The Pel Air Westwind Ditching
The facts of this event are quite straightforward. On 17 November 2009, an IAI Westwind aircraft was dispatched from Sydney to Samoa on an aeromedical retrieval flight (ATSB, 2012). Onboard were the crew of two pilots, a doctor and a flight nurse. The route included a stop at Norfolk Island for refuelling before continuing on to Samoa, where it arrived at about 7 am local time. Therefore, this was a flight through the night. The crew went to a hotel to rest prior to the return flight, scheduled to depart at 1830 local later that day. The crew had less than 12hours on the ground.
An hour before departure, the captain attempted to file the flight plan for the return journey but discovered that the internet service on the island was not working. He called the company’s main office but could not get through to the duty manager. He then called the briefing officer of Airservices Australia, and between them, they created a flight plan based on reversing the one submitted the day before. The pilot was given the latest weather forecast for Norfolk Island, which was acceptable for a landing at the planned time of arrival but the briefing officer did advise of a trend in the forecast from 1500 UTC, but the actual details were not requested. The captain then arranged for the aircraft’s main fuel tanks to be filled. The aircraft did have additional wing tip tanks, but these were left empty as the captain deemed that the weather at Norfolk Island did not require additional fuel for a diversion to an alternate. The first officer (FO) was not present during any of the planning but did look at the notes the captain made about the Norfolk Island weather. She was also advised about the fuel. There was a slight delay with loading the passengers, and the aircraft finally got airborne 15 minutes late.
The crew did not update the weather en route and, by the time they arrived at Norfolk Island, the weather was below minima. Four approaches were made before the captain decided that they should ditch the aircraft just off the coast at around 2200 local time. Both the crew and the four passengers were rescued. The aircraft wreckage and flight data recorders were not recovered on cost grounds, a decision partly influenced by the fact that all the witnesses to the event survived.
In the aftermath, the pilot was proclaimed a hero and comparisons were made with the ditching of the A-320 in the Hudson River in January of that year. He was also something of a celebrity, having been a finalist in an Australian reality TV show, ‘Bachelor of the Year’. Therefore, media attention was intense. However, in December 2009, barely a month after the crash, the Australian Civil Aviation Safety Authority (CASA) withdrew the pilot in command’s license. A preliminary report into the investigation was published by the Australian Transportation Safety Bureau (ATSB) in January 2010 stating that:
The investigation is continuing and will include further examination and analysis of the:
Within a matter of weeks, then, the regulator took punitive action despite the fact that the accident investigation was still ongoing. It is a tenet of safety investigation that its purpose is not to assign blame, but regulators do have a responsibility to consider possible safe courses of action after an adverse event. Pending the outcome of an investigation, any response runs the risk of being either premature, wrong or both. Nonetheless, CASA chose to act against the captain, but no action was taken against the FO.
The final 78-page full ATSB report was published in 2012 (ATSB, 2012). Having stressed that the findings should not be seen as attributing blame to individuals or organisations, the report contained the following findings as to cause:
Contributing safety factors
• The pilot in command did not plan the flight in accordance with the existing regulatory and operator requirements, precluding a full understanding and management of the potential hazards affecting the flight.
Other Safety Factors
Other key findings
The process for releasing an ATSB report includes an early sight of the text by interested parties before public release. The first draft was sent to those interested parties - the operator, the pilot and CASA - on 26 March 2012. Two rounds of consultation and drafting were completed and the final version was released to the public on 30 August 2012. Publication of the final report triggered something of a furore. A television documentary was broadcast on 3 September 2012 that was very critical of both the ATSB and CASA. The mere fact that the documentary was broadcast so soon after the publication of the report suggests that it had been in production, probably, since the initial draft was released. In particular, the apparent emphasis on the actions of the captain, with little discussion of the role of the operator or CASA, was considered biased. Suspicions had been raised when it emerged that a CASA audit conducted immediately after the crash had not been shared with the ATSB.
