AeroSafety WORLD
NO GO-AROUND The psychology of continued unstable approaches UNINTENDED CONSEQUENCES Fatal control movements in EC135 CEASE FIRE Suppressing cargo airplane fires
BENEFITS OF WAKE VORTEX MEASUREMENT TRAILING INDICATORS
THE JOURNAL OF FLIGHT SAFETY FOUNDATION FEBRUARY 2013 The Foundation would like to give special recognition to our BARS Benefactors, Benefactor and Patron members. We value your membership and your high levels of commitment to the world of safety. Without your support, the Foundation’s mission of the continuous improvement of global aviation safety would not be possible.
bars benefactors
benefactors
Snecma
patrons PRESIDENT’SMESSAGE
CHANGE AT The Helm
n Jan. 1, I assumed the helm of the Flight Traveling on behalf of the Foundation took Safety Foundation, the most respected countless hours away from the office and, most independent and impartial interna- importantly, away from his family. Without his wife, tional aviation safety organization in Carol’s, support and understanding, that would Othe world. Following in the sizeable footsteps have not been possible. For that, I would like to rec- of our founder, the late Jerry Lederer, and the ognize and thank her on behalf of the Foundation. Foundation’s most recent president and CEO, Throughout Bill’s tenure at the Foundation, he Bill Voss, will not be an easy task, but it is a was known for his insight into individual safety challenge that I sought and about which I am issues. He has a gift for shaping an issue so it can very excited. be understood by everyone, inside and outside Bill’s background as a pilot and air traffic of the aviation community. For that, we will be controller gave him a well-rounded perspective forever indebted. of the basics of aviation safety. His many years This spring, Bill plans to write one more at the U.S. Federal Aviation Administration column for AeroSafety World, and to provide and the International Civil Aviation Organiza- us with his unique perspective on global avia- tion gave him an appreciation for what can be tion safety and on his tenure at the Foundation. achieved when regulators and industry partners I look forward to reading that piece and seeing work together to improve aviation safety. When and consulting with Bill as he embarks on a new he joined the Foundation in November 2006, Bill flight path here in Washington. immediately began visiting state regulators, avia- As the new president and CEO, I am excited tion organizations and industry groups, resulting to share my thoughts with you each month in this in dozens of trips annually, hundreds of speeches column as I, along with the Foundation Board and presentations, and even more meetings, con- of Governors and a very dedicated and talented ferences and one-on-one sessions — all in quest staff, move the Foundation into a new generation! of advancing aviation safety. His appearances were not limited to the de- veloped world. Bill traveled to areas where help was needed most, where regulators, operators and industry needed to become better informed on safety issues. Sudan and the Middle East were among the places where Bill made a significant Capt. Kevin L. Hiatt difference in how aviation operations and safety President and CEO are handled. Flight Safety Foundation
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 1 AeroSafetyWORLD
contents February 2013 Vol 8 Issue 1 features
12 CoverStory | Wake Vortex Perspectives 12 18 2012Review | CFIT’s Unwelcome Return 22 FlightOps | Dissecting Go-Around Decisions
18 29 CausalFactors | Abrupt Collective Input
33 CargoSafety | Fire Detection and Suppression
37 AvWeather | Fathoming Superstorm Sandy
42 SeminarsIASS | Pan American Accomplishments
22 departments
1 President’sMessage | Change at the Helm
5 EditorialPage | The Right Decision
7 SafetyCalendar | Industry Events
8 AirMail | Letters From Our Readers
2 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 29 37
33
AeroSafetyWORLD telephone: +1 703.739.6700 9 InBrief | Safety News Capt. Kevin L. Hiatt, publisher, FSF president and CEO 47 DataLink | C-FOQA Trends [email protected] Frank Jackman, editor-in-chief, FSF director of publications 52 InfoScan | SMS Under the Microscope [email protected], ext. 116
Wayne Rosenkrans, senior editor 57 OnRecord | Fatal Flight Test [email protected], ext. 115
Linda Werfelman, senior editor [email protected], ext. 122
Rick Darby, associate editor [email protected], ext. 113
Jennifer Moore, art director [email protected]
Susan D. Reed, production specialist About the Cover [email protected], ext. 123 Operations at London Gatwick. © Steve Morris/Jetphotos.net Editorial Advisory Board
David North, EAB chairman, consultant
We Encourage Reprints (For permissions, go to
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 3 Select the Integrated Air Safety Management Software Solution... SMS IAMS FAA Airline Quality JSA IOSA Flight Control Safety Audits Risk Matrix Safety Integrated JSA Aviation Security IAMS SMS IOSA Environmental JSA FAA Risk Management IAMS SMS SMS IATA Operational Safety Audit Flight Control Job Safety Analysis IATA Risk Matrix Safety Management System Job Safety Analysis FAA IOSA Risk Assessment JSA Integrated Airline Management System IATA JSA Environmental Federal Aviation Administration IOSA IOSA Reporting Airline Industry IAMS Aviation Risk Assessment FAA Airline Safety JSA SMS Safety Audits Aerosafety Risk Management Flight Control Flight Control Safety Audits Safety Audits Reporting JSA Risk Matrix SMS Airline IOSA FAA SMS Airline IAMS IAMS Air Safety IATA Quality Integrated Airline Safety IOSA Reporting Flight Control Job Safety Analysis Aerosafety Safety Audits Risk Management IATA Airline Industry FAA SMS SMS JSA Aviation FAA
Best Practices: Integrated modules for Aviation Safety Management: ...with the most VALUE Document Control • Reporting • Audits • Training • Job Safety Analysis CAPA • Safety Incident Reporting • Risk Assessment ...and more! Safety Assurance: Handles daily events, hazard analysis, and controls related to Aviation Safety Management System Assessment & Corrective Action: Uses intelligent Decision trees to determine whether event conform to requirements, and take Corrective Action to known events Safety Risk Management: Intuitive process for determining potential hazards using Job Safety Analysis, FMEA and similar tools Risk Assessment: Identifies, mitigates, and prevents high-risk events in the Aviation Safety Management System Change Control: Implement Controls to mitigate risks and Change Management to change processes related to known hazards Integration: Integrates with 3rd party business systems Scalable: Readily adapts to enterprise environments, and multi-site Deployments Business Intelligence: Enterprise reporting tracks KPIs, aids in decision-making with hundreds of configurable charts and reports 800-354-4476 • [email protected] www.etq.com/airsafety EDITORIALPAGE
THE RIGHT Decision
everal years ago, my wife and I were flying to Ph.D.; and Capt. William F. Curtis of The Presage the U.S. Virgin Islands for a little mid-winter Group. I’d like to thank all three gentlemen for their R&R sans kids. Moments before anticipated work on the Foundation’s go-around project and touchdown at St. Thomas’ Cyril E. King Air- for the hours they put into crafting the article. I’m port,S it flashed through my mind that the airplane looking forward to the next installment. was landing long. Almost at that same moment, we As always, we welcome feedback from our heard and felt the roar of the engines as the pilots opt- readers. ed to go around and try again. A little spooked, my wife asked what had just happened. As I explained Kudos and Thanks the concept of a go-around she started to laugh. While I’m handing out thanks, I want to mention After years of patiently listening to me blather on long-time Director of Technical Programs Jim about this or that technology or process, she found Burin. Careful readers of ASW will note on p. 6 it funny that “go-around” really is a technical term. that Jim has a new title and is transitioning into But go-arounds are no laughing matter. Run- the role of Foundation Fellow. We won’t be seeing way excursions account for one-third of all acci- Jim in the office as much, but he will continue to dents, and the greatest risk factor for excursions be engaged in the Foundation’s activities. In my is the unstable approach. An unstable approach nearly 10 months at the Foundation, Jim has been should result in a go-around, but usually does an invaluable source of knowledge and ideas, and not. In fact, according to Foundation research, has never been shy about letting me know what only 3 percent of all unstable approaches result details he likes and doesn’t like in ASW articles. in go-arounds. Why is that? Sometimes the best medicine is the most difficult In 2011, the Foundation launched a Go-Around to swallow. Thanks, Jim. Decision Making and Execution Project, the intent of which is to mitigate runway excursions caused by unstable approaches by achieving a higher level of pilot compliance with go-around policies. With this issue of AeroSafety World, we are beginning a series of articles that will take an in-depth look at the results of the project’s work to date. The first Frank Jackman article in the series begins on p. 22 and was writ- Editor-in-Chief ten by J. Martin Smith, Ph.D.; David W. Jamieson, AeroSafety World
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 5 Serving Aviation Safety Interests for More Than 60 Years
OFFICERS AND STAFF light Safety Foundation is an international membership organization dedicated to Chairman the continuous improvement of aviation safety. Nonprofit and independent, the Board of Governors David McMillan Foundation was launched officially in 1947 in response to the aviation industry’s need President and CEO Capt. Kevin L. Hiatt F for a neutral clearinghouse to disseminate objective safety information, and for a credible General Counsel and knowledgeable body that would identify threats to safety, analyze the problems and and Secretary Kenneth P. Quinn, Esq. recommend practical solutions to them. Since its beginning, the Foundation has acted in the Treasurer David J. Barger public interest to produce positive influence on aviation safety. Today, the Foundation provides leadership to more than 1,000 individuals and member organizations in 150 countries. ADMINISTRATIVE
Manager of Support Services and Executive Assistant Stephanie Mack MemberGuide Flight Safety Foundation 801 N. Fairfax St., Suite 400, Alexandria VA 22314-1774 USA FINANCIAL tel +1 703.739.6700 fax +1 703.739.6708 flightsafety.org Financial Operations Member enrollment ext. 102 Manager Jaime Northington Ahlam Wahdan, membership services coordinator [email protected] Seminar registration ext. 101 MEMBERSHIP AND BUSINESS DEVELOPMENT Namratha Apparao, seminar and exhibit coordinator [email protected] Senior Director of Seminar sponsorships/Exhibitor opportunities ext. 105 Membership and Kelcey Mitchell, director of events and seminars [email protected] Business Development Susan M. Lausch Donations/Endowments ext. 112 Director of Events Susan M. Lausch, senior director of membership and development [email protected] and Seminars Kelcey Mitchell FSF awards programs ext. 105 Seminar and Kelcey Mitchell, director of events and seminars [email protected] Exhibit Coordinator Namratha Apparao Technical product orders ext. 101 Membership Namratha Apparao, seminar and exhibit coordinator [email protected] Services Coordinator Ahlam Wahdan Seminar proceedings ext. 101 Namratha Apparao, seminar and exhibit coordinator [email protected] COMMUNICATIONS Website ext. 126 Director of Emily McGee, director of communications [email protected] Communications Emily McGee Basic Aviation Risk Standard Greg Marshall, BARS managing director [email protected] GLOBAL PROGRAMS BARS Program Office: Level 6, 278 Collins Street, Melbourne, Victoria 3000 Australia tel +61 1300.557.162 fax +61 1300.557.182 Director of Global Programs Rudy Quevedo Foundation Fellow James M. Burin
BASIC AVIATION RISK STANDARD
BARS Managing Director Greg Marshall facebook.com/flightsafetyfoundation
@flightsafety Past President William R. Voss Founder Jerome Lederer www.linkedin.com/groups?gid=1804478 1902–2004
FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 ➤ SAFETYCALENDAR
FEB. 7–8 ➤ Emergency Response Planning MARCH 8 ➤ ABCs of SMS (free course). APRIL 15–19 ➤ OSHA/Aviation Ground and Crisis Management. Vortex Training Aviation Consulting Group. San Juan, Puerto Safety. Embry-Riddle Aeronautical University. Seminars. Denver. Stephanie Brewer,
FEB. 21–22 ➤ European Business Aviation APRIL 11–13 ➤ Internal Evaluation Program Aviation safety event coming up? Safety Conference. Aviation Screening. Munich, Theory and Application. U.S. Transportation Tell industry leaders about it. Germany. Christian Beckert,
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 7 AIRMAIL
Landing Weight at ‘Special’ Airport season and that high landing weights are [RB211-]535 series engine. note the article by Mark Lacagnina, common when operating at the air- The end note no. 54 in the “Double Whammy” [ASW, 9/12, port.” I am sure that great care is taken; NTSB report, page 21, states that P&W- Ip. 34] and the follow-up [“Speed however, would it not reduce the risk equipped airplanes are the only Boeing Brake Warnings,” ASW, 10/12, p. 12]. factor if it were arranged to arrive at this planes that have this potential, but One aspect is briefly mentioned airport well below the maximum land- that note appears also to be erroneous. but receives no attention from any- ing weight? We all know how quickly The photograph in the accompanying one thereafter. It is noted that Jackson the landing margins can be eroded. I am article clearly shows RR engines. Hole is a “special airport,” and that “the assuming that the high landing weight Darren Dresser runway is usually slippery during the ski is due to fuel and I have no doubt that there seem to be good reasons for oper- Mark Lacagnina replies: The only ating in this way, but if that is the case, I mention of P&W engines in the Causal do think that more attention should be Factors article is in the statement: “The given to this aspect. thrust-reverser system on 757s and 767s encourages AeroSafety World Richard T. Slatter equipped with Pratt & Whitney engines comments from readers, and will has a ‘sync-lock’ mechanism that is assume that letters and e-mails Engine (Identification) Failure? intended to prevent the translating sleeves are meant for publication unless enjoy your magazine and appreciate from extending accidentally due to a fault otherwise stated. Correspondence the opportunity to read it at my place in the system.” is subject to editing for length of employment. However, a recent That statement is included in and clarity. I article [“Double Whammy”] caught the capsulization of NTSB’s findings Write to Frank Jackman, director my attention. about why the incident crew and other of publications, Flight Safety Specifically, a causal factor is er- American Airlines pilots likely believed Foundation, 801 N. Fairfax St., roneously identified as a “sync-lock” thrust reverser lock was not possible and Suite 400, Alexandria, VA mechanism on 757s/767s equipped thus were not prepared to handle it. The 22314-1774 USA, or e-mail with Pratt & Whitney engines. In fact, sync-lock mechanism was not identified
8 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 INBRIEF Safety News Proactive Safety Proposed Penalties
he European Commission (EC) says it he U.S. Federal Aviation Administration (FAA) has proposed $633,000 is proposing “ambitious and compre- in civil penalties against Trans States Airlines for its operation of two Thensive” steps to develop a proactive, TEmbraer 145 regional jets on 3,660 passenger flights while the aircraft al- evidence-based aviation safety system, with an legedly were out of compliance with Federal Aviation Regulations. emphasis on comprehensive data-gathering. The FAA said that the airline operated two airplanes on 268 revenue pas- “The current aviation regulatory system senger flights while the airplanes were equipped with improperly installed is primarily a reactive system relying on tech- radio altimeter antenna cables. One aircraft was operated on 3,392 passenger nological progress, the adoption of legislation flights with improperly installed electrical wiring in its fuel supply system, overseen by effective regulatory authorities the FAA said. and detailed accident investigations leading to In an unrelated case, the FAA proposed a $275,000 civil penalty against recommendations for safety improvements,” Pinnacle Airlines for allegedly operating a Bombardier CRJ on 11 flights after the EC said in a December memo. maintenance personnel failed to install a required part when they replaced an “However, whilst the ability to learn engine. The FAA said that, because Pinnacle is being reorganized under U.S. lessons from an accident is crucial, systems bankruptcy laws, the notice of proposed penalty is not a demand for payment. which are essentially reactive are showing their limits in being able to drive further improvements in the accident rate.” The answer, the EC said, is to gather and analyze all available aviation safety information. The EC proposals include establishing “an appropriate environment to encourage aviation professionals to report safety-related informa- © Boeing tion by protecting them from punishment ex- 787 Grounding cept in cases of gross negligence” and ensuring that “the scope of mandatory reporting covers he U.S. Federal Aviation Administration (FAA) on Jan. 16 grounded all major potential risks and that the appropriate U.S.-registered Boeing 787s, citing an in-flight “battery incident” earlier means to capture any safety threat are estab- Tin the day on an All Nippon Airways (ANA) 787. The FAA said it would lished [through] voluntary reporting schemes.” issue an emergency airworthiness directive to address the risk of battery fires In addition, the proposals call for confi- in the airplanes. dential safety information to be made available Other civil aviation authorities worldwide immediately took similar only to maintain or improve aviation safety. action to keep 787s out of the skies. ANA and Japan Airlines had grounded The EC added that its intent is to “diminish the their 787s prior to the FAA action. negative effect that the use of such data by judi- “The FAA will work with the manufacturer and carriers to develop a cor- cial authorities may have on aviation safety.” rective action plan to allow the U.S. 787 fleet to resume operations as quickly Other proposals call for improving the and safely as possible,” the agency said. “Before further flight, operators of “quality and completeness” of occurrence U.S.-registered Boeing 787 aircraft must demonstrate to the [FAA] that the reports, developing a better exchange of in- batteries are safe.” formation among EC member states and im- Boeing Chairman, President and CEO Jim McNerney said the company is proving data analysis at the European Union “committed to supporting the FAA and finding answers as quickly as possible. (EU) level so that it complements analysis …We are confident the 787 is safe, and we stand behind its overall integrity.” performed at the national level. Published reports said the in-flight incident involved warning lights The EC proposal must be approved by indicating a battery problem in a 787 on a domestic flight in Japan and quoted the European Parliament and the Council of Yoshitomo Tamaki, director general of the Japan Transport Safety Board, as member states before it takes effect. saying there was a bulge in the metal case that housed the battery. EU transport ministers also called for an The grounding came five days after the FAA announced a review of the external aviation policy that will strengthen the 787’s critical systems, especially its electrical systems; that action came in the competitiveness of the European aviation indus- wake of a battery fire in a Japan Airlines 787 parked at Logan International try, in part by developing EU-level air transport Airport in Boston. agreements with neighboring countries. Fifty 787s had been in service worldwide.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 9 INBRIEF
SMS Start-Up
he U.S. Federal Aviation Administration (FAA) should years to accomplish the “cultural and procedural shift” in FAA consider asking Congress to provide additional protections internal operations and in the agency’s oversight of airlines, Tfor data gathered through safety management systems airports and other aviation stakeholders. (SMS), a U.S. government watchdog agency says. “Going forward, if FAA is to attain the full benefits of The Government Accountability Office (GAO) said in a SMS, it will be important for the agency to remain commit- December report on the FAA’s progress in SMS implementation ted to fully implementing SMS across its business lines,” GAO — both within the agency and throughout the aviation indus- said. “FAA has taken a number of steps that align with prac- try — that data protection concerns “could prevent aviation tices we identified as important to successful project plan- stakeholders from fully embracing SMS implementation, thus ning and implementation but has not addressed or has only hindering its effectiveness. partially addressed other key practices … [that] are important “Without assurance of protection from state [freedom of for large-scale transformative projects such as SMS.” information] laws, some aviation stakeholders may choose to GAO’s other recommendations included calls for devel- collect only the bare minimum of safety-related data or may opment of a data-collection system to be used in evaluating choose to limit the extent to which collected information is whether SMS is meeting designated goals and implementa- shared among aviation stakeholders.” tion of a system of evaluating employee performance as In addition, GAO said, “the ability of FAA to identify safety related to SMS. risks, develop mitigation strategies and measure outcomes is GAO also recommended developing a system to track hindered by limited access to complete and meaningful data.” SMS implementation and conducting a workforce analysis to GAO said that the FAA and the aviation industry are identify employee skills and strategies for addressing SMS- making progress in SMS implementation, although it will take related skills gaps.
Flight Simulation Goals Workplace Safety
he Australian Civil Aviation Safety Authority (CASA), urging orkplace safety standards for flight at- the increased use of flight simulators throughout the aviation tendants should be enforced by the U.S. Tindustry, has established six related goals for flight simulation WOccupational Safety and Health Adminis- over the next two years. tration (OSHA), the Federal Aviation Administra- The goals include adopting the International Civil Aviation tion (FAA) said in proposing a regulation to expand Organization framework for simulator classification, mandating that OSHA’s authority. simulators be used for training and checking of high-risk emergency “While the FAA’s aviation safety regulations take procedures in some aircraft types, and encouraging operators to precedence, the agency is proposing that OSHA upgrade and maintain their simulators. be able to enforce certain occupational safety and “Technological advances have seen significant improvement in health standards currently not covered by FAA the fidelity of flight simulation devices at all levels,” CASA said in oversight,” the FAA said. its Flight Simulation Operational Plan 2012–2014. “Flight simula- Under the proposal, flight attendants could tors provide more in-depth training, particularly in the practice of report workplace injury and illness complaints to emergency and abnor- OSHA, which would have the authority to investigate. mal operations, than Workplace issues could include exposure to noise and can be accomplished disease-causing microorganisms, the FAA said. in aircraft.” “The policy … [would] not only enhance the CASA said that health and safety of flight attendants by connect- Australia currently has ing them directly with OSHA but will by extension 34 full-flight simula- improve the flying experience of millions of airline tors; five flight training passengers,” said U.S. Labor Secretary Hilda L. Solis. devices, which do A final policy will be announced after au- not have motion; and thorities have reviewed public comments on the 91 instrument flight proposed regulation. The comment period was trainers. scheduled to end Jan. 22.
