Presentation to the Manitoba Aviation Council

Presentation to Manitoba Aviation Council
Kathy Fox, Board Member
Transportation Safety Board of Canada
Winnipeg, Manitoba
24 April 2012

Click here to see PowerPoint Presentation (PPT)  [4.92 MB]

Check against delivery

Slide 1: Title Page

Thanks for the introduction, and thank you very much for this opportunity to speak today.

Slide 2: Outline

First I'd like to give you some information about the TSB, our mandate and investigation processes.

After that, I will talk briefly about our Watchlist, and then I'll focus on one of its safety issues, runway safety, which is a topic of great concern to the TSB, operators, and Canadian airports. Then, after outlining what action Transport Canada is planning, and what the Board would like to see, I look forward to hearing your perspective and answering any questions you may have.

Slide 3: About the TSB

The Transportation Safety Board was formed in 1990 with the passing of the Canadian Transportation Accident Investigation and Safety Board Act. The Board itself currently consists of 5 Board Members, including the Chair. Overall, there are approximately 230 employees in nine offices across the country, plus a lab in Ottawa.

We're not the regulator; we don't have any powers of enforcement. We are not a court; we do not find fault or determine civil or criminal liability. We're an independent agency whose only mandate is to advance transportation safety by conducting independent, expert investigations of selected marine, pipeline, rail and air “occurrences.”

An “occurrence” can be an incident or accident. Incidents generally involve events such as engine failures or risks of collisions, while accidents involve serious injury, loss of life or significant equipment damage. The reportable types of occurrences are defined for each transportation mode in the CTAISB Regulations.

Slide 4: Our Investigations

The TSB as a whole receives notice of approximately 4000 occurrences every year, spread across all modes. Roughly 1300-1500 of these are in the Air Branch. (In 2011, they received notification of 1344 occurrences.) Once these are reported to us, we make a decision as to the depth of investigation warranted–a decision which is based on several factors. Chief among these is whether an in-depth investigation has the potential to advance transportation safety.

Over the last five years, the Air Branch has averaged 39 investigations (i.e., those culminating in a public report) per year. In 2011, we started 33 occurrence investigations (including 2 overruns, four veer-offs, and one runway incursion).

However, even if we do not launch a full investigation, we still assess each and every reported occurrence and track that information in our database. That way, we can analyze the information to spot trends, and then target issues that continue to be a problem.

During each investigation, TSB investigators identify safety issues by assessing the technical, operational and human factors related to an occurrence. Our investigators then determine what conditions were present, what actions the operator took, as well as any other underlying factors that might have a negative influence on safety. From there, they assess the risks, analyze the defences in place, as well as any other existing risk-control options.

Slide 5: Board Responsibilities

The TSB's practice is not to wait until an investigation is complete before making important safety-critical information public. When we identify an urgent safety deficiency, we communicate it to those who can make transportation safer–right away–thereby allowing for timely and appropriate safety action.

Every investigation report is reviewed by the Board. We review every report in draft before it goes out to official “Designated Reviewers.” These are people from whom the Board solicits feedback based on considerations such as their ability to comment on the technical accuracy of the report, and whether their interests may be affected by the report.

Once we receive this feedback, the Board reviews every comment and approves every report before it is made public.

For difficult, systemic issues, the Board may issue recommendations. These are meant to draw attention to particular safety deficiencies that involve substantial risk and require immediate action.

The TSB also issues Safety Advisories and Safety Information Letters. These notify industry and regulators as soon as possible when significant safety risks are found during an investigation.

Slide 6: Watchlist

The TSB also has another tool at its disposal, our safety Watchlist, which we launched in March 2010. This document highlights the nine transportation safety issues posing the greatest risk to Canadians–the ones that have proven toughest to solve, the most stubborn to eradicate, and which keep coming up in investigation after investigation.

These issues, three of which are related specifically to air transportation, are based on our review of accident investigation reports–and in particular on recurring incidents/ accidents, Canadian and worldwide statistics, TSB safety communications, and a total of 41 separate and outstanding Board recommendations.

I want to talk about one of those issues today.

