Aftermath: The Mystery of Colgan Air 3407

[SteveC]

On the CRJ we have to be at less than 60% N1 by 75 knots and idle thrust (30% N1) at 60 knots. Hence we used to call out 80 knots on landing as a cue to start stowing the reversers and then 60 knots to ensure they are stowed.

I've got some guys that I fly with that think that "idle thrust by XX%" means that they have to have them stowed by then. I always figured that as long as I was at idle thrust I could safely leave the buckets out all the way down to the point where I would have to start adding forward thrust just to keep us moving. I've seen enough comments throughout this thread that make me wonder which way is really correct.

 

[BobDDuck]

Per our book we can taxi around with the reversers deployed all day long if we want, as long as they aren't above idle thrust (30% N1). The only limitation pertains to avoiding operating them while on contaminated surfaces as they can kick up stuff onto the wings.

The expanded landing procedures section of the POH states that

Thrust reversers are most effective at high speed. At low speed, minimize the intensity and duration of reverse thrust. However, in an emergency situation, maximum reverse thrust may be used to a complete stop

The only place I've been told to actually stow them at 60 knots is in training and even then it's not written down anywhere. It was just part of the training flow shown to us from day 1. At 80 knots (now 90 knots) start moving the reversers to idle and then at 60 knots, stow them.

On a unrelated note, for a while we had a bunch of Jets For Jobs guys who insisted on popping the inside reverser when they were making turns while taxing to "assist" the nose wheel steering in turning the plane. I dunno if maybe there is some jet floating around out there where that actually works, or these guys were just "different" but that never made any sense to me.

EDIT:

On the 200 when you switch from idle reverse to idle forward thrust, the plane actually surges forward a little bit. In the 700 (and I'd guess 900 and 1000) this doesn't happen.

 

[SlumTodd_Millionaire]

"The crashes of a Singapore Airlines 747 in 2000 and of a regional jet at Lexington, Kentucky, in 2006 were due to pilots trying to take off from the wrong runway, a blunder so fundamental and seemingly obvious that no technical means are in place to guard against it."

I stopped reading the article after this point in the second paragraph, since the author is obviously clueless about technology that is already a decade old. Honeywell's RAAS software will alert the crew if they attempt to start a takeoff roll on a runway different than that programmed into the FMS.

What bothers me most about the Colgan accident is that it only turned out this way because the aircraft was lacking some basic technology that almost every modern airliner has: autothrottles. I couldn't have gotten myself into this situation in the B717 if I wanted to, because the airplane automatically kicks on the autothrottles and adds power to keep me from stalling, even if I turned the autothrottles off. Technology can save us from human error, which can never be eliminated. Personally, I'd like to see autothrottles with low-speed protection added as a basic requirement for Part 125 certification.

 

[BobDDuck]

Just for the sake of argument maybe they should take autothrottles out of all part 125 aircraft. The Turkish 737 probably wouldn't have crashed if it didn't have autothrottles.

 

[wrxpilot]

Just for the sake of argument maybe they should take autothrottles out of all part 125 aircraft. The Turkish 737 probably wouldn't have crashed if it didn't have autothrottles.

Exactly. I was just going to mention that accident.

Airdale is right... There should be no tolerance for poor airmanship. If somebody has a history of checkride failures and/or they clearly don't know what they're doing in an airplane, they need to be gone period.

 

[SlumTodd_Millionaire]

Just for the sake of argument maybe they should take autothrottles out of all part 125 aircraft. The Turkish 737 probably wouldn't have crashed if it didn't have autothrottles.

That's a bit of a stretch. The real problem in that case was a repeat maintenance discrepancy with the radio altimeter that had somehow gone unfixed over and over again. In addition, the flying pilot (the FO) initiated a go-around and applied max power, but the idiot captain took controls from him in the middle of his go-around and didn't bother to touch the throttles when he assumed controls. Had the FO been allowed to fully execute the go-around, everything would have turned out just fine, as simulator tests indicate that it was fully recoverable.

The Dutch investigators also determined that the autothrottle logic on the 737 is significantly different than other aircraft in how it deals with false radio altimeter signals. Boeing apparently knew about this, but didn't deem it a safety problem. Of course, because fixing it would cost dollars.

 

[Airdale]

Technology can save us from human error, which can never be eliminated. Personally, I'd like to see autothrottles with low-speed protection added as a basic requirement for Part 125 certification.

You're not doing pilots any justice for wanting higher pay here Todd.

Technology is certainly great, but should technology be a replacement for basic airmanship? I don't think so. Technology should reduce workload, not replace it. And it definitely shouldn't increase workload.

