Wrong Deck Landings
Occasionally, offshore helicopters do land on the wrong offshore installation. On 26 August 2014 the media picked up an such an event that occurred a few days earlier involving a Sikorsky S-92A. So is it a big deal? Well, yes and no! There are good reasons to avoid landing on a deck that is not standing by for an arriving aircraft. These range from the lack of fire cover, the unknown pitch, roll and heave on vessels, cranes may be operating adding to the collision risk, radio silence procedures may be on force on the installation (e.g. when explosives charges are being prepared) and so on. Of course if the deck is open and expecting an aircraft, then there is the risk of a second helicopter, which is actually heading for the right deck! While landing on the wrong deck does increase risk, in practice, with many similar installations in close proximity and various human factor challenges, misidentifications do occur a few times ever year in the UK sector alone. Each needs an appropriate investigation to identify any systemic improvements, but few in reality actually deserve press headlines.
We are not in the position offer an further informed comment on last week’s occurrence. We can however share the story of an investigation we conducted overseas after a wrong deck landing:
Our oil and gas customer had two mobile jack-up drilling rigs on contract from the same drilling company, at different sites in the same field. Although a few miles apart, they were just a few degrees different in heading from the airport onshore. The rigs were of the same type. They were both painted in the drilling company’s standard colour scheme and had similar multi word names. The equally similar rig callsigns were listed one after the other in the database of the aircraft’s navigation system.
Due to the wind direction that day, the approach was also in a direction that meant visibility of the rig and deck markings was dramatically reduced and the helideck crew would have had a limited view of the approach too.
Consequently all the ingredients were in place for a simple programming error and confirmation bias…
Our customer requested we review the air operator’s investigation when it was complete to ensure there was systemic, preventative learning. They were concerned because the ‘next deck’ was over the median line in a less than friendly neighbouring country, where a wrong deck landing would have been a major diplomatic incident!
Perhaps an unintended consequence of this (large) customer taking an interest in the (small) helicopter operator, was that the operator’s investigation was swift and had emphasised ‘who’ made the errors not ‘why’. Consequently, the actions taken consisted of ‘warning letters’ being put in personnel files. Rather than review the operator’s investigation, we therefore had to start from scratch and persuade the operator to focus on WHY (not WHO) and therefore the systemic issues (internal to them and, with our help, externally). To their credit they did and reversed their earlier action, removing the warning letters.
For more background, the Health & Safety Executive (HSE) published a research report in 2000. A valuable resource on investigating human error is The Field Guide to Understanding Human Error Paperback by Prof Sidney Dekker, which we highly recommend.
UPDATE 27 May 2015: Controls against wrong deck landings (‘Threat 12′) are included in the new Flight Safety Foundation (FSF) Basic Aviation Risk Standard for Offshore Helicopter Operations (BARSOHO). Note: UPDATE 1 February 2017: BARSOHO Version 3, fully aligned with the HeliOffshore Safety Performance Model issued in 2016, is now available.
UPDATE 4 April 2016: HeliOffshore is readying a commendably comprehensive study on wrong deck landings with a series of recommendations following an extensive human factors study and both simulator and flight observations.
UPDATE 9 June 2016: The UK Air Accidents Investigation Branch (AAIB) issue their report on their 22 August 2014 incident: AAIB Bulletin: 6/2016 G-VINL and a Southern North Sea WDL: AAIB Bulletin: 6/2016 G-CHBY
The helicopter was operating a multi-sector flight, between rigs located approximately 60 nm north-east of Aberdeen Airport, when it landed on the wrong helideck, which was unmanned. On the third sector, it was required to fly from the Paragon Midwater Semi-Submersible 1 (MSS1) rig to the Buzzard complex helideck, a distance of 7 nm on a track of 205°M. However, the crew misidentified the Golden Eagle complex, on a bearing of 354°M from the Paragon MSS1 at a distance of 3 nm, as the Buzzard complex and landed there instead.