Slide 7: Landing Accidents and Runway Overruns

Every year millions of aircraft land at Canadian airports. The landing is one of the most critical phases of flight. An accident can happen on the runway, or the aircraft may fail to stop in time and run off the side or end.

Although these runway excursions can occur both during takeoff and landing, I'd like to focus on landing overruns. That's because, according to a May 2009 article from FSF, Reducing the Risk of Runway Excursions, landing excursions outnumber takeoff excursions by a ratio of approximately 4:1.

Many of these accidents happen in bad weather, with crews typically faced with a dynamic and quickly changing environment.

Almost everyone, of course, remembers the Air France occurrence in 2005 in Toronto.

But … Barely two years ago, on March 24, 2010, Cargojet flight 620, a Boeing 727, overran a Moncton runway. The aircraft came to rest in deep mud, the nose wheel approximately 340 feet beyond the runway end and 140 feet beyond the edge of the paved runway end strip. (A10A0032)

And it's not just the big jets. These accidents happen to all sorts of aircraft, and they happen often. In fact, the TSB has investigated, or is currently investigating, at least nine other overruns–all of them within the last few years.

Slide 8: More Common Than You Might Think

A10H0004: On June 16, 2010, an Embraer 145, overran the runway upon landing at Ottawa's MacDonald-Cartier International Airport. The aircraft came to rest 500 feet off the end of runway and 200 feet left of the runway centerline.

A10Q0213: On November 30, 2010, a Boeing 737-800 left the runway upon landing at Montreal Pierre Elliott Trudeau Int'l, eventually stopping on the grass.

A11Q0052: On March 12, 2011, a Bombardier BD100 1A10 lost directional control upon landing at Iqaluit. The aircraft exited the runway to the right about 4300 feet from the threshold. The aircraft travelled 400 feet over frozen ground and stopped about 100 from the edge of the runway in a snow bank.

A11O0098: On June 17, 2011, a Skycharter Ltd. Falcon 10 was substantially damaged when it touched down to the right of the runway centerline at Buttonville airport before exiting the runway on to the grass infield. The aircraft entered the adjacent taxiway before again exiting on to the grass, where it struck a runway sign.

Slide 8 (continued) �

A11C0102: On July 4, 2011, a Missinippi Airways Cessna 208 Caravan was departing Pukatawagan, MB, when the pilot rejected the takeoff. The aircraft continued past the end of the runway and into a ravine. A post-crash fire ensued, killing one passenger and injuring the pilot and seven passengers.

A11A0035: On July 16, 2011, a Kelowna Flightcraft Air Charter Boeing 727-200 landed at St. John's International and overran the end of the runway by approximately 400 feet.

A11H0003: On September 4, 2011, an EMB-145 aircraft, slid off the west side of Runway 32 at Ottawa MacDonald-Cartier International Airport (CYOW) after landing in heavy rain.

A12W0004: On January 9, 2012, a Boeing 737-700 overran the runway threshold at Ft. Nelson, British Columbia, by approximately 250 ft.

A12O0005: On January 15, 2012, a Pilatus PC-12/45, powered by a PWC PT6A-67B turboprop engine, made a straight-in approach at Timmins, Ontario. The aircraft landed at higher-than-normal approach speed, became airborne before reaching the runway end, and touched down again approximately 1200 ft. beyond the runway end in about two feet of snow. The aircraft slid for about 300 ft. and came to a stop heading almost 90° to the left of runway orientation.

Slide 9: A Worldwide Challenge

Aircraft running off the end of runways is a problem worldwide. According to statistics from ICAO, there were an average of 32 overruns per year between 2000 and 2010, with 1038 fatalities.

According to a 2009 Flight Safety Foundation report, almost thirty percent of aircraft accidents between 1995 and 2008 were runway excursions–either overruns or veer offs.

Slide 10: A Worldwide Challenge (cont'd)

These graphs from Airbus, however, are even more recent–and they put that figure closer to 40 percent.