© CAE
10 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 INBRIEF
Information Sharing
light Safety Foundation (FSF) and the International Civil Aviation Organization (ICAO) have begun a new Fcooperative effort to promote and share aviation safety information and metrics. The new worldwide initiative is designed to support ICAO guidance for safety management systems, which calls for increased monitoring, analysis and reporting of safety data. “The establishment of this framework for enhanced cooperation with FSF is an important step in helping us © Konstantin Tyurpeko/RUSpottersTeam achieve the highest levels of aviation safety worldwide,” Engine Warning said Roberto Kobe González, president of the ICAO Coun- cil. “Aviation safety knows no borders, and these types of perators of aircraft with Rolls-Royce RB211-524 engines collaborative data sharing and risk mitigation efforts are have been warned of a potential for degradation of the essential to help states and industry address safety risks Oengines’ intermediate-pressure turbine blade interlocking before they lead to a serious incident or accident.” shrouds, which, if not corrected, could result in the cracking The memorandum of cooperation calls for ICAO and and loss of turbine blades, the Australian Transport Safety the Foundation to work together to encourage compliance Bureau (ATSB) says. with ICAO standards and recommended practices and The ATSB cited the May 9, 2011, malfunction of an engine related guidance material. on a Qantas Airways Boeing 747-400 during a flight from The memorandum also “promotes joint activities Sydney, New South Wales, Australia, to Singapore. The crew between the organizations in the areas of data sharing and observed abnormal indications from the no. 4 engine during analysis, training and technical assistance,” according to a climb from 36,000 ft to 38,000 ft. The crew shut down the the announcement of the agreement. “The joint analyses engine, continued the flight to Singapore and landed without developed will facilitate the harmonization of proactive further incident. and predictive safety metrics and the promotion of a just The ATSB investigation traced the problem to the “failure safety culture globally.” and separation of a single intermediate-pressure turbine William R. Voss, then FSF president and CEO, noting blade … [which] fractured following the initiation and that some U.S. air carriers and the U.S. Federal Aviation growth of a fatigue crack from an origin area near the blade Administration already operate under cooperative data- inner root platform.” sharing agreements, said the new cooperative agreement The cause of the blade failure was not immediately identi- would help other countries “establish models that are fied, but the manufacturer’s post-accident analysis revealed suited to their unique needs and constraints.” that “wear and loss of material from the turbine blade outer Regional forums will be convened soon to aid in estab- interlocking shrouds had reduced the rigidity and damping ef- lishing information-sharing goals. fects of the shroud and may have contributed to the high-cycle fatigue cracking and failure.” The manufacturer’s analysis was continuing. In Other News … The ATSB said that Rolls-Royce issued non-modification service bulletin 72-G739 in October 2011, directing opera- Michael Huerta was sworn in as administrator of the U.S. tors to inspect the intermediate-pressure turbine blades in the Federal Aviation Administration in early January, after serv- affected engines to determine if any shroud interlock material ing as acting administrator for more than one year. … The was missing. Qantas had completed the required inspections European Union and Eurocontrol have agreed to establish and found no instances of excessive wear, the ATSB said. a new framework for cooperation in implementing the The agency said three similar events have been reported Single European Sky program. … The European Com- in RB211-524 history and the probability of further events is mission has removed all air carriers certified in Mauritania “extremely low.” Blade separation probably will result in engine from its list of those banned from operating in the European malfunctions and an in-flight engine shutdown, but risks to Union. The December revision added to the list air carriers the safety of continued flight are minor, the ATSB said. certified in Eritrea.
Compiled and edited by Linda Werfelman.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 11 COVERSTORY
oday’s portable sensors and data-analysis with European counterparts, have used field techniques enable scientists worldwide to research to build safety cases verifying that risks visualize dimensions, measure velocities in proposed changes to air traffic control (ATC) and track positions of wake vortices gener- procedures are acceptable. . atedT by specific variants of large commercial jets. Essentially, the National Airspace System That’s a far cry from igniting elevated smoke pots has begun to see the results of a decision in 2001 for low-level overflights in the early 1970s, says that set near-term, mid-term and long-term Steven Lang, director of the U.S. Center for Air goals “to focus on operationally feasible solu- Traffic Systems and Operations at the John A. tions rather than just looking at wake science as Volpe National Transportation Systems Center. a solution,” Lang said. Flexibility was added, too, “Wake turbulence is an inevitable conse- to explore solutions to practical problems other quence of flight — aircraft lift generation,” Lang than encounters with heavy-jet wake vortices (see said during a Web briefing for news media in “Airbus Measures Relative Wake Vortex Char- November 2012. “Wake turbulence separations acteristics,” p. 14). Lang also credited clear-cut, in a sense reduce capacity at airports because you stakeholder advisory processes launched then have to add spacing behind the larger aircraft for under the FAA’s safety management system. safety mitigation.” In the past 30 years, various sensors and The evolving precision partly explains how techniques incrementally improved study of several redesigns of air traffic procedures have wake generation, transport and decay. The most been accomplished recently, he said, summa- radical change came from pulsed lidar, which rizing a paper published in October.1 In the Lang described as “a radar-laser type of device United States, Volpe and the Federal Aviation that actually measures the vortex as it’s generated Administration (FAA), often in partnership from the aircraft [and] shed from the aircraft. …
Outmaneuvered AIRFLOW BY WAYNE ROSENKRANS U.S. wake vortex science safely updates approach and departure concepts essential to NextGen capacity gains.
12 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 © Steve Morris/AirTeamImages.com ATC standards within NextGen be will full potential.” Many concepts would not realize their you’re stuck with what you have. … … Unless wake turbulence is addressed, lence gets solved for intime NextGen. it’sSo important very that wake turbu- you can’t put aircraft together. closer … but ifyou don’t solve wake the problem, that else along goes everything with that, improvementand surveillance and ofdid all that navigation improvement lence separation. …It’s that good you spacing is which wake needed, turbu - butful, last the piece …is maximum the Lang said, “All things are those wonder minimum that enable aircraft operated to be with navigation advances and surveillance from satellite-based communication, systemcreased capacity come will partly Transportation System (NextGen), in- centerlines. their (762m)between ft ways as runways that have less than 2,500 FAA- closely run parallel defines spaced wakein-trail separation operations. The runway operations and single-runway into closely parallel two spaced types: had orally to fly operate system.” the involvement from people the that actu- wanted to research and there was little what wanted they to study, what they fort,” scientists decided he “The recalled. “Before that, it was purely ef- ascience community science the and regulators. controllers,safety air specialists, traffic bringinghas been together pilots, airline wake vortex mitigation. Another factor development of solutions practical for of scientists has accelerated also the lidar.”pulsed concernedbe with is now measurable by The entire safety region that we have to FLIGHTSAFETY.ORG For example, one of long-term the In planning Next the Generation Air The ATC innovations fall discussed Cooperation among networks global spacing needed for spacing needed safety. But |
WORLD AEROSAFETY | FEBRUARY 2013 -
or ‘No, you cannot doit.’” controller that [says] ‘Yes, you can do it,’ spacing, ayea-or-nay spacing, to the it’s… So asystem that …delivers a and [meteorological] the conditions. ofbecause configuration, the weight the aircraft the behind 3nm[5.6km] be an aircraft; next the day you might you behind might 4nm[7.4km] be scenario,”specific he said. “So one day, develop separation the standard for that intotaken account and [ATC then will] of arrival, route the of are flight all tion, weather the condition time …the where aircraft the weight configura- separation.dynamic “That’s pairwise Figure 1 to Parallel Runways Spaced Less Than 2,500Feet Apart,” Policy 30,2012. JO7110.308 CHG3,effectiveOct. Source: Steven R.LangandU.S. Federal Aviation Administration Air Traffic Organization, Dependent Mile Approaches “1.5-Nautical conditions. from 2.75to GSAs 3.1degrees. vary controllers to safely applythisminimum1.5-nmwithin-pair spacingregardless orwind ofwake vortices Note: This ATC procedure IMChasbeenimplemented during at eight majorU.S. enabling airports, GSA =glideslopeangle CSPR =closelyspaced parallel runways ATC =airtraffic control Staggered ILSApproaches to Closely Spaced Parallel Runways GSA for higherapproach; single runways) applyfor weight type, andusesa an aircraft following #2. separation rules(asfor Aircraft #2may beany ATC in-trail standard separation (762 m) <2,500 ft 1.5 nm(2.8km) pair spacingat least Diagonal within- IMC =instrument meteorological conditions ILS =instrument landingsystem encounter for parallel-runway the you reduce actually risk of the awake authorized as of 2012. October for such which runway pairings were to addtwo more to eight the airports Safety now cases are prepared being runways done could be safely (Figure 1). arrivalstheir to closely parallel spaced from(2.8 km) any larger aircraft during tioning aircraft asmaller at least 1.5nm proved with safety-case data that- posi wake vortex measurements. One effort onthat now-outdated based had been FAA and Volpe revisited also procedures “By using runways, parallel the With that on still far the horizon, stagger Threshold lower approach. withGSAfortypically procedure,arrival apply thisstaggered CSPR weight for type ATC to torestricted large orsmall separation pair, is aircraft ofthereduced- Aircraft #1,thelead COVERSTORY |
13 COVERSTORY
Airbus Measures Relative Wake Vortex Characteristics
ake vortex encounters severe enough to threaten termed an incorrect assumption that wake vortices from a 747 an upset of one large commercial jet flying behind do not descend more than 800 or 900 ft. Wanother have been rare for simple reasons, suggest Airline pilot knowledge and training to correct an un- recent presentations of data from experiments by Airbus. expected roll remain sufficient mitigations for wake vortex Benign encounters are very common, however, says Claude encounters involving one large commercial jet behind Lelaie, senior vice president and product safety officer, another, he noted. “In the vortex … you can get strong vertical Airbus, and a former Airbus test pilot and airline captain. acceleration, positive or negative,” Lelaie said. “For the vortex “The probability to have a severe encounter is in fact very encounter, what we clearly recommend [to Airbus flight crews] low,” Lelaie said. “Why? Because you have to enter the vortex, is please do nothing. Release controls and do nothing, and a very small tube … about 6 m [20 ft] diameter. You have to once you have passed the vortex, nothing will happen. … enter exactly in the center, and you have to enter with the The roll [response] is just normal roll control.” International proper [10-degree] angle. … If you have turbulence and so guidance on airplane upset prevention and recovery has been on, everything disappears. … Even when trying to have a published by government and industry.2 strong encounter every time, we did not manage to have a One part of the Airbus study focused on measuring the strong encounter every time.” rate of descent of wake vortices from each generator aircraft. Nevertheless, the Airbus analysis also has concluded that Another focused on effects on the follower aircraft. The most “there is a possibility to have a severe encounter in flight important effect was roll acceleration, the direct indicator of where there is a type of generating aircraft at a distance vortex strength (Table 1, p. 16). Less interesting to research- [more than] the standard minimum 5 nm [9 km] separation ers in practical terms were altitude loss, bank angle, vertical and with 1,000 ft vertical separation,” he said. acceleration and roll rate, he said. Scientific instruments and Airbus presented these data and conclusions to the video cameras also documented the bank, buffeting and the Wake Vortex Study Group of the International Civil Aviation pilot’s correction of uncommanded bank. Organization (ICAO), which has been updating recommen- Regarding the rates of descent of vortices while flying at dations for flight crews and air traffic controllers. Lelaie also Mach 0.85, there was no difference between the A380 and briefed Flight Safety Foundation’s International Air Safety 747-400, Lelaie said. He noted, “There was a slight difference Seminar in October 2012 in Santiago, Chile. with the A340-600 flying at Mach 0.82, but at the end of the The 200 encounters Airbus studied were carefully orches- day, all vortices [had descended] 1,000 feet at around 12, 14, trated missions — at a cruise altitude of about 35,000 ft — to 15 nm [22, 26, 28 km]. … This showed clearly that … at 15 insert a follower aircraft into the center of the strongest/ nm behind any of these aircraft, you can find a vortex. … The worst wake vortices/contrails to induce effects associ- [strength/roll rate acceleration] decrease with the distance is ated with in-flight upset, Lelaie said. The missions involved rather slow. At 5 nm, you have a good encounter; at 15 [nm] precisely positioning the generator-follower pairs in ideal, you have decreased [strength of] maybe 30 to 40 percent, it’s repeatable calm-weather conditions. An Airbus A380 with an not a lot.” adjacent A340-600 or a Boeing 747-400 on a parallel flight As expected, lateral-acceleration maximum load factor path were used as the wake vortex–generators. The A340- and minimum load factor were significantly different in the 600 and an Airbus A318 took turns as follower aircraft. He forces recorded at the back of the follower-aircraft fuselage described one test protocol. versus those felt by occupants because the airplane’s turning “Two aircraft were flying side by side [into the wind], the point actually is in front of the aircraft. “These load factors are A380 and the reference aircraft, which was either a 340-600 or not what the passenger or what the pilot can feel,” he said. the 747,” Lelaie said. “An A318 was flying behind and below at “[They’re] much higher.” Nevertheless, occupants may feel a distance between 5 and 15 nm, and we had above a Falcon strong lateral acceleration on the order of 2.5 g, 2.5 times 20 from the DLR [German Aerospace Center] with an onboard normal gravitational acceleration. “Even at 18 nm [33 km], we lidar.”1 A 10-degree entry angle was considered the most criti- have with all aircraft 2 g, again at the back,” he added, and cal case. “If you are almost parallel, you will be ejected from the data in some cases showed small negative-g values. vortex,” he said. “If you cross perpendicularly, [the encounter] “One which is interesting is this one, 747 and A318,” he will be very short and almost nothing will happen.” said. “Look at that: –0.7 [g],” he said. “In the middle of the fuse- Some findings ran counter to conventional assumptions lage it would have been –0.4 or –0.3 [g] but the [unrestrained about wake vortex effects on the existing design of reduced person] in that seat will bump on the ceiling.” Cases of the vertical separation minimums operations, notably what he A380 followed by the A318 and the A380 followed by the Continued on p. 16
14 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 COVERSTORY
aircraft versus going in-trail,” Lang said, 900,000 lb (408,233 kg) and a 767 that pairwise separation will be supported explaining that “by placing an aircraft in weighs about 320,000 lb (147,417 kg) in weather-based phase 3. Lang said a staggered position, it has less risk of a were in the heavy category B. that such changes typically take time wake encounter than if you put it single “These two aircraft have to be 4 nm to generate predictable and measur- file to the same runway.” apart because they are in that same able capacity benefits while the local Data collection and building of category, regardless of [which] is in ATC personnel become accustomed to safety cases for arrivals positioned front, [and that] doesn’t make a lot of new procedures. FAA/Volpe to pursue similar concepts sense,” he said. “The [767] behind [the Related applications of wake vortex to make simultaneous departures of 747] probably needed 4 nm but the 747 data have enabled the FAA to divide disparate-size aircraft on closely spaced following [the 767] did not need 4 nm.” three variants of the 757 within U.S. parallel runways feasible mainly by tak- The resulting program, implemented in ATC separation and in separation stan- ing into account the effect of a favorable Memphis in November, is called Wake dards of the International Civil Avia- wind direction and velocity through RECAT phase 1 and includes additional tion Organization. Another example a new wake turbulence mitigation for safety buffers for the lightest aircraft he cited was Volpe’s wake data collec- departure (WTMD) system. types.3 Preliminary reports estimate at tion for Boeing during testing of the To mitigate the risk of a wake least a 10-percent capacity boost, and 747‑800 for standards development. encounter, “physics tells you that if [one possibly 20 percent. aircraft] is a heavy jet, you would have “In Memphis, the one Concept of Wake Turbulence Mitigation to stop this [other, lighter] aircraft from observation that FedEx has for Departures departing for three minutes in this ge- made is they used to have ometry (Figure 2) or two minutes if this backups at the runway both for [runway end is staggered by] less than arrivals and departures, and 500 ft [152 m],” he said. “If the wind is now they find themselves ‘dry- Wind direction blowing this direction, this wake for the ing up,’ as they call it,” he said. most part cannot transport against the “Recategorization has now 1,300 ft wind and get over to that [parallel] run- made it [so] that there is no way. … The controllers have a system queue, and now they’re having in the control tower at … three airports to rethink how they get the Wake vortices — going live in January at Houston and aircraft out of the ramp areas, then in San Francisco and Memphis.” out to the runway to be able to 30L The system advises the controller with take advantage of the empty a red light/green light display when the runway.” This system operates required conditions exist. independently of meteorologi- 1,500 ft When fully available in Houston, “we cal conditions. envision [WTMD] will increase their The main reason that 30R capacity significantly [by] three, maybe other airports cannot imple- four departures an hour,” Lang added. ment Wake RECAT phase 1 STL = Lambert–St. Louis International Airport The third focus of practical solutions in the same time frame has IAH = Houston Intercontinental Airport MEM = Memphis International Airport derived from advanced measurement involved local variations in SFO = San Francisco International Airport has been single-runway solutions. Essen- ATC automation systems, he Note: An operational demonstration at three U.S. airports with closely spaced parallel runways (IAH, MEM and tially, this program recategorizes aircraft said. Wake RECAT phase 2, SFO; not including STL used here for illustration) permits from their legacy ATC-spacing catego- also under way, supports ATC upwind-runway departures to occur simultaneously with downwind-runway departures that meet specified real- ries, based on wide ranges of maximum static pairwise separation — time wind criteria with conditions of approximately 3 mi takeoff weights and wingspans, to a new that is, separation based on (5 km) visibility and a minimum 1,000 ft ceiling. Source: Steven R. Lang, John A. Volpe National Transportation Systems set of six categories based on different airport-specific categories of Center, U.S. Department of Transportation parameters. Under the legacy system, aircraft. As noted, the long- both a Boeing 747 that weighs about term move to ATC dynamic Figure 2
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 15 COVERSTORY
Airbus Measures Relative Wake Vortex Characteristics (continued)
A340-600 also showed that “you can have something quite Notes strong in terms of g,” he said. Lelaie also pointed to ongoing work by a Eurocontrol– 1. Lidar means light detection and ranging, and pulsed lidar combines laser and radar sensor technology to visualize and Delft University of Technology study, looking at the correla- measure wake vortex characteristics. tion between actual wake vortex encounters and mapped hot spots, areas where encounters were predicted based on 2. One such resource that discusses wake turbulence is the European air traffic data, as a promising path to further risk Airplane Upset Recovery Training Aid, Revision 2 (November reduction. 2008) available at
Wake Vortex-Induced Main Upsets for Selected Cases in Encounters Tested by Airbus
Vertical Separation <1,000 ft Vertical Separation >1,000 ft
Generator airplane A380 A340-600 B747-400 B747-400 A380 A380 B747-400 A380 Follower airplane A318 A318 A318 A318 A340-600 A318 A318 A340-600 Horizontal separation (nm) 12.2 12.3 5.32 14.9 13.5 18.1 15.8 19.3 Vertical separation (ft) 838 608 432 832 851 1,015 1,038 1,168 Roll acceleration (deg/s2) 49 75 69 146 24 68 109 12 Roll rate (deg/s) 24 35 18 36 5 20 31 7 Bank (degrees) 46 38 35 31 10 29 34 10 deg/sec2 = degrees per second per second deg/sec = degrees per second Note: The A318, A340-600 and A380 are Airbus aircraft types; the 747-400 is a Boeing aircraft type. Airbus also reported the lateral and vertical accelerations of the follower aircraft; these are not shown. Source: Claude Lelaie
Table 1
Rethinking wake turbulence risk has involved Notes
more than the research capability. For example, 1. Tittsworth, Jeffrey A.; Lang, Steven R.; Johnson, meteorological and short-term wind nowcasting Edward J.; Barnes, Stephen. “Federal Aviation have improved significantly. “One thing FAA is has Administration Wake Turbulence Program — Recent been pursuing, and we have been supporting, is Highlights.” Paper presented to Air Traffic Control getting wind [data] off the aircraft [in real time],” Association Annual Conference and Exhibition, Oct. 1–3, 2012.
16 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 Theme: Building an Accident Free Legacy: Predictive Safety to Avoid ‘the Inevitable’
18th - 20th March 2013 Westin Bayshore Resort & Marina, Vancouver, BC, Canada
Improving Safety in aviation
Exceptional Value World-Class Speakers Outstanding Networking Unprecedented Training Opportunities
registration is now open
www.chcsafetyqualitysummit.com 2012REVIEW
he year 2012 set records globally for the But the increase in controlled flight into fewest major accidents involving com- terrain (CFIT) accidents continued. Three of mercial jets and commercial turboprops. the seven commercial jet accidents were CFIT. The decreasing trend in the commercial Commercial turboprops also set a record low for jetT accident rate was extended. The 2011 record the number of major accidents, although CFIT rate, 0.28 major accidents1 per million depar- again dominated their fatality numbers. Busi- CFIT claimed the lives tures for commercial jets, was reduced by 50 ness jets had 13 major accidents, slightly above of all 127 occupants percent to a record low of 0.14. For the second their 12-year average. of a Boeing 737 in a year in a row, there were no commercial jet There are now more than 22,000 commer- crash on approach to upset aircraft accidents. cial jets in the world. Of these, approximately Islamabad, Pakistan.
