On 12 March 2009 a Cougar Helicopters Sikorsky S-92A on a flight to the Hibernia oil production platform had a total loss of oil in the transmission’s main gear box. The flight crew descended to 800 feet and headed towards St. John’s. Approximately 35 nautical miles from St. John’s, during an attempted ditching, the helicopter struck the water in a high rate of descent. One passenger survived with serious injuries and the other seventeen occupants of the helicopter died of drowning. Mike Reyno Photo
Sticking with a policy it has had in place for some time now, the Transportation Safety Board of Canada (TSB) has refused to cite “pilot error” as a factor in the March 12, 2009, loss of a Cougar Helicopters Sikorsky S-92 off the coast of Newfoundland.
“That phrase . . . is not in our lexicon,” TSB chair Wendy Tadros replied flatly when the question of pilot error was raised by a reporter after the agency had released its final report on the accident.
Investigator-in-charge Mike Cunningham pointed out in a subsequent interview with Vertical that the TSB, like many of its counterparts in other countries, has essentially phased out the term. “The art and science of accident investigation had determined a long time ago that these complex occurrences often have a lot of underlying factors – in particular, organizational issues which are beyond the control of the flight crew.”
Instead, Tadros, during the news conference for the report, said 16 different factors contributed to the crash and its outcome: “You can’t look at any single one.”
What Happened
All but one of the 18 people on Cougar Flight 91 (also referred to as Flight 491) drowned almost immediately after impact in near-freezing water, 11 minutes after the aircrew advised air traffic control they were issuing a mayday and returning to base (see p.30, Vertical, June-July 2009).
Cunningham told reporters that the aircraft had descended to 800 feet above sea level, from its normal cruising altitude of 9,000 feet, at which point: “something dramatic happened that triggered the crew to decide to ditch . . . .” The flight data recorder stopped working at an altitude of 90 feet; by then the aircraft had a rate of descent of 2,300 feet a minute. The impact seconds later was so severe it compressed the cabin roof down to the headrest level and caused extensive damage to the entire cabin structure. With the emergency floats disabled, the aircraft sank rapidly. While all aboard survived the initial impact due to the S-92’s crashworthiness features, only two escaped, one drowning shortly afterward.
The instigating factor of the crash was the failure of two of the three titanium mounting studs on the main-gearbox oil filter bowl. This resulted in a catastrophic oil leak that caused overheating of the pinion gears, which then led to a loss of power to the tail rotor. “A tail rotor failure is one of the most difficult emergences to control and this was the first documented case of one in an S-92,” said Cunningham. “The crew had no option but to shut the engines down and begin an emergency power-off landing. They did their best, faced with an extremely challenging situation.”
The stud failures were confirmed by the TSB just a week into its investigation. That information was then immediately relayed to other regulators and the aircraft manufacturer. Since then, said Tadros, the titanium studs have been replaced with steel ones on all S-92s worldwide.
But, she added, “This tragedy was about more than failed titanium studs,” explaining that as the investigation proceeded, TSB “found many underlying problems.” Among other things, these discoveries led to aviation safety advisories on crew helmets, survival suits, the automatic oil bypass system and the emergency flotation systems. In all, said Tadros, “We have identified 16 factors that contributed to this accident. Take any one of them out of the equation and we likely would not be here today.”
One factor was the potentially confusing descriptions in the flight manual regarding gradual and total oil loss, which Cunningham suggested had delayed the crew’s response to the warning light. “In addition, the loss-of-oil procedure emphasized that the crew should expect to see an increase in oil temperature. Instead, the crew saw a normal temperature indication.” Appropriate procedures were not covered in the manual – which has since been revised – or during simulator training, which led the crew to suspect a sensor or pump problem. Also, there was not any secondary indication of trouble, such as the vibration they had experienced in simulator training. “The Flight 491 crew was seeing something for the first time,” said Cunningham.
Training and Experience
The captain of Flight 91/491 had flown for another local company on single-engine helicopters from 1996 to 2003, joining Cougar as a first officer in January 2005. He had 5,997 flying hours in total, including 1,061 on type, and had been a captain on the S-92 since September 2007. The first officer, who flew Sikorsky CH-124 (civil S-61) Sea Kings for the Canadian military before joining Cougar in April 2008, had 2,584 total flying hours, including 94 on the S-92.
During the news conference at which the report was released, Cunningham acknowledged that the pilot and co-pilot had done their utmost in trying to deal with a cascade of issues that the TSB said contributed to the crash. He then echoed that in the interview with Vertical, saying that the captain and co-pilot clearly had used all “the tools and knowledge and skills that they had,” including the rotorcraft flight manual and their simulator and cockpit/crew resource management (CRM) training.
Interestingly, despite a TSB recommendation in 1995, Canadian Aviation Regulation (CAR) subparts 703 (air taxi) and 704 (commuter) do not require CRM training for services such as those offered by Cougar. Transport Canada so far has mandated it only for airlines (CAR subpart 705) and the latest word is that although it does plan to extend the requirement to CAR 703 and 704 operators, there is no indication how long the regulatory process might take.
