The FAA has responded to some questions submitted by this writer about the latest advisory circular mentioned in that earlier blog, AC 20-53B. This document deals with lightning protection in fuel tanks. The earlier version, AC 20-53A contained a precise discussion of the waveforms to be tested against (e.g. “a 25kA/µs rate-of rise for at least 0.5 microseconds…’). The new “B” version of the AC strips out the specific waveforms to be tested against. “B” does not stand for “better.”

The FAA was asked why the specific waveforms – updated or left intact – were removed.

The FAA responded, in part, as follows:

“In the early 1990s, the SAE [Society of Automotive Engineers] Lightning Committee as well as the [FAA] Aviation Rulemaking Advisory Committee [ARAC] Electromagnetic Effects Harmonization Working Group noted that the lightning environment and test waveform definitions existed in several reports, including AC 20-53A … These differences caused conflict between the reports, so the Lightning Committee and the ARAC working group recommended to the FAA that the lightning environment and test waveform definitions be documented in a single report, and removed from the advisory circulars and SAE reports … So the … lightning committee prepared report ARP [Aerospace Recommended Practice] 5412 to document the lightning environment and test waveform definitions …

“With SAE ARP5412, only one document has to be revised, not four or five.”

In other words, the change was made for efficiency and consistency.

All well and good. But, ARP5412 (some 600 pages long) is only available to the public by purchase from SAE. Unless one is willing to pay, the standards in ARP5412 remain a mystery. Are they an improvement on the requirements formerly articulated within AC 20-53A? The FAA says the new standards are the same. This is hardly comforting. AC 20-53A was issued in 1985, before the use of composite materials in aircraft structures became as fashionable as it is today.

There may be a better way to make the requisite waveforms available to the public. Instead of a suspiciously incestuous and secretive SAE document, the FAA could publish a single technical standard order (TSO) outlining the lightning protection standards to be applied. The FAA has published numerous TSOs — for airport lighting, for terrain warning systems, etc.

A TSO for lightning protection could easily be published, available for look-up online at the FAA’s website.

The critical issue here is not that the standards have been consolidated into one document. The issue is that that document is published by SAE. This organization is not the regulator and therefore is not responsible for the currency of the test standards and is not accountable to the public, as the FAA would be if the requisite lightning protection waveforms were in an FAA-published TSO.

The difference is one of government standards lurking in the shadows or open and available for all to see.

The two programs are essential and complementary safety building blocks. As such, one would think they would be required as part of an airline’s FAA-issued operating certificate. They aren’t; why they are not is a separate discussion. For the moment, both are voluntary safety programs.

In its January 2010 report, “Answering the Call to Action on Airline Safety and Pilot Training,” on page 14 of this 200+ page report, we see the 94% figure touted by the FAA (remarks in parentheses are my comments):

“• Twenty-two operators currently have both FOQA and ASAP programs in place (Note: this is 22 of 98 Part 121 scheduled airlines, or slightly fewer than 25%, which is considerably less enthusiastic than the FAA’s 94%).

• Twenty-eight operators have ASAP and state their intention to implement FOQA (Note: not now, but sometime in the future).

• Ten operators with ASAP did not state their intention to implement FOQA. (Note: we can presume they are not).

• Nineteen operators who have neither program at present stated that they will establish one or both. (Note use of weasel words).

• Three operators stated that they do not intend to establish either FOQA or ASAP (Thereby making a mockery of the FAA’s oft-proclaimed “one level of safety” by not planning to have in place precisely those programs needed to routinely evaluate the safety of flight operations. Note also that the responses total up to just 82 of 98 carriers asked to provide responses to the FAA).

“To summarize, the responses show that:

• Ninety-eight percent of aircraft operating under Part 121 are flown by operators that have, or intend to implement, ASAP (Note: this is based on 82 airliners queried, but the percentage under the most generous interpretation is closer to 80% given that there are 98 carriers. Note also that the metric has subtly shifted from operators to the percentage of aircraft).

• Ninety-four percent of aircraft operating under Part 121 are flown by operators that have, or intend to implement, both programs (When will these ASAP/FOQA programs be in place? Not stated; could be in 5 or 10+ years).”

A typical FAA report; hyperbole, not clarity.

It’s important to note that this “action” plan was launched after the National Transportation Safety Board (NTSB) last spring identified endemic, industry wide safety deficiencies that were not limited to the circumstances surrounding the fatal crash of Colgan Air flight 3407, a Dash 8-Q400, in February 2009.