The autothrottles on the E170 earned my distrust when I was fairly new to the aircraft. Cleared for a visual into PHL. We were descending, and I had the autopilot in FLCH mode with autothrottles on. FLCH is a speed on elevator mode, so the throttles remain idle. I clicked the autopilot off to fly the visual, but left the throttle engaged and the aircraft in FLCH mode. Last time I do that again. I started to slow the descent rate up as we got on glide path and started calling for flaps. I had my hand on the throttles the whole time. As I started to level I instinctively wanted to add power, the throttles would have stayed idle the entire time because of FLCH mode. If I wasn't paying attention, we could have easily stalled, and the AT's would have let it.

Technology is great, but it can back you into a corner. I don't care what aircraft you fly, how much experience you have, what the weather is or how tired you are and how much technology is on board. First and foremost - fly the airplane. Something we as professionals, should do and do well everytime we operate. This is not the job for poor performance, because people can die. Just ask any family affected by 3407.

 

[grnclvrs]

So, the A/T did what you told them and you don't trust them?

 

[400A]

I've got some guys that I fly with that think that "idle thrust by XX%" means that they have to have them stowed by then. I always figured that as long as I was at idle thrust I could safely leave the buckets out all the way down to the point where I would have to start adding forward thrust just to keep us moving. I've seen enough comments throughout this thread that make me wonder which way is really correct.

I might have the "documentation" somewhere. The stowed by 60 knots is to prevent the hot exhaust air from being reingested by the engine. IIRC, the actual speed where this starts to happen is 55 knots.

The bizjet style airframes I have flown also have a time limit of TR deployment and that usually comes pretty close to the time (relatively) to the time it takes to deploy slow to 60 knots and get them stowed.

 

[SteveC]

I might have the "documentation" somewhere. The stowed by 60 knots is to prevent the hot exhaust air from being reingested by the engine. IIRC, the actual speed where this starts to happen is 55 knots.

The bizjet style airframes I have flown also have a time limit of TR deployment and that usually comes pretty close to the time (relatively) to the time it takes to deploy slow to 60 knots and get them stowed.

If you have the documentation I'd be curious to see it. I'm only rated in two aircraft, but both of them allow TR deployment at any speed, with the Citation requiring idle thrust only below 60 knots, the Lear limit is 40 knots below which you must be at idle.

From the Citation Bravo Pilot Training Manual:

Reverse thrust power must be reduced to idle reverse detent position at 60 KIAS on the landing roll.

Maximum allowable thrust reverser deployed time is 15 minutes in any one hour period.

Engine static ground operation is limited to idle power (if thrust reversers are deployed).

The use of thrust reversers to back the airplane is prohibited.

(I think that the 15 minute per hour limitation is hydraulic related.)

The Lear 40/45 limitations include:

Limited to ground ops on paved surfaces, do not attempt to deploy in flight.
No backing aircraft.
TR’s must not be used during touch and goes
Max reverse down to 40 KIAS
Limited to idle reverse with computer in manual mode

I could not find any limits on the amount of time that the TR's can be deployed on the Lear.

 

[400A]

I will have to dig, the study was a NASA thing and was pretty old, 1960ish.

The citation II had a 5 minute limit on the ground, I thought, but it has been a couple of years. IIRC it was a hydraulic thing, the fluid would get so hot and become corrosive. That was either a by-product of the type of fluid or the open center hydraulic system, I can't recall.

The BeechJet is limited to 30 seconds above idle reverse and only 5 seconds with the buckets open during taxi ops. 55 knots is the FULL reverse limit.

 

[C150J]

Dale,

First of all, the Captain failed 3 PC's? He also failed numerous checkrides before the airlines? I don't have the information on his failures, but was there a trend here?

Definitely a trend, and more critically, a failure to rectify it. That being said, I know plenty of people that failed GA checkrides and became great pilots, akin to your statement below:

People who in most missions could do just fine, but raise the stakes and raise the pressure and they completely fell apart. These people would show you signs early on of weakness. Some would develop beyond it and make excellent crew members, others not so.

This is where we need a much better system in place. Primary flight training should weed people out if they can't meet standards after REAL retraining (not just checking a box), much like UPT in the military. This retraining may need to come from another instructor, making sure that the applicant was trained properly by their original CFI.

I know people will defend the checkride failures, but I will not. I passed every FAA checkride and I can say, that they are not hard. I know some folks who got stuck with a tough fed and got a pink slip. No big deal. But if you can't pass checkrides without failing more than 2, I think further evaluation needs to be done.