Accident numbers and rates decreased further in 2012, but CFIT is still a concern.
CFIT’sUnwelcome Return BY JAMES M. BURIN
18 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 2012REVIEW
5 percent are Eastern-built. The world’s com- Major Accidents, Worldwide Commercial Jets, 2012 mercial turboprop fleet is 20 percent Eastern- built. About 9 percent of the total commercial Date Operator Aircraft Location Phase Fatal jet fleet is inactive, including almost 50 percent April 20 Bhoja Airlines 737 Islamabad, Pakistan Approach 127 of the Eastern-built aircraft. Fifteen percent of May 9 Sukhoi Su-100 Mount Salak, Indonesia En route 45 the 6,012 turboprops are inactive. For the third June 2 Allied Air 727 Accra, Ghana Landing 0 year in a row, there were inactive business jets, June 3 Dana Air MD-83 Lagos, Nigeria Approach 153 including 3 percent of the inventory. Nov. 30 Aero Service IL-76 Brazzaville, Congo Landing 6 The commercial jet inventory grew about Dec. 25 Air Bagan F-100 Heho, Myanmar Landing 1 1 percent from the 2011 numbers, while the Dec. 29 Red Wings Airlines Tu-204 Moscow, Russia Landing 5 commercial turboprop inventory decreased 2 percent. The business jet inventory continued Controlled flight into terrain (CFIT) accident Runway excursion to lead in growth, with the current inventory Source: Ascend of 17,642 aircraft representing a 2.5 percent Table 1 increase from the previous year. Seven major accidents involving scheduled Commercial Jet Major Accidents, 2001–2012 and unscheduled passenger and cargo opera- tions, for Western- and Eastern-built commer- Western-built Eastern-built 20 cial jets, occurred in 2012 (Table 1). Six of the seven were approach and landing accidents. 15 Three of the seven were CFIT, and there were two runway excursion accidents. 10 Figure 1 shows the total number of major ac- cidents, including those involving Eastern-built 5 aircraft, for commercial jets during the past 12 Number of accidents years. The overall number of accidents in 2012 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 20122011 was down dramatically. Even though only about Year 3 percent of the active commercial jet fleet is Source: Ascend Eastern-built, they accounted for 43 percent of the major accidents. Figure 1 Figure 2 shows the commercial jet major accident rate and the five-year running average. Western-Built Commercial Jet Major Accident Rates, 1999–2012 This rate is only for Western-built jets because, . even though we know the number of major Major accident rate 5-year running average accidents for Eastern-built jets, we do not have . reliable worldwide exposure data (hours flown . or departures) to calculate valid rates for them. After a decade of an almost constant major . accident rate for commercial jets, we now see a . trend of improvement. . Business jets had 13 major accidents in 2012
Accidents per million departures . (Table 2, p. 20). This is slightly greater than their 2010 2011 12-year average of 10.5. Calculating accident Year rates for business jets is difficult due to the lack Note: Total departure data are not available for Eastern-built aircraft. of reliable exposure data. One rate that can be Source: Ascend calculated is the number of major accidents
© Muhammed Muheisen|Associated Press © Muhammed Muheisen|Associated per 1,000 aircraft. Using that metric shows the Figure 2
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 19 2012REVIEW
improvement in the business jet accident Major Accidents, Worldwide Business Jets, 2012 rate over the past eight years (Figure 3). The 17 major accidents involving Date Operator Aircraft Location Phase Fatal Western- and Eastern-built com- Feb. 2 Extrapoint Lear 35 Pueblo, Colorado, U.S. Takeoff 0 mercial turboprops with more than Feb. 12 Trident Aviation Svcs. Gulfstream IV Bakavu-Kavumu, DRC Landing 3 14 passenger seats in 2012 (Table 3) March 1 Asia Today Citation X Egelsbach, Germany Approach 5 were well below the 12-year average of Franklin-Macon, March 15 Private Citation I SP North Carolina, U.S. Landing 5 25.9. The past 12 years of turboprop June 18 Triple C Development Beech 400 Atlanta, Georgia, U.S. Landing 0 accident numbers show the record July 13 Universal Jet Aviation Gulfstream IV Le Castellet, France Landing 3 low in 2012 (Figure 4). Unfortunately, Santiago de CFIT continues to dominate the fatality Aug. 2 Airnor Citation 500 Compostela, Spain Approach 2 numbers for commercial turboprops. Sept. 15 Private Lear 24 Bornholm, Denmark Approach 0 In 2012, four of the 17 major accidents Sept. 18 Dewberry Air Beech 400 Macon, Georgia, U.S. Landing 0 (24 percent) were CFIT. Over the past Nov. 11 Tropic Air Taxi Aero Citation 525 São Paulo, Brazil Landing 0 six years, 28 percent (more than one Greenwood, in four) of the commercial turboprop Nov. 17 U.S. Customs Citation 550 South Carolina, U.S. Landing 0 major accidents have been CFIT. Dec. 9 Starwood Management Lear 25 Iturbide, Mexico En route 7 CFIT, approach and landing, and Oklahoma City, Dec. 21 U.S. Customs Citation 550 Oklahoma, U.S. Landing 0 upset aircraft accidents continue to ac- count for the majority of accidents and Source: Ascend cause the majority of fatalities in com- Table 2 mercial aviation. There were only seven commercial jet accidents in 2012, but Major Accidents, Worldwide Commercial Turboprops, 2012 six of the seven (86 percent) were ap- proach and landing accidents, and three Date Operator Aircraft Location Phase Fatal of the seven (43 percent) were CFIT. Jan. 30 TRACEP AN-28 Namoya, DRC En route 3 The upward trend of CFIT accidents April 2 Utair ATR-72 Tyumen, Russia Takeoff 31 for all commercial jets since 2009 (Fig- April 9 Air Tanzania DHC-8 Kigoma, Tanzania Takeoff 0 ure 5) is disturbing, particularly because April 28 Jubba Airways AN-24 Galkayo, Somalia Landing 0 more than 95 percent of commercial jets May 14 Agni Air DO-228 Jomsom, Nepal Approach 15 have been equipped with terrain aware- June 6 Air Class Líneas Aéreas SW Metro III Montevideo, Uruguay Climb 2 ness and warning systems (TAWS) since June 10 Ukrainska Shkola Pilotov LET-410 Borodyanka, Ukraine En route 5 2007. During the past six years, there June 20 ITAB Gulfstream I Pweto, DRC Landing 0
have been 37 commercial aircraft CFIT Aug. 19 ALFA Airlines AN-24 Talodi, Sudan Approach 32 accidents (14 jet, 23 turboprop). In the Aug. 22 Mombassa Air Safari LET-410 Ngeredi, Kenya Takeoff 4 past two years, more than 50 percent of Petropavlovsk- the commercial jet fatalities have been Sept. 12 Kamchatsky Air Enterprise AN-28 Palana, Russia Approach 10 caused by CFIT accidents. Sept. 28 Sita Air DO-228 Kathmandu, Nepal Climb 19 In 2006, upset aircraft accidents took Oct. 7 Azza Transport AN-12 Khartoum, Sudan En route 13 over from CFIT as the leading killer in Oct. 19 Air Mark Aviation AN-12 Shindand, Afghanistan Landing 0 commercial aviation. Over the past two Nov. 27 Inter Iles Air EMB-120 Moroni, Comoros Climb 0 years, commercial jets have suffered six Dec. 17 Amazon Sky AN-26 Tomas, Peru En route 4 CFIT accidents and no upset aircraft ac- Dec. 22 Perimeter Aviation Metro III Sanikiluaq, Canada Approach 1 cidents. Because of this, CFIT is about to Controlled flight into terrain (CFIT) accident regain its title as the leading killer. Source: Ascend But until then, upset aircraft ac- cidents still are the leading killer Table 3
20 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 2012REVIEW
in commercial aviation. In keeping with the Major Accidents, Business Jets, 2005–2012 terminology in the 1998 Airplane Upset Recov- 2 1.25 ery Training Aid, an aircraft is considered upset if one of the following conditions is met: pitch 1.00 attitude greater than 25 degrees nose up; pitch attitude greater than 10 degrees nose down; bank 0.75 8-year average: 0.67 angle greater than 45 degrees; within the previous 0.50 4-year parameters, but flying at airspeeds inappropriate average: 0.52 for conditions. “Upset aircraft” accidents include 0.25 accidents involving related terms such as loss of Accidents per 1,000 aircraft 0.00 control, lack of control, unusual attitude, stall, 2005 2006 2007 2008 20092010 2011 2012 Year extended envelope and advanced maneuvering. An upset aircraft accident is one in which Source: Ascend the aircraft is upset and unintentionally flown Figure 3 into a position from which the crew is unable to recover due to either aircrew, aircraft or en- vironmental factors, or a combination of these. Major Accidents, Worldwide Commercial Turboprops, 2001–2012 Another term used to describe these accidents is 40 “loss of control.” This is a somewhat misleading 39 Western-built Eastern-built term, because in 48 percent of the “loss of con- 33 29 30 31 31 trol” accidents over the past 10 years, there was 25.9 (12 years) no literal loss of control — the aircraft respond- 24 21.4 (5 years) 20 22 ed correctly to all control inputs. However, in 24 21 20 23 15 100 percent of the currently classified “loss of 10 control” accidents, the aircraft was upset. There Number of accidents currently are more than 15 international efforts 0 under way to address airplane upset prevention 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Year and recovery. The lack of any commercial jet upset accidents over the past two years indicates Source: Ascend that these efforts may be seeing some success.
Figure 4 James M. Burin was the director of technical programs at Flight Safety Foundation. CFIT Accidents, Worldwide Commercial Jets, 1998–2012 Notes
10 1. The termmajor accident was created by Flight 9 CFIT accidents 5-year running average 8 Safety Foundation in 2006. It refers to an accident in 7 which any of three conditions is met: The aircraft is 6 considered destroyed, as calculated by dividing the 5 estimated cost of repairs to the hypothetical value 4 3 of the aircraft had it been brand new at the time of 2 the accident; or there were multiple fatalities to the
CFIT accidents per year CFIT accidents 1 aircraft occupants; or there was one fatality and the 0 2001200019991998 2002 2003 2004 2005 2006 2007 2008 2009 2010 20122011 aircraft was substantially damaged. Year This criterion ensures that the categorization of an accident is not determined by an aircraft’s age or its CFIT = controlled flight into terrain insurance coverage. Source: Flight Safety Foundation 2.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 21 FLIGHTOPS
FailureBY J. MARTIN SMITH, DAVID W. JAMIESON to AND WILLIAM Mitigate F. CURTIS
Studying the psychology of decision making during unstable approaches and why go-around policies are ineffective.
he Flight Safety Foundation has analyzed the almost complete failure to call go-arounds as the past 16 years of aircraft accident data a preventive mitigation of the risk of continuing and found that the most common type of to fly approaches that are unstable constitutes accident is the runway excursion, which the number one cause of runway excursions, and Taccounts for 33 percent of all aircraft accidents.1 therefore of approach and landing accidents. If The highest risk factor for runway excursions our go-around policies were effective even 50 is the unstable approach.2 Unstable approaches percent of the time, the industry accident rate occur on 3.5 to 4.0 percent of all approaches, would be reduced 10 to 18 percent. There is no but only 3 percent of these unstable approaches other single decision, or procedure, beyond call- result in a go-around being called in the cockpit: ing the go-around according to SOPs that could almost all aircrew in this state — 97 percent — have as significant an effect in reducing our ac- continue to land. It can be argued, therefore, that cident rate. Why, then, is compliance so poor?
22 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 © Ismael Jorda/AirTeamImages.com the implicitthe incentive structure for flying prior to making decisions?”;“What their is information didpilots solicit to assess risk with go-around “What decision making?”; of pilot characteristics, ifany, are associated of instabilities than with “What others?”; sorts go-arounds associated more with some kinds attempted to answer such questions as: “Are leading up to adecision. The experiments we factors inwhich pilot assessed the survey results of two experiments conducted within ance. The remaining include articles will the to help mitigate risk and increase compli- applied other operational inseveral contexts Model (DSAM),that we successfully have Situational Dynamic the we call Awareness for understanding psychology, this which unstableto fly approaches. decisions to go around rather than to continue tois designed understand psychology of the Foundation by The Presage Group. The survey conducted pilotwide on survey of behalf the ect’s work, to which date includes aworld- latest the describe we results will of proj the - AeroSafety World over course the of year, this go-around maneuver itself. as well as risks associated the the with flying tions management’s role phenomenon, inthe contributionschosocial opera- flight behind iswhich ongoing, examine psy- will the also dations on findings. the The based project, approaches?” and making then recommen- notally made being during many so unstable should made according be to policy, actu - question, “Why are go-around decisions, that pliance to result from answering research the This enhancedpolicies. project com- expects level ofhigh pilot compliance with go-around caused by unstable approaches by achieving a to designed mitigateect runway excursions around Decision Making and Execution Proj- FLIGHTSAFETY.ORG This first article describes anovelThis first article describes strategy In of published articlesto aseries be in The Foundation in2011initiated aGo- |
WORLD AEROSAFETY | FEBRUARY 2013 environment, of inall its facets, and it is this or cognitivelogical is aware being act of their aware ance, must they first and foremost fully be shape decision their making around compli- and risks to aircraft stability, would which then operationalthe landscape for potential threats prior to pilots’ the ability to accurately assess pilot’s decision making? Well, put simply, very ness avaluable for method understanding a Why is an investigation into situational aware- Situational AwarenessDynamic Model with company policy. on abetterbased alignment of pilot psychology to offer about ideas how to mitigate risks these compliance with unstable approach SOPs, and to on bear thinking topicnew the of non- answers questions, to these our is goal to bring rants ago-around decision?” themselves inan to be unstable state that war parametersflight dopilots personally define beyond what thresholds of instability on key and “Apart from company’s their definitions, organizations have out set to handle them?”; standard operating procedures (SOPs) their andefines unstable condition, as well as the definitions by set organizations their for what ing protocols?”; “Do pilots accept basic the compliance with go-around decision- mak anyexperience post-decisional regret for non- risk assessments and decisions?”; “Do pilots sions?”; “What are key drivers true of the their ences that were inputs actually - to deci their approach,- experi the all and reflect dothese to go around or continue with an unstable what factors dopilots attribute decisions their with go-around “In policies?”; hindsight, to crew interactions that support compliance zation’s culture?”; “What is nature the of the approach that pilots perceive organi intheir - go-arounds versus continuing unstable the In other words, pilots’ the first psycho- very By understanding more completely the of objective the world around them. FLIGHTOPS - |
23 FLIGHTOPS
awareness that shapes and molds subsequent other pre-cognitive, intuitive, emotional and perceptions of operational risks and threats, and implicit knowledge comes into play? And how of the manageability of those risks and threats. does their immediate cockpit environment These perceptions and judgments in turn — the social realm, including the important inform decision making around risk appetite contributions of crewmembers, of communica- and compliance. From a psychological research tions styles and of interpersonal dynamics — point of view, it makes sense to test whether low influence their decisions? Within the DSAM situational awareness does in fact equate with In Figure 1, we present the simplified poor risk assessments and increased rates of sequence of events leading up to a decision be- model, situational non-compliance during the unstable approach. tween continuing an approach that is unstable in In order to fully picture how DSAM is lived by this scenario versus deciding to fly a go-around awareness comprises pilots during an unstable approach, consider the maneuver. Changing objective flight condi- nine distinct but following description. tions and developing instabilities (step 1 in the Imagine, a pilot-in-command and his/her sequence) must be noticed via the pilot’s senses, interconnected and crew are flying a routine approach when they registered and mentally processed in light of experience a late-developing instability below their always-developing expectations about their seamless sub-aspects stable approach height (SAH; as defined by their current situation. of awareness. company). How might we describe the psychol- Situational awareness of the environment, in ogy of this situation up to and including the all its facets and continuous state of flux (step 2), moment when this pilot decides to go around or is the psychological prerequisite state for a pilot not, and the experiences of both the pilot flying to judge risk at any moment (step 3), and then and other crewmembers as they handle this to make a subsequent decision to maintain com- rapidly changing situation under time pressure pliance and safety in light of that judgment (step and heavy workload? 4). This state of awareness must be continuously In the cognitive realm, what thoughts, updated and refreshed based on a stream of beliefs, expectations and information factor sensory inputs and knowledge provided by the into their situational appraisal of the instabili- pilots’ instruments, the kinesthetic and other ties and their manageability? By what cognitive senses, the crewmembers’ inputs and so forth. calculus do they assess risks of both choices, This study employed DSAM for measur- continuing to land or going around? What ing and interpreting the psychological and social factors that collectively make up situ- How Situational Awareness Plays a Role in Decision Making ational awareness. Within this model, situational awareness comprises nine distinct but intercon- nected and seamless sub-aspects of awareness (Figure 2). Much of the following discussion
Objective Situational Risk Decision will be framed around how each of these sub- conditions awareness assessment making aspects influences a pilot’s risk assessment and decision-making processes, singly and in concert with one another, to remain compli-
Psychological ant versus non-compliant in the face of aircraft instabilities while on approach.
Source: The Presage Group In a typical response to the unstable approach event we have imagined, a typical situational Figure 1 awareness profile (SAP) emerges for the pilot
24 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 FLIGHTOPS
flying the aircraft so as to maintain compliance Breaking Down Situational Awareness with company SOPs. The “sequence” of processes might look something like the following. Imagine again our late-developing instability below SAH, Gut feeling Company and consider the pilots’ phenomenological expe- for threats support (a ective) for safety riences of it, as it is lived, through its description (environmental) as an SAP. (Note: While we have serialized the Situational Knowing awareness subsequent description in steps to easily explain the the Knowing the instruments procedures various DSAM awareness concepts, these aware- (functional) (hierarchical) nesses actually exist in a mutually interdependent whole of causation, with rapid feedback loops and Relying on Keeping interactions. Changes to one type of awareness experience each other (critical) safe quickly influence the others in a psychological (relational) Seeing the process called “spreading activation.”) Knowing Adjusting threats to threats (anticipatory) the limits (task (compensatory) Example of a Go-Around Experience empirical)
1. At a point immediately above SAH, the pilot’s “gut,” or what we refer to as affective aware- Source: The Presage Group ness , subtly signals him or her to confirm that Figure 2 the aircraft’s flight characteristics and profile are normal. In a near-instantaneous and with the reactivated gut, expert instrument seamless fashion, this might be followed by … knowledge and experience — all awarenesses 2. A visual check, or what we refer to as a check that are now signaling a non-normal event to provide functional awareness, which would — and there arises an immediate need for a be made where the pilot’s expert knowledge signal from … and ability to understand the instruments 5. Task-empirical awareness, the pilot’s expert plays a key role in confirming whether the knowledge of the safe operational enve- cue from their gut was, in fact, correct. lope limits of the aircraft. Imagine further Simultaneously, there is … that this expert knowledge confirms that 3. An immediate and confirmatory statement although the aircraft is now unstable, it from the pilot’s network of past experiences, or still remains within the safe operational critical awareness, occurs, in which professional envelope. However, before concluding that experience confirms the presence of a “normal” parameters are now safe or unsafe, manage- flight profile. Seconds later, however, imagine able or unmanageable, this developing event that in continuing its descent below SAH, the requires immediate input from another aircraft encounters significant turbulence with awareness competency … headwinds shifting to tailwinds and down- 6. Compensatory awareness, or the ability to un- drafts altering VREF (reference landing speed) derstand how to compensate correctly for non- by +21 kt, accompanied by a vertical descent normal events, occurs by referencing through now greater than 1,100 fpm. Instantly, … functional awareness whether the aircraft and 4. The pilot’s anticipatory awareness, the ability the instruments will direct the flight state back to see these threats, registers in harmony to a normal condition if acted upon. Whether
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 25 FLIGHTOPS
the answer, not yet fully formed but informed needs to be used repeatedly and quickly to keep by critical awareness, is likely to be “yes,” “no” the entire spectrum of potential threats in the or uncertain, imagine that the pilot is also situation and their possible causes and resolutions simultaneously receiving … alive in awareness at all times, and that there is a natural system of mutual causation among them 7. Through relational awareness — the pilots’ that also must be sustained to maximize safe deci- knowledge of how they use their relation- sion making. ships to protect safety — input from a Table 1 summarizes the complete system crewmember that re-enlivens a memory of nine DSAM concepts (called Construct 1 to trace of a prior verbal signal, based on a Construct 9). The table shows a working name conversation and agreement earlier in the for each concept, its scientific name within our approach initiated by the pilot monitoring, system and a brief definition. that a go-around might be necessary should the aircraft become unstable at or below What’s Ahead? SAH, which … So far, we have summarized the problem and 8. Informs and motivates the pilot to engage described the psychological theory behind our hierarchical awareness, or the individual’s experimental survey framework. Robust exper- expert knowledge of operational procedures imental survey work, which we will describe under specific operational conditions, so in the next articles in this series, provides valid as to confirm their ability to safely fly a go- data and meaningful insight and understand- around if necessary. Finally, with the pilot- ing into this critical issue for flight safety, and in-command and other crew rapidly coming from which we can begin to define necessary to a common assessment of, and agreement corrective actions. We will describe the about, the risks inherent in continuing with In these articles, we will describe the results the unstable situation that faces them, in from analyzing data from more than 2,300 pi- results from more comparison with the inherent risks of any lots asked to recall in detail the last instance in go-around maneuver, and which they experienced an unstable approach than 2,300 pilots that either led to a go-around decision or a 9. Confident that their company would sup- asked to recall decision to continue in the unstable state and port a decision to initiate a go-around, and land. In addition to reporting about various in an expression of their environmental in detail the last aspects of the DSAM model that differ between awareness concerning the wider organi- pilots going around and those continuing the instance in which zational reward structures surrounding unstable approach, we will also describe the support for safety, the pilot flying puts all kinds of objective flight conditions and pilot they experienced an of these elements of awareness together to characteristics that are associated with these judge that the risks confronting the flight unstable approach. decisions, post-event perceptions pilots had crew are not fully manageable, and so about the causes of their decisions and hind- decides to call for a go-around. sight judgments they made about the wisdom Again, this description is not in any way intended of their choices. to be prescriptive, that is, to suggest the way the We will also describe the results of a dynamic situational awareness processes should second experiment conducted using the same work in this situation (i.e., the sequence or inter- survey, which was designed to investigate the actions among the awareness types). But it does il- personally held thresholds for instability that lustrate that each of these awareness competencies pilots believed would necessitate a go-around
26 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 FLIGHTOPS
Constructs in the Dynamic Situational Awareness Model
DSAM Construct Name Description
“Gut feeling for threats” Pilot’s gut feelings for threats; seat of the pants experience, which is characterized by an Affective awareness (C1) emotional, sensory experience that triggers further cognitive analysis.
