Unpacking the Last Flight of Pawan Hans Dauphin VT-PWA

What happened?

On Jan 13, 2018, at about 0443 UTC (1013 IST), VT-PWA, an AS365N3 Dauphin helicopter belonging to Pawan Hans Ltd (PHL) took off from Juhu (VAJJ) for NQO process platform in Bombay High. In the cockpit were two seasoned offshore captains. It was a ‘VIP flight’ – five DGMs from Oil & Natural Gas Commission (ONGC) visiting offshore facilities were on board. Weather was typical wintry Mumbai; hazy with poor slant visibility (3000 metres). The helicopter departed under “Special VFR” dispensation that allows Visual Flight Rules (VFR) flights to operate below VFR minima for the purpose of entering / leaving an aerodrome control zone.

Helicopters operating ex-Juhu have to follow VFR routes (known as ‘Kilo’ or ‘Kopter’ routes) promulgated by Airports Authority of India (AAI). Kopter Route K17 catering for Runway 09 at Mumbai (VABB) was in use that morning. VT-PWA tracked out on radial 330 for 25 nm (330-25) as per standard procedure. At 330-25, it changed over from Juhu to offshore frequency, entering uncontrolled airspace between sea level to 3000 feet used by offshore traffic. The next reporting point was point PAPA, west-south westerly from 330-25. As per standard procedure, the helicopter turned left towards PAPA and commenced climb to cruise altitude of 3000 feet.

When the helicopter became overdue at PAPA, the first level of alert was raised by ONGC radio officers through aircraft in vicinity. There was no reply from VT-PWA. Soon, it was evident that the helicopter was missing when its AIS return dropped off the ‘Ship Plotter’ display of ONGC monitoring stations. (AIS = Automatic Identification System, used for tracking merchant ships through secondary radar principle; also used for ONGC helicopters).

Screenshot of VT-PWA on Ship Plotter display of ONGC shortly before it went missing (Pic from VT-PWA accident report)

Investigations revealed that it took less than 40 seconds for VT-PWA to plunge from cruise altitude of 3000 feet to the Arabian Sea below. The surface of water can be hard as rock when you arrive at a rate of descent in excess of 6000 feet per minute. The crash was unsurvivable – it was instant death for all seven onboard. The final moments reveal a doomed scenario with both crew fighting for controls. Investigation revealed that the pilot-in-command (PIC) was under the death grip of an old killer: spatial disorientation (SD).

SAR was activated through the Maritime Rescue Coordination Centre (MRCC). ONGC, navy and coast guard assets swung into action. All seven dead bodies were recovered. Analysis of the wreckage revealed high-energy impact with water in a nose-down, right-banked attitude. There was no evidence of pre-impact damage, critical unit failure, fire or in-flight disintegration. The autopilot (AP) upper modes were disengaged and the aircraft was hand-flown in an unusual attitude into the sea.

Accident Report Reveals Glaring Similarities

Indian Aircraft Accident Investigation Board (AAIB) published the ‘accepted report‘ in Jul 2019, almost 18 months after the accident. The 87-page report holds nothing new or unforeseen in the accident-ridden history of PHL. In Nov 2015, another Dauphin N3 (VT-PWF) had crashed during a night training flight off Bombay High. Both pilots sustained fatal injuries (there were no pax onboard). The PIC’s body was never found.

The similarities between both ill-fated flights VT-PWF and VT-PWA are hard to miss. To name a few:

  • Both flights were operated by highly experienced aircrew from PHL.
  • Spatial disorientation (SD) was cited as ‘probable cause’ in both crashes.
  • Break in flying for Pilot Flying (PF) was noted in both cases. PF of VT-PWF was flying by dusk/night after almost a year. PF of VT-PWA had numerous breaks in flying before the accident.
  • Calls by Pilot Monitoring (PM) went unheeded by PF during the critical moments before crash.
  • Both flights were operated under VFR in low visibility / Instrument Meteorological Conditions (IMC) where SD is an ever present danger.
  • Inappropriate or insufficient use of available automation was a predominant factor in both crashes.
  • Standard patter and recovery drill for incipient SD in a multi-crew cockpit was missing in both instances.
  • Recovery action was either not initiated or initiated late, leading to a “too little too late” situation in both cases.
  • Both helicopters smashed into the sea in a high speed, banked attitude, hands-on with upper modes of AP disengaged.

