nose of the aircraft made a significant dip. He said that he “stared straight down onto the black sea” and saw a red light on a islet below.
The first officer said that he grabbed the control wheel and pushed the engine controls all the way until they stopped (approximately 118%). He remembers pulling the stick back with both hands and believed that they were still likely to crash into the sea. The aircraft started to climb. Once at a safe height of 3,000 feet, the first officer and the captain agreed that the captain should resume his role as pilot flying.
They diverted to Leknes for a safe landing and then continued to Bodø to debrief there.
Based on the FDR data, we know as fact that the aircraft stopped its descent at 83 feet (25 metres) above the sea and, eight seconds after the stick shaker activated, began to climb again.
The first officer wrote:
A total of 38 people were on board of flight WF 814 on Dec 2nd 2010. The luck and coincidences that night averted that this evolved to be the worst accident in Widerøe’s history.
However, he did not receive a reply to his direct submission for six months. It was the 14th of June, 2013 when he received a formal response from the AIBN, signed by the director as well as the Aviation Department Head.
The accident investigation did not find the basis for a change of the status of the event from incident to serious incident
There is a significant discrepancy between what has been initially reported to the CAA and how you describe the event today.
The situation was not seen critical as reported by [Svolvær] Tower.
The letter went on to state that although, the AIBN does get contacted by passengers reporting “unpleasant” experiences, they had not received any reports from the passengers on the flight that day.
This does not necessarily mean that the passengers did not experience the event as scary. Such passenger reports are a safety net for the AIBN in case reporting requirements are not being met. Widerøe’s last internal event report and conversation with you, cabin crew and others in Widerøe suggest, that a lot could have been done better after the event had occurred. The long time since the event took place would hamper an investigation by the AIBN.
They also confirmed that the print-out from the flight data recorder matched that submitted by the airline in 2010. In the end, they stated that the operator (Widerøe) had room for improvement for occurrence reporting and should ensure that pilots were aware that they could independently report to the CAA and the AIBN.
There does not appear to have been any mention of the conflicting reports from the captain and the first officer, nor is there any mention that the report from the first officer was missing from the operator’s report.
And so, the occurrence was deemed again not to need an investigation – at least until this year.
In February of 2015, Norwegian newspaper Nordlys reported that an aircraft had lost height on approach to Svolvær and was recovered at 27 metres over the sea, accompanied by charts based on the read-out from the flight data recorder. The mainstream Scandinavian media didn’t pick up on the story.
On the 16th of March, the AIBN reported they had (finally) opened an investigation into the occurrence. They stated that flight WF-814 was in the final stages of the approach to Svolvær when the airspeed abruptly slowed and the stickshaker activated. The crew aborted the approach and diverted to Leknes where a safe landing occurred.
On the 5th of June, a source contacted the Aviation Herald. The source told Aviation Herald that the aircraft had entered a full stall on visual approach to Svolvær.
The aircraft was turning base when the airspeed sharply dropped and the stickshaker activated. The engines were accelerated to maximum power available, the pitch continued to increase until the aircraft entered full stall however and the captain basically froze. The first officer took control of the aircraft, pushed the nose down, managed to recover the aircraft from the stall and pulling +2.7G arrested the descent at 25 meters AGL. The aircraft subsequently diverted for a safe landing.
Following the occurrence the captain provided a report about the occurrence to the safety department of the airline which prompted both engines to undergo checks for overtorque before the aircraft was returned to service. Our source said: “The captain drastically understated the severity of the situation. The AIBN cannot investigate what it does not know.” The first officer left the company and joined another airline.
Aviation Herald verified the information received, including the FDR data, and contacted the Statens Havarikomisjon for Transport (SHT, includes the AIBN) seeking clarification. They asked specifically why the investigation had not been opened in 2010. They also stated that they planned to release the story on the afternoon of 18 June 2015.
On the morning of 18 June, they received a statement from the Chief Inspector of the Aviation Department of the AIBN.