The AAIB report that distracting circumstances occurred that would have raised the stress level on the crew for that sector:
Whilst on the Paragon MSS1 helideck, the co-pilot had some difficulty obtaining the return payloads to Aberdeen from the Buzzard Log operator, but learned that another company helicopter was due to arrive at the Buzzard helideck at about the same time as G-VINL. He eventually established that their anticipated return payload, of three passengers, was cancelled and there would be no passengers on the flight to Aberdeen. Meanwhile, the other helicopter elected to slow down to allow G-VINL to land on the Buzzard helideck first. The discussion between the commander and co-pilot in G-VINL, and the high level of RTF transmissions regarding this issue, were described by the crew as ‘busy and confusing’, with contradictory information coming from Buzzard Log regarding the return loads for each helicopter. The crew discussed the next sector (to the Buzzard complex) and identified the large grouping of platforms ahead of them, in their 12 o’clock, as the Buzzard complex. It looked large enough and appeared to conform to the picture provided in their Helicopter Limitations List (HLL). They were unaware that it was the Golden Eagle complex and not the Buzzard complex.
The crew choose to fly the short sector manually.
At no point was the compass, area Rigmap or the FMS used to confirm the position of the next landing point, although both flight crew were sure that the Buzzard was selected as the next waypoint on the FMS.
While on the approach there were both further unnecessary external distractions and challenges due to the poor positioning of signage:
The crew then carried out the abbreviated Final Approach checklist, the first item of which states ‘Landing Point…..IDENTIFIED’. At this stage, the name painted on the helideck was not visible and the commander advised the co-pilot that confirmation of the helideck name would occur later in the final approach. As the co-pilot performed item 2, arming the flotation equipment, the crew were interrupted by the Paragon MSS1 Traffic operator asking if they had switched to the Buzzard Log. The co-pilot asked him to standby because they were at a critical stage of flight. Whilst the co-pilot was on the radio, the commander recalled identifying the helideck at the far end of the complex, as they had expected, but with its name unreadable at that stage. The superstructure of the complex passed down the right side of the helicopter but, due to the orientation and position of the helideck, they could not see it fully until they were almost alongside.
The operator carried out a prompt internal investigation into the incident and identified a number of potential safety actions for internal consideration. The following are of relevance to this report:
1. Section 8.3.2 of the [company] Operations Manual should be reviewed with regards to using multiple sources of information to confirm navigation.
2. All Checklists (normal and abbreviated) should be reviewed in order to incorporate an action to positively select appropriate navigation aids.
3. The S92 Operations Manual Final Approach Checklist (abbreviated) should be reviewed with a view to incorporating the words ‘HELIDECK NAME…..confirm’ as part of the checks.
4. The S92 Operations Manual Final Approach Checklist should be reviewed with a view to reordering the checks and making the Landing Point check the final action.
5. All offshore Radio Operators should be informed of the significance of landing on the wrong helideck with regards to the fire, crash and rescue cover. They should also be informed of the recommended actions which should be carried out post an unexpected helicopter landing on their helideck.
6. All current [company] pilots are to be made aware of the significance of landing on the wrong helideck by the means of a Flight Safety Circular using this incident as an example. The Flight Safety Circular should include the potential hazards and state the actions required as per the Operations Manual.
7. The training department should review the adequacy of teaching Unintentional Deck landings in both the Initial Line Training Lectures and also in the Command Course Syllabus.’
AAIB report that the AW139:
…was carrying out a scheduled flight, transporting ten passengers to the Normally Unmanned Installation (NUI) A2D, in the Amethyst Field in the Southern North Sea. The flight crew were carrying out line training and inadvertently landed on the NUIA1D, which was a similar platform located 2 nm ahead of the A2D in the direction of flight.
With 2 to 3 nm to go to the platform, level at a height of 400 ft, the Finals checks were completed and the helicopter was turned towards the platform, onto an into wind heading of 060°M. As briefed, the landing was to be carried out by the left seat pilot.
The crew then identified the platform ahead as the A2D, whereas it was, in fact, the A1D. The two platforms appear almost identical and it was reported that, given the distance, they may have misread the name plate on the side of the platform, which was a large rectangular yellow board with the name ‘AMETHYST A1D’ in red letters, mistakenly transposing the number 1 on the name plate for a number 2.