Findings of both investigations reveal many areas of overlap, highlighting grey areas of PHL that is neither novel nor surprising. The reports are even comparable in volume (VT-PWF:84 pages, VT-PWA:87 pages), indicating that little, if at all, has changed at PHL between the two fatal accidents. Two out of three members of the AAIB were common to both investigation teams.

The two crashes claimed a total of 9 lives (4 PHL +5 ONGC personnel). ONGC holds 49% stake in PHL. Indian government’s efforts to privatise PHL have not yielded any results so far.

Unpacking the Accident Report

The board that investigated VT-PWA has borrowed a quote from J. Moloney’s paper, ‘Error modelling in anaesthesia: slices of Swiss cheese or shavings of Parmesan‘, published by the British Journal of Anaesthesia, Volume 113, Issue 6, December 2014, and twisted it awkwardly to state: “Accidents occur through the concatenation of multiple factors, where each may be necessary and at the same time are sufficient enough to produce the accident.

The actual line from Moloney’s abstract however reads as follows:

Accidents in complex systems occur through the concatenation of multiple factors; each may be necessary, but only when all are present are conditions sufficient to produce an accident.

This is consistent with James Reason’s “Swiss Cheese Model” of accident causation quoted by Moloney wherein “accidents could be seen as the result of interrelations between real time ‘unsafe acts’ by front-line operators and latent conditions. Holes can appear in multiple levels of the system. When these holes line up, as in multiple slices of Swiss cheese, an accident can occur.”

Screenshot of Swiss Cheese model of accident causation from VT-PWA accident report

The corollary that flows out of this theory is that if any of these “holes” were to be removed or the slices of cheese realigned, an accident could well have been prevented. If successive accident reports are analysed, PHL’s model would yield more holes, less cheese. A sickening rate of accidents is but expected from such a flawed system.

Here’s a selected list of cheese slices with their gaping holes that aligned that day to bring down VT-PWA.

Latent Failures of Management & Regulatory Oversight

The report notes that “recommendations of earlier investigations having serious operational/ safety implications are yet to be implemented by PHL in true spirit”. AAIBs have repeatedly pointed out that “there was no supervision of the operational activities on day to day basis as no senior person having adequate authority was available” at PHL. The complicated ‘organogram’ of aviation safety at client ONGC’s end “could not capture any latent hazard or contribute towards safe operation as viewed from the SMS point of view”, the board observed.

Even when interim observations from VT-PWA crash were passed on to PHL for corrective action two months after the crash, the organisation “went into denial mode instead of taking these observations in a positive sense”, the board noted. Supervisory functions were being carried out as a formality. SMS, SOPs and rule books had turned into ‘shelfware’ nobody really cared about. Results of an audit of PHL undertaken by DGCA in Oct 2017 were not intimated to the operator even till the date of accident, three months later. This incestuous relationship between three government bodies set the stage for a calamitous outcome.

Slop Chit Syndrome

VT-PWA had a history of defect in the autopilot (AP) system that was neither documented nor addressed. The helicopter was grounded through the month of Dec 2017 till Jan 6, 2018 for 600-hourly inspection. The AP defect, wherein the aircraft would keep drifting slowly with ‘heading’ (HDG) upper mode engaged, existed prior the inspection till final test flights post the 600-hourly inspection. Yet, nothing was done to rectify the defect.

The aircraft crashed on the first revenue flight after the test flight, with the autopilot defect acting as tripwire.

A system of verbal reporting of defects was in force at PHL that subverted operational and maintenance checks and balances. The board noted that “the test flight was carried out and test performa was filled in casual manner.” Such a “Slop chit syndrome” was highlighted by this author in a 2016 post. The board noted that “investigation was carried out in a shallow manner without any root cause analysis” for incidents preceding the ill-fated flight. FDR data analysis and monitoring was weak, incapable of capturing hazards or unsafe conditions.