Yes it is correct that the mentioned aircraft was involved in an incident on December 2010. This incident was reported, both by the captain, the company and the air traffic controller, as an incident to the CAA and thus never reached the accident investigation board of Norway (SHT). Two years later the SHT became aware of the incident. Based on the information at that time the SHT chose not to reclassify the incident as a serious incident. This together with the fact that the incident had occurred two years ago the SHT chose not to open an investigation. It is also true that it was public expectation that made the SHT to open an investigation this year. This is a full investigation, as any other. The available information from FDR is analyzed in cooperation with Bombardier. At this stage in the investigation, the SHT will not release any information from FDR. Based on this we are sorry to say that we will not confirm your assumptions. The SHT expects the report to be public by the end of this year.”
Aviation Herald’s information is that the captain submitted a report of a stick shaker activation and stated that the aircraft lost some height but never descended below 300 feet. The captain’s report also included the fact that both engines were overtorqued well above engine limitations and that the period of overtorque may have lasted up to 20 seconds. Their sources told Aviation Herald that Widerøe decided that, as they could not reconcile the two conflicting reports, they decided to only forward the captain’s report to the Civil Aviation Authority.
On 7th July 2015 Richard Kongsteien, Vice President Public Relations of Widerøe, wrote to Aviation Herald:
Because the incident is under an investigation, we will not comment on the story apart from saying that your insinuations that Widerøe deliberately withheld information from the authorities is utterly wrong, and contradicts every safety policy we live by in our company. A clerical filing error was made when one of the reports was delivered to us some time after the incident. This is something we are sorry happened. The error was discovered by us, and corrected by us. We look forward to the closing of the investigation by the authorities, which we hope will end all rumors and speculations about this incident.
The Norwegian investigators now definitely have have both crew member reports along with Widerøe’s operational report and the FDR data. They released a preliminary report on the 21st August in which the occurrence is reclassified as a serious incident: Preliminary report on serious aviation incident at Svolvær airport Helle, Norway 2 December 2010 involving Bombardier DHC-8-103, LN-WIU operated by Wideroes Flyveselskap AS | aibn
The preliminary report states that, based on the FDR data, the most probable time when the control wheel was taken over was approximately two seconds before the engine power was increased beyond maximum. This supports the first officer’s versions of events, however they go on to say that, as a result of inertia, the captain could have already implemented the measures that the first officer was waiting for.
The aircraft lost approx. 270 feet over the course of 8 seconds. Preliminary analyses of the graphs from the flight recorder indicate that during these seconds, the aircraft was exposed to significant external influences. The Commander’s reaction and correction of both attitude and engine power seem to have prevented stalling. The subsequent pull-up most likely prevented a crash into the sea.
It is correct that the nose was below the horizon and that the aircraft was still losing altitude when the engine controls were moved completely forward until stop, but at the time AIBN believes the First Officer intervened, the stick had already been moved significantly back and the tendency had turned.
The report mentions the visual conditions (bad weather and in darkness) which are typical for creating spatial illusions. However, the report concludes only that the crew successfully averted crashing into the sea after being exposed to a major wind shear at low altitude.
For further reading, I recommend the sequence of events and the Aviation Herald’s investigation on the incident page:
Incident: Widerøe DH8A at Svolvær on Dec 2nd 2010, aircraft rapidly descended on base turn, recovered at 25 meters AGL
The first officer was the only person to push for an investigation; this fact alone has me firmly on his side. I find it incredibly worrying that the AIBN don’t seem be dealing with the issue at all that a serious incident was not reported, instead focusing on explaining why it was not their fault that they didn’t respond in 2012. Quite frankly, I would be pushing back hard at the CAA and Widerøe for not informing the AIBN, rather than some nonsense about how none of the passengers reported it. That’s not their job.
I hope that the final report takes on the issues without prevarication or excuses. A good investigation does not seek to allocate blame but to understand how the situation happened and how to avoid it happening again. In this case, they need not only to consider the incident but also the response.