“Knowing the instruments and equipment” Pilot’s expert knowledge of knowing how to read and translate what his/her instruments are Functional awareness (C2) telling him/her.
“Relying on experience” Pilot’s ability to draw from his/her personal and professional experience bank as a means to Critical awareness (C3) assess here-and-now events as “normal” or “abnormal.”
“Seeing the threats” Pilot’s ability to see and/or monitor real and potential threats as they move and change over time Anticipatory awareness (C4) and through space.
“Knowing the limits” Pilot’s expert knowledge of the safe operational envelope of his/her equipment. Task-empirical awareness (C5)
“Adjusting to threats” Pilot’s ability to know how and when to compensate or adjust correctly for present and Compensatory awareness (C6) anticipated future operational conditions to ensure safe SOP-compliant operations.
“Keeping each other safe” Pilot’s ability to accurately assess and engage crewmember relationships in a manner that Relational awareness (C7) protects safety and compliance.
“Knowing the procedures” Pilot’s expert and comprehensive knowledge of operational procedures, their order and correct Hierarchical awareness (C8) sequencing.
“Company support for safety” Pilot’s experience of how his/her company supports and encourages safety and how this in turn Environmental awareness (C9) shapes his/her commitment to safe and compliant behavior.
DSAM = Dynamic Situational Awareness Model; SOP = standard operating procedure Note: The informal construct names in quotes appear above the corresponding standardized terms from the Presage researchers’ glossary.
Source: The Presage Group
Table 1 call, considering deviations in several flight The Presage Group specializes in real-time predictive analyt- parameters. Later articles will report on other ics with corrective actions to eliminate the behavioral threats of employees in aviation and other industries. Further aspects of the FSF project, including the re- details of the methodology of their survey, experiments and sults of the management survey on this topic results are described at
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 27 PRISM Complete Safety Management Solution
PRISM’s unique internet-enabled solutions coupled with our dedicated in-house Helicopter Safety Experts address the challenges of setting up a functional and effective Safety Management System (SMS) PRISM Affiliations: PRISM Services Include: • SMS Briefings for Executives • Training Classes available for Helicopter Operators (Delivered in English or Spanish) • SMS Facilitation Services (Develop, Plan, Implement) • On-site SMS Evaluation and Guidance • Manual Review and Development • Certification Consultant • “ARMOR” a Web-based Safety and Quality Management Tool Compatible with iPad®
[email protected] +1 513.852.1010 www.prism.aero © Jason R. Fortenbacher/Fight to Fly Photography Unintended FLIGHTSAFETY.ORG A (NTSB) says. U.S.the National Transportation Safety Board a Eurocopter EC135inCave Creek, Arizona, U.S., probable cause of Feb. the 14,2010,fatal crash of | AEROSAFETYWORLD tently stepped on collective—were the the moved from her father’s lap and inadver likely” resulted a5-year-old when girl brupt control movements “highly —which | FEBRUARY2013 - led tothefatalcrashofanEC135,NTSB says. (241 km) to(241 km) south. the Parks, about Arizona, Airport, to 150mi Scottsdale planned from flight Whispering Pines Ranch near conditionscal around during time a 1505local meteorologi invisual crash,the occurred which - ofsence proper cockpit discipline from pilot.” the The pilot and four all passengers were in killed The NTSB cited as acontributing factor “ab the - ‘Abrupt andunusual’control movements BY LINDAWERFELMAN CAUSALFACTORS
|
29 30 | CAUSALFACTORS of ScottsdaleAirport. 14 nm(26km)north access road about wash onagravel crashed nearariver The helicopter owner or child were the and secured restrained that he could not tell ifeither helicopter the man replied ‘occasionally.’ The foreman stated front cockpit seat with her father, ranch the fore- asked how frequently child occupied the left the tioned on her father’s lap,” report the said. “When frontleft cockpit- seat, “with posi girl small the forward cockpit door. estimated at 42lb (19kg),board through left the ownerthe and his daughter, weight whose was dogs were boarded before foreman the observed ing engines. the Two adult passengers and two ing into right the front cockpit seat and- start conducted preflight the before inspection climb- accident,the pilot the helicopter the loaded and telling NTSB investigators that, on day the of movements.” would not intentionally make such control trol inputs,” report the said. “A helicopter pilot collective andfull-up con cyclic near- full-aft followed by asimultaneous reaction of nearly ering of collective to the near lower the stop, have was as aresult occurred of asudden low- preceded crash. the drive shaft, resulting loss inthe of control that horizontal left the endplate and rotor tail the tion that revealed arotor blade had struck November NTSB 2012,the said its investiga - The owner and his daughter sat both inthe The report quoted ranch the foreman as only way“The that condition this could In accident final the report approved in pilot for certificate single-engine airplanes, is - formed satisfactorily and noted no deficiencies. training records indicated that pilot the had per trainingflight in2003,2004,2006and 2008.All completed in2002,with recurrent ground and Vietnamthe War. pilot had U.S. flown Army helicopters during time, flight butmilitary SGApersonnel said the pilot’sthe and logbook found no record of certificate. Investigatorsmedical didnot obtain owned helicopter. the He had asecond-class also Groupnel at of Services (SGA),which America accident, according to operations flight person- EC135 T1and 13hours 90days inthe before the hours,11,000 flight including 824hours inthe The 63-year-old helicopter pilot had about 11,000 Flight Hours area and was consumed by fire. The helicopter ground the struck inariver wash “circled and dove to ground,” the report the said. separate seconds of final inthe before flight, it sawSome of said they parts helicopter the licopter and descended crashed into ground. the heard popping or banging sounds before he the - said they north of Airport, (26 km) Scottsdale strapped his daughter inon top of him.” on previous helicopter the flights, owner had helicopter.in the The foreman that revealed spoke of pressure the he felt inhis job. that, during training sessions, accident the pilot ter instructor pilots told accident investigators at of time the accident. the tors could not determine man which was flying controls. the take The report said that investiga- and to“liked fly” that it was common for himto tigators, SGA chief the pilot said that owner the but inpost-accident comments to accident inves- background. and little other information about his aviation no indication that he certificate held amedical ofsued in1967.Areview FAA records revealed The owner of SGAwas 64and held aprivate His training initial EC135T1was inthe Witnesses near crash the site, about 14nm The report said that two American Eurocop- The owner didnot have ahelicopter rating, FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY2013 -
U.S. National Transportation Safety Board CAUSALFACTORS
One instructor said the accident pilot “dis- played an abnormally high degree of perceived Eurocopter EC135 T1 pressure to accomplish flights from the owner of the helicopter” and was “visibly shaken when discussing the amount of pressure he received.” The instructor said that, during initial tran- sition ground school training in 2002, the ac- cident pilot had told him “that it would not be uncommon to fly the helicopter’s owner from Seattle to his home of Vashon Island when the weather conditions at night were so poor that they would follow the ferryboat lights to navi- gate across the bay under foggy conditions.” The other instructor said that, during a
2008 training session, the accident pilot had © Łukasz & Maciek Golowanow Hypś, Konflikty.pl/Wikimedia commented “about the owner dominating the he EC135 is a twin-turbine light helicopter first flown in 1988 with cockpit duties prior to a flight. two Allison 250-C20R engines. “I emphasized the importance of following T The T1, first delivered to a U.S. customer in 1996, is the the checklist and always performing the hydrau- Turbomeca engine version. The accident helicopter was equipped with lic check. He commented that when the owner two TM USA Arrius 2B1 turboshaft engines. The helicopter can be equipped to seat up to eight people. It has flies, he gets in the cockpit and ‘flips switches a maximum normal takeoff weight of 5,997 lb (2,720 kg), maximum and goes.’ I felt [the accident pilot] was intimi- cruising speed at sea level of 139 kt and a maximum range at sea level dated by the owner and would not insist proper with standard fuel of 402 nm (745 km). aircraft procedures be followed.” Source: Jane’s All the World’s Aircraft, U.S. National Transportation Safety Board Accident Report No. In information submitted by SGA for the WPR10FA133 accident investigation, the company’s chief pilot questioned the instructors’ accounts. He wrote that he considered the accident pilot as “not a owner in 2002 and had accumulated 1,116 pilot who would be intimidated” and “a consci- operating hours. It had been maintained in ac- entious and professional pilot, in every sense of cordance with the manufacturer’s recommenda- t h e w ord .” tions, and its most recent annual inspection had Noting that the instructors had “inferred been conducted Oct. 30, 2009. that [the accident pilot] feared for his job if he The helicopter had two Turbomeca USA did not perform his trips regardless of risk,” Arrius 2B1 turboshaft engines. At the time of the chief pilot said, “After 24 years of service the October 2009 inspection, the left engine had with Services Group of America, there could be recorded 1,103 hours total time since new, and nothing further from the truth. I do not believe the right engine, 227 hours. that an individual could stay at any company The helicopter had been involved in three that long if they felt such pressure from their incidents before the crash, the report said. employer.” The accident pilot had left SGA in In the first incident, in May 2003, the the late 1990s but returned three years later and helicopter’s owner was at the controls when the remained with the company until his death, the left seat — reportedly “not in the proper detent chief pilot said. position” — slid aft, the report said. “The helicopter dropped about 50 ft but Three Incidents was recovered by a quick collective input,” the The accident helicopter was manufactured in report said. “In an incident report submitted 1999, was purchased by SGA from its original by American Eurocopter, it was reported that
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 31 CAUSALFACTORS
a loud bang was heard, followed by the touch- last radar return was recorded at 1503:37, about down of the helicopter.” two minutes before impact. An NTSB radar The impact damaged the horizontal stabi- study said the helicopter’s last known position lizer, and pieces of the engine were found on was about 14 nm (26 km) north of Scottsdale the ground. The helicopter was repaired and Airport, above the accident site at 3,700 ft above returned to service in August 2003. mean sea level. The second incident, with the accident pilot The NTSB investigation found that a single flying, involved a January 2004 hard landing at impact of one of the main rotor blades had dam- a grassy heliport on Vashon Island, Washington, aged the tail rotor drive shaft. U.S. After repairs, the helicopter was returned to “No pre-impact failures or material anoma- service in April 2004. lies were found in the wreckage and component In September 2007, an engine chip light examinations that could explain the divergence illumination occurred, followed by a yaw, an of the … blade from the plane of main rotor engine shutdown and a single-engine landing; rotation,” the report said. the engine was replaced in January 2008. The most likely explanation, the report add- In most helicopters, In addition, one of the helicopter’s main ed, was that “all of the main rotor blades were including EC135s, rotor blades was removed in November 2009 be- following a path that would have intersected the the collective pitch cause maintenance personnel could not balance tail rotor drive shaft as a result of an abrupt and control lever is it correctly, and a temporary replacement blade unusual control input.” on the left side of was installed. The replacement was still in place The report said investigators had conducted the pilot’s seat. when the accident occurred. a biomechanical study that showed that “it was feasible that the child passenger … could fully depress the left-side collective control by step- ping on it with her left foot” to stand up from her place in her father’s lap. “It is highly likely that the child inadver- tently stepped on the collective with her left foot and displaced it to the full down position,” the report said. “This condition would have then resulted in either the pilot or the helicop- ter owner raising the collective, followed by a full-aft input pull of the cyclic control and the subsequent main rotor departing the normal plane of rotation and striking the left endplate and the aft end of the tail rotor drive shaft.”
This article is based on NTSB accident report no. WPR10FA133 and accompanying docket information. U.S. National Transportation Safety Board Safety Transportation National U.S. Note
Clear Skies 1. The collective pitch control is the part of a heli- Visual meteorological conditions prevailed at copter’s flight control system that simultaneously the time of the accident, with clear skies, 10 mi changes the pitch angle of all main rotor blades. In (16 km) visibility and no wind. the EC135, and in most other helicopters, the collec- tive is on the left side of each pilot’s seat. The cyclic, Air traffic control facilities had no contact located in the EC135 between the pilot’s legs at the with the pilot on the day of the accident. Radar center of each pilot’s seat, changes the pitch of the showed the helicopter flying south toward rotor blades one at a time, as each blade rotates past Scottsdale from Whispering Pines Ranch; the the same point in the rotor disk.
32 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 CARGOSAFETY Detect and
SuppressBY LINDA WERFELMAN
The NTSB urges improved fire detection, suppression and containment systems to prevent injury and damage in cargo airplane fires.
xisting fire-protection regulations for p. 35) — in issuing three safety recommenda- cargo airplanes are inadequate, and action tions in late November to the U.S. Federal Avia- is needed to improve the detection and tion Administration (FAA). suppression of blazes in cargo containers, “These fires quickly grew out of control, Ethe U.S. National Transportation Safety Board leaving the crew with little time to get the air- (NTSB) says. craft on the ground,” NTSB Chairman Deborah The agency cited information gathered in A.P. Hersman said. “Detection, suppression its recent cargo container fire study and the and containment systems can give crews more investigations of three in-flight cargo airplane time and more options. The current approach fires — a February 2006 fire on a United Parcel is not safe enough.” Service (UPS) McDonnell Douglas DC-8- The NTSB’s recommendations call on the 71F; the fatal September 2010 crash of a UPS FAA to: Boeing 747-400F; and the fatal July 2011 crash “Develop fire detection system perfor-
©Kris Klop of an Asiana Cargo 747-400F (“In-Flight Fires,” mance requirements for the early detection of
WWW.FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 33 CARGOSAFETY
fires originating within cargo containers and In addition, “the effect of the use of contain- pallets and, once developed, implement the ers and pallets to contain cargo is not factored new requirements; into the current overall fire protection strategy “Ensure that cargo container construction or certification process,” the NTSB said, noting materials meet the same flammability require- that the certification process is conducted using ments as all other cargo compartment materi- empty cargo compartments. als in accordance with [U.S. Federal Aviation In a letter to then-Acting FAA Administra- Regulations (FARs) Part 25.855]; and, tor Michael Huerta that accompanied the safety “Require the installation and use of active recommendations, the NTSB discussed the fire suppression systems in all aircraft cargo findings of accident investigations and a series compartments or containers, or both, such that of tests conducted in August 2011 to develop fires are not allowed to develop.” a better understanding of cargo container fires Cargo aircraft currently are subject to the and the most appropriate prevention strategies.1 same FAA fire-protection regulations that The tests — designed in part to examine govern all transport category aircraft, the NTSB the burning characteristics of cargo container said. fires — prompted researchers to conclude that “Although these regulations limit the flam- “container design has a significant effect on mability of construction materials used in cargo the time it takes for an internal fire to become compartments and also specify minimum fire detectable to a smoke detector outside the resistance requirements for cargo compartment container” and that “container construction liners, there is limited regulation concerning fire materials have a significant effect on the total protection associated with cargo containers,” the fire load2 and energy release rate of a cargo NTSB said. fire,” the NTSB said. Smoke billows from For example, the agency noted that materials In the two accidents in 2010 and 2011, an aluminum and selected for the construction of cargo contain- investigators found “a relatively short interval polycarbonate cargo ers undergo a horizontal Bunsen burner test, between the time a fire warning indication container during “which does not prevent the use of highly com- was delivered to the flight crew and the onset flammability tests. bustible materials.” of flight control and aircraft system failures,” the NTSB said. In the fatal UPS crash, about 2 minutes 30 seconds elapsed between the first fire warning and the loss of some aircraft systems; timing information has not been released in the ongoing Asiana investigation, the NTSB said. The NTSB’s report on the 2011 tests, pub- lished in a report in March 2012, concluded that “the time it takes for a fire detection system to detect a fire originating within a cargo container may easily exceed the one-minute time frame specified in … Part 25.858(a)” and that “the growth rate of container fires after they become detectable by the aircraft’s smoke detection system can be extremely fast, precluding any mitigating action and resulting in an over- whelming fire.” In tests of cargo containers, the NTSB found
that the time between fire initiation and fire Board Safety Transportation National U.S.
34 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 CARGOSAFETY
detection ranged from 2 minutes 30 seconds to 18 minutes 30 seconds — longer than the one- In-Flight Fires minute detection time currently required. “The fires grew very large, capable of caus- hree in-flight cargo airplane fires were cited by the U.S. National ing significant damage to an aircraft, shortly af- Transportation Safety Board (NTSB) in its recommendations for ter becoming a detectable fire,” the report said. Timproved fire safety. “The NTSB is concerned that, when fires inside The first was a Feb. 7, 2006, fire in a United Parcel Service containers become detectable to the aircraft’s (UPS) McDonnell Douglas DC-8-71F, which landed at Philadelphia smoke-detection system, there is little time International Airport after the crew smelled smoke and then — 20 min- until the fires reach levels that can compromise utes later — the “CARGO SMOKE” light illuminated (ASW, 4/08, p. 28). All three crewmembers were treated for minor injuries from the integrity of the cargo compartment and smoke inhalation, and the airplane was destroyed. The NTSB said the then threaten the structure and systems of the cargo fire began “from an unknown source,” probably inside one of aircraft. … the DC-8’s cargo containers; contributing factors were the “inad- “If the fire were to be detected while generat- equate certification test requirements for smoke and fire detection ing smoke inside the container, valuable time systems and the lack of an on-board fire suppression system.”1 would be gained for alerting flight crews and Deborah Hersman, a member of the NTSB and now its chair- man, said during the agency’s public hearing on the accident that mitigating the effects of the fire.” the flight was “seconds from disaster.” Because existing regulations dealing with The second fire broke out on a UPS Boeing 747-400F flammability limits are “very limited” for cargo that crashed Sept. 3, 2010, inside an army base near Dubai container materials, those materials can sig- International Airport (DXB) in the United Arab Emirates. The two nificantly increase the fire load within a cargo flight crewmembers were killed, and the airplane was destroyed. compartment, the NTSB said. An interim report by the UAE General Civil Aviation Authority (GCAA) said that a fire warning light illuminated about 22 minutes For example, the agency cited collapsible after takeoff from DXB while the airplane was in cruise at 32,000 ft. containers made from corrugated polypropylene The crew declared an emergency, and the airplane crashed as they as “significant contributors” to fire intensity. maneuvered to land at DBX. The investigation is continuing.2 The third fire occurred July 28, 2011, on an Asiana Cargo Fire Suppression 747-400F, which crashed 70 nm (130 km) west of Jeju Island, Most current practices base fire suppression Republic of Korea, as the flight crew attempted to divert to Jeju International Airport because of the fire. Both pilots were killed, in main deck cargo compartments on pas- and the airplane was destroyed. The investigation by the South sive suppression systems, such as the use of Korean Aircraft and Railway Accident Investigation Board (ARAIB) fire-resistant materials and oxygen depriva- is continuing.3 tion. Because the compartments are so large, — LW however, fires can become very large before Notes oxygen deprivation slows their growth, the 1. NTSB. Accident Report No. NTSB/AAR-07/07, “Inflight Cargo Fire; United NTSB said. Parcel Service Company Flight 1307; McDonnell Douglas DC-8-71F, In the 2006 UPS blaze, the agency said, N748UP; Philadelphia, Pennsylvania; February 7, 2006.” Dec. 4, 2007. “the aircraft did not achieve depressurization 2. GCAA. Accident Reference 13-2010, “Air Accident Investigation [which aids in suppressing flames] until after Interim Report: Boeing 747-44AF, N571UP; Dubai, United Arab Emirates; September 03, 2010.” system failures and flight control issues began to occur.” 3. ARAIB. ARAIB/AAR1105, “Aircraft Accident Investigation Interim Tests by FAA researchers have found that, Report: Crash Into the Sea After an In-Flight Fire; Asiana Airlines, B747-400F/HL7604; 130 Km West of Jeju International Airport; July although depressurization contributes to fire 28, 2011.” suppression, when an aircraft descends to a more oxygen-rich environment, the fire again begins to grow. that passive fire suppression in large cargo com- “Hence, experience from the UPS [Dubai] partments due to oxygen deprivation may not accident, as well as FAA experiments, suggest be effective,” the NTSB said.