Having made this early identification, the crew then concentrated on flying the approach as part of the training element of the flight. The subsequent landing on the helideck, on which the name is also displayed, was uneventful and the helicopter departed for Humberside after the passengers were clear of the deck. After the helicopter’s departure from the platform, the mistake was identified.
Another technological complexity is that:
There are four platforms in the Amethyst field and the position of each platform is stored in the helicopter’s FMS as ‘user waypoints’. Unless added to the ‘active route’, ‘user waypoints’ do not appear on the Primary Flight Display (PFD). With the radar operating, the PFD shows raw radar returns (from the platforms) and inputs from the FMS, only showing the locations of the platforms loaded as ‘user waypoints’ in the ‘active route’. With the A2D loaded as the destination ‘user waypoint’, the A1D appeared as a raw radar return ahead of it. When the radar was selected to ‘standby’, as part of the Finals checks, the radar return disappeared and only the A2D ‘user waypoint’ remained. With the descent complete and the helicopter level at a height of 400 ft, the A2D was visually obscured behind the A1D.
At this point, the crew were focussed on flying the final approach, as part of the line training, having earlier ‘confirmed’ the platform as their destination. The final opportunity to identify the platform was in the final stages of the approach, when the name was displayed on the helideck. However, the mistake was not noticed. The crew concluded that the wrong deck landing was the result of early identification of the A1D as the A2D, either through not reading the name plate or misreading it. They also considered that a recent intense period of offshore simulator training, where the name on the simulated platforms and vessels was not read, may have been a contributory factor.
Both reports illustrate a range of human factors behind WDLs and emphasise our previous point on the need to do careful human factors evaluations on this type of incident. Hopefully the HeliOffshore recommendations will be enacted in full by all stakeholders.
UPDATE 10 June 2016: Although the UK AAIB reports show some of the complexities that can induce errors, it is exceptionally disappointing to read a new report of the International Oil and Gas Producers Association (IOGP) which includes details of a WDL in Malaysia on 9 July 2015 were the one, so called, corrective action was to ‘terminate’ the crew. In this case in hazy conditions the crew misidentified the destination and were distracted by a “recurring nuisance alarm” on approach.
UPDATE 4 October 2016: HeliOffshore Wrong Deck Landing Report
UPDATE 28 May 2017: For context, a US National Transportation Board (NTSB) report on a Delta Airbus A320 N333NW that landed at Ellsworth AFB rather than Rapid City Regional Airport, South Dakota on 7 July 2016, comments:
A review of wrong airport landing data provided by ASRS revealed that in the previous 20 years approximately 600 wrong airport landings or near landings had been voluntarily reported.
In April 2014 the NTSB had issued a Safety Alert for landings at the wrong airport.
UPDATE 13 July 2017: AAIB issue their report into AW139 G-VINB, landing on wrong platform, Ravenspurn North Platform, North Sea, 20 January 2017. Again multiple human factors combined:
After reporting for duty, a change in the flying programme resulted in a requirement for the crew to fly a seven-sector shuttle at short notice. Whilst the commander reviewed the technical log and discussed some issues with the engineering department, the co-pilot completed the flight planning. The initial plan was to refuel on West Sole Alpha platform but the flight crew surmised that the seven sectors could be completed with round trip fuel.
While re-planning the flight, the fourth destination was incorrectly inserted as Ravenspurn North platform rather than Ravenspurn Alpha platform. The error was not noticed and the flight proceeded to land at Ravenspurn North platform, whose helideck was not manned.
There were a number of occasions when the error could have been picked up; briefing before departure from Norwich, crosschecking the flight plan and payload information, during the radio calls with the Ravenspurn HLO and in programming the FMC. These opportunities were missed through perceived time pressure, differences in codes and possibly confirmation bias in the crew.
Once the mistake was realised, the crew correctly followed the operator’s procedures, waiting on the deck with rotors running until the helideck was manned and they received permission to depart.
