On 12 June 2011 a flight-crew were preparing an A321-231 – registered – for a scheduled passenger flight from Darwin, Australia to Bali, Indonesia.
The load-sheet provided the flight-crew with a planned TOW which was approximately 10 tonnes below MTOW. Accordingly, the flight-crew were able to plan for an intersection departure from runway 11, via intersection with taxiway Bravo, in accordance with the operators SOPs. This provided a TODA of 2316m. The flight-crew also elected to use a flex-thrust (reduced thrust) setting for the take-off, as standard procedure that reduces internal engine temperatures, increasing the engine-lifespan.
The Captain (PIC) acting as PNF for the sector prepared the final TOW calculation, and made performance calculations using the Aircraft Performance Manual. He then left the aircraft to complete the pre-flight external aircraft inspection. Whilst outside, the First Officer, PF for the sector, cross-checked the final TOW and performance calculations made by the PNF. During this cross-check PNF detected an error in the calculations made by PF.
Subsequently PNF corrected the error, and recalculated new performance calculations, including critical V speeds for the departure. Unfortunately the ‘corrected’ calculations were correct figures but applied to the wrong performance table. PF had used a chart for a full length departure, as TODA od 3444m as opposed to the intersection departure as planned. A difference of 1128m from the Brave intersection departure point.
On return to the flight-deck, PF advised PNF of the original error and his actions to rectify it.
PNF then cross-checked the amended calculations prepared by PF, receiving the Aircraft Performance Manual from PNF with a card inserted into the manual to
bookmark the performance chart he’d used. PF checked the revised data on the page presented, but failed to notice the error of page selection by PNF.
The crew completed then pre – flight preparations and taxied for runway 11 via taxiway Bravo.
On departure it became apparent to PF passing 120-130 knots that there was insufficient distance left to complete a Rejected Takeoff, even though the (incorrectly) calculated V1 was higher. He accordingly decided not to reject if conditions suggested he should beyond that point.
The aircraft rotated at the (incorrectly) calculated VR with approximately 450m remaining. The flight-crew kept the flex-thrust setting for departure, not electing to selected TOGA. Later in the flight, the flight-crew reflected on the situation and recalculated the performance figures, discovering the error in their planning.
Planned TODA: 3444m V Actual TODA: 2316m
Planned V1: Approx 160 KIAS V Actual V1: 138 KIAS
If the crew had rejected at erroneously calculated V1 of 160KIAS the aircraft would have required an additional 1000m of runway for the required accelerate and stop distances required. If an engine failure occurred at the calculated V1 the obstacle clearance would have been compromised.
What was the error?
The first error was a slip committed by PF in miscalculating the final TOW was well identified by PNF independently cross checking the calculation.
However PNF then used the wrong performance table, creating the second error, an additional slip as intended action was not completed correctly.
The third error, either a slip (not identifying the use of the wrong performance chart during his independent cross-check) or a lapse (forgetting to check the performance chart in use), was made by PF not identifying the error made by PNF.
It had become common practice for many line-pilots to ‘bookmark’ the Aircraft Performance tables for calculations. This involved the pilot completing the performance calculations handing the performance table book to the other pilot, with the relevant performance table marked for their attention.
This relatively simple and at face value, innocent action is flawed. By presenting the ‘cross-checking’ pilot with the ‘right’ performance table to use, they are statistically less likely to check that the presented performance table is actually the correct table to be used for the circumstance.
By passing the book to the cross-checking pilot closed, they must analyse which performance table to use independently. This reduces the risk of both pilots using an incorrect data set for making their performance calculations.
Following the incident the operator changed its SOPs to specifically prohibit the book marking culture, ensuring that independent calculations were made for a cross comparison between flight-crew members.
The Airbus operator used a paper based performance data table book for runway performance calculations. This tried and tested method has been used in the industry for decades. A performance table is provided for each airport the operator flies to, with various tables for different runways and intersection departures.
Many airlines now utilise the Airbus led Less Paperwork Cockpit (LPC) philosophy and use electronic devices (laptops or tablets) with custom performance and mass and balance software to calculate performance data.
Although this software is designed carefully to reduce human error entry input to a maximum, the risk cannot be wholly eliminated while humans are still in the loop.
