Accident Investigations

A few days ago, most were saying it’s too early to tell what brought down the Russian Airbus that crashed on Egypt’s Sinai Peninsula, killing all 224 aboard.  Now, there’s talk of the aircraft being downed by a bomb.  

Why a bomb?  The best explanation comes an article written by former NTSB investigator Doug Herlihy, appearing in the Goldendale Sentinel:

First, it’s likely that the aircraft came apart in flight.

The pieces of aircraft and persons on board are being found spread over 20KM (over 15 miles) in the mountainous region of the southern Sinai. The spread of wreckage is the most critical piece of the accident puzzle.    

When aircraft break up in flight, the parts are spread by two phenomena: the “ballistics” of the pieces and the wind.  “Ballistics” refers to the shape and weight of the object (like bullets or feathers) and how they will fly to earth.  The second factor is the wind at various altitudes as the parts fall to earth.  Like tearing a pillow in the wind, the parts are widely spread.

Second, aircraft seldom blow up because of a defect lurking within.  History shows that they almost always blow up because of an outside force.

Rarely, have any system or fuel supply or tank ever exploded a modern airplane.  Jet aircraft jet fuel, like kerosene and diesel fuel is not prone to explosion.  And, though it is not uncommon for an airliner to be hit by lightning, it’s almost unheard of that it has caused an explosion. Investigators know that either an on-board bomb or a hit by an explosive device is very high on the list of clues to search for.

NTSB preliminary reports do not draw conclusions as the cause of a crash.  But the NTSB’s preliminary report of the Turbine Otter crash that killed 9 near Ketchikan on June 25 suggests a weather-related “CFIT” crash, exactly as described here.

First, the report indicates that the flight was conducted under Visual Flight Rules.  That means that that pilot was supposed to stay out of the clouds and avoid the terrain by looking out the window rather than by relying on instruments.Chelton disply

Second, the report indicates that the closest reported weather was “marginal” for flying under visual flight rules.    (“The closest weather reporting facility is Ketchikan Airport (KTN), Ketchikan, AK, about 24 miles southwest of the accident site. .  .  few clouds 800 feet, broken clouds 1,200 feet, overcast clouds 2,700 feet. . .”) 

Third, and most significantly, a helicopter pilot searching for the aircraft minutes after the crash was unable to get to the crash site because the terrain was obscured by clouds and fog.  

The NTSB noted that the Otter was equipped with a moving map display that is designed to depict the aircraft’s position with respect to hazardous terrain.  When first introduced, such displays were seen as a boon to safety, making it easier for pilots to avoid terrain that they might not otherwise be able to see.  But some argue that such technology doesn’t increase safety at all, because pilots use the technology to fly closer to the edge than they otherwise would.  The phenomenon coming into play is called “risk homeostasis.” And in fact, the NTSB has previously found that aircraft equipped with moving maps and the other technology comprising modern "glass cockpits" have a higher rate of fatal accidents than those that aren’t.

We don’t know much yet about the plane crash in Alaska that killed the pilot and 8 tourists from the MS Westerdam.  But the crash looks eerily similar to the Alaskan crash that killed Senator Ted Stevens and three others in 2010.

Like the plane that was involved in the Westerdam crash, the plane that crashed with Stevens Otter Senator Stevens Crashaboard was a de Havilland Otter retrofitted with floats and a turboprop engine. Both tour pilots encountered adverse weather that is common in Alaska:  Low Ceilings. Fog. Gusty winds.  

In the Steven’s crash, instead of turning around when he encountered the low clouds, the pilot pressed on.  Unable to see where he was going, he inadvertently flew into the side of the mountain. (The local papers were calling the pilot a "hero" because not everyone aboard was killed. I had to disagree.)

In last week’s crash at Ella Lake, the weather conditions were similar.  It looks as though the pilot,Promech Air Otter employed by tour operator Promech Air, inadvertently flew into the clouds and struck the side of a cliff.  

This sort of accident is not uncommon, particularly in Alaska and Hawaii.  The type of accident is called "controlled flight into terrain."  It is almost always due to pilot error. 

Blue Hawaiian helicopters was probably the last tour operator that flew a perfectly good aircraft into the side of a mountain due to low clouds.  Compare the photo of the weather conditions that contributed to the Blue Hawaiian crash (left) with the photo of the weather conditions that the Promech Air pilot tried to fly through.  Note how, in both photos, the clouds obscure the mountain tops.

Otter Crash near Ella Lake Alaska weather

 

 

Let’s get it out of the way: there is little in common between the apparent loss of AirAsia Flight QZ 8501 and the disappearance of Malaysia Airlines Flight MH 370.  