Slide 11: Approximate Runway Overrun Accident Rates (1990 to 2006)

According to a 2008 study commissioned by Transport Canada,1 Canada's rate of runway overruns per million landings is almost twice the world average. Compared with the United States, we have almost 3 times the rate of overruns. [Note that these statistics are based on transport category aircraft only–jet and large turboprop aircraft over 5700 kg/12 500 lbs.]

The numbers don't get any better when it comes to wet runways; in those conditions, Canada's rate of overrun accidents per million landings is 4 times the the world average.

Slide 12: Overrun Accidents Involving Airplanes Over 5700 kg in Canada (1985-2011)

In the TSB database, there are a total of 17 accidents involving aircraft over 5700 kg between 1985-2011 where there was at least substantial aircraft damage and, in many cases, injuries.

In Canada, including incidents and accidents, runway overruns occur at a rate of about 3.3 per year for these larger aircraft. So even though the numbers are fairly small, we are all exposed to the possibility of a runway overrun, with potentially catastrophic consequences.

[NB: Although “only” 17 accidents involved substantial aircraft damage, there were 88 accidents in total.]

Slide 13: A Complex Problem

I want to be clear about something here: The Board recognizes that runway length is not the sole determinant of whether a plane will have an overrun, nor whether there will be any damages or injuries to passengers and crew. In fact, the top 3 risk factors for a landing excursion are: go-around not conducted, touchdown long, and ineffective braking (e.g.,hydroplaning, contaminated runway).2

Clearly, then, runway overruns are a complex problem, one often involving human factors, and the TSB acknowledges this. Therefore, numerous lines of defence are needed to both:

  • prevent overruns from occurring in the first place
  • prevent injury, loss of life, or damage to property or the environment in case an overrun occurs.

Slide 14: Air France

Let me refresh your memory with a recap of the Air France accident: On August 2nd, 2005, an Air France Airbus A340-313, with 297 passengers and 12 crew, overran the runway at Toronto's Pearson Airport.

The crew conducted their approach and landing in the midst of a severe and rapidly changing thunderstorm. There were no procedures within Air France related to distance required from thunderstorms during approaches and landing, nor were these required by regulations. Approaching the threshold, the aircraft entered an intense downpour, and the forward visibility became severely reduced. When the aircraft was near the threshold, the crew members became committed to the landing and believed their go-around option no longer existed. The touchdown was long because the aircraft floated due to its excess speed over the threshold and because the intense rain and lightning made visual contact with the runway very difficult. The aircraft touched down about 3800 feet from the threshold of Runway 24L, which left about 5100 feet of runway available to stop. The aircraft overran the end of Runway 24L at about 80 knots and was destroyed by fire when it entered the ravine. Although the area up to 150 m beyond the end of Runway 24L was compliant with Aerodrome Standards and Recommended Practices (TP 312E), the topography of the terrain beyond this point, contributed to aircraft damage and to the injuries to crew and passengers.

Slide 15: Previous Recommendations

To prevent runway overruns from happening in the first place, the Board has previously issued 3 recommendations to enhance operational standards, pilot training, and operational procedures as follows:

  • That regulators establish clear standards to limit landings in bad weather (A07-01); [Update: TC has taken action to raise the awareness of Canadian operators, and initiated international discussion on the issue.]
  • That pilots receive mandatory training to better enable them to make decisions about landing in deteriorating weather; and ; [A07-03: Update: TC initially said it would include a training requirement for pilots operating under Subparts 703, 704 and 705 of the Canadian Aviation Regulations (CARs), but the Board is now concerned that TC has placed its action related to pilot decision-making training on hold.]
  • That crews be required to establish the margin of error between landing distance available and landing distance required before conducting approaches in deteriorating weather. [A07-05: Update: TC has proposed six NPAs back in 1999. In the long-term, these would reduce or eliminate the deficiency. However, TC's latest response is silent on any short term solution.]

These were formal Board recommendations. But there are other ways to help prevent overruns.

Slide 16: Surface-condition reporting

Reporting of runway surface conditions, for example, is explicitly highlighted in our Watchlist, as this is something that can significantly affect landing-distance calculations.