The “bookmark factor” – whereby the other pilot is passed the performance table book with a marked page, implying the page has been identified and selected as the correct one by the pilot passing it – is removed from the equation, the pilot entering the data can still select the wrong start of takeoff run in the software.
Ultimately paper or paperless performance is an infallible design which requires particular care of this involved in calculating figures. While humans are in the loop, human error risk will exist requiring mitigating behaviour and actions to minimise the risks of incorrect data being used.
Don’t ‘bookmark’ documents for colleagues. Doing so increases the probability of confirmation bias, and reduces the probability of effective independent cross-checking.
Take care to really check all data prior to using it. Pages, charts, dates, intersections, temperatures, pressure settings etc. Don’t assume your colleague is correct.
Passing through 1300′ during final approach into Leeds Bradford Airport (LBA) PF called for flap full. At the same time ATC issued a wind check, PNF acknowledged it and simultaneously selected park brake on. Shortly afterwards, PF checked the flap indicator and became aware that flap full was not selected. PF called again for flap full selection, PNF selected it correctly. The crew completed the landing checklist then landed. Immediately after touchdown, the flight crew noticed that the brakes appeared to take effect immediately with a greater deceleration than normal.
The Captain noticed that the AUTOBRAKE blue caption remained illuminated, but with no DECEL indication. The first officer then ‘dabbed’ the brakes in an attempt to disengage the auto-brake, but this had no effect.
The aircraft came to rest slightly left of the centre-line. After coming to a complete stop the Captain asked the First Officer to set the park brake on. The First Officer reached for the park brake and became aware that it was already set. During the landing all four main gear tyres deflated, leaving the aircraft stuck on the runway.
Most Airbus operators include a check in the Landing Checklist that requires the pilots to check to the upper ECAM display to ensure that pre-landing tasks have been completed. The AMBER message ‘PARK BRK’ was generated but not noticed by the crew.
Airbus classify this ECAM caution as a ‘Level 1’ caution, therefore there the Master Caution light doesn’t illuminate and there is no audible tone. The aircraft – G-DBCI – had a pre-‘H2F3’ standard Flight Warning Computer (FWC) fitted.
In the same circumstances in an aircraft fitted with the ‘H2F3’ standard FWC, the master caution light will illuminate and the ‘attention getting’ audible tone will be generated. The system upgrade by Airbus, would have made the situation far more obvious to the flight-crew.
What was the error?
The main causal factor was an error of commission made by the PNF. The PNF incorrectly selected Park Brake On rather than selecting Flap Full. This error is commonly referred to as a Slip, where the operator makes the wrong action. A contributing factor was the distraction from ATC of the spot wind radio call. Leeds Bradford Airport can be a challenging airport to fly into, primarily because of its location on a hillside. It can often be windy with turbulence on approach. During the approach ATC gave the crew six wind reports, considerably more than usual.
Experience does not necessarily reduce the risk of a Slip. It’s been proven over the years that the more experienced an operator becomes, regularly performed actions may become motor actions. Motor actions that become almost subconcious or require little thought. At lower levels of experience, operators rely more on knowledge or rule based active-thinking to perform actions, relying more rigidly on action and response frameworks which may mitigate the risk more effectively.
It’s easy to say, but as we gain more experience we need to think back to our earlier stages of flying and adopt the more robust knowledge or rule based method for conducting actions in the flight-deck. For example, following Airbus Philosophy by visually identifying a control prior to using it, then checking after operation that the indication matches our expectation or not. This robust paradigm reduces our natural tendency to rely on motor-functions, that may fail us without highlighting the dangerous they carry.
At the time of the incident in 24 January 2007 Airbus confirmed that the same incident had occurred previously five times.
Adopt good flight-deck discipline. Visually identify a control before moving it. Check displays after moving a control, to see if the planned outcome has occurred or not.
Remember that motor-skills are not infallible.
Experience: The Captain (PNF) was 49 year old Airline Transport Pilot Licence holder with 9,500 flying hours of which 950 were on type.
Onboard: 5 crew and 53 passengers.
Route: London Heathrow (LHR) to Leeds Bradford Airport (LBA).