But Flight 8501’s disappearance does have at least some resemblance to the 2007 loss of Adam AirAirAsia A320 Flight 547.  Both Indonesian airliners disappeared shortly after contact was lost in bad weather. Both disappeared in Indonesian airspace — the AirAsia flight over the Java Sea; the Adam Air Flight over the Makassar Strait.

Bad Weather vs. Pilot Inputs

The speculation after the Adam Air crash was that the flight was brought down by severe weather — weather that the crew had been warned about.  But that turned out to be wrong. Adam Air Flight 547 went down because the crew fixated on troubleshooting a problem with the aircraft’s navigation system, not because of weather.  The crew became so preoccupied withthe navigation system that they allowed the aircraft to slowly roll into a steep bank.  They allowed the nose to point down and the aircraft to build too much speed.  When the pilot figured out was going on and tried to recover, his control inputs broke the wings.

But the AirAsia Crew Had Requested a Deviation for Bad Weather

Unlike the Adam Air crew, the AirAsia pilots had requested from Air Traffic Control a clearance to climb to a higher altitude but didn’t immediately get it.  A short time later, all contact with the airliner was lost. Isn’t that a strong indication that rough weather may have been a factor?  

First, while small aircraft are often brought down by rough weather, it’s extremely rare for an airliner to be.  Airliners avoid rough weather largely for comfort rather than for safety.  Second, although Air Traffic Control delayed in giving the AirAsia flight a clearance to climb, the pilots were free to do so immediately in the unlikely event the weather posed a risk to the aircraft’s safety.

But while it’s rare for an airliner to be brought down by turbulence, it’s quite possible for an airliner to be brought down by a pilot’s reaction to that turbulence.  

Airbus Rudder System

That’s exactly what happened to American Airlines Flight 587 in 2001.  The aircraft encountered turbulence climbing out of JFK.  The co-pilot tried to correct by pushing on the rudder with his foot. He pushed too hard and the aerodynamic forces caused structural failure.  The airliner crashed and killed all 260 aboard and 5 on the ground.

American Airlines 587 was an Airbus A300. More then 10 years after that crash, the FAA required all 300 series aircraft to be modified to warn the pilot to "stop rudder inputs" when structural damage becomes a risk, a modification that I felt was inadequate.  Flight 8501 was an Airbus A320.  That’s the same model which the NTSB called flawed because its rudder system was too sensitive:

The Airbus A320 family is . . .susceptible to potentially hazardous rudder pedal inputs at higher airpeeds. 

Airport fire trucks must get to a burning plane within three minutes if they are going to save any lives. That’s the maximum response time allowed by the National Fire Protection Association, the organization that sets the standard for airport firefighters, including those working at U.S. Air Force bases. 

The survivable atmosphere inside an aircraft fuselage involved in an exterior fuel fire is limited to approximately 3 minutes if the integrity of the airframe is maintained during impact. This time could be substantially reduced if the fuselage is fractured. . . rapid fire control is critical. . .

Aircraft flown in air shows are usually smaller and less fire resistant than transport category aircraft.  At air shows fire trucks need to get to crash sites even quicker – within 60 seconds or less.

The key to getting fire trucks to a crash quickly is to station the trucks near to where an accident is most likely to occur.  Normally, that might be the end of the active runway.  But most air show crashes occur at “show center” rather than the end of the runway.  As one Travis Air Force witness put it, show center is where ‘the majority of dangerous events focus.”  At air shows, that’s where fire trucks should be waiting.

Eddie’s Accident

On May 4, Eddie Andreini was flying a routine at the Travis Air Force Base open house.  He was attempting a stunt known as an inverted ribbon cut.  Something went wrong.  Eddie’s Stearman slid upside down along the runway, coming to a stop at smack dab show center. Eddie was uninjured but was trapped inside.  A fire started almost immediately.  Air Force personnel say that they saw Eddie struggling to get out as he waited for the fire trucks to save him.  One minute passed, then two, then three.  But the crash trucks didn’t come.  When they finally did, it was too late.

What happened?

The Air Force refused to explain why it took so long for its fire trucks to reach Eddie.  So we sued it under the Freedom of Information Act.  We now have internal Air Force documents showing that the brass didn’t understand the Air Force’s own regulations.  They mistakenly believed regulations prohibited them from stationing fire trucks near show center.  So instead, the Air Force positioned the fire trucks more than a mile and a half away. 

The Travis speed limit for fire trucks is 45 mph.  So it took the first fire truck (a “Rapid Intervention Vehicle”) more than four minutes to get to Eddie.  Had the Air Force positioned even one truck at show center–as it was supposed to–firemen would have gotten to Eddie within a minute and Eddie would have been saved.