On 2 February, the NTSB held its final hearing on the crash investigation, in which some two dozen recommendations were issued on airplane design, crew training, and safety programs needed at carriers across the industry. The FAA’s “Call to Action” report issued on the eve of this hearing, metaphorically speaking, was an attempt to blunt criticism of lax FAA oversight.

It didn’t work. NTSB Member Robert Sumwalt groused, “I don’t know where this 94% came from, and I don’t believe it.”

Colgan Air was a so-called regional carrier, operating short hop flights for Continental Airlines.

Sumwalt said, “Only 2 of 27 regional carriers have FOQA programs.”

That’s a scant 7%.

Sumwalt went on, “Even if it was 98% [instead of 94%], the 2% [not participating] would mean the bottom feeders” who might most benefit from FOQA programs.

NTSB Chairman Deborah Hersman recalled, “Colgan last May said it would have FOQA by last July.”

Still no FOQA at Colgan.

NTSB staffer Roger Cox noted that the Dash 8-Q400 twin-turboprop “has no FOQA data being acquired in the U.S.”

The airplane is operated in Europe, but FOQA data acquired there is not shared here.

The 94% figure pushed by the FAA takes two separate measures – those that already have and those that intend to have ASAP/FOQA – and lumps them together. This creative addition yields a comforting situation that does not apply today and who knows when it will apply in the future?

Better to have started with all airlines subject to FAA oversight, then listed those that have ASAP/FOQA and the number of planes operated by each carrier. Then do the same for those carriers without ASAP/FOQA who plan to implement in the next two years. And, finally, those carriers who are more than 2 years away or not planning ASAP/FOQA programs at all.

From these simple, straightforward numbers, one can tabulate far more meaningful rates of participation. The present tortured but conveniently obscure method reminds one of the aphorism, “You can drink freely from a clear stream and suspect poison in muddy water.”

The FAA has dumped a heap of mud into the public waters. Inevitably, questions would be raised about the credibility of the agency’s rosy 94% estimate.

The document in question is Advisory Circular (AC) 20-53B, “Protection of Aircraft Fuel Systems Against Fuel Vapor Caused by Lightning.” This document is especially relevant to the B787. The airplane is the first airliner to be built primarily out of composite materials, in this case carbon fiber-reinforced polymer (CFRP). In a nod to the airplane’s vulnerability to lightning strikes, and their potential to ignite flammable fuel-air vapors in fuel tanks, Boeing has decided to provide an inerting system, whereby a nitrogen-enriched flow of air will be pumped into the void spaces of the B787’s fuel tanks to render them less susceptible to lightning energy penetrating the tanks and triggering an explosion that could blow a wing off.

Think about this for a minute. Boeing is using CFRP to save weight over traditional aluminum construction, thereby (hopefully) making the airplane 17% more fuel efficient compared to the B767 – which the new B787 will replace in the product line up. Yet the danger of lightning is so serious that Boeing is adding the weight of an inerting system to the new airplane.

This is because lightning behaves differently against a composite structure. Against traditional aluminum, lightning has a tendency to play along the surface of the wings and fuselage.

Lightning strikes a conventional aluminum jetliner shortly after takeoff and is diverted from interior mechanical, electronic and human contents.

Against CFRP, lightning will penetrate through the skin and play havoc with vital mechanical, electronic and fuel systems. Boeing is doing a number of things to minimize this vulnerability on the B787. In addition to inerting the fuel tanks, it is embedding a fine metal mesh in the composite. The purpose is to disburse the lightning strike around the airframe to prevent concentrated damage.

Gap are a problem. A slight gap between a wing-skin fastener and the hole it goes into could be a source of sparking as current jumps the gap. Boeing will install each fastener precisely and seal it on the inside to ensure a snug, spark-free fit. There are 40,000 fasteners on the B787, and they must be perfect for the life of the airplane to assure protection against lighting.

Lightning will penetrate even the slightest gap where fasteners are used.

Any gap inside the wings, where the wing skin meets internal structural spars, could cause a spraying out of electrons in a lightning strike. This phenomenon is called “edge glow.” Boeing will seal all edges with a nonconducting goop.

The question is what are the energy levels, intensities and durations or waveforms that must be protected against?

The 2006 AC doesn’t say. This is significant for three reasons: 1) this AC was issued by the FAA at a time when the B787 program was starting in earnest; 2) the 2006 AC replaced AC 20-53A, issued in 1985, that contained a useful and precise discussion of the waveforms to be tested against (e.g., “a 25kA/µs rate-of-rise for at least 0.5 microseconds …”) that is missing from the new AC; and 3) the British Air Accidents Investigation Branch (AAIB) recommended after a glider accident from lightning that the certification standards in AC 20-53A need to be upgraded because they’re not adequate for protecting airplanes, “Particularly those which utilize significant amounts of composite material in their primary and control structures.”