On 17 November 1980 a Nimrod MR2 departed RAF Kinloss with a crew of 20 to complete a training sortie. Shortly after departure, at approximately 20’ above ground level, the aircraft flew through a dense flock of seabirds, suffering a multiple bird-strike.
The four engine aircraft rapidly lost thrust. Immediately after the multiple bird-strike, engine number one suffered a catastrophic failure. Simultaneously engines number two and three were damaged and produced very little thrust. A single engine effectively powered the aircraft.
Despite the best efforts of the flight-crew, the loss of thrust was too great for the aircraft to remain airborne. 27 seconds after take-off, the aircraft crash landed into a forest, quickly becoming engulfed in flames.
18 members of the crew escaped, five with injuries. Unfortunately both flight-crew members incurred severe head injuries during the impact, and subsequently died.
After the accident, the investigators found the remains of 77 Sea birds on or near the runway. It is not known how many more birds were ingested by the engines or impacted the airframe.
Prior to departure the MOD ground crew conducted a search for roosting birds on the airfield. This was a Standard Operating Procedure (SOP) within the MOD at the time. The search was conducted in semi-darkness, and did not reveal any roosting birds.
The RAF deemed ‘Conventional protective helmets unsuitable’ for use in Nimrod aircraft prior to the accident. After investigating the accident, the Ministry of Defence (MOD) raised a safety action and tasked the Institute of Aviation Medicine to explore the feasibility of ‘comfortable lightweight head protection’ being used by flight-crew in Nimrods.
On January 15 2009 US Airways Flight 1549, an Airbus A320‐214, registered N106US with 150 passengers and 5 crew ditched into the Hudson River in New York a mere six minutes after take-off from LaGaurdia Airport. This accident has allowed the aviation world to learn some great lessons, for us to improve on in the future.
Above is an interesting image produced by the National Transportation Safety Board -NTSB – depicting the emergency exit routes used by passengers and crew onboard Flight 1549. A note for understanding the illustration, those boxes slip into two colours show the initial exit used by a passenger who then re-entered the aircraft and exited through a second exit.
It’s interesting to see how they got on. When boarding an A320 it can often take anywhere between 10 – 25 minutes depending upon the number of passengers, their composition of families, large groups, individuals etc, their mobility, the size and volume of cabin baggage. Manufacturers have to demonstrate that aircraft can be rapidly exited for certification issue prior to use as commercial aircraft. The tests are arguably not reflective of a real flight, as the test passengers for certification are well aware of why they are there, and what they have to do. Forward thinking, planning and improving their exit times. These tests do not necessarily reflect the true time it would take for a ‘normal’ passenger load to evacuate a ‘normal’ schedule flight.
On Flight 1549 the rear of the cabin flooded when the aircraft ditched. Passengers in the front exited via 1L and 1R, while those in the centre naturally exited via the over wing exists. Although those exited from the front vacated with ease, those in the centre of the aircraft encountered congestion, apparently according to passenger feedback due to the ‘slow’ exit time of passengers.
Understandably in the heat of the moment, passengers may find it harder to remain calm and exit in a swift, orderly manner.
At the rear of the cabin passengers initially moved reward to exit, however a Flight Attendant ordered them to move forward as the rear cabin was flooding as it sat below the water line. This cause congestion in the rear of the aircraft, and apparent concern for many of the passengers to scramble over chairs towards the centre wing exits to avoid the queue in the aisle.
On seeing the congestion in the centre of the cabin, the flight attendant at the front of the aircraft encourage some passengers to move forward and exit via 1L and 1R.
Passenger interviews indicated that about 70 percent of the passengers did not watch any of the preflight safety briefing, indicating that passenger attention to the preflight briefings was generally low.
Amazingly about 70 percent of passengers decided to not watch any of the preflight safety briefing. Travelling as a passenger we are probably all to aware that some of our fellow passengers may not be giving the briefing their full attention to say the least. It would appear that even though many in the cabin may look alert and interested, that their minds may be elsewhere.
2.10.4 Passenger Education
In addition, more than 90 percent did not read the safety information card before or during the flight. The NTSB believes that these responses clearly indicate that passenger safety information is still routinely ignored by most travellers.