Regulations can be confusing. Was the Air Force’s mistake understandable?  Not really. The manual that Travis show organizers had in hand–and agreed to follow–makes clear that fire trucks belong at show center.  According to that manual, the personnel who were permitted in the “aerobatic box” (the area in which performers fly) included “demonstration teams and fire/rescue.” (Page 28.) The manual goes on to direct that fire trucks should be located “with immediate access to the show line” (page 34) – not a mile and a half away.

To the extent the Air Force brass was confused, the FAA cleared things up for them when, a week before the air show, it told Travis that crash trucks did indeed belong “in the box” near show center.

Our team, specifically the air ops staff, was led to believe that we could not put an emergency vehicle (or anything else) inside the Show Box at Show Center, because it was sterile and protected.  We learned that this was not correct about a week before the show after [name redacted] discussed it with [name redacted] of the FAA.  We learned that we could place airshow official vehicles or people in the aerobatic box.”

Travis had time

The Air Force’s own documents prove that Travis officials had a week before the show was to begin to correct their mistake and arrange for the trucks to be stationed at show center. But the Travis officials had already decided that the fire trucks were going to be positioned where they couldn’t be of any use to a performer.  Having made a plan, they weren’t going to change, even if it put lives at risk unnecessarily.

“I’ll say it again, I need the trucks on the runway!  I need the trucks on the runway now!”

The Travis Command Post recording is difficult to listen to. After hearing it, it’s hard to believe that Travis still tells the public that its fire department responded to the crash in an “exemplary” fashion.

(Notes:  At 2:14, one of Eddie’s crew tried to fight fire with a hand-held extinguisher.  The extinguisher was too small and was expended in seconds.  By that time, the Rapid Intervention Vehicle had not yet even left its station.  The Air Force documents do not explain why it took so long for the truck to roll.  It finally arrives on scene after the 4 minute mark.  The time stamps were placed on the photos by Air Force.)

Someone changed the course of Flight MH370 and turned off the aircraft’s transponder.  Turning off an aircraft’s transponder makes it more difficult for the plane to be tracked by radar.  A hijacker with even minimal flight training would have known that. 

But there is one wrinkle.  The transponder was reportedly turned off when air traffic control was in the process of a “handoff” from Malaysian Air Traffic Control to Ho Chi Minh City Control in Vietnam.  At that moment, the aircraft was in the shadows: on the outskirts of Malaysian radar coverage and just entering Vietnam radar coverage.  The crew had said goodbye to Malaysian air traffic control, but hadn’t yet established contact with Ho Chi Minh City Control.  If a crew wanted to disappear, that would be an ideal time to pull it off.  Only the most sophisticated hijacker would know that.

Airline’s Obligation to Compensate Family Members

An airline’s obligation to compensate the families of those lost in the crash of an international airliner is governed by an international treaty known as the Montreal Convention.  The Montreal Convention requires the airline to compensate the families of those lost whenever the crash was the result of an “accident.” An “accident”  is defined as “an unexpected or unusual event or happening that is external to the passenger.”  Whether the crash was caused by a pilot’s wilful misconduct, a hijacking, or even a terrorist attack — it doesn’t matter.  The crash counts as an accident and the airline is liable.   

Cap on Airline Liability

An airline is strictly liable for a family’s loss up to 113,100 “Special Drawing Rights,” an amount equal to about $175,000. The airline can avoid liability for sums exceeding that amount only if it can prove it was totally “free from fault.” That is usually an impossible task for an airline, even if the crash was caused by a terrorist.  The air carrier can seldom show that there was nothing it could have done to avoid the accident.  It’s the problem of proving a negative.  Thus, if in fact flight 370 was lost in a crash, it’s unlikely the Convention’s “cap” on liability will come into play. 

More in my interview appearing in the Malaysian press

This animation compares what Asiana 214’s approach should have looked like to what it did look like. From the data we have, the animation appears to be fairly accurate, except the audio is not properly synchronized. (The initial transmissions are from when the aircraft was 7 miles from the runway, not several hundred feet.)

If the audio were fixed, would this animation be admissible in court?

Not in it’s current state.  It relies too much on guesswork. But once the data from the black boxes is available and the animation modified accordingly, it’s exactly the type of thing the lawyers would want to show to a jury.

As described here, passenger claims against Asiana Airlines are limited by the Montreal Convention.  But any claims the victims’ may have against a manufacturer of the aircraft or its component parts are not.  

NTSB Chairman Deborah Hersman reported that evacuation slides opened inside the passenger cabin. The slides are, of course, designed to open outside the cabin.  Passengers (or crew) who were injured by the slides may be entitled to compensation for those injuries from the appropriate manufacturers, if it is proven that the slides malfunctioned because of a defect rather than an error on the part of the flight crew.  Those sorts of claims would be governed by U.S. product liability law, not by the Montreal Convention.