It is thus plausible that PHL helicopters were flying revenue sorties with known anomalies hiding under a parallel, devious snag reporting culture.

System of Liquidated Damages

Helicopters are provided to ONGC through contracts by companies such as PHL, Global Vectra Helicorp Ltd (GVHL), Heligo Charters, etc. Almost all these operators have raised concerns about the hourly liquidated damages (LD) clause that imposes heavy penalties on operators not being able to provide helicopters as contracted. The intensity of offshore flying combined with limited number of choppers, essential downtime for maintenance, snag rectification and global supply chains for spares puts enormous strain on operators. Operators resort to suppressing minor defects or cannibalising spares and rotables from other aircraft under such circumstances. The board grimly notes that “there is preponderance of unsafe situations due to this (LD) clause”.

Incentivising Wrong Metrics

Though not highlighted in the case of VT-PWA, a system of incentivising pilots on ‘per flying hour basis’ has been noted as a contributory factor in many PHL accidents. Compensation based on number of landings and flying hours per month allures pilots to fly when they shouldn’t. This was compounded by a steady attrition of pilots who left PHL for greener pastures. Throwing money at a problem without solving complex underlying issues often leads to unsavoury outcomes, as PHL history will show.

Maintenance Shortfalls

The AAIB’s observation on the prevalent maintenance culture at PHL is self-explanatory and needs no further elucidation:

“The helicopter had undergone 600 hours inspection which is one of the major inspections. One of the B1 engineers of the operator was told verbally to carry out the 600 hrs maintenance of the aircraft. There was no briefing by Continuous Airworthiness Manager (CAM). No evidences were available that CAM had carried out regular or random supervision. So, there was lack of due diligence on ground in the maintenance process especially to identify defects.”

Company Resource Management

The VT-PWA accident report reads like a perfect “what not to do” checklist for resource management and rostering. The test flight after major inspection was flown by a different crew with a copilot who had recently qualified on type. There was no briefing by engineering neither did the crew ask for anything. A blank test performa, duly signed, was returned at the end of the sortie. The board noted that “entries were made and signed for the checks which were not carried out during that test flight”. The seemingly innocuous defect pertaining to ‘HDG’ mode of autopilot was neither reported nor rectified. It slipped through the cracks to meet the ill-fated crew of VT-PWA another day. The PF that day had frequent breaks in flying and had not flown VT-PWA for over a month. It is quite likely that the crew were unaware of the latent but recurrent defect in autopilot and had no handover, either verbally or through documentation.

Weather Conditions

All offshore flying ex-Juhu is VFR and goes on unabated 365/7 till visibility drops below 1000 metres or the airfield gets flooded. Winter and monsoon months routinely see visibility at Juhu fall below 5 kms – the minimum for VFR flights. Seeking “special VFR” departure or arrival with a 10-minute separation is not unusual at Juhu. Once the helicopter departs Juhu control zone under this clearance, it may have to negotiate patches of weather and visibility well below that minima. A helicopter that was 10 minutes ahead of VT-PWA that morning reported haze and no discernible horizon when radio contact with PWA was lost.

The final holes in the Swiss cheese were falling into alignment.

Spatial Disorientation

There are only two types of pilots: those who have experienced SD and those who will. In a recent, hard-hitting article, Elan Head from Vertical Magazine wrote why helicopter crashes like Kobe Bryant’s keep happening.

The phenomenon of spatial disorientation is taught to every pilot in basic flight school. SD does not go away with flying experience. In fact, experienced crew may well be more susceptible to SD owing to their dogged reluctance to accept and acknowledge onset of SD in time. When SD happens, it is like “toppling off the edge of a cliff”, in Elan’s words. It is rare that both pilots would experience SD at the same time. The only safety is in slavishly believing instruments over illusions, constantly scanning instrument indications, appropriate use of available automation and, if required, handing over controls to the other pilot.