Furthermore over 90% of passengers didn’t bother to read the safety information card onboard. During post accident interviews, the NTSB discovered that the most common reason for choosing to not bother, was that they were frequent travellers and didn’t feel they need to remind themselves. A clear case of complacency.
The NTSB had previously proactively conducted research to find out how many people were paying attention to safety briefings onboard US operator flights. The study – Safety Study 00/01 – concluded that of 377 respondents 68% didn’t read the safety information card.
NTSB issued Safety Recommendation A-00-86, which recommended that the FAA do the following:
Conduct research and explore creative and effective methods that use state-of-the-art technology to convey safety information to passengers. The presented information should include a demonstration of all emergency evacuation procedures, such as how to open the emergency exits and exit the aircraft, including how to use the slides.
The NTSB had previously reported the apparent lack of attention paid by passengers to safety briefings to the FAA, however their concern was not acted upon.
… the FAA did not take many of the recommended actions because operators expressed concerns about the financial burden.
Many operators had expressed concern about the administrative and financial burden of overcoming this major issue. At face value this might sound harsh, but the reality of finding creative, interactive and effective ways of engaging a higher percentage of travellers to give their full attention to a safety briefing on every flight is a significant challenge. A challenge that is unfair for operators to face alone, manufacturers, regulators and passenger associations should participate. Perhaps a challenge for universities, institutes and behavioural psychologists to consider to work with industry to find a solution.
CVR data indicated that the preflight safety briefing provided by flight attendant B included information about the flotation seat cushions but that it omitted information about the location, removal, donning, and inflation of the life vests. This omission was not in accordance with federal regulations or company procedures, which stated that this information should be provided.
The investigation revealed that about 77 passengers retrieved and evacuated with floatation seat cushions, whereas only about 10 passengers retrieved and evacuated with life vests. The omission of information about the life vests may have resulted in the stark difference between the uptake of those remembering about their floatation seat cushions and those that remembered to take their live vests.
Many found it difficult to locate their life vest
Many decided to not bother finding it, to not delay vacating the aircraft
It is estimated that eventually when outside the aircraft about 33 passengers eventually wore a life vest, thanks to the persistence of the flight attendants who issued them.
Official NTSB Report: http://www.ntsb.gov/doclib/reports/2010/AAR1003.pdf
Flying in an A320 during the initial climb out ADR 1 fails, upon reaching cruising altitude, ADR 3 also fails. We now have a double ADR failure of ADR 1 & 3.
Initial considerations –
Attitude & Speed – Are we flying a safe attitude? IMC or VMC, if in any doubt about attitude or speed information, perform the Unreliable Airspeed memory items.
Auto Flight – ATHR, AP1 & AP2 become unserviceable, leaving you to fly manually.
Flight Controls – Aircraft defaults to ALTN law, protections lost and max speed 320 KIAS.
PFD – Air data information is lost on the Captains PFD.
Landing Gear – landing gear safety valve is closed, leaving gear retraction inoperative and gear extension must be performed by the gravity extension QRH. This is unique to ADR 1 & 3 failing.
EGPWS – ADR 1 failure results in the loss of the EGPWS.
ECAM: Initially a singular ADR failure will result in ECAM requiring the Air Data selector to be turned to the spare, and for the affected ADR to be switched off. When the second ADR fails it also requires the second failed ADR to be switched off. The second failure results in downgrade to Alternate Law.
Flight Directors – Turn Off. By turning them off, it gives the PF the track line to steer to. Remember to turn the Bird on.
Performance – Calculate Landing Distances and VAPP for the failure.
Approach – Make an approach expecting to go into Direct Law when the Landing Gear is deployed.
What did I learn?
The double failure is relatively easy to manage from a technical perspective, but creates non-technical challenges to continue implementing effective CRM skills.
Monitoring: The Captains PFD is left blank, leaving the FO with the only functioning PFD. The creates difficulties for the Captain – the default PNF – to effectively monitor the FO’s raw data flying.
Getting ahead of the game: FCOM 3 provides useful information about the failure, allowing you to expect several consequential failures. For example, by reading it prior to making the approach you are aware that the APP Phase cannot be activated – leaving you without GS Mini and having to fly selected speeds – and that on selection of landing gear you will observe a LGCIU failure. Without expecting these issues, they could cause a distraction on approach.