Pel Air had been subject to an audit by CASA prior to the accident and deficiencies had been found in a number of areas. Immediately after the crash, a special audit of the company was initiated and report produced on 1 August 2010. An additional review of ‘lessons learnt’ from a regulatory perspective was commissioned at the same time. CASA oversight procedures called for a variety of checks and audits to be conducted against a schedule. For example, a site inspection was supposed to be completed each year, a systems audit was required every 3 years, check and training and supervisory pilots were to be observed on initial award and subsequent renewal of approvals. Ramp checks and observations of operational flights were to be done on an ad hoc basis.
A system audit was done in 2006 when Pel Air merged its air operator certificate with another company. Audits were conducted again in 2008 and 2009 because of safety concerns. A site inspection was undertaken in 2006 but not again after that. Six-monthly ‘Safety Trend Indicator’ exercises were completed on 60% of the required occasions between 2006 and 2010. These exercises are supposed to alternate between a desktop telephone interview and then combined with a site inspection. Observation flights were done with management and training pilots, but no observation of line pilots was done. During this period, 31 requests for corrective action were issued, together with one safety alert.
The management report observed that the inspector assigned to Pel Air was not current on the aircraft type although specialist Westwind Inspectors were based at other offices. A team of 16 CASA inspectors descended on Pel Air for the special audit after the ditching, but the audits done in 2006 and 2007 made use of just three CASA staff. The available inspector establishment was reduced by two after a reorganisation in early 2009 and, in addition, the local office was understaffed by two. It was 25% behind on its surveillance targets at the time of the review. The report makes two interesting observations about the behaviour of inspectors. First, the process of signing off on requests for corrective action was based on the operator submitting a plan for future remedial action and not necessarily on an inspection of actual remedial action. The second observation related to the unwillingness of inspectors to create tension by digging too deep and, therefore, asking too many difficult questions. The Pel Air inspector was considered to have a good relationship with the company and, on that basis, it was assumed that CASA had clear oversight of the operation. The reality was much different. In fact, the inspector did not know that Pel Air was routinely conducting medical flights to Samoa. Oversight of the operation, then, had been suspect for a number of years prior to the accident.
Within days of the broadcast, a Senate Inquiry was convened, which reported in May 2013 (Commonwealth of Australia, 2013). The inquiry report recommended, among other things, that the aircraft wreckage be recovered so that the flight data recorders might be examined. An independent audit of the ATSB procedures by a foreign investigation agency was requested and duly conducted by the Canadian Transportation Safety Board (TSB, 2014). The TSB report was published on 1 December 2014 and was broadly supportive of the ATSB.
The Senate Inquiry drew attention to an earlier coroner’s investigation into a fatal accident at Lockhart River on 7 May 2005. The 2 pilots and 13 passengers were killed when the aircraft struck the ground. The Queensland coroner’s report (2007) challenged the relationship between CASA and the ATSB and recommended that a new MoU be drawn up between the two organisations. The Pel Air investigation was being conducted as this MoU was being finalised. The TSB audit report observed that a structural reorganisation that increased the independence of the ATSB as an agency was in progress but that, coupled with the fact that the MoU was still in negotiation, the lines of accountability between CASA and the ATSB were blurred. The Senate report suggested that there was collusion between the two agencies in the drafting of the first accident report in order to minimise possible criticism of CASA’s role in the oversight of Pel Air, although the Senate report did state that no evidence to this effect could be found; the suggestion was, nonetheless, left in play. One recommendation of the Senate report was to, again, redraft the MoU.
The ATSB report of the ditching was reissued on 23 November 2017 (ATSB, 2017), now running to over 500 pages. In relation to the accident, it stated:
Contrary to the consistent practice of the operator’s Westwind fleet, the longdistance flight to a remote island aerodrome departed without uploading the maximum possible amount of fuel prior to departure. Had the flight departed with the maximum amount of fuel, it is very likely the aircraft would have had sufficient fuel to divert from the top of descent or to hold at the remote island for a significant period of time.
The captain’s pre-flight planning did not include many of the elements needed to reduce the risk of a long-distance flight to a remote island or isolated aerodrome. Limitations included:
The original report identified 12 factors associated with the event. The final report identified 30 additional findings. However, at the core of the report, the performance of the captain was still considered fundamental to the outcome.