The new AC has stripped out the energy levels, referring the searcher instead to a third party document, the Society of Automotive Engineers (SAE) Aerospace Recommended Practice (ARP) No. 5412A. The AC says this document “offers methods you can use in determining lightning strike zones and the aircraft lightning environment.”

These technical requirements could easily have been included in an annex to the new AC. But this course wasn’t followed, and the technical requirements are spelled out in ARP 5412A instead. This document is not available to the public. One must be a registered member of SAE and a fee is required for access to ARP 5412A.

Thus, a simple, straightforward comparison of the lightning protection energies in the original AC and the current edition is not possible. One should also note that compliance with ARP 5412A is not essential; whatever standards it contains should be considered guidelines. One is left to speculate:

– The new standards are the same inadequate (to the AAIB) standards of yore (AC 20-53A), only now they’re concealed in a third party document.

– The new requirements relax those of yore, on the assumption that fuel tank inerting, with its nitrogen enriched air filling the void space in the tanks will make up for the decreased protection against lightning by limiting the chance that a bolt of electricity will ignite any fuel-air mixture. If this is so, then the “belt and suspenders” approach to safety is out the window.

– The new standards are improved, but not to the extent called for by the AAIB.

– The new standards are upgraded 8 or 9 fold, as called for by the AAIB, but this development is not easily discerned because the new waveforms are in an SAE document (which, by the way, the industry is not obligated to follow).

Even though lightning protection is supposedly improved on the B787 through internal metal wire or strips to divert the energy, through inerting the fuel tanks, through improved fasteners and such, the average passenger has no way of knowing if the airplane is more resistant to lightning strikes. The waveforms that must be protected against are the critical measure – and the FAA is now effectively treating them as a state secret, held by a trusted third party for use by insiders. The FAA is not practicing openness, and as the title of this essay suggests: hiding requirements = suspicion they’re inadequate.

A few days ago I asked my two nephews how they were spending the last few days of their holiday vacation. The response was an enthusiastic, “snowboarding!” Of course, as you might imagine, my next question was, “Are you wearing your helmets?” They looked at each other and (with a somewhat quizzical and mildly guilty tone) responded, “No” — but quickly added, “We wear them when we ride our bikes!” I almost knew what their answer would be and was quick to explain how important it is to protect your head and brain when engaging in any activity that puts you at risk for falls or a bump on the head. Wearing a helmet is not just for bike rides in the summer. While I wasn’t aware that they received snowboards for Christmas, I was especially concerned about safety in light of the recent news of Kevin Pearce, the champion snowboarder who is hospitalized in critical condition after sustaining a severe traumatic brain injury while preparing for the qualifying events for the upcoming Olympics in Vancouver. Kevin is obviously a well trained and well equipped athlete, and accidents happen — especially at the professional level where the risks are much greater. According to the Brain Injury Association of America, falls are the number one cause of traumatic brain injuries in the U.S., and helmets are 85% effective in reducing the risk of brain injury. It is important to remember, whether engaging in professional sports or just having some recreational fun, we must take every precaution to protect our brains.

Snowboarding, skiing, sledding, — these and many other winter activities offer the well trained, as well as the casual participant, a great deal of fun and enjoyment. Remembering to be well equipped with the right protective gear, including a properly fitted helmet, is a must if you want to minimize your risk of injury. My nephews knew that it’s important to wear a helmet when riding a bike — that message came through loud and clear. Let’s continue to spread the word and reinforce the message that any activity that puts your brain at risk, no matter what the season or sport, is worth “gearing up” for!

For more information and resources on brain injury prevention, please visit the following websites: Brain Injury Association of Illinois, www.biail.org., Brain Injury Association of America, www.biausa.org, and ThinkFirst, www.thinkfirst.org.

First flight of Boeing's B787 in December 2009, with employees in foreground cheering the event.

It may be useful to recount an A to Z of failures of design and failures in service on other airplanes. To be sure, there is an unsubtle distinction between systems that are prone to failure and those that are of a faulty design. And some failures are, indisputably, the result of systems that are simply both.

The message here is that history is replete with shortcomings on airplanes whose design was not nearly as advanced as the 787. Here’s a dear hope that Boeing and the FAA thoroughly wring out the 787 before the airplane, two years behind its original schedule, is pressed into widespread use by the airlines.