To give Cougar credit, Cunningham pointed out that the company does provide CRM training, which he said is “actually pretty good,” but the accident report highlighted weaknesses in how the crew responded, and were eventually overwhelmed by cascading events. “Modern CRM instruction will provide the tools and knowledge which are intended to mitigate those types of breakdowns.”
Inherent Issues
While the crew was experiencing the situation for the first time, Mark Clitsome, TSB director of air investigations, told reporters it wasn’t the first time a problem with the titanium studs on the S-92 had arisen. An oil supply problem on an aborted Australian offshore flight in July 2008 was reviewed by the TSB’s Vancouver, B.C., office at the request of the aircraft’s operator, Richmond, B.C., based CHC Helicopter. An Australian engineering firm concluded that two of the titanium studs had fractured due to metal fatigue probably caused by galling (metal wearing). Clitsome said that happens when the nuts are removed and re-installed during oil changes, which have been more frequent than the original S-92 design called for. “Titanium, for all its advantages – lightness and strength – is relatively poor at withstanding wear and tear from sliding contact.”
The results of the TSB’s review were forwarded to the United States Federal Aviation Administration (FAA) – six months before the Newfoundland accident – and the FAA circulated a warning notice to all operators. The damaged studs were shipped to Sikorsky, which issued a January 2009 service bulletin requiring the titanium fasteners be replaced with steel ones within 1,250 flight hours or one year.
Clitsome said that although Cougar had ordered the steel studs before the accident, the company “had not, however, effectively performed enhanced inspection and maintenance procedures,” which meant that “damaged studs were neither detected nor replaced.”
Tadros said that while Sikorsky, Cougar and the regulatory bodies have all taken action and that “the causes of this specific accident have largely been addressed,” a look “at the bigger picture” disclosed that other risks still remain. One of those involves the certification standards, specifically: for how long should helicopters be able to fly after the loss of main gearbox oil? Said Tadros, “The chances of this happening were once considered ‘extremely remote.’ Tragically, this was not the case.”
Clitsome said one of the FAA certification requirements for Part 29 helicopters such as the S-92 is that the main gearbox be able to run dry for 30 minutes, a time window “to enhance landing opportunities after a massive oil loss.” While this originally was a military requirement, the FAA opted to require it of civil aircraft, too. However, when early run-dry tests on the S-92 yielded a catastrophic failure after only 11 minutes, Sikorsky and the FAA decided a total loss would only happen if the oil cooler system failed, a prospect they saw as “extremely remote.”
Consequently, rather than redesign the gearbox, they opted for an oil cooler bypass valve. “What they did not consider was a failure in the main gearbox oil filter bowl – or its titanium studs,” explained Clitsome. The fault was “with the rule and this hasn’t changed,” he said, noting that the S-92 was the only helicopter to be certified by the FAA with this “extremely remote” provision.
European regulators have since certified four helicopters as capable of meeting the 30-minute run-dry rule and the FAA and Transport Canada, one each. This means, said Clitsome, that “it’s possible for civilian aircraft to meet this requirement.”
The Recommendations
In her closing, as one of four recommendations, Tadros said the “extremely remote” provision needs to go, and that all Category A helicopters should be able to fly for at least 30 minutes following a massive loss of main gearbox oil. Next, she said regulators should look at hostile operating environments such as the Atlantic offshore region, the Arctic and the North Sea “and decide whether even 30 minutes is enough time.”
Third, in a rough-water ditching, Tadros said the emergency flotation system should keep the aircraft afloat long enough to enable aircrew and passengers to evacuate safely. “If it can’t do that – if a helicopter isn’t up to the task – it shouldn’t be operating.”
Finally, the TSB also addressed the issue of drowning, the leading cause of death in ditchings. Tadros said that while underwater breathing apparatuses now are available on all East Coast flights where survival suits are mandated, the policy should be extended nationally.
The full TSB report is available at http://www.bst-tsb.gc.ca.
Unresolved Issues
Following the release of the Transportation Safety Board of Canada’s report on Cougar Flight 91/491, the sole survivor of the crash and the families of the victims killed have asked federal Transport Minister Chuck Strahl, to investigate his department’s certification of the Sikorsky S-92.
The letter, sent by survivor Robert Decker and 15 others, calls for an investigation of the Transport Canada certification process. It contains a detailed chronology of the certification process and the inability of the S-92 main gearbox (MGB) to meet the 30-minute run-dry standard. The letter stated, “Despite the fact that the S-92 MGB design was unproven and had catastrophically failed during certification testing, the FAA [United States Federal Aviation Administration] accepted Sikorsky’s conjecture that the risk of a complete loss of MGB oil was extremely remote. Transport Canada also accepted this conjecture and certified the aircraft in Canada on that basis.”
The letter then requested that Strahl answer a key question: “Did Transport Canada succumb to pressure from the FAA or Sikorsky, or did they simply fail to recognize a serious safety/certification issue?”
The complete text of the letter can be found at: http://ecostrategy.ca/wp-content/uploads/2011/02/letter.pdf.