WWW.FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 35 CARGOSAFETY
The agency noted that, in 2007, as a result of high as 1,200 degrees [F (649 degrees C)], the its investigation of the 2006 fire, it had recom- fire was suppressed for four hours and 95 per- mended that the FAA require fire-suppression cent of the packages were undamaged. Even the systems for the cargo compartments of all FARs laptops worked.” Part 121 cargo airplanes. The NTSB reported that UPS said that it also is installing quick- the FAA’s response had been that the cost of in- donning integrated oxygen masks and smoke stalling “compartment-flooding fire-suppression goggles in all aircraft, and the VisionSafe Corp. systems, as those used in Class C cargo compart- Emergency Vision Assurance System (EVAS), ments,3 was not justified for the main deck cargo designed to displace smoke in a pilot’s vision compartments of aircraft of any weight.” path to allow him or her to see basic flight However, the NTSB said that the fires in 2010 instruments and the flight path, as well as and 2011 “continue to demonstrate the critical emergency checklists and navigation charts.5 need to suppress cargo fires.” As an alternative to the compartment- Automatic Suppression Systems flooding system evaluated by the FAA, the FedEx Express began installing on-board auto- NTSB suggested alternatives, including the matic fire-suppression systems in its aircraft in “aircraft-based system” used by FedEx and in- 2009, the same year it won the FSF-Honeywell container suppression systems being developed Bendix Trophy for Aviation Safety for develop- by the industry. ing the devices (ASW, 11/09, p. 39). The system incorporated infrared heat sen- ‘Multi-Layered Approach’ sors, foaming-agent generators and an overhead The NTSB’s issuance of the safety recommenda- cargo-container injector. If the sensors detect tions coincided with an announcement by UPS heat in a cargo container, the overhead fire- that it had developed a “multi-layered approach suppression equipment activates, piercing the consisting of matched solutions that include container and injecting argon foam. At the checklists, training and new technologies” to same time, the crew is alerted. mitigate in-flight cargo fires. Among those new technologies are fire- Notes resistant fiber-reinforced plastic containers, ex- 1. NTSB. Materials Laboratory Study Report No. 12- perimental fire-suppression units that “smother 019. March 21, 2012. In addition to addressing the a fire with potassium aerosol powder and can burning characteristics of container fires, the study save 95 percent of packages in the container” also examined the fire-load contribution of lithium and lithium-ion batteries. The NTSB noted that the and fire-containment covers for palletized cargo. involvement of these types of batteries “has come The approach was developed by a UPS/In- into question” in both the 2006 fire and the 2010 fire. dependent Pilots Association task force that had 2. Fire load is defined by the NTSB as “the amount of com- identified as its first step “increasing the time a bustible material that can become involved in a fire.” crew had to manage a smoke or fire event,” said 3. Class C cargo compartments have smoke or fire Capt. Bob Brown, a task force member. detector systems that provide warnings on the flight The group’s goal was to contain a fire inside deck; built-in, pilot-controlled, fire-suppression a unit load device (ULD) for four hours. In a systems; methods of excluding hazardous amounts test in October, a ULD containing 215 packages, of smoke from any occupied portions of the airplane; including “20 working laptops with batteries, 50 and methods of controlled compartment ventilation. working cell phones with batteries and 300 bulk- 4. Brown, Bob. “UPS/IPA Safety Task Force Pioneers shipped lithium ion batteries, was set on fire by Advancements in Aviation Safety.” Leading Edge (Fall six lithium ion batteries,” Brown wrote in Lead- 2012): 2. ing Edge, the UPS flight operations and safety 5. VisionSafe Corp. EVAS.
36 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 U.S. National Aeronautics and Space Administration “S BY EDBROTAK Sandy: York City area. LaGuardia Airport, Wednesday, 31. Oct. Oct. 28,2012,andOct. continuing through which iswhich located on awaterfront, suf- tarmac flooded, and didnot reopentarmac flooded, near $500million. More than 20,000 until Nov. 1.In anticipation of Sandy’s flights wereflights canceled starting Sunday, major Sunday closed airports night. fered significant damage, with the exceed $50billion,exceed and estimated losses forlosses airline the are industry strong winds, New York City’s three FLIGHTSAFETY.ORG Particularly hard hit was New the United States. Total may losses uperstorm” Sandy go down will as one of most the destructive storms of history inthe the | AEROSAFETYWORLD
| FEBRUARY2013 Airport outsideAirport Washington, 47kt; Weather conditions steadily deterio- (Maine) International Jetport, 48kt. All U.S. International Dulles Coast: East Philadelphia International 59 Airport, International 52kt; Portland Airport, wind gusts near 70kt were recorded, reported at various along airports the rated overnight. Winds were gusting and at wind-driven times, rain severely of these peak gusts occurred within gusts occurred of peak these over 40kt by morning. In evening, the port inNewport York, 69kt; Boston Logan kt; John F. Kennedy International Air lowered visibility. several hoursseveral on evening the 29. of Oct. seen byseen wind looking gusts at peak the The magnitude of storm the can be Type of Storm A Different Different A - when itwhen first reached tropical storm was referred to as Sandy, name the given was coming ashore near Atlantic City, New York and New Jersey 29, on Oct. New Jersey, it amidlatitude became or day, 30,and conditions Oct. flying im- ricane, eventhough it maintained the to buildings and significant damage to meteorologists stopped it calling ahur aircraft on ground the were reported. operations.airport damage Structural extratropical cyclone, losing its tropical proved, widespread power outages and a lack of transportation surface impeded same intensity. Technically, just as Sandy status 23.As on Oct. Sandy was ravaging Throughout its duration, storm the Even after winds subsided on Tues-
AV WEATHER - |
37 AVWEATHER
characteristics. The U.S. National Weather Ser- area above the surface low. This removal of air vice continued to use the name Sandy to avoid lowers the surface pressure. For extratropical public confusion. The results — the wind, the cyclones, the divergence aloft is produced on rain and the massive storm surge — were the the east side of a pre-existent upper-level trough same, regardless of the nature of the storm. of low pressure. When this upper trough moves The storm was So, why should we be concerned if a storm over a surface front, cyclogenesis — the process is tropical, extratropical or something in be- by which the low pressure area develops — oc- 1,000 mi (1,609 km) tween? Consider these elements of Sandy: The curs. For tropical cyclones, there are no pre- across, more than storm was 1,000 mi (1,609 km) across, more existent upper features. However, over time, the than twice the size of the large and extremely towering cumulonimbus clouds release enough twice the size of the destructive Hurricane Irene that affected this heat aloft to develop a high pressure area over same region in August 2011. When Irene came the low-level cyclone. This self-developed high, large and extremely up the East Coast, it weakened considerably miles above the surface, provides the divergence destructive Hurricane over the cooler waters, typical of a true tropical aloft needed to maintain the surface storm. system. Sandy did not weaken, even though it It is not unusual for tropical cyclones to be- Irene that affected traversed the same waters nearly two months come extratropical. If the tropical system moves later in the year. In fact, it strengthened. The into higher latitudes, especially in the late fall, it this same region central pressure fell to 940 millibars (mb; 27.76 can merge with a midlatitude frontal system and in Hg), 20 mb lower than the famed superstorm its attendant upper trough. Some of these con- in August 2011. that moved up the East Coast in March 1993. verted storms can be very strong. The “textbook What are the differences between tropical case” was Hurricane Hazel, which moved up the cyclones such as hurricanes and extratropical East Coast in 1954. Hazel came ashore along cyclones, the typical winter storms? the extreme southern coast of North Carolina Tropical cyclones only develop over warm on Oct. 15. It was a powerful Category 4 (ASW, waters, usually in the lower latitudes. Extra- 7/12, p. 29) hurricane with maximum sustained tropical cyclones can develop over land or water winds of 110 kt and a minimum central pressure where the air is colder and have even occurred of 937 mb (27.67 in Hg). It almost immediately in Arctic regions. Extratropical cyclones require joined a strong cold front and began to ac- a temperature contrast to develop. They usually celerate to the north. Cold air poured into the form along fronts that separate warm and cold system from the west, quickly transitioning air masses. Tropical cyclones develop within the storm into an extratropical system. Unfor- a single warm, humid air mass with no fronts tunately, weakening was limited, and by the involved. Tropical cyclones get their energy next day, it passed Washington with a central from the warm ocean water. Evaporation puts Figure 1 vast amounts of water vapor in the air. When the air is lifted in the storm’s circulation, the water vapor condenses in the towering cumulo- nimbus clouds, releasing latent heat that drives the storm. Extratropical cyclones derive their energy from the temperature contrast between warm and cold air masses. Energy is released as the warm air is lifted over the cold. An earlier article (ASW, 2/12, p. 48) described how a surface low pressure area is produced. Air is removed from above in the process called divergence. Air is lifted by the
low pressure and then spreads over a larger and Space Administration National Aeronautics U.S.
38 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 AVWEATHER
pressure of 970 mb (28.64 in Hg) and sustained winds of 100 kt and a central a “hybrid cyclone,” part tropical system, the transformed Hazel produced a gust pressure of 954 mb (28.17 in Hg; Figure part extratropical system. at Washington National Airport (now 1, p. 38). Tropical storm or gale force Meteorologists have known about Ronald Reagan Washington National winds (34 kt or greater) spanned a hybrid storms for years. They have even Airport) of 98 mph, a record that still diameter of 200 mi (322 km). By the classified one type of hybrid storm, the stands. Hazel continued hundreds of next day, a weakened Sandy continued subtropical cyclone. These low pressure miles northward into Canada, still pro- to move northward toward the Baha- areas develop only over ocean areas and ducing hurricane force winds, although mas. The central pressure had risen to have characteristics of both extratropical its track was well inland. 969 mb (28.61 in Hg), and maximum and tropical cyclones. Most subtropical Sandy made the transition to fully winds were barely hurricane force (64 cyclones develop from midlatitude, deep extratropical, probably a few hours kt), but the storm had doubled in size upper-level troughs or closed lows. They before landfall on Oct. 29. Of more with gale force winds covering 400 mi actually develop downward and eventu- interest was what was happening with (644 km). And by late in the day on Oct. ally produce a surface low. The cloud Sandy in the days prior to this. Sandy 26, Sandy no longer looked like a true pattern resembles a comma, very notice- started as a pure tropical system. It tropical system on satellite imagery. It able on satellite imagery. The strongest formed in the Caribbean Sea on Oct. was still warm core with convection winds, which can exceed hurricane force, 22. Three days later, as it was com- near the center, but it now had a long are found well away from the center of ing ashore on the south coast of Cuba, frontal-looking cloud band associated the storm, unlike tropical systems. If this Sandy was at its maximum strength with it. Forecasters at the National Hur- system sits over warm water, convection as a purely tropical system — a strong ricane Center said in their technical may develop near the center. The storm Category 2 hurricane with maximum discussion that they were dealing with can become warm-core and tropical in
Figure 2 Figure 4
Figure 3 Images: U.S. National Oceanic and Atmospheric Administration Images: U.S.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 39 AVWEATHER
nature. The convection and warm core Carolina. A buoy in the open ocean low off the Southeast coast. A power- are often confined to a small central measured a wave height of 100 ft. At one ful trough (cold core) is located in the region, surrounded by the extratropical point, convection developed near the middle of the United States. Not only is part of the storm. So you can have a trop- storm center, and the inner core took on the trough steering Sandy to the north, ical cyclone embedded within a larger the structure of a tropical cyclone. but the divergence on the trough’s east- subtropical or even extratropical system. Sandy began to resemble a subtropi- ern side is causing the pressure to fall Subtropical cyclones are not limited to cal cyclone (Figure 2, p. 39), but Sandy in the storm, 10 mb in one day. Cooler the Atlantic. The “Kona storms” that was a tropical system that was acquir- water and increased wind shear should sometimes affect the Hawaiian Islands in ing extratropical characteristics, not have weakened the storm. Sandy’s peak winter are subtropical. Other subtropical vice versa. It featured a warm, tropical winds remained the nearly the same, cyclones have occurred in the Mediter- core embedded within a much larger but the storm continued to grow in size. ranean Sea and the Indian Ocean. non-tropical cyclone. It had two wind Sandy moved parallel to the coastline Prior to Sandy, probably the most maxima, one near the center and one on Oct. 28, while its central pressure famous of the hybrids was the so-called over 100 mi (161 km) north of the continued to fall and the storm grew. The Perfect Storm of 1991. Developing in center. The surface map for 0000 coor- 1200 UTC Oct. 29 surface chart (Figure the North Atlantic, south of Nova Scotia, dinated universal time (UTC) on Oct. 5) shows Sandy well off the Virginia coast. Canada, in late October, this cyclone 28, 2012 (Figure 3, p. 39) shows Sandy The 500 mb chart for the same time had peak sustained winds of 65 kt and off the Southeast coast. The front it will (Figure 6) shows a large upper-level high a minimum central pressure of 972 mb eventually merge with is to the west. over the Canadian Maritimes, blocking (28.70 in Hg). Although it never came The 500 mb (~18,500 ft, 5,500 m) chart Sandy’s northward march. At the same ashore, severe beach erosion occurred for the same time (Figure 4, p. 39) de- time, a closed, cold core low has formed from the Canadian Maritimes to North picts Hurricane Sandy as a warm core over North Carolina. Sandy is being
Figure 5 Figure 6
Figure 7 Images: U.S. National Oceanic and Atmospheric Administration Images: U.S.
40 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 AVWEATHER
Figure 8 pulled to the west into the stronger system. Con- vection near Sandy’s center continued to develop (Figure 7). The central pressure also continued to fall, reaching 940 mb by 2100 UTC. Sandy’s hybrid nature was a double-edged sword. Had Sandy been a pure tropical hurricane, with such low pres- sure, it would have been a Category 4 hurricane with maximum sustained winds of 114 to 135 kt. Instead, winds were still holding near 70 kt, but the wind field was huge, with gale force winds now covering nearly 1,000 mi. Images: U.S. National Oceanic and Atmospheric Administration Images: U.S. Sandy likely became a true extratropical Figure 9 cyclone just before it came ashore in southern New Jersey with maximum sustained winds of 70 kt and a central pressure of 946 mb (27.94 in The precise forecasts helped prevent flight Hg). The surface map for 0000 UTC on Oct. 30 incidents associated with Sandy. It was easy (Figure 8) shows a fully transformed Sandy now to cancel flights and even close airports based associated with an array of fronts. In terms of on the accurate predictions. But these hybrid pressure, Sandy was the strongest storm ever to storms may not always be forecast that well. Er- make landfall this far north. Although winds rors in track or intensity predictions could result of 70 kt ordinarily wouldn’t produce an exces- in little warning of dangerous flying conditions. sive storm surge, because of Sandy’s huge size, it How is climate change involved in all this? brought devastatingly high tides to the New Jer- Obviously, the earth is getting warmer. As the sey and New York shorelines. The 0000 UTC 500 air and water warm, there will be more energy mb chart (Figure 9) shows that the two 500 mb available for all types of storms. Another way to lows have basically merged over eastern Mary- look at this: The purpose of storms (cyclones) is land. Sandy’s residual pool of warm air can be to transport energy on the earth, basically from seen over eastern Pennsylvania and New Jersey. the equator, where it’s hot, to the poles, where Can these hybrid or transitioning cyclones it’s cold. A warmer earth would mean more be forecast? In the case of Sandy, the answer storms and potentially stronger storms. Sandy is yes. Computer models accurately forecast could just be a harbinger of things to come. Sandy’s intensity and even its point of landfall Edward Brotak, Ph.D., retired in 2007 after 25 years days ahead. Advance warnings saved lives. How- as a professor and program director in the Department ever, the property destruction and disruptions of Atmospheric Sciences at the University of North to airline service were unavoidable. Carolina, Asheville.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 41 42 | SEMINARSIASS F Civil ofCivil (DGAC). Chile That means, ineffect, general of Dirección the de Aeronáutica General of aviation, said Jaime Alarcón Pérez, director oneveryone public the both and private sides responsibility the owner; for safety belongs to significant safetyfurther improvements. edented sharing of information hold key to the say data also flight who analysis and unprec - world, according to regional aviation specialists Flight operational safety has no single those facing airlinesthose inother of parts the more than alike different compared with and regulators, their today’s threats are far or American Latin and airlines Caribbean Pan American safety intelligenceonmanagingrisksofprojected growth. Regional aviationleadersshare expertiseandoperational rule forrule aviation maintenance centers and an (SMS) an for rule, SMS an rule airports, SMS providers including asafety management system setting upgraded standards for service air traffic its integrated for policy aviation management, in aviation, recently publishing version 2.0of ing anational program for operational safety cepted road map for states all inimplement- Air Safety inSantiago, Seminar Chile. 2012 at Flight Safety Foundation’s International than ever. He was among presenters inOctober working together with greater unity of purpose initiative new that every implies sides both Chile has followedChile internationally the ac- FLIGHT FOUNDATION SAFETY Style BY WAYNE ROSENKRANS |
WORLD AEROSAFETY | FEBRUARY 2013
© Anton Balazh - Fotolia.com SEMINARSIASS
SMS rule for other types of aviation businesses. accidents, in 2011. The five-year moving average Alarcón said phase one elements of the road for the period ending in 2011 showed “basically, map have been completed successfully, the sec- globally, that [the trend is] going down slightly ond phase is in progress, and the final phase is and very slightly in the Pan America region,” slated to be done by the end of 2015. More than Martin said. RASG-PA currently focuses on 50 SMS courses have been taught in Chile since “three main killers” that account for 73 percent 2006, and 32 percent of the 1,050 DGAC staff of all accidents worldwide — runway excur- have attended so far, with 80 percent expected to sions, loss of control–in flight and controlled be SMS-trained by 2015. flight into terrain (CFIT) — because of their David McMillan, then director general of equal regional relevance. Eurocontrol and new chairman of the FSF Board RASG-PA has promoted the use of stan- “It’s extremely of Governors, quoted Alarcón’s characterization dardized CFIT awareness training, tool kits for of the situation as an “explosion of growth” in runway excursion reduction and runway safety important not to both Chile and the region. “It’s important to be teams; conducted research on go-arounds and sure that we deliver the safety which is necessary,” mitigation of unstable approaches; added to fall into the trap McMillan said, comparing these circumstances flight training an advanced maneuvers manual of complacent with those in some other regions, such as Europe, and tool kits on pilot monitoring; conducted now facing tough cost-cutting among airlines and safety workshops; issued a runway-maintenance thinking or to air navigation service providers alike. manual in conjunction with Airports Council “The issue is how you … make sure that International; and issued the first in a series of think that excellent safety gets the resources it needs at a time when safety advisories, covering airplane automation tough action is indeed being taken to address mode awareness and energy state management safety practices those costs,” he said. “Europe has a great safety risks. can be sustained record. But as you know, it took a lot of work to “We had air navigation safety and aviation get there, and it’s extremely important not to fall security, but until this group was established, without effort.” into the trap of complacent thinking or to think we never quite had a forum for states together that excellent safety practices can be sustained with industry to [focus on operational safety] without effort.” — RASG-PA is it,” said Oscar Derby, director general of the Jamaican Civil Aviation Author- Regional Aviation Safety Group ity and government co-chair of RASG-PA. The The Regional Aviation Safety Group–Pan group especially has been strong in providing America (RASG-PA), a government-industry states with data-driven guidance on compli- partnership, was formed in 2008 in Costa Rica ance with eight critical elements defined by under the framework of the International Civil ICAO and making the world’s best information Aviation Organization (ICAO) Global Aviation resources readily available — and mostly free Safety Plan and Global Aviation Safety Road- of charge — through the group’s website
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 43 SEMINARSIASS