Malfunctioning of a particular channel of autopilot (HDG in the case of PWA) does not preclude use of other channels and modes. Flying hands-on or use of SAS mode in IMC conditions comes with its own perils as described in another article this author wrote for Vertical Magazine last year. The deadly phenomenon of SD, exacerbated by the tendency to fly hands-on in a debilitating situation, sent VT-PWA spiralling all the way down from 3000 feet into the sea below.

The Final Moments

The last holes in Swiss cheese aligned this way:

On Jan 13, 2018, after slipping through many weak defence mechanisms and latent failures, VT-PWA departed Juhu with two crew and five VIP passengers, under “Special VFR” clearance, with a PIC who had just returned from a break in flying, a defective autopilot nobody onboard knew (or cared) about, on a less-than-perfect day where they needed automation and crew coordination the most.

The first signs of AP malfunction showed up at 700 feet on the outbound leg to 330-25 where it is likely that some horizon reference or ground contact was still available. The heading drift was corrected. Once the helicopter left 330-25 and climbed to cruise level of 3000 feet, Instrument Meteorological Conditions (IMC) replaced Visual Met Conditions (VMC).

Onset of spatial disorientation was aided by the slow-drifting autopilot. Once the illusions set in, PF acknowledged the Pilot Monitoring’s (PM) alerting calls, froze on controls, and failed to use appropriate automation or handover controls to PM in time (CVR transcript, pg 48 to 52 of accident report refers). The helicopter spiralled down within seconds, both pilots fighting over controls, not following any standard protocols for SD / unusual attitude recovery. In the steep spiral descent, airspeed, load factor and rotor RPM limits were breached. The helicopter slammed into the sea in a high-speed, nose-down, right-banked attitude about 15 miles off the coast.

It was all over in less than 40 seconds.

In Conclusion: An Avoidable Accident

John F Kennedy Jr could still be pardoned for the crash of his single-engine Piper Saratoga that killed him, his wife Carolyn Bessette Kennedy and her sister Lauren Bessette on July 16, 1999. He was the only pilot onboard that hazy, dark night with a meagre flying experience of 310 hours. Spatial disorientation was named as the most probable cause of that accident by NTSB. Contrast this with VT-PWF and VT-PWA where the total in-cockpit experience of  crew exceeded 20,000 hours.

Despite the grim statistics, India’s largest helicopter operator continues to flourish and beseech investors. Every tenet in aviation safety has been systematically undermined at PHL to a point where it ceases to have any meaning. A former employee of PHL who did not wish to be named had this to say:

“Indifference from top to bottom has become a way of life at PHL; one that often breeds a nonchalant or ‘cowboy’ attitude at the frontline, sometimes almost unrecognisable from a death wish”.

To put the lid on VT-PWA crash by attributing it to spatial disorientation is to go for the low hanging fruit. It does nothing to mitigate or prevent recurrence of such accidents. Compared to the gross negligence at various echelons that culminated in this accident (as many in the past), AAIB’s safety recommendations at the end of VT-PWA report appears too watered down, insipid and repetitive to be even taken seriously at PHL.

Culture cannot be mandated. Nothing short of a purge, a deep and wide cleansing through rank and file of PHL, is required if matters are to improve. Any entity that ventures to invest in PHL cannot afford to ignore this elephant in the room.

Truth be told, if such an accident were to happen with a private operator, they would sooner be shuttered by the powers-that-be (like Mesco after the Aug 11, 2003 offshore crash of a Mi-172 that killed 23 ONGCians and 4 crew). But this is PHL. It is business as usual. And that, to me, is the biggest shame.

True to an old aphorism, you can only wake up a sleeping person, not one who is pretending to be asleep.

Blue skies.

***********

(An edited version of this story was published by VERTICAL magazine on Apr 23, 2020. It can be accessed here.)

©KP Sanjeev Kumar, 2020. All rights reserved. I can be reached at kipsake1@gmail.com. Cover photo of PHL Dauphin used for representation only. References are linked below for further reading.