Herewith, a forget-me-not of fatal flaws:

A. Attitude sources (instruments) scattered around the cockpit with no readily available visual correlation with each other. The hazard here is that when a pilot’s attitude instrument fails, the resulting unusual attitude needs to be very quickly compared to two other widely scattered instruments on which the display attitudes are also rapidly changing. In other words, correlation is soon impossible and the situation quickly goes beyond recoverable limits. Main and standby attitude displays should be twinned (i.e., placed close to one another so that non-harmonious movements will become immediately apparent).

B. Circuit breakers (CBs) that were never designed as switches necessarily being used as such by both flight crew and mechanics. Continual tripping/reset of CBs changes their critical tripping threshold and can ultimately prevent CBs from performing their designed crucial function – tripping as a result of thermal overload, resulting in no circuit protection and a system failure or fire.

C. Aromatic polyimide wiring, which has the fault embedded within its name. Known generically as Kapton wiring, the aromatic moniker describes the fact that the insulation dries out with age, becomes hydroscopic (water loving), cracks and takes up water, becoming increasingly vulnerable to arcing and flash-over. The suitability of Kapton for aeronautical applications was never properly tested. Once in general service, it was too expensive to replace, except when it began knocking out or bringing down large numbers of military aircraft, thereby affecting defense preparedness. Although Kapton wiring is no longer being installed in its naked form, it is still being used with a Teflon outer coating (e.g., TKT wiring insulation).

Kapton wiring arcing.

D. Leading edge de-icer boots that allow ridges of ice to build up behind the boot, affecting the basic aerodynamics of the wing and the ailerons.

E. Turboprop propellers that rotate in the same direction on port and starboard wings. The complication is that in heavy icing conditions the asymmetric build-up of ice on wings, empennage and fuselage leads to a premature “one wing first” stall, followed by rapidly a rapidly unrecoverable autorotation.

F. Flawed flap actuation on the whole range of CRJs (commuter regional jets) – a design feature that’s almost impossible to rectify. The Transportation Safety Board (TSB) of Canada said:

“(D)espite best efforts by the industry and regulators alike to reduce the number of flap failures in the CRJ fleet, that number is increasing. A CRJ flap failure clearly has the potential to lead to a much more serious incident or an accident.”

G. Forward-facing inlets on tail-mounted APUs (auxiliary power units) on a range of jets, making the APUs prone to ingestion of anti-icing fluids.

H. Transponders that don’t alert sufficiently (aurally and visually) upon inadvertent switch-off that can also disable TCAS (Traffic Alert Collision Avoidance System) and lead to midair collisions.

I. “Triply redundant” critical systems (e.g., ADIRU’s, or Air Data Inertial Reference System) based upon multiple simultaneous raw data feeds that are vulnerable to simultaneous disabling (such as the ubiquitous three pitot heads and internal ice blockage when pitot heaters are overcome by supercooled droplets during protracted cruising in clouds).

J. Auto-throttles that rely upon singular radar altimeters, which in some intermittent failure modes can close the throttles to idle without generating any alerts, per the Turkish Airlines B737G crash in Amsterdam.

K. Automated flight systems that can auto level off, but not increase power via an autothrottle (the Buffalo low-altitude stall crash of a DHC8-400).

L. Throttle levers that do not move and alerts that do not continue (RETARD calls) when a lethal situation develops (Congonhas A320 pilot left the engine with the deactivated reverser up in the flight range after touchdown and reverse selection.

M. The failure to provide leading edge devices on the wings of the entire early range of CRJ’s and Challenger business jets, leading to numerous take-off accidents due to light icing/hoar frost on supercritical wing sections.

N. Confusing ON/OFF oxygen valves and insufficient warning devices of this; pressurization failure/failure to pressurize led to a number of hypoxia death crashes (e.g., Payne Stewart Learjet crash).

O. Use of an identical warning horn to signify takeoff configuration warnings on the ground and pressurization warnings airborne (Helios B737 accident, Athens).

P. Failure to assess the flammability risk of ullage fumes in center-section fuel tanks located above heat-producing air conditioning packs (TWA Flight 800).

Q. Failure to provide pulsing highly-directional deterrents (strobes) for birds along projected flight paths, which would prompt the flocks to take “dive and avoid” escape measures. Such directional strobes should be mandatory for twin-jets on climb – and on descent/early approach – below 20,000 feet.