information.” This work has enabled carriers are about the world average, information. Data-sharing initiatives so the group to conduct training sessions actually slightly below the world aver- far are flourishing regardless of those and seminars that equip states to roll age. But where you see a big jump is fears “because at the CEO level of the out this legal framework. in the non-IOSA carriers, and this is an airlines of ALTA, they strongly feel Derby cited regional versus global area that concerns us all. … We have that the risk is much smaller than the data on air transport accident rates.1 a number of governments that have benefits that we can get by sharing that “The 10-year moving average from already taken IOSA as part of their data,” he explained, and the gaps in 1990 through 2000 for all regions … certification and requirements; Chile protection often have been overcome was 1.2 [accidents per million depar- is one of those countries, Brazil [is an- by sheer creativity. “ALTA gathers tures],” he said. “The [Latin America other]. However, we still need to make the data, puts it together and shows it; and Caribbean region’s comparable] sure that we do not create two levels of we don’t print it, we don’t give it to 10-year moving average was 3.8. … In aviation in the region.” the authorities because we’re still not the 2010 10-year moving average, Latin RASG-PA and IATA in 2012 col- protected,” he said. “They look at it, America has made huge strides in mov- laborated on data sharing and shared we work together, and then we go and ing that accident rate down [to 2.3], trend analysis. As of November, data we do our work.” with the world rate going down to 1.0 from flights reflecting more than 80 per million.” percent of available seat kilometers Panama’s Data-Sharing Emphasis Derby noted that RASG-PA has have been collected in a new data Despite an iceberg-size volume of ad- been acutely aware of the disparity in exchange program, de Gunten said. A vice floating around about SMS for air safety performance among operators of small related program has brought to- carriers, difficulty in practical imple- large commercial jets versus operators gether a trusted regional team “working mentation of the theories and processes of turboprop airplanes in some parts of and sharing information, trend infor- can leave an airline with the sense that the region, and the group’s issue analy- mation to again identify opportunities, something essential is still “hidden be- sis team soon will determine whether identify risks and mitigate them, and low the waterline,” said José Eduardo new targeted mitigations are warranted. we have already had some very sig- Rodríguez, a captain and director of The group nevertheless has a few nificant results in terms of changes of safety and quality assurance for Copa areas of concern. “One of them is infra- procedures, reductions of [traffic-alert Airlines. A year-long project at his structure,” said Alex de Gunten, execu- and collision avoidance system alerts], company recently reviewed elements tive director of the Latin American etc., and a similar program is now also of its SMS — including nonpunitive and Caribbean Air Transport Associa- working in Brazil [and] Chile.” safety reporting methods — in consul- tion (ALTA) and industry co-chair of Among other issues vying for atten- tation with the Autoridad Aeronáutica RASG-PA. “We’ve got a major concern tion, despite ICAO’s standards for pilot– Civil of Panama (AAC) and Flight as to where … are we going to land [a air traffic controller phraseology, “we Safety Foundation. much larger fleet of] airplanes in the are not where we should be in the region The current focus of the project is next 20 years, because our airports are based on a recent RASG-PA survey [of working closely with the pilot union, already saturated.” the two professions],” he said. “We … and subsequent phases will involve the None of the region’s airlines that asked them if they knew the standard remaining unions to encourage a strong participate in the International Air ICAO phraseology; about 31 percent voluntary reporting culture. “Trying to Transport Association (IATA) Opera- said ‘no.’ We asked those who knew build [this] within the company is not an tional Safety Audit (IOSA) program [it], ‘Do you apply it 100 percent of the easy step,” Rodríguez said. “It’s some- has had a fatal accident in four years, time?’ and about another 25 percent said thing that takes time. It takes a lot of de Gunten said. “We’ve got a few ‘no.’ … This is an area of concern.” training from the organization, a lot of priorities in the region, number one is Contrary to other presenters, reception and trust from the rest of the IOSA. … If we look at the accident de Gunten downplayed the anxiety coworkers.” The only precedent had rate of IOSA versus non-IOSA [car- seen among regional lawyers and been mandatory occurrence reports. riers] for Latin America over the last some safety professionals about Nonpunitive reporting also involves four years … the Latin American potential abuses of confidential safety safety action groups in operational
44 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 SEMINARSIASS
areas, with ultimate oversight by a safety review board, which keeps the CEO apprised of how risks are being managed inside the company and how accidents are being prevented. Goal- setting and a regulatory and internal structural framework for a related ini- tiative called Pan American Voluntary Safety Information Program — involv- ing a memorandum of understanding among the civil aviation authority, pilot union and Copa — recently have been in progress. First on the agenda will be analysis focused on a set of 300 events Barbosa/jetphotos.net Vicente © Alex Sandro involving a loss of required air traffic Colombia. A good fit will be possible Up-to-date FOQA technology and separation, and steps to ensure that em- partly by replacing traditional manage- methods will help LATAM to meet ployees will trust the system enough to ment “silos” with horizontal, process- publicly declared safety commitments voluntarily submit reports to specialists driven organizational structures under and aspirations, Rosende said, while who can update risk assessments. The SMS, he said. becoming the largest air carrier in agreement empowers the civil aviation the Latin America and Caribbean authority to resolve disagreements. FOQA at LATAM Airlines Group region. The new holding company The second phase of the project will The 2012 merger of Chile’s LAN Group has nine affiliate airlines operating extend this voluntary reporting to ramp and TAM Linhas Aéreas of Brazil re- 309 aircraft among 116 destinations operations staff, maintenance techni- quired an intensive four-month process in countries such as the United States, cians and flight attendants. Protec- to use flight operational quality assur- Mexico, Colombia, Ecuador, Peru and tive measures have yet to be added to ance (FOQA) to quickly identify and Argentina. Plans call for fleet expan- regulations, and for that reason, the mitigate new risks associated with the sion to approximately 500 airplanes company has partnered with the Pana- gradual changes in flight operations, around 2015. manian authority to reconfigure the said Enrique Rosende Alba, corporate Human errors will be understood regulatory structure. director of safety and security, LATAM as opportunities to improve operational Currently, the airline presents a Airlines Group. safety, but constant emphasis will be monthly report of safety data trend analysis, based on internal flight data analysis, to the AAC. As Copa is Panama’s only air carrier, sharing of data or trend analysis within the coun- try has not been possible. The company also participates in industry-level flight data sharing — for example, unstable approach data for six Central American airports through a program based in Costa Rica. In November 2012, flight data monitoring specialists from Copa Airlines were scheduled to visit their counterparts at Copa Colombia and officials of the Aeronáutica Civil of C. /jetphotos.net Valle © Carlos P.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 45 SEMINARSIASS
placed on avoiding violations of company norms look at parameters over a longer period of time and standard operating procedures. He called during each approach. operational safety the one “non-negotiable This more accurate, TAM-derived process value” among four in the new holding com- — beginning at 1,000 ft above ground level — is pany’s values statement. best suited for cross-affiliate comparison and Given its mode of “permanent growth” aligned with industry best practices, Rosende in fleets and new routes, LATAM Airlines said. Analysts already have seen improvement Group has been submitting each contemplated in the incidence of unstable approaches — a change to formal risk analysis, Rosende said. noticeably inverse proportion to the increasing This was recently completed for Airbus A320 level of FOQA program monitoring and associ- Analysts already have operations and was set to begin for Boeing 767 ated training of pilots. The process inherently operations. The risk analyses extend to issues encourages pilots to improve, he said, and LA- seen improvement such as violations of regulations and opera- TAM also is willing to share the lessons learned in the incidence of tional policies on alcohol and drug consump- with other airlines through ALTA. tion. During 2012 alone, more than 10,000 No related operational changes have had to unstable approaches employees received corporate-level training on be introduced to pilots, however. Adherence to safety aspects of the holding company–related existing SOPs and education of pilots about the — a noticeably changes under way. more precise measurement have been sufficient. inverse proportion to Over the years, FOQA has been valuable Beginning in January, LATAM expected to have from an operational efficiency viewpoint as well this version 2.0 measurement process fully in the increasing level as in risk management, he said. One example place to help reduce unstable approaches. has been monitoring the rollout of required Next on the agenda is concentration on of FOQA program navigation performance instrument approaches hard landings, deep landings, rejected takeoffs and verifying that flight crews use them as and normal go-arounds — often involving monitoring and intended. Another example has been verifying operation of aircraft to/from relatively complex associated training crew compliance with “lean fuel” practices that airports in the region, Rosende said. A pre- the company desires and airframe manufactur- liminary look at one set of 46 hard landings, of pilots. ers recommend. 34 unstable approaches and 59 deep landings Confidentiality of FOQA data is assured actually found no variable in common between partly by a team that has centralized gatekeeper one event and another, reflecting the analytical responsibility at an office at the holding compa- challenges. Other current interests are mitiga- ny, collecting data from all the affiliate airlines. tion of bird strike risks and in-flight shutdown A few conditions in which data confidentiality of engines. can be terminated are specified by policy, such “We firmly believe that this information is as a crewmember’s repeated responsibility for valuable not only to operators but valuable to the same type of event. the aviation system,” Rosende said. “Therefore, given the conditions we’re in, we’re predis- Almost Infinite Information posed to voluntarily deliver this information Systems to identify and mitigate unstable ap- with the ultimate purpose that we all will win proaches have been refined significantly in the from the viewpoint of operational safety.” context of LATAM changes, Rosende said. A Note previously effective LAN Group method was judged unsuitable for meeting new demands. 1. Data represent ICAO-defined hull loss accidents, by airline domicile, involving Western-built transport With multiple affiliates in mind, version 2.0 of airplanes with maximum takeoff weight of 60,000 lb the unstable approach program has been pilot- (27,200 kg) or greater, and using known departures tested at the holding company level. Essentially, coupled with indirect estimates of missing depar- this FOQA analysis takes a deeper and finer tures data from maintenance logs.
46 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 DATALINK
BY RICK DARBY C-FOQA Data Show Continued Improvement As program participation increases, trends become more meaningful.
he annual rates of flight operations events, operations, risk monitoring, aircraft limitations or predefined exceedances of selected and aircraft systems. parameters, continued to decrease in the During 2010 and 2011, flight operations event most recent analysis of corporate flight rates have been under 10 per 100 flights, with Toperational quality assurance (C-FOQA) busi- an improving trend since 2007 (Figure 1). The ness jet data for the years 2006–2011 by Austin length of the error bars has also been decreasing.2 Digital C‑FOQA Centerline.1 Also notable was a For risk reduction in flight operations — as reduction in the rate of unstable approach events in other event categories — it is important not in 2011, to the lowest level in the six-year period. only to know the rate of events, but what kind The trend was particularly encouraging be- of events they were. This offers a clue to the cause of growth of participation in the program, relative prominence among the event types with these flights and events making the data that might be accident precursors. In 2011, the even more statistically relevant to this industry highest rate — nearly double the next highest sector. While event rates continue to decline over- — involved GPWS glideslope alerts below 3,000 all, there is still concern that go-arounds follow- ing unstable approaches are not being conducted, C-FOQA Annual Flight Operations Event Rates, 2006–2011 and there is persistent evidence of high-energy 20 approaches, both of which are strongly associ- ated with runway excursion events as shown in the 2009 briefing of the Runway Safety Initiative 15 coordinated by Flight Safety Foundation. Exceedances represent cases when an event’s 10 parameter data are considered less than optimal for safe operation. For example, during ap-
proach, C-FOQA analysis flagged an exceedance per 100 ights Events 5 if the flight data recorder showed that the air- craft was above or below the glideslope, or was out of alignment with the localizer, in each case 0 2006 2007 2008 2009 2010 2011 by a stipulated deviation. It also tagged an event, Year for another example, if the ground-proximity C-FOQA = corporate flight operational quality assurance warning system (GPWS) produced a “sink Note: Error bars are calculated with a 90 percent confidence interval. rate” or “pull up” warning at certain altitudes Source: Austin Digital C-FOQA Centerline and rates of descent. Events were categorized under the headings of unstable approach, flight Figure 1
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 47 DATALINK
ft radio altitude (Figure 2). Analysis discovered final. This was supported by the average thresh- that most of the glideslope excursions revealed old crossing height of 36.1 ft, somewhat lower a tendency to “duck under” during very short than the target height of 50 ft. Other relatively common events included bank angle exceedance, altitude excursion, a traffic-alert C-FOQA Flight Operations Event Rates, by Type, 2011 and collision avoidance system resolution advisory Events per 100 ights lasting more than two seconds, and exceedance of 0 0.5 1.0 1.5 2.0 2.5 3.0 expected deceleration during rollout. GPWS: glideslope High bank angle for this height Annual unstable approach event rates in Altitude excursion 2011 decreased notably from those of 2010 and TCAS resolution advisory Strong deceleration during rollout were the lowest in the 2006–2011 data set (Fig- High rotation rate ure 3).3 These events, unlike flight operations Excess ground speed: taxi in Low-level wind shear events, showed no discernible trend during the GPWS: unknown warning type five previous years of the C-FOQA program. For Excessive ground speed: taxi out Large lateral acceleration on the ground all program years aggregated, the highest rate High rate of descent for this height GPWS: sink rate of unstable approaches occurred in the third Passenger comfort limits exceeded quarter — July, August and September. High roll rate GPWS: don’t sink Considering types of unstable approach GPWS: too low terrain events, the highest rate in 2011 was for flying Below desired ight path during initial climb GPWS: terrain above the glideslope, only slightly less frequent Not in takeo con guration than events involving being fast on the approach GPWS: pull up (Figure 4). Of those flights when the approach was flown above the glideslope, nearly half were C-FOQA = corporate flight operational quality assurance; GPWS = ground-proximity warning system; TCAS = traffic-alert and collision avoidance system between 0.00 and 0.25 dots high. About 0.3 per- Source: Austin Digital C-FOQA Centerline cent of the flights were between 2.00 and 2.25 dots high, past the caution limit, and the greatest Figure 2 deviation — about 0.1 percent of the flights — was 3.50 dots high or more. Of greater concern C-FOQA Unstable Approach Event Rates, 2006–2011 is that four of the top five unstable approach 6 event causes indicated high-energy approaches, strongly associated with runway excursions. Preliminary analysis of unstable approach events correlated with both time of day for 4 the event and length of the runway have been introduced. The aggregate data are statistically irrelevant at this point; however, individual op- 2 erators now have the ability to look more closely
Events per 100 ights Events into their own operations, where event rates may be more statistically relevant. Risk monitoring events concerned alerts 0 2006 2007 2008 2009 2010 2011 or cautions for threats such as fuel exhaustion, Year controlled flight into terrain (CFIT), stall, land- ing overruns, hard landing and tail strike. The C-FOQA = corporate flight operational quality assurance Note: Error bars are calculated with a 90 percent confidence interval. highest rates in 2011 were for CFIT risk and risk of a landing overrun (Figure 5). Source: Austin Digital C-FOQA Centerline C-FOQA Centerline says that it is now Figure 3 working to provide pilots with more detailed
48 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 DATALINK
information concerning the risk of landing C-FOQA Unstable Approach Event Rates, by Type, 2011 overrun events by offering analysis of landing performance for the first time. That includes Events per 100 ights monitoring threshold crossing height, airspeed 0 0.2 0.4 0.6 at threshold, float distance, tailwind at threshold Above glideslope on approach Fast approach [CAS-VAPP] and runway remaining when slowed to 80 kt. Below glideslope on approach These events will be combined to provide analysis Late gear extension Late nal ap extension of the newly drafted stabilized landing concept. Not aligned with runway (localizer deviation) Event rates in 2011 that exceeded the air- High rate of descent on nal approach Slow approach [CAS-VAPP] craft’s recommended operating limits tended Approach unsteady in roll Approach unsteady in pitch to be low; those events primarily consisted of Final ap position not valid for landing calibrated airspeed beyond the aircraft model’s flap speed limit, based on the aircraft’s reference C-FOQA = corporate flight operational quality assurance; CAS = calibrated airspeed; V = approach speed flight manual. That occurred slightly more than APP Source: Austin Digital C-FOQA Centerline 0.7 times per 100 flights. C-FOQA Centerline says its data suggest that flap overspeed events Figure 4 generally decline markedly after the first two years in the program when flight departments C-FOQA Risk Monitoring Event Rates, by Type, 2011 take measures to reduce them. All the other measured operating limits Events per 100 ights 0 0.1 0.2 0.4 exceedances occurred less than 0.2 times per 0.3 Risk of controlled ight into terrain 100 flights. Risk of runway overrun during landing Aircraft system events for 2011 were negligi- Risk of wing or pod strike during landing ble in number except for selecting or maintain- Risk of fuel exhaustion ing reverse thrust while decelerating at relatively slow speed. C-FOQA = corporate flight operational quality assurance Participation in the C-FOQA program has Source: Austin Digital C-FOQA Centerline grown steadily since it was initiated in 2006 (Fig- Figure 5 ure 6, p. 50). In 2011, more than 10,000 flights contributed data, for a total of more than 30,000 the report also suggests that fatigue often goes flights since the program’s origin. Twenty-five unreported in accidents and incidents — first, operators participated in 2011, with the data rep- because pilots are reluctant to admit flying resenting 73 aircraft of 16 types or variants. “under the influence” of fatigue out of concern it could provoke punitive action by an employer or Pilot Fatigue Barometer even criminal prosecution; second, if the pilots The European Cockpit Association, which repre- happen to be killed in an accident, fatigue leaves sents national pilot associations of 37 European no material evidence. states, has summarized the results of surveys con- The main potential consequences of fatigue ducted by some of its members in a report titled during flight duty include degradation of “Pilot Fatigue Barometer.”4 The surveys were thought processes, perception and reaction carried out between 2010 and 2012 in Austria, time; periods of unintended sleep; and mo- Denmark, France, Germany, the Netherlands, mentary “micro-sleep.” Percentages of pilots Norway, Sweden and the United Kingdom. Some who reported having experienced fatigue in the 6,000 pilots responded to queries about how cockpit ranged from 93 percent of those in the fatigue affected their flying performance. Denmark survey to 45 percent who responded Mentioning several well-known accidents in to the U.K. survey. Pilots who said they had which pilot fatigue was cited as a causal factor, dozed off or had a spell of micro-sleep ranged
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 49 DATALINK
that night flights are one of the major causes of C-FOQA Participation, 2006–2011 pilot fatigue. …
35,000 “The study among British pilots shows that Total flights in the program New flights fatigue prevalence is associated with the number 30,000 of sectors, flying and duty hours or [the com- 25,000 mander’s decision making] frequency.”
20,000 Other identified causes included a series of morning departures; insufficient rest between
Flights 15,000 duty periods; being recalled from standby status; 10,000 and inadequate rest accommodations.
5,000 Nevertheless, the report says, only 20 to 30 percent of the pilots polled reported that they 0 2006 2007 2008 2009 2010 2011 had acknowledged feeling unfit for duty. “Such Year under-reporting of fatigue has been confirmed by an independent survey of 50 U.K. aviation C-FOQA = corporate flight operational quality assurance medical examiners in April 2011,” the report Note: As of end of year 2011, 73 aircraft of the following types contributed to the data set: Boeing BBJ, Bombardier Challenger 300, Bombardier Challenger 604, Bombardier Challenger says. “The vast majority (70 percent) of the avia- 605, Bombardier Global Express, Bombardier Global Express XRS, Cessna Citation X, Dassault Falcon 2000EX, Dassault Falcon 900DX, Dassault Falcon 900EX, Dassault Falcon 7X, Embraer tion medical examiners believe that pilots are re- ERJ-135, Gulfstream G450, Gulfstream G550, Gulfstream GIV and Gulfstream GV. luctant to report fatigue within their company.” Source: Austin Digital C-FOQA Centerline About a third of the pilots who chose not to file fatigue reports gave as their reason that they were Figure 6 too tired at the end of an exhausting workday. from 54 percent of respondents from Sweden to 10 percent of those from France. The report Notes did not speculate on the reasons for the national 1. The C-FOQA User’s Group — comprising all pro- differences among responses. gram participants — is led by a steering committee “More than three out of five pilots in Sweden plus Austin Digital, Flight Safety Foundation and other external parties. (71 percent), Norway (79 percent) and Denmark (80–90 percent) acknowledge [having] made 2. Austin Digital explains the meaning of error bars: mistakes due to fatigue, while in Germany it was “When displaying event rates (e.g., events per 100 flights), it is appropriate to compute proportion con- four out of five pilots,” the report says. fidence intervals — error bars — along with the raw Responses indicated that 92 percent of Ger- event rate. These bars indicate a range, based on the man pilots reported that they had felt “too tired” number of flights sampled, within which the true rate or “unfit” for duty on the flight deck at least once likely falls with high confidence. In general, the larger in the previous three years. In the Austrian pilot the sample of data (i.e., more flights), the smaller the association, 85 percent of respondents reported error bar will be and the more confident you can be that the resulting rate is a statistically significant value.” that they had been too fatigued for flight duty but nevertheless had reported for their assignments. 3. Unstable approach criteria are aligned with the ele- Two-thirds of those said they had flown under ments published by Flight Safety Foundation, which include nine requirements, all of which must be that condition more than once. Swedish and satisfied. The criteria also specify that flights must Danish pilots reported similar percentages. be stabilized by 1,000 ft above airport elevation in “According to the surveys among pilots, instrument meteorological conditions and by 500 night flights or a series of night flights are major ft above airport elevation in visual meteorological contributors to fatigue,” the report says. “For conditions. FSF ALAR Tool Kit, Briefing Note 7.1, example, in France, almost 70 percent of the
50 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 © Chris Sorensen Photography contact us ightsafety.org Fax: +1703.739.6708 Tel.: +1703.739.6700 Alexandria, Virginia U.S. 22314-1774 801 N.Fairfax Street, Suite 400 Headquarters: FoundationFlight Safety CONVERSATION LEADING THE — astrong, e culture. globalsafety ective Membership intheFlight Foundation Safety isyour visiblecommitment to theaviation community’s core value each member. work Our isexempli ed inthefollowing areas: The Foundation that achieves challengingprojects make itsgoalsby undertaking aviation safer, thereby bene tting Individual membersrange from pilotsto accident investigators to regulators andbeyond. manufacturers, educational institutions, airports, non-pro t andgovernment companies. service organizations andsupport The Foundation membershipcomprises organizations from around theworld —aircarriers, businessaviation operators, umt n eiashl rudtewrd BARS– The Aviation Basic Standard Risk Globaltraining initiatives andseminarsheldaroundSummits theworld Humanitarianefforts Approach (ALAR) andLandingAccident Reduction for safetydata confidentiality Support Media outreach membership@ ightsafety.org Tel.: ext. 102 coordinatormembership services Ahlam Wahdan Member Enrollment for more than65years. JOIN THEDIALOGUE AeroSafety World lausch@ ightsafety.org Tel.: ext. 112 and businessdevelopment managing director ofmembership M.Lausch Susan Donations/Endowments/Membership … join theFlight Foundation. Safety Email: bars@ ightsafety.org Fax: +611300557182 Tel.: +611300557162 Melbourne VIC 3001Australia GPO Box 3026 Melbourne VIC 3000Australia Level 6|278Collins Street ProgramBAR Standard O ce INFOSCAN
A systematic review of theInsert effectiveness document titleof SMS Under safety management systems the Microscope DrLocation Matthew | Date J W Thomas | Westwood-Thomas Associates Has the safety management system concept been ‘oversold’?