This issue, in effect, boils down to pilot information: what they get, and when they get it. According to TP 312, the publication in which Transport Canada sets out aerodrome standards and recommended practices, the following is required: Standard - The condition of the movement area and the operational status of related facilities shall be monitored and reports on matters of operational significance or affecting aircraft performance given, particularly in respect of the following:
  • 3. snow, slush or ice on a runway, a taxiway or an apron;
  • 4. standing-water on a runway, a taxiway or an apron;

In winter, obviously, the focus is on snow, slush, or ice on the runway–and many airports already have a good system in place for this.

In summer, however, there is a different concern �

Slide 17: Wet Runways

According to that 2008 TC-commissioned study I mentioned earlier,3 degraded aircraft performance on wet runways has been identified as a factor in the majority of aircraft accidents on landing. Studies have shown that jets and large turboprop aircraft are seven times more likely to overrun when landing on a wet un-grooved runway versus one that is dry. The risk of overrun increases during periods of heavy rainfall, which is often associated with other adverse weather conditions such as strong and gusty winds, wind shear, and poor visibility. Also according to TC's study, “the risks associated with landings during heavy rainfall on un-grooved runways are currently much higher than acceptable levels in commercial aviation.” In particular, the high number of overruns is due to hydroplaning, which causes a loss of directional control and can reduce braking to nil. Once hydroplaning has started, it may persist to a significantly slower speed.

Obviously, then, information about braking action–or, to use the term preferred by ICAO, “friction characteristics”–needs to be reported to pilots, because each runway is unique in how water builds up, depending on factors such as its construction and level of contaminants present.

Slide 18: Friction Testing

Likewise, if airports carried out more frequent periodic friction testing, this would give a better handle on the trend of the runway surface condition. And that would facilitate early action if it appears that a runway's surface friction is deteriorating.

Slide 19: What Else Can We Do?

But even with all of that: training, procedures, timely information to pilots, friction testing� Overruns will not be eliminated entirely. James M. Burin of the FSF said that:

The severity of runway excusrsion accidents depends primarily on the energy of the plane as it departs the runway, and the airport's layout, geography and rescue capability.

So what else can be done?

Slide 20: Recommendation A07-06

In 2007, the Board recommended an additional line of defence. When all else fails, adequate runway end safety areas, or RESAs, will greatly increase the chances for an aircraft to stop safely in case of an overrun. At some airports, moreover, the terrain beyond the end of the runway could contribute to aircraft damage and injuries to passengers and crew. Therefore this safety area must be sufficiently clear of obstacles, appropriately graded, and have a surface that will aid in stopping the aircraft as quickly as possible.

Recommendation A07-06 urges TC to require all Code 4 runways (those longer than 1800 m) to have a 300 m runway end safety area (RESA) or some other means of stopping aircraft that provides an equivalent level of safety.

I'd like to take a moment to repeat something here, to clarify, because I think there's a misperception that's sprung up since we issued this recommendation.

The TSB is not demanding the construction of a 300 m RESA on every runway in Canada. That kind of one-size fits all approach wouldn't be feasible, nor do we feel it would be necessary to address the risk. The key is a balanced approach. We know that all runways do not present the same risks; that is why the Board's recommendation focuses on Code 4 runways at Canada's largest airports. And we've also given careful consideration to the phrase “a means of stopping aircraft that provides an equivalent level of safety.” But more on that in a few minutes.

Slide 21: Why 300 m?

In 1987, a study by the FAA showed that almost all aircraft that overrun runways–90 percent of them–stop within 1000 feet of the runway end. That's approximately 300 metres. In 2009, a study by the Australian TSB reconfirmed this figure when it came to the same conclusion.

In those runway overrun accidents that were deemed the most serious, a key factor was an insufficient overrun area. This was a factor in 4 overrun accidents since 1999 in the U.S., and has resulted in 12 fatalities and 185 injuries.

So, if almost all aircraft stop within approximately 300 m of the runway's end, it only makes sense to make that area safe for aircraft to stop should there be an overrun.

Note that the 300 m RESA also provides additional protection in the event of an under-shoot on final approach or a rejected take-off at high speed.