R. Failure to adopt fail-safe jackscrew designs for critical applications (e.g., non-redundant structural assemblies such as the MD-80 T-tail’s horizontal stabilizer, the Alaska Air crash scenario). For that matter, the commercial industry is far behind the U.S. Air Force, which implemented automatic lubrication, different long-lasting allows, and protective sleeves to avoid contamination on horizontal stabilizer jackscrews found on jet transports.

S. Certification of rudder controls that are either susceptible to unwanted reversals (B737) or are overly sensitive to rudder inputs (A300). Together, these two flaws have killed hundreds.

T. Decades of certifying aircraft insulation blankets for fire resistance that actually burn (FAA had used Q-tips soaked with alcohol and only on samples laid flat). Alcohol is a relatively “cool” burning fuel, favored by magicians for flame effects. No tests were performed for vertical mounted insulation or the arcing effects of wiring faults under much hotter temperatures.

U. Failure to provide “stay awake alerts” for two pilot crews (dead-man buttons that must be pressed at least every 20 minutes or sound an alarm, a seat vibrator and a flight attendant alert). This feature is even more important now that cockpit doors are locked and must be released from the inside.

V. Failure to provide a software enabled credibility check to stop fatigued pilots inadvertently entering Zero Fuel Weight into their flight management computer, thereby causing low powered/premature rotation take-off accidents.

W. Failure to provide take off acceleration warning systems to guard against inadequate reduced power take-offs leading to take-off overruns (B747F at Halifax).

X. Failure to guard against leading-edge slat components punching holes in wing fuel tanks at the leading edge (e.g., the China Airlines B737-800 burn-out at Narita).

Y. It should be impossible for an aircraft to remain pressurized on the ground and prevent escape from a burning fuselage (Riyadh L-1011/Chicago DC-10 with electrical failure).

Z. It should be impossible to interchange potentially lethal parts between different marks of the same model (e.g., ATR-42/ATR-72 fuel quantity indicators).

One could go on, probably doubling this list. The point is that aircraft designs have gotten into production with shortcomings that should have been identified – and addressed – during certification. Entirely too much is “fixed” after the fact by the airworthiness directive process, when hundreds of aircraft are already in service.

The small cargo plane was registered to Michigan-based Royal Air Cargo and was empty at the time of the accident. The flight left Waterford, Michigan around 1: 00 p.m. and had been hired to pick up a load in Wheeling, Illinois later that afternoon.

Emergency personnel arrived on the scene shortly after the accident and had to travel on-foot to reach the wreckage, which was submerged in about four feet of water. Authorities from several neighboring communities — Wheeling, Mount Prospect and others — sent crews to the scene of the crash. Members of the National Transportation Safety Board arrived around 4 p.m. and said the investigation would begin Wednesday morning.

Royal Air is a family run business which owns and operates both passenger aircraft and cargo planes. They have approximately two dozen aircraft and are no stranger to regulatory scrutiny, accidents and operational violations. In 1999, a Royal Air aircraft was involved in a crash in Pittsfield, Mass., which has some similarities to Tuesday’s accident. On March 25, 1999, a Royal Air plane plummeted almost 12,000 feet in less than a minute before hitting the ground. In both accidents there were sudden losses of communication just before the planes crashed. Pilot Brian Templeton, of Waterford, Mich., was killed in the 1999 accident.

A lawsuit related to the 1999 accident accident was filed by Nolan Law Group on behalf of the family of pilot Brian Templeton. According to the lawsuit, Royal Air Freight.was negligent in performing maintenance on the aircraft, autopilot and de-icing system and in supplying information to support an alternate means of compliance for an MU-2 Airworthiness Directive. Other Defendants include Mitsubishi Heavy Industries, Honeywell and Mid-Continent Instruments.

Royal Air was also sued by federal authorities in 1999 for cutting corners on engine maintenance and inspections. Violations listed in the lawsuit included failure to conduct scheduled inspections of engines, propellers and wing flaps and failure to produce maintenance records. The company ultimately agreed to pay $250,000 in fines for maintenance and record-keeping violations as part of an agreement with the U.S. attorney for the Eastern District of Michigan. Less than a year later, the FAA proposed $60,000 in additional fines against Royal Air for allegedly failing to investigate the backgrounds of 13 newly hired pilots.

2009 has proven an interesting year for airline pilots and the flying public. In January, we witnessed the heroism of Captain Sullenberger averting disaster and gracefully landing US Airways Flight 1545 in the Hudson River. Cockpit voice recordings reveal a calm and measured reaction to a bird strike, as well as a calculated decision to land the plane in the Hudson. His professionalism, training, experience and judgment prepared him to successfully and artfully land a plane under trying circumstances.