Safety actions The operator carried out a prompt internal investigation into the incident and identified a number of potential safety actions. The following are of relevance to this report: 1. Carry out a Flight Planning Software review for robustness and ease of use. 2. Carry out a review of the destination nomenclatures used for planning applications and software. 3. Reiterate to all crews the importance of clear and unambiguous communications. 4. Reiterate to crews the importance of re-briefing all aspects of the flight when a significant change has been applied.’
UPDATE 6 December 2017: AAIB issue the report on an incident when AW139 G-CIPW landed on wrong offshore platform, namely the Viscount Platform in the Southern North Sea on 9 June 2017
The crew approached and landed on what they mistakenly believed to be the correct offshore platform. However, the crew had not appreciated that the planned destination had been changed and, in part, this happened because certain platforms can be referred to using different identifiers. When the crew realised they had landed on the wrong, unmanned platform, they lifted off and continued to the correct platform without further incident.
UPDATE 8 August 2019: The AAIB release their report on a Bristow S-92A that made a wrong deck approach to Brae Bravo, Northern North Sea, 23 August 2018:
The pilots were operating the S-92A helicopter on a multi-sector route between platforms in the Brae field in the northern North Sea, approximately 150 nm north-east of Aberdeen.
This was the 3rd line training flight for the co-pilot following conversion to type. Wind direction dictated an approach flown from the LH seat, so the co-pilot was Pilot Flying and the Commander was Pilot Monitoring.
On the third sector from the East Brae platform to the Brae Alpha platform, the pilots mis-identified the Brae Bravo platform as the destination and made an approach to the hover above the deck of the platform. The radio operator on the Brae Bravo platform told the pilots that they had made an approach to the wrong deck; following clearance to depart, the pilots continued the flight without further incident.
The analysis section of the report looks at the human factors in detail. In conclusion:
This incident arose from the pilots initially misidentifying and selecting the Brae Bravo platform, instead of the Brae Alpha platform, as the destination and subsequently not detecting this incorrect selection. Several prevention controls that should have alerted the pilots to the incorrect platform selection and subsequently aided them in identifying the incorrect-selection proved ineffective.
Contributing factors included platform alignment and characteristics, coupled with inadequate identification by the pilots of the key features and differences of the platforms in the Brae field. The choice to fly the short sector manually and to navigate visually, which was appropriate for the good in-field visibility and this stage of line training, resulted in an increased workload for the PIC, as PM, and reduced the attention given to the electronic cues that existed in the cockpit. The short sector provided a very small window of time for the pilots to identify, select and confirm the destination platform with little subsequent opportunity to review. The inherent nature of the early stages of line training increased the workload on the PIC, as PM, and effectively nullified the protections afforded from operating in a multi-crew environment. This high workload, combined with the overriding influences of expectation and confirmation biases, undermined the ability of the pilots to make the correct identification in the first place and then, subsequently, to trap this incorrect selection.
This event highlights the challenges that exist while flying off-shore, even during benign conditions, during a typical sortie flown in the early stages of line training. Many of these factors highlighted above have previously been identified as typical factors that contribute to wrong-deck landing
The operator stated that it would conduct additional training addressing the task management requirements and complexity during shuttling to prevent a recurrence.
Controls in place at the time proved inadequate to break the confirmation bias of the pilots. Discussion between the pilots about the position of the crane on the 40C platform probably resulted in them switching their attention incorrectly to this platform and away from the 40D platform. The crane was not stowed on either platform, so did not serve as a distinguishing feature. The pilots’ familiarity with the Forties field, the physical similarity of the platforms, and the identical approach and landing flight path to each of them served to reinforce their selection of the wrong deck. The pilots did not verify they were approaching the correct platform by cross-checking the position of the platform against the FMS bearing and distance to the destination or reading the platform name on the helideck before committing to land.
Aerossurance is pleased to be supporting the annual Chartered Institute of Ergonomics & Human Factors’ (CIEHF) Human Factors in Aviation Safety Conference for the third year running. This year the conference takes place 13 to 14 November 2017 at the Hilton London Gatwick Airport, UK with the theme: How do we improve human performance in today’s aviation business?