Slide 22: ICAO RESA Standards

There are a number of different standards for runway end safety areas.

  • The current ICAO STANDARD is for a 90 m RESA starting at the end of a 60 m strip (red area). Their RECOMMENDATION is for a 240 m RESA at the end of the 60m runway strip (blue area).
  • The current FAA STANDARD is for a 300 m RESA from the end of the runway.
  • Today, TC's TP312 REQUIRES only a 60 m strip at the end of Code 3 and 4 runways. However, TP312 RECOMMENDS an additional 90 m RESA starting at the end the 60 m strip. (red area).

Slide 23: Recommendation A07-06 (Update)

It's been more than four years since the Board issued this recommendation. Yes, the regulatory process takes time, sometimes a long time, but let's be clear: TC does not yet meet the current International Standard.

TC has reviewed airport certification standards and proposes harmonizing Canadian standards with ICAO's current RESA standards (90 m) beyond the runway strip end. The latest version of their proposed regulatory amendment would require that all runways of 1200 m or greater, or those under 120 m where the runway is certified as precision or non-precision, and used by scheduled passenger-carrying operations of an air carrier operating aircraft designed for more than 9 passenger seats as determined by the aircraft type certificate –to have a 150 m RESA, or another acceptable alternate means providing an equivalent level of safety. This requirement would not apply to airports located north of the 60th degree parallel that only serve air carrier operations utilizing small aircraft.

In the spring of 2010, TC also committed to individual risk assessments for each of the Code 4 runways in Canada within a year–a move that had the potential to address the key safety deficiency. Instead, however, TC decided to conduct an aggregate risk assessment, rather than an assessment of individual runway ends. This, however, may not identify or effectively mitigate risks due to hostile terrain beyond the planned 150 m RESA on other runways at some Canadian airports.

Slide 23 (continued) �

While this regulatory process appears to be progressing, even if fully implemented as described in TSB recommendation A07-06, it will not eliminate the safety deficiency at many Canadian airports.

While we appreciate that TC is taking some action, for Canada to lag the world in safety is, frankly, disappointing.

Slide 24: Engineered Materials Arresting System (EMAS)

OK, now I'd like to return to the phrase I mentioned a few minutes ago: a means of stopping aircraft that provides an equivalent level of safety. I'm talking about an Engineered Materials Arresting System, or EMAS. As I said, in locations where terrain features don't permit a 300 m RESA (inclusive of 60 m runway strip), this can be an alternative.

On screen you'll see two pictures of an EMAS that stopped a Bombardier CRJ-200 at Yeager Airport in Charleston, West Virginia, (KCRW) in January 2010.

Slide 25: EMAS (cont'd)

According to the manufacturer, EMAS is installed at 67 runway ends at 45 airports worldwide.

To date, there have been eight occurrences in the United States where the technology has worked successfully to stop aircraft which overran the runway and in several cases has prevented injury to passengers and damage to the aircraft.4

Slide 26: Cost vs Safety

Cost is certainly factor in all of this. The Board realizes that RESAs or EMAS involve a significant capital investment, particularly in the current economic climate.

Doing nothing, however, costs even more: A Flight Safety Foundation study found that between 2005 and 2007, runway excursions cost the aviation industry as a whole $1.5 billion, or $506 million per year. Some of this cost is also shouldered by airports to repair damage, in delays due to downtime and in litigation.

Slide 27: Cost vs Safety (cont'd)

This slide from a 2011 Airbus presentation uses different figures, but makes the same point.

From 1985 to 2010, there were over 1000 incidents, and more than $5.4 billion in damages. Do the math: that's an average of $5 million per incident!

Slide 28: What's Needed?

At the TSB, we are sometimes afforded an opportunity to see problems before they become evident to everyone. We're a world leader in accident investigation in part because, as I said earlier, the Air Branch alone carries out dozens of full investigations each year–and access to that much data sometimes lets us spot trends in advance. But on the matter of overruns, Canada lags behind the rest of the world when it comes to implementing much-needed safety measures.

So what's the fastest way to improve safety? Operators need to take safety action that is tailored to the particular risks of their individual operations.