A mere month later, Continental Air Flight 3407, operated by Colgan Air, crashed into a house during approach near Buffalo, NY, killing all 49 passengers and crew as well as one person the ground. Unlike Captain Sully, the pilots operating this regional flight were sleep deprived, sick, distracted and flying in inclement weather. They lacked sufficient training and resources, and were thus unqualified to be flying a plane under those circumstances.

In October, two Northwest pilots missed their destination by over 150 miles and failed to respond to air traffic controller attempts to reach them. The pilots claimed they “lost situational awareness” because they were distracted, reviewing a new company policy on a laptop. Speculation surrounding this incident has focused heavily on the theory that the pilots were in fact sleeping, again highlighting the issue of pilot fatigue.

Which brings me to the events of Tuesday night in Jamaica. The facts as they unfold have many similarities – both from an operational standpoint, as well as the aircraft type and runway environment – to Southwest Flight 1248 overran its runway in December 2005. In the Southwest accident investigation, the NTSB looked at factors such as decision to land, calculation of landing distance on a contaminated runway, company braking procedures, as well as pilot training.

Reports indicate that Tuesday’s flight in Jamaica had sufficient fuel to return to Miami, yet decided to land on the contaminated runway rather than turn around. The pilots were near timing out for their flight hours for the day, which raises the possibility of pilot fatigue impacting their decision-making process and their operation of the aircraft. Was the decision to land made based on the safety of the passengers or – considering the pressure of holiday travel, passenger frustration, pilot fatigue and cost – did the pilots decide that the safety risk was worth it?

The numerous accidents and incidents of 2009 raise serious questions about what is going on in the cockpit. The over arching question is a serious one: during these economic times, is aviation industry creating a culture of undervaluing risk to save money?

Make no mistake, there are numerous technical issues that may have contributed to the scenario that unfolded on Tuesday night, as well as the lack of preventative measures that could have mitigated damages. Moreover, the risk of human error is everpresent, and for that reason we must advocate also for additional safety measure that minimize the impact of such errors. Nonetheless, the events and mistakes outlined above are not discrete individual incidents; rather, they are evidence of a deteriorating safety culture. We are entrusting the safety of passengers to tired, overworked, and often under paid pilots who have insufficient training and distractions in the cockpit. Congress must act to ensure that the business interests of airlines do not outweigh the safety of our passengers. In 2009, Captain Sullenberger’s “Miracle on the Hudson” was an exception in a year fraught with serious safety hazards. But the reality is, he was not lucky – he was prepared. In 2010, let’s make his example the rule.

The airplane’s wheels touched down, but speed was not appreciably abated and the B737-800 with 154 passengers and crew aboard roared off the end of the runway, ripped through a perimeter fence, split apart and came to a halt just a few feet from the ocean surf.

Ninety people were taken to hospitals with broken bones, contusions and bruises; four people were hurt seriously, although none of the injuries was considered life threatening. Injuries to crew are unknown.

Ten years ago, on 1 June 1999, the captain of an American Airlines MD-82 twin jet was killed in similar circumstances landing in a thunderstorm at Little Rock, Arkansas. The first officer was badly injured, as were a dozen or so passengers, some of whom escaped the wreckage though splits in the fuselage as the airplane ground to a halt just short of the Arkansas River.

The two cases, then and now, raise a simple question: what has the aviation industry learned in the intervening decade to prevent a repeat?

Kingston, Jamaica, 2009 -- going off the end of a wet runway and stopping just short of the ocean.

Little Rock, Arkansas, 1999 -- roaring off the end of a wet runway and stopping just short of the Arkansas River.

The following issues will doubtless be examined by the Jamaica Civil Aviation Authority, assisted by a team from the U.S. National Transportation Safety Board (some of whom, hopefully, were participants in the Little Rock crash investigation – their memory of the particulars of that crash will be especially pertinent):

– The role of the dispatcher. Every airline flight is a collaborative effort between the cockpit crew and the dispatcher at the airline’s flight operations center. Was this flight dispatched into an area of known thunderstorm activity? At certain U.S. airlines, it is policy to avoid thunderstorms entirely. The lack of such policy at American Airlines in 1999 had much to do with the set-up for the accident. If the dispatcher recommended delayed landing until the thunderstorm passed – say about 20 minutes – the accident might have been avoided as well. And of course there is the subject of alternate airports if the field at Kingston, Jamaica, was not usable because of severe thunderstorms.