References:

  1. Accepted report VT-PWA, issued by AAIB, India.
  2. Accepted report VT-PWF, issued by AAIB, India
  3. 2018 chopper crash: Pilots disagreed over control, shows probe, Times of India, Nov 23, 2019
  4. Aviation Safety Network, Jan 13, 2019
  5. J. Moloney, Error modelling in anaesthesia: slices of Swiss cheese or shavings of ParmesanBJA: British Journal of Anaesthesia, Volume 113, Issue 6, December 2014, Pages 905–906, https://doi.org/10.1093/bja/aeu223
  6. Article on plane automation crashes, Popular Mechanics, Mar 18, 2019
  7. William Harris “How Autopilot Works” 10 October 2007. Linked here.
  8. NTSB news release: Final report on investigation of crash of aircraft piloted by John F. Kennedy Jr.
  9. NTSB Says Disorientation Likely Caused JFK Jr. Crash, The Washington Post, Jul 7, 2000.
  10. Slop Chit Syndrome, by KP Sanjeev Kumar, Kaypius.com, Jan 27, 2016
  11. Why do helicopter crashes like Kobe Bryant’s keep happening? by Elan Head, Verticalmag.com, Apr 7, 2020
  12. News article ‘ONGC Chopper Hiring Process Was The Culprit‘, Financial Express, Dec 11, 2003
  13. News article ‘Pawan Hans stake sale: Government extends EoI deadline by a week‘, Economic Times, Sep 11, 2019
  14. News article ‘3rd EoI date extension casts doubt on Pawan Hans disinvestment‘, OutlookIndia.com, Sep 24, 2019
  15. Perneger TV. The Swiss cheese model of safety incidents: are there holes in the metaphor?BMC Health Serv Res. 2005;5:71. Published 2005 Nov 9.
  16. Brochure on Spatial Disorientation issued by FAA
  17. The issue of automation in aviation, by KP Sanjeev Kumar, Vertical Magazine, Oct 23, 2019

8 thoughts on “Unpacking the Last Flight of Pawan Hans Dauphin VT-PWA

  1. Sir, you have the gift of putting technical jargon into simpler words. Not to mention, the superb analysis.
    Non aviator here yet I read it in one go.

  2. In one of the old movies of the late actor Rajesh Khanna, “Amar Prem”, there is a timeless classic song “chingari koi bhadke…. ” sung by the legendary Kishore Kumar. A line from that song’s lyrics “मझधार में नैया डोले, तो मांझी पार लगाये, मांझी जो नाव डुबोये, उसे कौन बचाये ?” seems to resonate with the organisational culture that has been identified by the Investigation Board to be responsible (amongst other factors) for the series of fatal accidents.
    Whilst the above analogy may seem extreme as one could argue that “it’s not as if the organisation planned these accidents”, the very act of not learning from previous accidents (and taking corrective action) amounts to setting up the foundations for another accident to occur/ ensuring that holes in the cheese start aligning.
    Hats off to you Sir for the wonderful analysis !!

  3. I used to work in a factory for a Japanese car manufacturer. You could see strict adherence to safety procedures from the lowest line members to the top managers and everyone in between. That’s when you realize how important safety culture is. It must stem from the top management.

  4. Many Army & Airforce Pilots failed to recognize the danger of flying in Offshore whereas Naval Pilots don’t go overboard when flying in hilly terrain. Learn to accept the comfort zone of a profession,wherein experienced Pilots from land with less than 5 years of flying over Sea start taking it very easy & lightly. Dangers & perils of Ocean flying is like Chakrvuyuh wherein entry is easy but getting away is sealed mostly like Abhimanyu’s fate…SAD it’s for attractions to pecuniary benifts of Offshore that kills.

    1. Very true. I have done both, SEA flying with Navy & HILL flying with Army. Did two tenures with Army flying over Sikkim. Second time I had maximum qualification on helicopters with close to two thousand hours and highest Instrument rating of MASTER GREEN held for last six years. However I could not get all helipads cleared though the instructors were more than keen to clear me seeing my loads of experience. As Hill flying was a very difficult and different ballgame all together I refused to rush and got cleared only when I was confident. This attitude helps me see yet another morning day after day.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.