This is where our opinions diverge a little. It all depends on where you trained and what "school" your DPE came from. Some believe that a failure is "character-building," as I was once told by an examiner. Others are mercenaries, only failing a percentage to stay under the radar. Some are cookie-cutter, whereas some are a complete curve ball. Our GA training and testing programs are not standardized, not thorough, and NOT working. This is why I really try to judge each pilot's history on a case-by-case basis. Am I going to watch the new-hire with multiple training failures like a hawk initially? Yes. Am I going to watch the 141 guy with no failures like a hawk? Yes. Unfortunately, our primary training is too subjective to put my trust in training histories.


I'm not disagreeing with you, but I just can't correlate busting primary checkrides with crashes, and neither can the FAA (check out its response to NTSB recommendation A-05). Now, 121/135 events are generally a whole different story, as training is much more uniform. If I interview a guy with multiple failures as a current professional pilot, I'm probably going to take a closer look.

 

[SlumTodd_Millionaire]

If I wasn't paying attention, we could have easily stalled, and the AT's would have let it.

I've never flown an EMB-170, so I'm not familiar with your systems. Sounds like your autothrottles don't have automatic low-speed protection, though. Notice what my post said: "Personally, I'd like to see autothrottles with low-speed protection added as a basic requirement for Part 125 certification."

Is it a replacement for basic airmanship? Of course not; but it is a safety backup for when pilots let their basic airmanship slip, possibly because of fatigue.

 

[Cruise]

Mixed-mode automation was absolutely a factor in this accident.

I didn't list it previously....but my list was just a few of the factors. Like I said, those were just a few of the factors involved. I could keep going.......

 

[Cruise]

Mixed-mode automation was absolutely a factor in this accident.

I didn't list it previously....but my list was just a few of the factors. Like I said, those were just a few of the factors involved. I could keep going.......

 

[amorris311]

I can vouch for cruise.

 

[WacoFan]

I've stayed out of 3407 threads because I have no experience...but I have read them and am deeply interested. I also like reading other accident reports and such.

Has there ever been an instance where fellow 121 pilots will simply say "Yeah...that was an egregious screw-up and shouldn't have happened at all...simply pilot error"? Eastern 401, or maybe the Columbian airplane that ran out of fuel around NY in the 80's comes to mind. It would seem, from a layman, that 3407 COULD be in that realm. I know there are a lot of factors so I don't feel that way - but at what point do you simply throw up your hands and say "Yeah, that was pilot error pure and simple".

I understand that there were many factors - automation, training, fatigue, and others. However, no other Colgan crew has done this and I would imagine that many other Colgan crews were more tired, in much worse weather, etc. So, my question is to people close to the investigaion - at what point to you simply have to place blame on the individual pilot? I think it is important to look at EVERYTHING else first - and other pilots are the only ones that will aggressively do that (Airline wants pilots blamed, Manufacturer wants pilots blamed, etc) but at what point do you throw up your hands and say "You know...there was like 80% pilot error here"?

Just curious more as to the mindset of those investigating - that is what interests me. It would seem to be a brutal job to investigate something like this and I'm really interested in some of the things you guys look at and go through.

 

[jynxyjoe]

I've stayed out of 3407 threads because I have no experience...but I have read them and am deeply interested. I also like reading other accident reports and such.

Has there ever been an instance where fellow 121 pilots will simply say "Yeah...that was an egregious screw-up and shouldn't have happened at all...simply pilot error"? Eastern 401, or maybe the Columbian airplane that ran out of fuel around NY in the 80's comes to mind. It would seem, from a layman, that 3407 COULD be in that realm. I know there are a lot of factors so I don't feel that way - but at what point do you simply throw up your hands and say "Yeah, that was pilot error pure and simple".

I understand that there were many factors - automation, training, fatigue, and others. However, no other Colgan crew has done this and I would imagine that many other Colgan crews were more tired, in much worse weather, etc. So, my question is to people close to the investigaion - at what point to you simply have to place blame on the individual pilot? I think it is important to look at EVERYTHING else first - and other pilots are the only ones that will aggressively do that (Airline wants pilots blamed, Manufacturer wants pilots blamed, etc) but at what point do you throw up your hands and say "You know...there was like 80% pilot error here"?

Just curious more as to the mindset of those investigating - that is what interests me. It would seem to be a brutal job to investigate something like this and I'm really interested in some of the things you guys look at and go through.

I really think it's important to note these pilots were literally drunk with fatigue.

I think the answer to your question will have to come from the final report Waco. You'd really need an NTSB investigator cornered for a few hours to find out.

Although I am a highly trained "armchair" Internet Aviation Accident Investigator, (a)IAAI, and I tell you all blame squarely resides with the company in all matters. It's always their fault and I'm always perfect.