– The utility of weather radar in the cockpit. In the 1999 crash investigation, questions were raised about the adequacy of the flight crew’s knowledge and training in the use of the airplane’s on board weather radar. The practice of “avoiding the red” on the radar display in the Little Rock crash was clearly insufficient. The areas depicted in red may show the most intense rainfall, but not necessarily the most turbulent convective activity.

– Time of day. As at Little Rock, the crash at Jamaica was late at night. Not only is poor visibility a factor, but also the matter of crew fatigue. At Little Rock, the captain was clearly fatigued, as evidenced from the cockpit voice recorder (CVR) indicating his non-responsiveness to the first officer (FO: “Want 40 flaps?” Captain: “Oh yeah, I thought I called it.”).

– Support from the tower. At Little Rock, the tower controller did not suspend operations, even though the airport was experiencing a Level 6 thunderstorm, considered “severe” on a six-point scale. The wind shear detectors at Little Rock were also not placed optimally for assessing the wind shear hazard.

– Touchdown point. Exactly how long is the runway at Kingston and where did the B737-800 touch down? Reportedly, the runway is 8,786 feet long, and the B737-800 landed long — about halfway down, due in part to a tailwind component. At Little Rock, the airplane touched down long on the runway. Given the circumstances of aircraft configuration, the airplane did not have enough paved length remaining to stop, no matter what procedures were employed by the crew.

– Configuration. At Little Rock, the crew had failed to arm the spoilers, so they did not deploy on landing. The spoilers kill lift and put maximum weight on the main landing gear. With about 95% of the airplane’s weight on the main landing gears, braking effectiveness is optimal. This wasn’t the case at Little Rock, where only 10% of the airplane’s 127,000 lb landing weight was on the wheels, and the airplane hydroplaned down the runway. Hydroplaning also occurred at Kingston, from early accounts.

– Reverse thrust. Did the crew at Kingston apply maximum reverse thrust to aid in deceleration? The flight data recorder (FDR) should reveal much. At Little Rock, the crew did not apply maximum reverse thrust. With a late touchdown, no spoilers, insufficient reverse thrust and delayed braking, plus hydroplaning, the Little Rock aircraft roared off the end of the runway at more than 100 mph. One passenger on the Kingston jet recalled, “The airplane did not seem to be slowing down when it landed. There was a loud sound, then a huge thud, and then we started to feel rain through the top.”

– Cabin safety. When the airplane landed hard at Kingston, overhead bins popped open, spilling carry-on bags and articles. The spillage contributed to the general sense of passenger confusion and frustration during the subsequent emergency evacuation. This bin spillage was a problem in the Little Rock crash, too. The flimsy overhead bin doors were a problem in 1999, and they’re still problematic in 2009, when passengers are carrying even more items aboard the aircraft. The locking mechanism on the doors needs to be strengthened. The manner in which ceiling panels are affixed also is an issue.

– What safety area? Every runway is supposed to have 1,000 feet of unpaved ground as a safety area off the end of the runway. This was not the case at Little Rock, where the Arkansas River floodplain was right off the end of the runway. The airplane hit runway lighting poles, which were not frangible, badly damaging the airplane but the poles helped stop it short of the water. At Kingston, the airplane was damaged by dunes, which had the effect of stopping it just 10 feet short of the ocean. At Little Rock, after the accident the airport authority installed an Engineered Materials Arresting System (EMAS), a porous form of concrete, to assist in stopping airplanes that cannot stop on the paved runway. The wheels bury themselves in the porous material and the airplane is quickly brought to a halt (without damaging the airplane itself and with only minor degradation to the landing gear). It is interesting that the Little Rock installation was hastily completed after the accident. Had EMAS been installed at Kingston, the airplane would have been stopped before plowing into dunes and splitting open, damaging the almost-new B737-800 beyond repair.

– The response. At Little Rock, the airport rescue and fire fighter (ARFF) response was delayed because of uncertainty about the crashed aircraft’s location. The fire fighters originally went to the wrong end of the runway. The Kingston airport ARFF response, however timely or not, will surely get close scrutiny. As at Little Rock, ARFF personnel were trying to locate a wrecked aircraft on a dark and stormy night.

So once again investigators will have issues to address that were not fully resolved in the ten years after the 1999 accident. And that raises an ugly question: why is the industry so tardy on correcting obvious, basic safety deficiencies? It could well take ten additional years to redress similar shortcomings revealed by the crash at Jamaica.

American’s timetable shows flight AA331’s scheduled arrival time is 21:10, but the carrier says the aircraft landed at 21:22CST, equating to 22:22 local.

Meteorological information from Norman Manley International Airport indicated heavy rain and possible thunderstorm activity at this time.