BY RICK DARBY ATSB Transport Safety Report InvestigationResearch [InsertCross-modal Mode] Research Occurrence Investigation Investigation XX-YYYY-####XR-2011- 002 Final
REPORTS normal operations. Accident causal factors can Opinion Versus Evidence be anticipated and, as far as possible, mitigated A Systematic Review of the Effectiveness of Safety before they do their worst. Management Systems It is an exciting prospect, with a touch of Thomas, Matthew J.W., and Westwood-Thomas Associates. magic. We can take charge of the future rather Australian Transport Safety Bureau (ATSB). 46 pp. Figures, tables, references, appendix. November 2012, updated Dec. 10. Available than just waiting to see what it throws at us. at
52 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 INFOSCAN
evidence to support these claims, nor legal — one way of looking at aircraft accidents evidence with respect to due diligence. Indeed, — “with the remaining 23 studies relating much of the regulatory effort with respect to to work health and safety,” the report says. the adoption of SMS as the primary regulatory “In addition, very few of these studies were platform has been characterised on uncritical undertaken in transport domains, and many acceptance, and based on expert opinion and studies only measured subjective perceptions face validity, rather than subjected to formal of safety rather than objective measures. The scientific validation. limited [amount of] quality empirical evidence “Previous published reviews of SMS available relates to the difficulty of measuring research do not appear to provide strong objective safety improvements in industries empirical evidence to support the specific where the SMS is aimed at avoiding LP-HC ac- benefits of adopting an SMS. For instance, cidents and the relative recency of the applica- the summary of a 2006 review of evidence tion of SMS.” for the effectiveness of SMS across a wide Even among the 37 papers accepted for cross-section of industries suggests that there analysis, the study’s authors were less than ful- has been a ‘less than expected’ reduction in ly satisfied with the quality of evidence. Only accident occurrence since the implementa- a single study met the scientific “gold stan- tion of SMS.” (References can be found in the dard,” a randomized, controlled trial. Of the 37 original report.) articles included in the systematic review, 19 ATSB commissioned Matthew Thomas used objective measures of safety performance. and Westwood-Thomas Associates to under- And 15 of the 19 related to workplace health take a meta-analysis of SMS research. The and safety, using such metrics as occupational authors began with a comprehensive search injuries to workers. “Of these studies, the ma- of the literature and found 2,009 articles, a jority demonstrated significant positive effects promising start. However, the great majority with respect to dimensions of SMS,” the report of the sources washed out because of rigorous says. “A number of studies found general re- inclusion criteria. Among other requirements lationships between SMS implementation and were that only peer-reviewed articles pub- safety performance.” lished between 1980 and 2012 were accepted; Eighteen of the 37 articles analyzed in the studies “must have clearly defined a research systematic review used only subjective, self- ‘Previous published question that related to the effectiveness reported measures of safety performance, most of safety management systems, or specific with a survey-based methodology in which both reviews of SMS components of a safety management system”; individual perceptions of effectiveness of SMS research do not studies must have defined effectiveness in components and safety metrics were subjective. terms of safety-related outcomes, rather than The report notes, however, that across mul- appear to provide other standards such as improved productiv- tiple studies, there was scant agreement about ity; and studies must have reported quan- which components of an SMS individually strong empirical titative measures. There was also a quality caused change in safety performance. evidence to support appraisal based on published guidelines for The four studies of L-P/H-C industries, methodological soundness. probably the most relevant to aviation, demon- the specific benefits Ultimately, 37 papers were determined to strated “no consistent findings … with respect be directly relevant to the objectives of the to performance on various dimensions of an of adopting an SMS.’ investigation. However, “only 14 [studies] SMS and poor safety outcomes. … involved an SMS designed to avoid low-prob- “Several studies explored the relation- ability/high-consequence (LP-HC) accidents” ships between components of SMS and
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 53 INFOSCAN
safety performance in the context of major between a number of different factors, and hazard facilities. The first of these studies their relationship to a particular outcome — in from an oil refinery environment established this research context. Using such a methodol- a relationship between self-reported safety ogy, it says, “to tease out the inter-relationships performance and the two components of between components of safety management (1) management commitment and (2) safety systems, safety climate factors and safety perfor- communication. A second study, undertaken mance might not assist in clarifying the complex by the same authors, found no direct effect set of factors influencing safety performance, of management commitment, but rather (1) and does not really assist in enhancing our supervision, (2) safety reporting and (3) team understanding with respect to establishing the collaboration as the immediate drivers of safe effectiveness of SMS.” work practices. A particular problem with surveys and self- While an association “Slightly different findings were obtained reporting is that they “fail to utilise a standard in another study, whereby (1) management set of instruments, thus leaving the industry may be found between commitment and (2) safety rules and pro- unsure of exactly what is being measured. Fur- cedures were found to be directly associated thermore, there is a tendency to infer causality an SMS ‘model’ and with safe work practices in major hazard from the findings of these models, inasmuch better safety-related facilities in India.” as increased management commitment leads One study seemed to offer some evidence to reduced rates of safety occurrence. No such behavior, it is not of what factors were effective in improving directional causality can be inferred through safety performance. “This study, within the these study designs, and … each of these stud- clear whether a causal maritime domain, found that safety behaviour ies is limited from the perspective of common relationship exists. was influenced by safety policy and perceived method variance.” supervisor behaviour rather than other com- In other words, while an association may ponents of safety management systems,” the be found between an SMS “model” and better report says. The authors of that study conclud- safety-related behavior, it is not clear whether ed that “shipping companies should therefore a causal relationship exists. And if one does, it invest large amounts of money in developing has been impossible to determine if the causal and implementing safety rules, procedures factor is one element of SMS, more than one ele- and training.” ment or the SMS in its entirety. Another way of The report says, “In perhaps one of the looking at the data, as suggested by the maritime most important studies [published in 2008] in study, is that management commitment rather terms of relevance to high-risk transport indus- than SMS is the active ingredient. tries (using a cross-section of industries), there Textbooks about worker behavior invari- was no real relationship established between ably discuss the “Hawthorne effect,” derived everyday safety performance and L-P/H-C from a series of studies conducted at the events. This finding from the U.S. highlights Western Electric Hawthorne Works in Ci- the lack of clarity in what might actually be cero, Illinois, U.S., beginning in the 1920s. driving ultra-safe performance, and in many Experimenters tested the effect of increasing respects, the question as to SMS effectiveness is or decreasing the lighting in the employees’ unable to be adequately answered by even the work environment, as well as other variables, most recent research.” on productivity. The researchers found that The report questions the validity of surveys productivity improved with any change, even and structural equation modeling — a statisti- if it was only reversion to a previous condition. cal technique used to explore the relationship Their eventual conclusion was that the output
54 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 INFOSCAN
improved either because the employees were If so, an SMS is a codification of principles aware that they were being studied or because learned through experience, an evolution rather managers seemed to care about the quality of than a revolution. their working environment. The report suggests another concern, which Although, as with most scientific studies, the is that “it just might be the case that the ever- conclusions about the Hawthorne effect have growing list of components of a safety manage- since been questioned, it is generally accepted ment system may well result in dilution of effort that observation affects behavior and doesn’t across the spectrum of safety management merely measure it. So perhaps part of the reason activities. This dilution of effort may well result for safety improvement attributed to SMS — if in poorer safety performance as the critical there is objective improvement — is that the components receive less time and effort at the SMS is on everyone’s mind, more than any theo- expense of yet another ‘good idea’ dressed up retical content of the system. as a legitimate safety program. Given that, at The report concludes with some thoughts present, there is no clear objective empirical about the “frameworks,” “models” and “strate- evidence as to whether there are any critical ele- gies” that have been upgraded in status to SMS. ments, this is a real possibility.” It says: Scientists, however, have a saying that “ab- “There is a well-known axiom that states, sence of evidence is not evidence of absence.” ‘there was never a randomised control trial for Given the practical methodological difficul- the effectiveness of the parachute.’ This is to say ties of studying SMS, it is not surprising that there has never been a study in which one that demonstrating its effectiveness remains group jumps from an aeroplane with a para- beyond current findings. It would be a rare chute, and their survival is compared with a flight operations department that would agree group that jumps in exactly the same conditions, to deliver the presumed benefits of an SMS to but without a parachute. half its operations while denying them to the “The argument here is simple: Some inter- other, “control” half — particularly because ventions just do not require large-scale experi- the system represents a continuing process, ments to establish their effectiveness. Many not a quick fix. interventions are based on first principles, that The report concludes, “Even within a vac- are things that we already know to be true, and uum of evidence, the precautionary principle logic. Safety management systems contain many states that we must not fail to take precaution- of these elements. For instance, logic simply dic- ary action. To this end, it is likely that the tates that if you are to prevent the reoccurrence current regime of an aggregate set of compo- of an event, you need to understand what caused nents assembled into something, which we the event, and put in place strategies such that call a ‘safety management system,’ remains an those causes are prevented from occurring important tool in the management of safety.” again. Hence, the need for accident investigation is a simple logical necessity that requires no em- From Strategy to Action pirical evidence to support its use within safety 2012 European Strategy for Human Factors in Aviation management processes. European Human Factors Advisory Group (EHFAG). First issue, Sept. 1, 2012. 8 pp. “This review of the scientific literature suggests that this logical necessity, which European Human Factors Strategy many might call ‘common sense,’ has driven has been developed by the EHFAG in much of the development of safety manage- A conjunction with the European Aviation ment systems.” Safety Agency (EASA). “The strategy sets
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 55 INFOSCAN
out to achieve two principal functions,” this [Material] will help regulators oversee report says. “First, to foster consistency in the the effective implementation of human integration of human factors principles in the factors by industry and in incident and regulation, governance, system design, train- accident investigations.” ing, licensing, audit and assurance of aviation activities. Second, it outlines how the practi- • “Promoting the importance of human cal understanding and application of human factors — At a European level through factors can serve in enhancing safety perfor- the EASA website, regular newsletters and mance across the aviation safety system. The bulletins, and EASA conferences; at a na- strategy serves as a framework document to tional level through the cascading of EASA support the European Aviation Safety promotion and national conferences.” Plan (EASP).” • “Coordinating activities — across organ- The strategy encompasses Europe’s aviation isations, including regulatory organisa- system as a whole — “rule makers, authorities, tions, to avoid the transfer of risk from investigators, researchers, service providers, one domain to another. This coordina- industry and other stakeholders.” tion should be across both European Lessons should be The EHFAG has significant input to EASA and non-European aviation systems (e.g., rules and advisory materials, and in turn, the with FAA [U.S. Federal Aviation Ad- learned and shared EASA rules affect operators around the world. ministration]). This should also include For example, there are about 6,000 repair sta- from many sources, coordination with other safety organisa- tions with U.S. Federal Aviation Regulations tions and initiatives such as the Euro- including accident Part 145 certificates. About 1,300 of those, pean Strategic Safety Initiative; Advisory mainly the larger U.S. repair stations, are EASA- Council for Aviation, Research, Innova- investigations, certificated. Many technicians working in the tion in Europe; Eurocontrol; and the United States are following EASA rules. data analysis implementation of safety management The EHFAG describes its guiding principles: systems. EASA and the EHFAG should and operational seek opportunities to influence and coor- • “Providing appropriate governance dinate human factors with international and leadership — This would include experience. bodies such as ICAO [International Civil consideration of appropriate HF [human Aviation Organization].” factors] expertise within the regulatory organisations.” Lessons should be learned and shared from • “Developing a balanced regulatory struc- many sources, including accident investigations, ture — Human factors principles will data analysis and operational experience, the be addressed in all the aviation regula- report says. tions, whilst recognising the need for the The EHFAG’s next step will be to develop regulation to be proportionate with an an action plan from the strategy, converting it appropriate balance between rule, accept- into a detailed human factors program by the able means of compliance (AMC) and end of June 2013. “Priority of tasks and actions guidance material.” will be based on the impact to the overall im- provement of safety performance,” the report • “Providing guidance and interpretive says. An appendix lists specific components of material — Adequate tools, guidance and the action plan under several headings, such as AMC material to help industry apply hu- “training and competency” and “regulation and man factors principles will be provided. rulemaking.”
56 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 ONRECORD Fatal Flight Test G650 pilots received no warning of an incipient stall during a takeoff-performance evaluation.
BY MARK LACAGNINA
The following information provides an aware- manufacturer’s target for V2, basically the mini- ness of problems that might be avoided in the mum speed that a transport category airplane future. The information is based on final reports must attain at a height of 35 ft to meet the required
by official investigative authorities on aircraft OEI climb gradient to 400 ft. The V2 speeds accidents and incidents. achieved during the tests had consistently been too high to meet the manufacturer’s goal of providing a balanced field length of 6,000 ft (1,829 m). JETS Various rotation techniques and angle- ‘Aggressive Schedule’ Cited of-attack (AOA) targets had been tried by Gulfstream G650. Destroyed. Four fatalities. Gulfstream Aerospace’s flight-test crews, but none had succeeded in meeting the target V2 he manufacturer’s “persistent and increas- without exceeding the 20-degree pitch angle
ingly aggressive attempts to achieve V2 that had been set to assure passenger comfort. T[takeoff safety] speeds that were erroneously The most recently developed technique in- low” and its “inadequate investigation” of previ- cluded an abrupt and rapid rotation, using the ous uncommanded rolls during takeoff perfor- maximum allowed 75 lb (34 kg) of pull force on mance tests were among the probable causes of the control column, to an initial 9-degree pitch the accident that killed all four crewmembers attitude, then a further increase in pitch attitude
during certification flight testing of the Gulf- to achieve V2. This technique had produced the stream G650, according to the U.S. National best results, exceeding the target by only 3 kt. Transportation Safety Board (NTSB). However, in the course of about a dozen The accident occurred at Roswell (New Mex- test flights earlier the morning of April 2, the ico, U.S.) International Air Center the morning pilot-in-command (PIC) had decided that of April 2, 2011. The flight crew was conducting a smooth rotation, pausing only briefly at 9 a simulated one-engine-inoperative (OEI) takeoff degrees before increasing pitch to about 16 and was not able to correct an uncommanded degrees, might be a better and more repeatable roll that occurred when the right wing stalled technique. “I’m not doing that jerk stuff,” he on liftoff. The wing tip struck the runway, and told a flight test engineer. “It doesn’t work … the experimental ultra-long-range, fly-by-wire and I don’t think the FAA [U.S. Federal Avia- business airplane veered right, struck a concrete tion Administration] is going to like it, either. platform housing electrical equipment and was It’s such a great-flying airplane, you shouldn’t consumed by a fuel-fed fire. The pilots and both have to abuse it to get [it] flying.” flight test engineers succumbed to smoke inhala- The NTSB report noted that both flight tion and thermal injuries. crewmembers had extensive experience as test Flight testing that day, as well as during several pilots. The PIC had 11,237 flight hours, includ- previous tests, had focused on achieving the ing 263 hours in G650 certification testing. The
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 57 ONRECORD
second-in-command (SIC) had 3,940 flight induced by over-rotation. “If Gulfstream had hours, including 140 hours in type. performed an in-depth aerodynamic analysis of During some of the earlier test flights that these events shortly after they occurred, the com- morning, the use of a smooth rotation with a pany could have recognized before the accident
brief pause at 9 degrees produced V2 speeds with- that the actual in-ground-effect stall AOA was in 4 kt of the target (135 kt). The pilots agreed to lower than predicted,” the report said. try an even briefer pause at 9 degrees during the Investigators found that the stall precipitat- next takeoff, which was conducted on Roswell’s ing the accident had occurred at an AOA of Runway 21 with flaps extended 10 degrees. As 11.2 degrees, or about 2 degrees lower than the planned, the SIC moved the right thrust lever to predicted stall AOA in ground effect. idle at 105 kt, simulating an engine failure. At 127 The report said that contributing to the Because of a kt, the PIC began rotating the airplane for takeoff. accident was the manufacturer’s “aggressive” Recorded flight data indicated that there was no flight-test-program schedule, which was designed miscalculation of pause when the pitch attitude reached 9 degrees. to achieve certification of the G650 by the third AOA quickly exceeded 11 degrees, the outboard quarter of 2011. “The schedule pressure … led to the G650’s stall AOA section of the right wing stalled, and the airplane a strong focus on keeping the program moving in ground effect, rolled right. The PIC attempted to level the wings, and a reluctance to challenge key assumptions.” but the bank angle increased beyond 16 degrees. The report noted that after the accident, the threshold for The pilots had received no warning of the Gulfstream suspended performance flight test- asymmetric stall. Because of a miscalculation ing and implemented several corrective actions. activation of the of the G650’s stall AOA in ground effect, the The target V2 was increased by 15 kt, while stick shaker and the threshold for activation of the stick shaker and maximum takeoff thrust was increased by 5 the primary flight display pitch-limit indicators percent to meet the takeoff performance goals. primary flight display had been set too high. “Ground effect refers to Certification of the new airplane eventually was changes in the airflow over the airplane resulting achieved in September 2012. pitch-limit indicators from the proximity of the airplane to the ground,” The accident investigation generated 10 the report explained. “Ground effect results in safety recommendations, including the NTSB’s had been set too high. increased lift and reduced drag at a given [AOA], call for the FAA to work with the independent as well as a reduction in the stall AOA.” Flight Test Safety Committee to develop detailed The stick shaker activated and the indicated guidance for aircraft manufacturers on flight test pitch attitude reached the limit shown on the operations (ASW, 11/12, p. 15). primary flight displays only after the stall oc- curred. The PIC pushed the control column Illness Prompts Diversion forward, applied full left control wheel and Boeing 777-200. No damage. No injuries. rudder, and called for “power.” The SIC already ore than an hour after the airplane de- had moved the right thrust lever full forward. parted from Paris for a flight to New York Despite these inputs, the airplane remained in Mthe morning of Jan. 17, 2011, the captain a stalled condition. The sound of an automatic became ill. A physician among the passengers warning when the right bank angle exceeded 30 diagnosed gastroenteritis and applied basic an- degrees was captured by the cockpit voice re- tispasmodic treatment, after which the captain corder shortly before the recording ceased about decided to continue the flight, said the report by 24 seconds after the takeoff was initiated. the French Bureau d’Enquêtes et d’Analyses. The report noted that uncommanded rolls re- About 90 minutes later, the captain felt sulting from right outboard wing stalls had been abdominal pain. “The doctor observed that the encountered during two previous test flights. In captain was very pale, with stiffness, shaking and both cases, the pilots recovered by reducing AOA. severe pains in the abdominal region,” the report The events subsequently were attributed to stalls said. The copilot declared an emergency and
58 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 ONRECORD