In short, don't wait for the regulator to prescribe a standard. Be proactive.

  • There's no reason companies can't require landing distance calculations and SOPs regarding landing in deteriorating conditions.
  • Pilots, meanwhile, need timely information about surface conditions, in all seasons.
  • We'd also like airport operators to evaluate their runways for RESA requirements, without waiting for TC's aggregate assessment data. And then take appropriate action to reduce the potential for serious injuries or deaths.

It is our hope that doing all of this will help ensure that the next time we have an aircraft overrun–and we will–that the aircraft will stop safely without passenger injury or death.

Slide 29: Pukatawagan Update

I'd now like to provide a brief update on one of our current investigations, one of great interest to the people in this room, and very relevant to this discussion: the fatal overrun and fire that took place last July at Pukatawagan. The TSB has not yet released its final report into this investigation, so I can't provide you with any findings or conclusions, but I do think it's worth looking at that day's events in the context of today's topic.

Just after 4pm on July 4, the pilot of a Cessna 208B began the take-off roll from Runway 33. The aircraft did not become fully airborne, and the pilot rejected the takeoff. The pilot applied reverse propeller thrust and braking, but the aircraft departed the end of the runway and continued down an embankment into a ravine. A post-crash fire ensued. One of the passengers suffered fatal injuries; the pilot and the 7 other passengers egressed from the aircraft with minor injuries. The aircraft was destroyed.

In this case, the runway was 2998 ft long (914 m), with an additional 230 ft (71 m) turnaround area. But beyond that, there was hostile terrain: the prepared surface gives way to an embankment that descends into a ravine. The slope of the embankment is approximately 30° to 45° and is comprised of gravel, rocks, and large boulders. The slope descends about 20 feet vertically, and then reverses sharply into the contour of a ravine.

Slide 30: Pukatawagan Update (cont'd)

According to both the current TP 312 and what's being proposed under the NPA for RESAs, the runway at Pukatawagan would not be required to have a RESA.

However, one person died on July 4 last year. Given the larger aircraft that use this runway, and others like it, the risk exists for an even bigger accident with greater loss of life.

Slide 31: Conclusion

Back in 2010, we first issued our Watchlist, we envisioned it as a “blueprint for change,” a way to stimulate action and agents to make transportation safer by targeting nine key safety issues. Runway overruns made that list because they were persistent, they kept coming up in investigation after investigation, and the issue needed immediate action.

The question now is: how far down the safety path can we proceed? TC has taken important steps forward, but we know these outstanding safety deficiencies are complex, and as I've said, overruns are not just about runway length. A blanket approach, or a one-size-fits-all solution isn't what we're urging. However, we also know from hard experience that if a persistent safety problem is not addressed, there will be another accident.

The issue of overruns requires a multi-faceted approach from multiple stakeholders. For instance, it may not be feasible to install a 150 m RESA (let alone a 300 m RESA), but perhaps there are other steps to mitigate hostile terrain.

Slide 31 (continued)�

  • From airport operators: risk assessments of individual runways can offer an opportunity to find which ones pose the greatest risk, and then develop appropriate mitigation measures–including the consideration of an arrestor system
  • From the regulator: we'd like to see clear standards to limit landings in bad weather. And pilots, meanwhile, need mandatory training to better enable them to make decisions about landing in deteriorating weather
  • And again, there's no reason operators can't require crews to establish the margin of error between landing distance available and landing distance required before conducting approaches in deteriorating weather.
  • Pilots, meanwhile, they require timely information from airports about surface conditions, especially in bad weather, year-round.

Slide 32: Canada wordmark

Thank you again for the opportunity to speak to you today. I'd be pleased to take any questions
  1. TP 14842E–Risk and Benefit-Cost Analyses of Procedures for Accounting for Wet Runway on Landing (2008).
  2. This is according to the May 2009 FSF article: Reducing the Risk of Runway Excursions.
  3. TP 14842E–Risk and Benefit-Cost Analyses of Procedures for Accounting for Wet Runway on Landing (2008).
  4. These stats have been updated from: http://www.esco.zodiacaerospace.com.