The airport has a single runway, designated 12/30, which has a length of 2,716m (8,910ft) but its virtually-offshore location – on a thin strip of land south of Jamaica – leaves little overrun margin at either end.

There is no confirmation of which runway the aircraft was using. While there is an instrument landing system for runway 12, the weather data indicates that this would have required landing with a tail wind.

NOTAM information, dated today, shows that the airport has restated the runway distances available to aircraft, and introduced a displaced threshold on runway 30.

American states that two of the 148 passengers were admitted to hospital for observation, but all others have been released. The jet, arriving from Miami, was also carrying a crew of six.

Damage to the 737 is substantial. Its fuselage has fractured aft of the wing, its right-hand CFM International CFM56 engine has separated and the left wing-tip has snapped.

By David Kaminski-Morrow

Workers sift through debris surrounding the fuselage of American Airlines flight AA331 which crash landed overnight on a flight from Miami to Jamaica, just beyond the runway of Norman Manley International Airport, in Kingston Jamaica, Wednesday, Dec. 23, 2009. More than 40 people were injured, at least 4 seriously, and there were no fatalities, according to officials, after the plane overshot the runway in Jamaica when it landed in heavy rain

Panicked passengers screamed and baggage burst from overhead bins as Flight 331 from Miami careened down the runway in the capital, Kingston, on Tuesday night, one passenger said.

The impact cracked the fuselage, crushed the left landing gear and separated both engines from the Boeing 737-800, airline spokesman Tim Smith said.

Crews evacuated dazed and bloodied passengers onto a beach from a cabin that smelled of smoke and jet fuel, passengers said. Rain poured through the plane’s broken roof, one said.

Some 44 people were taken to hospitals with broken bones and back pains and four were seriously hurt, airport and Jamaican government officials said. American Airlines said two people were admitted to the hospital and nobody suffered life-threatening injuries.

Heavy turbulence on the way to Jamaica had forced the crew to halt the beverage service three times before giving up, Pilar Abaurrea of Keene, New Hampshire, told The Associated Press by phone. The pilot warned of more turbulence just before landing but said it likely wouldn’t be much worse, she said.

“All of a sudden, when it hit the ground, the plane was kind of bouncing. Someone said the plane was skidding and there was panic,” she said.

U.S. investigators will analyze whether the plane should have been landing in such bad weather, Smith said, adding that other planes had landed safely in the heavy rain.

Passenger Natalie Morales Hendricks told NBC’s “Today” that the plane began to skid upon landing and “before I knew it, everything was black and we were crashing.”

“Everybody’s overhead baggage started to fall. Literally, it was like being in a car accident. People were screaming, I was screaming,” she said.

“There was smoke and debris everywhere,” after the plane halted, she said. “It was a mess. Everybody could smell jet fuel.”

Passenger Robert Mais told The Gleaner newspaper of Jamaica that he had heard the engine’s reverse throttle but that the plane didn’t seem to slow as it skittered down the runway.

The plane came to a halt about 10 to 15 feet (3 to 5 meters) from the Caribbean Sea and passengers walked along the beach to be picked up by a bus, Mais said. Rain came through the roof of the darkened jet and baggage from the overhead compartments was strewn about the cabin, he said.

The plane originated at Reagan National Airport in Washington and took off from Miami International Airport at 8:52 p.m. and arrived in Kingston at 10:22 p.m. It was carrying 148 passengers and a crew of six, American said. The majority of those aboard were Jamaicans coming home for Christmas, Jamaican Information Minister Daryl Vaz said.

Smith said there were two “significant” cracks in the fuselage, and the engines are designed to separate from the wings during an accident as a safety measure.

A team of six investigators from the National Transportation Safety Board was traveling to Jamaica from Washington on Wednesday morning to assist a probe led by the island’s government, agency spokesman Keith Holloway said.

The airport reopened early Wednesday after officials had delayed flights because of concerns that the plane’s tail might be hindering visibility.

Four hundred passengers waited for their flights to be cleared for takeoff, Security Minister Dwight Nelson told Radio Jamaica.

Heavy rains that have pelted Jamaica’s eastern region for four days are expected to dissipate by Thursday. Authorities said the rains washed away a 7-year-old girl on Tuesday and led to a bus crash in which two people died.

By KIRK WRIGHT, Associated Press

Associated Press Writers Danica Coto and Ben Fox in San Juan, Puerto Rico; Howard Campbell in Kingston, Jamaica; Carol Druga in Atlanta, Georgia; and Sofia Mannos in Washington contributed to this report.