diverted the flight to Keflavik, Iceland. “During encountered a downburst, an intense localized the descent, an improvement in the captain’s con- downdraft. The data showed that just before dition allowed him to assume the duties of PNF touchdown, the wind had changed from a 9-kt [pilot not flying],” the report said. The airplane headwind to a 6-kt tailwind. “At touchdown, the was landed without further incident. tailwind was recorded at 18 kt, and the rate of The captain was taken to a hospital, where descent was 783 fpm,” the report said. he was observed and released the same day. “The investigation could not determine the ex- Fuel Leak Causes Fire act nature of the captain’s pains,” the report said. Boeing 767-300. No damage. No injuries. he 767 was climbing through 9,000 ft after Microburst on Short Final departing from New York’s John F. Kennedy Airbus A340. Substantial damage. No injuries. TInternational Airport (JFK) for a flight to isual meteorological conditions (VMC), with Haiti with 201 passengers and 12 crewmembers winds from 360 degrees at 5 kt, had been the morning of Feb. 8, 2012, when the flight Vreported at Darwin (Northern Territory, crew heard a bang that was immediately fol- Australia) Airport the night of Feb. 28, 2012, but, lowed by warnings of a fire in the right engine. while completing the instrument landing system The crew shut down the engine and dis- (ILS) approach to Runway 29, the flight crew saw charged a fire bottle into the right nacelle. The heavy rain close to the threshold. fire warning persisted until the crew discharged The crew asked the airport traffic control- the second fire bottle, the NTSB report said. ler for an update on the weather conditions. They declared an emergency and turned back to The controller replied that there was a storm JFK. The first officer flew the airplane while the extending to the east from the runway threshold captain and standby first officer completed the but that the reported wind was still from 360 associated checklists and communicated with During maintenance degrees at 5 kt. The Australian Transport Safety the flight attendants. The crew then landed the Bureau (ATSB) report noted that the wind data airplane without further incident. the night before the were derived from sensors located about 2.3 km Subsequent examination of the 767 revealed incident, a bracket (1.2 nm) from the threshold. no damage from the fire or the overweight land- “The crew briefed the possibility of a missed ing. Investigators found that during mainte- and spray shield for approach if the conditions deteriorated,” the nance the night before the incident, a bracket report said. and spray shield for the integrated drive gen- the integrated drive Rainfall increased as the aircraft neared the erator’s fuel-oil heat exchanger had been reas- generator’s fuel-oil runway. The crew set maximum continuous sembled incorrectly. “A seal under the fuel tube thrust to arrest an increased sink rate encoun- flange that is held in place by the bracket had the heat exchanger had tered at 55 ft above ground level (AGL) but then O-ring partially missing, which was the source reduced thrust to idle shortly thereafter. “As the of the fuel leak,” the report said. “Contribut- been reassembled aircraft entered the flare, the rain intensified, ing to the cause of the fire was the 767 Aircraft significantly reducing visibility,” the report said. Maintenance Manual’s lack of any graphical or incorrectly. The A340 touched down hard, with a re- pictorial displays of the correct assembly of the corded vertical acceleration of 2.71 g. None of two-piece bracket and spray shield.” the 116 passengers and eight crewmembers was hurt, but subsequent engineering inspections Head-On Over the Atlantic disclosed a broken engine mount and damage Airbus A319, Boeing 737-800. No damage. No injuries. requiring replacement of several main landing controller’s loss of awareness of the airplanes’ gear components. flight paths resulted in the issuance of a climb The report said that analyses of recorded Aclearance that placed the A319 and the 737 flight data indicated that the aircraft might have on a head-on collision course off the eastern coast
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 59 ONRECORD
of the United States the evening of Nov. 11, 2010, by starting one engine only he would encounter according to an NTSB incident report. control problems taxiing in the slippery condi- The flight crews of both airplanes received, tions,” said the NTSB report. and followed, traffic-alert and collision avoid- Snow was falling, and the ramp area was ance system (TCAS) resolution advisories that covered with snow and ice. “During pushback, resulted in the A319 and the 737 passing about the tug was unable to gain enough traction to 1,800 ft vertically and 2.8 nm (5.2 km) later- move the airplane and was subsequently re- ally of each other about 66 nm (122 km) east of placed with a larger tug,” which initially was able Hobe Sound, Florida. to move the CRJ, the report said. However, the The A319 had been southeast-bound at airplane, with both engines at idle power, began Flight Level (FL) 360 (approximately 36,000 ft), to overpower the tug. en route from Washington to Bogotá, Colombia. “The captain … experienced a sensation of The 737 was northwest-bound at FL 370, en unusual movement [and] asked the tug driver if route from Oranjestad, Aruba, to Atlanta. the driver still had control of the airplane,” the The report said that the 737 was still “well report said. “The tug driver confirmed that he within” Miami Air Route Traffic Control Center had control; however, the airplane subsequently (ARTCC) Sector 21 when the sector controller moved forward while still attached to the tug, handed off the flight to the controller of an ad- which rotated to the right, striking the airplane’s jacent sector, Sector 2, which the 737 eventually fuselage.” The collision damaged the CRJ’s lower would enter — and which the A319 was transit- fuselage skin and several stringers. ing. About a minute later, the Sector 2 controller The report said that a factor contributing to handed off the A319 to the Sector 21 controller. the accident was that “no guidance existed for The Sector 21 controller had the 737’s data either the flight or ground crew regarding push- tag from his radar display and “did not maintain back procedures in low-traction ramp condi- awareness” of the airplane’s position after hand- tions with an inoperative APU.” ing off the 737 to the Sector 2 controller, the report said. Unaware of the conflict he was cre- TURBOPROPS ating, the Sector 21 controller cleared the A319 to climb to FL 370, “which placed the flight in Altitude Deviation Unnoticed direct conflict with the 737,” the report said. Bombardier DHC-8-100. No damage. No injuries. Shortly thereafter, the ARTCC’s radar data nadequate monitoring of flight instruments re- processing system generated a conflict alert, and sulted in the continuation of a gradual descent the sector controllers radioed traffic advisories and Ithat placed the Dash 8 on a head-on collision instructions to resolve the conflict. However, the course with another aircraft off the east coast of flight crews of both airplanes replied that they were the Hudson Bay the afternoon of Feb. 7, 2011, following TCAS resolution advisories. The 737 said a report by the Transportation Safety Board crew also reported that they had the A319 in sight. of Canada (TSB). The TSB also faulted the ab- sence of simulator training on TCAS maneuvers Tug Slides on Slippery Ramp for the initial incorrect reactions by the pilots of Bombardier CRJ200. Substantial damage. No injuries. both aircraft to TCAS resolution advisories. he airplane had been dispatched with its The aircraft, both Dash 8s operated by the auxiliary power unit (APU) inoperative per same airline, were flying in opposite directions Tthe minimum equipment list and was two between Puvirnituq and La Grande-Rivière, hours behind schedule for departure from Salt Quebec, in VMC but in airspace not covered by Lake City the night of Nov. 23, 2010. “The cap- air traffic control (ATC) radar. The DHC-8-100, tain stated that he started both engines [using with 28 passengers and three crewmembers, was an external power cart] due to a concern that northbound to Puvirnituq at FL 230; the other
60 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 ONRECORD
aircraft, a DHC-8-300, was southbound at FL two pilots and three company engineers for a 220 with three crewmembers. functional check flight the afternoon of Feb. 12, The -100 crew was using the autopilot’s 2011. The wreckage was found on Bintan Island, vertical-speed mode to maintain altitude be- and subsequent examination revealed that the cause of pitch oscillations that often occurred propeller on the left engine was not rotating when the altitude-hold mode was selected. “The on impact, said the report by the Indonesian use of [vertical-speed] mode is neither intended National Transportation Safety Committee. for nor evaluated for this function, but noth- Investigators found signs that the left engine ing prohibited the flight crew from using it to had erroneously been placed in reverse before Examination of the maintain altitude,” the report said. the pilots lost control of the aircraft. They found The captain was alone in the cockpit, after the no check flight plan or authorization for the PIC windshield revealed first officer left for “physiological reasons,” and to conduct the check flight following replace- a peel-chip fracture did not notice when the -100 began to descend ment of the left engine. The PIC, 61, had 13,027 at about 50 fpm, the report said. “Since the rate flight hours, including 3,311 hours in type. The of the inner ply. of descent was very slow, the speed and attitude first officer, 50, had 2,577 flight hours, including of the aircraft were very similar to those of cruise 152 hours in type. flight. It was therefore impossible to note the de- A postmortem examination of the PIC scent without reference to the flight instruments.” revealed lesions that had caused paralysis of During the next 14 minutes, the -100 descend- the vestibular organs in his right inner ear. The ed about 700 ft. The captain apparently did not report said this condition meant that the pilot notice the altitude-warning light. The first officer “could not [respond] normally to three-dimen- was re-entering the cockpit when the TCAS gener- sional motion or movement” and may have ated a traffic advisory, then a resolution advisory induced spatial disorientation. to climb. The captain disengaged the autopilot and began a 38-degree-banked right turn. “During Windshield Shatters this turn, the aircraft lost just over 50 ft in altitude Beech C90 King Air. Minor damage. No injuries. before beginning to climb,” the report said, noting he airplane had departed from Champaign, that the right turn might have been an automatic Illinois, U.S., and was climbing through reaction to opposite-direction traffic. T23,400 ft the afternoon of Feb. 9, 2011, Meanwhile, the pilot flying the -300, the first when the flight instructor and commercial pilot officer, had begun a shallow left climbing turn receiving instruction heard a loud bang as the after misinterpreting the TCAS “descend” reso- inner ply of the copilot’s windshield shattered. lution advisory displayed on the vertical speed “The [instructor] noted that the flight was indicator. “It is possible that the appearance of in clouds and that the temperature aloft … was the [-100 target symbol] in the upper right-hand minus 23 degrees C [minus 9 degrees F],” the corner of the display may have caused the [pilot] NTSB report said. “Additionally, he stated that to turn left,” the report said. there was no visible structural icing present and During these maneuvers, the aircraft passed the electric window heat was on.” within 1,500 ft vertically and 0.8 nm (1.5 km) The instructor declared an emergency and laterally. Both aircraft then continued to their requested and received vectors from ATC to destinations without further incident. the nearest suitable airport, Evansville, Indiana, where the airplane was landed without further Lesions Induce Disorientation incident. CASA 212-100. Destroyed. Five fatalities. Examination of the windshield revealed a TC radar and radio contact with the peel-chip fracture of the inner ply. This type aircraft were lost about 26 minutes after of fracture “has historically been an issue on Ait departed from Batam, Indonesia, with the King Air” and prompted the windshield
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 61 ONRECORD
manufacturer in 2001 to incorporate a layer runway might have caused the pilot to experi- of urethane “that relieves stresses on the inner ence “task overload” that contributed to his glass ply and prevents peel chip-type fractures,” improper fuel system management. the report said. “This airplane did not have the improved windshield.” Descent Beneath Low Clouds Cessna 340. Destroyed. Two fatalities. PISTON AIRPLANES he pilot was en route under visual flight rules from Henderson, Nevada, U.S., to Task Overload Suspected TCompton, California, the afternoon of Jan. Beech B55 Baron. Substantial damage. Two serious injuries. 18, 2010. He held a private pilot certificate with he landing gear warning horn sounded a multiengine rating but did not have an instru- shortly after the airplane departed from San ment rating. His total flight time was 474 hours. TLuis Obispo, California, U.S., the morning The airplane was cruising at 10,500 ft when of Feb. 7, 2011. “The pilot diagnosed the prob- it encountered instrument meteorological lem and determined that the landing gear had conditions, and the pilot initiated a descent. Re- retracted successfully and that the indication corded ATC radar data showed an “erratic and system was in error,” the NTSB report said. “He circling flight path” before radar contact was lost continued the flight with the horn intermittently at 4,800 ft. “It is likely that the pilot was having sounding.” difficulty determining his location and desired Nearing the destination — San Bernardino, flight track,” the NTSB report said. California — the pilot received an unexpected A glider pilot, who was driving on a high- clearance from ATC to navigate directly to the way in the area, saw an airplane matching the airport. “As a result, he rushed through the description of the 340 flying at 200 ft AGL just descent checklist items,” the report said. below the clouds and in “bad” visibility. About The pilot then requested and received eight minutes after the witness lost sight of the clearance from ATC to make a low pass over airplane, the 340 struck the slope of a gully at the runway so that the airport traffic control- about 2,490 ft near Lytle Creek, California. lers could perform a visual check of the landing gear. During the low pass, the pilot began to Engine Fails, Control Lost have difficulty controlling the Baron and did not Piper Twin Comanche. Substantial damage. One serious injury. realize that the right engine had lost power due he airplane recently had undergone main- to fuel starvation, as confirmed by the airplane’s tenance that included overhaul of both en- on-board engine-monitoring system. Tgines. The pilot had called the maintenance The pilot subsequently lost control of facility after landing in Big Bear City, Califor- the airplane, which crashed in a nose-down, nia, U.S., on Jan. 29, 2011, to report that the inverted attitude in a recreational vehicle stor- right engine was running rough. “A mechanic age facility. was not available to help him, and he was told Examination of the Baron revealed no that he should not fly the airplane,” the NTSB airframe or engine anomalies that would have report said. precluded normal operation. Investigators found The pilot apparently had no further contact that the pilot had not switched from the auxil- with the maintenance facility before attempting iary fuel tanks to the main tanks during descent to depart from Big Bear City for a flight to Pa- or approach, and that the right engine had lost coima, California, the next morning. A witness power after the fuel in the right auxiliary tank heard the engines popping and backfiring before was exhausted. the airplane began the takeoff roll. The report said that the unexpected ap- The pilot told investigators that he was proach clearance and the low pass over the turning onto a left crosswind when the right
62 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 ONRECORD
engine lost power. “He continued the left turn to able to escape, but the instructor and the pas- downwind and made sure to keep the airspeed senger succumbed to thermal injuries. above the single-engine control speed of 90 Investigators found that the bolt securing mph,” the report said. “The pilot’s last recollec- the lower flight control push-pull tube to the tion was turning to final approach and seeing left-front hydraulic servo had detached while the runway.” the helicopter was on the ground following the The report said that the pilot allowed simulated hydraulic failure. “The ‘feel’ of the airspeed to decrease below the single-engine flight control fault [would have] mimicked a control speed; information recovered from the hydraulic system failure,” the report said. Twin Comanche’s global positioning system Although the precise cause of the bolt indicated that the groundspeed was 76 mph detachment was not determined, investigators about 1,400 ft (427 m) from the runway thresh- found that “a number of self-locking nuts from old. “The airplane subsequently impacted the other aircraft, of the same specification as that roof of a private residence located about 900 ft used to secure safety-critical fasteners in [the [274 m] from the runway threshold,” the report accident helicopter] were identified to have said. “The airplane came to rest inverted in the cracked due to hydrogen embrittlement.” front yard.” The report noted that the aluminum fuel Examination of the wreckage revealed noth- tanks in the helicopter had not been replaced ing to explain the loss of power, but the report with bladder tanks, as recommended by a ser- said that a contributing factor in the accident vice bulletin issued by the manufacturer in De- was “the pilot’s decision to fly with a known cember 2010 to “improve the R44 fuel system’s deficiency in one engine.” resistance to a post-accident fuel leak.”
HELICOPTERS Control Lost After Tail Strike McDonnell Douglas 369FF. Substantial damage. One fatality, two Control Fastener Detaches serious injuries, one minor injury. Robinson R44 Astro. Destroyed. Two fatalities, one serious injury. he police helicopter was being used to s part of a biennial flight review at Cess- scout a 3,600-ft mountaintop near Marana, nock Aerodrome in New South Wales, TArizona, U.S., for installation of emergency AAustralia, the morning of Feb. 4, 2011, communications equipment the morning of Jan. the instructor simulated a failure of the flight 31, 2011. The NTSB report said that the weather control hydraulic boost system. After land- conditions, which included 10- to 15-kt winds, ing the helicopter, the pilot told the instruc- were “within the helicopter’s and the pilot’s tor that the hydraulic system would not performance capabilities.” re-engage. The pilot circled the peak before attempting The passenger, who normally flew the a pinnacle landing. “The passengers reported helicopter, told the instructor that the system that during the landing attempt, they felt a had been leaking and that he had replenished bump, the helicopter rose a few feet, then the the reservoir that morning. “The instructor an- nose pitched down and the helicopter began to nounced that he would reposition the helicopter spin to the right,” the report said. “The helicop- to the apron to facilitate examination of the ter tumbled and slid about 120 ft [37 m] down a hydraulic system,” the ATSB report said. shallow canyon … before it was halted by rocks The instructor lost control of the R44 shortly and scrub vegetation.” after becoming airborne. The helicopter was in Investigators determined that the pilot, who a steep left bank when it struck the runway and was killed in the crash, had lost control of the came to rest on its left side. A fire erupted and helicopter after the tail rotor struck the ground rapidly engulfed the helicopter. The pilot was during the attempted pinnacle landing.
FLIGHTSAFETY.ORG | AEROSAFETYWORLD | FEBRUARY 2013 | 63 ONRECORD
Preliminary Reports, December 2012
Date Location Aircraft Type Aircraft Damage Injuries Dec. 5 Drakensberg, South Africa Aero Modifications C-47TP destroyed 11 fatalities The turbine-modified Douglas DC-3, operated by the South African air force, struck a mountain while being flown in severe weather conditions. Dec. 7 La Yesca, Mexico Britten-Norman Islander destroyed NA The Islander was departing for an air ambulance flight when it encountered a strong crosswind and drifted sideways into trees. All three occupants survived the crash. Dec. 9 Iturbide, Mexico Learjet 25 destroyed 7 fatal The Learjet was on a charter flight from Monterrey to Toluca when it entered a high-speed descent from 28,000 ft and struck mountainous terrain. Dec. 10 Compton, Illinois, U.S. MBB BK-117A-3 substantial 3 fatal The helicopter crashed during an emergency medical services positioning flight from Rockford to Mendota, Illinois. Dec. 12 Zulia, Colombia Piper Seneca destroyed 4 fatal The Seneca was en route under visual flight rules when it struck rising terrain in instrument meteorological conditions. Dec. 14 Amarillo, Texas, U.S. Beech King Air E90 destroyed 2 fatal Shortly after departing from Amarillo and receiving clearance to deviate from course to avoid weather, the pilot lost control of the King Air, which then broke up in flight. Dec. 15 Ely, Nevada, U.S. Piper Cheyenne destroyed 2 fatal The Cheyenne was cruising at 24,000 ft in visual meteorological conditions (VMC) when it entered a right turn and descended within 30 seconds to 14,500 ft, where radar contact was lost. The airplane broke up before crashing in a canyon. Dec. 17 Tomas, Peru Antonov 26-100 destroyed 4 fatal The An-26 crashed in the Andes during a cargo flight from Lima to Las Malvinas. Dec. 18 Payson, Arizona, U.S. Piper Chieftain substantial 1 fatal The cargo airplane was nearing Payson when the pilot requested and received clearance to divert to Phoenix due to poor visibility at the destination. Shortly thereafter, the Chieftain struck a mountain at about 7,000 ft. Dec. 18 Libby, Montana, U.S. Beech King Air B100 substantial 2 fatal Night VMC prevailed at the airport when the King Air struck trees and crashed in mountainous terrain during a visual approach. Dec. 20 Holtanna Glacier, Antarctica Basler BT-67 substantial 2 minor, 13 none The air-tour airplane struck a snow drift while lifting off from an unprepared airstrip. The turbine-modified DC-3 then stalled and landed hard, collapsing the main gear. Dec. 21 Oklahoma City, Oklahoma, U.S. Cessna 550 Citation substantial 1 minor, 1 none The Customs Service airplane overran the runway after the nose landing gear collapsed on touchdown. Dec. 22 Sanikiluaq, Nunavut, Canada Swearingen Metro III destroyed 1 fatal, 2 serious, 6 minor The flight crew was attempting a second approach in adverse weather conditions when the Metro touched down hard and overran the runway. Dec. 24 Leesburg, Florida, U.S. Piper Chieftain NA 1 fatal Witnesses said that the engines were not functioning properly when the Chieftain stalled and crashed on approach. Dec. 25 Shymkent, Kazakhstan Antonov 72-100 destroyed 27 fatal Visibility was about 800 m (1/2 mi) in heavy snow and the ceiling was at 400 ft at Shymkent when the An-72 crashed on approach, about 20 km (11 nm) from the airport. Dec. 25 Heho, Myanmar Fokker 100 destroyed 2 fatal, 8 serious, 62 minor One person on the ground was killed when the Fokker struck power lines on approach in fog and crashed about 1 km (0.5 nm) from the runway. Dec. 29 Moscow, Russia Tupolev 204-100V destroyed 5 fatal, 3 serious The surface winds were from 280 degrees at 16 kt, gusting to 29 kt, when the Tu-204 overran Runway 19 on landing and struck a highway embankment during a positioning flight. Dec. 31 San Pedro Sula, Honduras British Aerospace Jetstream 31 substantial 1 minor, 18 none VMC with light winds prevailed when the Jetstream veered off the runway on landing and struck a ditch.
NA = not available This information, gathered from various government and media sources, is subject to change as the investigations of the accidents and incidents are completed.
64 | FLIGHT SAFETY FOUNDATION | AEROSAFETYWORLD | FEBRUARY 2013 CASSIOPEE PUT YOUR FLIGHT DATA TO WORK - Photo : Getty Images singulier et associés - Photo :
Cassiopee: Airline & Operator Services by Sagem
Airlines and operators of all shapes and sizes can boost financial and environmental performance with Cassiopee services, an integrated package of flight data management hardware and software. Improve your safety (FOQA), contain your maintenance costs (MOQA) reduce your fleet’s fuel consumption and CO2 emissions, optimize your crew management and boost your organization. You can count on Sagem’s proven solutions to raise your business profile. www.sagem-ds.com Registration Now Open! flightsafety.org/BASS2013
58th annual business aviation safety seminar BASS 2013 april 10–11, 2013
fairmont queen elizabeth hotel montreal, canada
supported by
Keynote Speakers: Raymond Benjamin Secretary General, To register or exhibit at the seminar, contact Namratha Apparao, International Civil Aviation Organization tel: +1.703.739.6700, ext. 101, [email protected]. Merlin Preuss To sponsor an event, contact Kelcey Mitchell, Vice President of Government and Regulatory Affairs, tel: +1.703.739.6700 ext. 105, [email protected]. Canadian Business Aviation Association To receive membership information, contact Susan Lausch, tel: +1.703.739.6700 ext. 112, [email protected].
To advertise in AeroSafety World magazine, contact Emerald Media, For full agenda visit Cheryl Goldsby, tel: +1.703.737.6753, [email protected] flightsafety.org/BASS2013 Kelly Murphy, tel: +1.703.716.0503, [email protected]