Real experts say:

I believe that this accident investigation has refocused the FAA’s and the industry’s attention on the many issues that the Board has raised over the last several years: issues like flight crew monitoring, pilot performance, sterile cockpit violations, fatigue, training, record keeping, use of personal electronic devices, and safety alerts. These are critical issues that require the attention AND action of the FAA and industry.
[...]
If we want to raise the bar, we can’t simply require industry to meet a safety ‘floor.’

Those are closing remarks from Deb Hersman.

We do things very well, however there is room for improvement from all parties. Even pilots. That's her take, as head of the (a)IAAI I say she is wrong and the fault lies squarely with the company.

 

[Firebird2XC]

This accident was a mixed-cause event, in my opinion.

While yes, end pilot input to the airplane is what put the airplane in the ground, understanding the circumstantial causal factors underlying the scenario does apply.

Safety culture is exactly that- a 'culture' that is cultivated within the company itself, at large. This is why we standardize in operations, training, work rules, etc. Creating a consistent work environment with given performance expectations promotes safety both in events and in aircrew decision making.

NASA Ames researches have done extensive studies with roots as far back as 1980.

Excerpt From a 1999 NASA authored report:

Source: http://www.hq.nasa.gov/congress/mann8-3.html

[FONT=Arial,Helvetica]Throughout the course of this outstanding research program, it has been evident that pilot fatigue is a significant safety issue in aviation. Rather than simply being a mental state that can be willed away or overcome through motivation or discipline, fatigue is rooted in physiological mechanisms related to sleep, sleep loss, and circadian rhythms. These mechanisms are at work in flight crews no less than others who need to remain vigilant despite long duty days, transmeridien travel, and working at night when the body is programmed for sleep.
[/FONT]

[FONT=Arial,Helvetica]
[/FONT]

[FONT=Arial,Helvetica]Evidence regarding the existence and extent of fatigue in aviation has been gathered from several different sources and environments, including aviation operations, laboratory studies, high-fidelity simulations, and surveys. Studies have been consistent in showing that fatigue is an issue with complex, diverse causes and potentially critical consequences. Field studies specific to different aviation environments and using a range of measures (e.g., performance, physiology, and behavior) have revealed a number of factors related to fatigue. For example, in long-haul operations, the non-24-hr duty/rest cycles, the circadian desynchronization associated with transmeridien flights, and the sleep loss accompanying nighttime flying are all associated with fatigue (1). For short-haul operations, long duty days, sleep loss as a result of short nighttime layovers, and shortened sleep episodes due to progressively earlier report times across trips serve to create flight crew fatigue (2). In overnight cargo crews, even regular nighttime flying often results in incomplete circadian adaptation. Additionally, duty periods ending in the morning hours lead to sleep loss due to an increasing signal for wakefulness from the biological clock during this time. The problem is compounded by daytime layovers that can sometimes be too short for an adequate sleep opportunity (3). [/FONT]

Fatigue has been demonstrated to be intoxicating. Airlines are aware of this. Aircrews are aware of this. Unfortunately, once you're 'intoxicated' your judgement is impaired. Reacting to additional external pressures in life may cause you to err in judgement.

Judgment has long been cited as one of the most crucial skills- if not THE most crucial skills- an aviator can have. When placed into a work environment that increases the likelihood of eroded or incapacitated judgment, an external control must be applied in order to correct for things.

Given that the company has failed to exercise due diligence in created a work environment in which a penalty-free scenario exists for a pilot to exercise proper judgment when they question their ability, or, furthermore, to create an environment in which judgment is not so eroded to begin with, a certain degree of culpability rests with the 'company' (any company in this sort of scenario) for failure to adequately promote a given safety culture.

This information has literally been around for decades, and those with access to it and prior knowledge have thumbed their noses at a prudent course of action. Any aviation accident is the cause of a sequence of errors that creates the end crash sequence. If any one of those errors were to be corrected, yes, it would prevent that accident.

In this case, there were many scenarios in which the company and the air crew both could have broken the chain of errors. The company has been allowed to rely so long on the aircrews to not fail that they have failed to oversee the links in the chain they can and should be accountable for. End result? When the pilots fail, the company pays lip service to the event and their contributing factors and blame the pilots.

Yes, the pilots screwed up. Yes, the company put them in a work environment that failed to encourage better decisions to prevent screwing that up. In this particular case, the corporate culture was on that actually could be perceived as goading pilots to continue regardless of the situations in which they found themselves.

To quote a former Colgan pilot, the philosophy there was "Move the rig."

Fault exists on both sides of the equation. Responsibility for correcting those faults exists on both sides as well.