A very in-depth look at Colgan Air (Continental Connection) Flight 3407.
A Colgan Air (as Continental Conneciton) Dash 9 Q402 makes its landing approach to DCA in Washington DC. (Photo by Fernando Montalvo)
Almost a year ago (February 12, 2009), Colgan Air (aka Pinnacle Airlines, aka Continental Connection) Flight 3407 plowed into a Clearance Center, New York residence while on an instrument landing system (ILS) approach into Buffalo-Niagara International Airport (BUF) in Buffalo, NY. Fifty people were killed that night: two pilots, two flight attendants, forty-five passengers, and one person on the ground. While the Hudson River accident involving US Airways Flight 1549 would show how one flight crew was very cool under pressure, the Clearance Center incident would show the extreme other end of flight crew performance. A year later, we know not only that the flight crew practically created the accident scenario by not paying attention to their flight displays, but that they failed to recover from one of the most practiced situations in the aviation world.
On Tuesday (Feb. 2, 2010), the National Transportation and Safety Board (NTSB) held a public meeting to discuss the aircraft accident and release the findings and conclusions of the accident investigation. While the actual final report is not out as of this post’s publication, the hearing and supporting documents present most of what will be in the report (if not all of it). First, let us look at what happened that night so that we’re all on the same page (you can also look at Velozia Air’s original post on the accident HERE).
Continental Connection Flight 3407 was a flight operated by Pinnacle Airlines and, in turn, operated by Colgan Air. The aircraft for the flight was a Bombardier Dash 8 Q400 (DHC-8-400) with registration number N200WQ (The plane in the picture above is another Colgan Air Dash 8 Q400 operating as Continental Connection). The aircraft was on an ILS approach into BUF when things went all wrong. For those of you not in the aviation community, an ILS approach is a type of radio-signal guided landing approach in which ground equipment sends out signals picked up by on-board avionics which help the pilot determine where he is on an approach when visibility is limited. Since the ILS has nothing to do with the accident, I will leave the explanation on ILS there. While on approach to the airport, the flight crew had noticed that some ice was forming on the windshield and flight surfaces of the high-wing, twin engine aircraft. De-icing equipment (in this case rubber de-icing boots) was turned on since about 11 minutes into the flight to deal with most of the ice on the wings, but some was still building up (the de-icing boots can be seen as the black “line” on the leading edge of the wings in the picture above). As the aircraft continued to descend, the airplane’s airspeed continued to be decreased for landing, while the minimum speed at which the aircraft can fly (which is a somewhat rough definition of the term “stall speed”) increased a bit from some of the remaining ice in the wings. While ice on the wings is a term passengers relate to accidents, it is important to point out that a lot of flights deal with ice on a daily basis and it is not a particularly dangerous condition, just something you have to keep in mind. Eventually, the Dash 8 got too slow and the plane began to stall (as in stop flying, not an engine stall like in a car). The flight crew failed to recover properly from this (we’ll go into details in a second) and the plane crashed into the ground five miles northeast of the airport.
Now let’s look at the flight with a lot more detail and let us do so not from the moment the plane departed Newark Liberty International Airport (EWR), but from about a day before. Captain Marvin Renslow spent the previous night sleeping in EWR’s crew room which, according to other Colgan pilots, is not a very good place to sleep. The atmosphere can be noisy and not all pilots are even able to sleep there. The crew room was definitely not the place to get a great night’s sleep. Sleeping at a hotel was an option and Renslow opted for that sometimes, but not the night before the accident. Also, it is known that shortly after 3:00 AM, Renslow was using the company computer in the crew room, so he obviously wasn’t sleeping all night. The captain regularly commuted from Florida and did not have a place in Newark to stay in. He had just flown in from Florida the previous night. Clearly, this wasn’t a well-rested pilot, especially when you take into consideration that the accident happened at what would have been the pilot’s normal bedtime on the next day (10:17 PM EST). While it is unknown what role fatigue played in this accident, being tired certainly has an implication as to how attentive you are while flying and how well you react to a dangerous circumstance. I think passengers would be surprised to know how crazy some pilots’ schedules can be sometimes, especially in the commuter airlines.
First officer (co-pilot) Rebecca Shaw was not doing any better with her rest either. Not only was she slightly sick, but she had just commuted from Seattle, Washington earlier that same day. Her “sleep before duties” procedures were not too hot either: sleeping on the transcontinental flight and at the previously mentioned crew room. Now we can safely say the entire flight crew was not well rested when the events of February 12 occurred.
As far as the plane the day of the accident flight, it was determined by the NTSB that everything in the plane that was supposed to be working was, so this is a purely human factors related accident. The flight itself departed Newark uneventfully, except that the aircraft was in icing conditions and soon after takeoff the de-icing system was activated. These boots work by inflating themselves and sort of dislodging the ice off the leading edge of the wing. Ice accumulation can occur at other parts of the wing under certain circumstances, but as long as it isn’t a serious amount, the aircraft will still be flyable. I should point out that the NTSB determined that the aircraft was definitely within the “flyable” parameters even though some minimal ice had built up on the wings and windshield. The only important adverse effect that was occurring was that the increased friction by the ice on the wing had increased the stall speed of the aircraft a bit so that the crew needed to fly the plane at a higher speed to avoid an aerodynamic stall.
As ironic as it is, the crew was talking about the ice on the wings and flying in ice moments before the crash occurred, including a comment by FO Shaw that due to her limited northeast winter flying experience she would have thought they would have crashed if she saw as much ice as the aircraft had accumulated (NTSB determined the ice accumulation as minimal). That particular conversation probably played a very important role in the crash as it probably kept Shaw and Renslow from focusing on their duties as flight crew members. Under federal aviation regulation (FAR) Part 121.542, non-essential conversations should not take place below 10,000 feet or any other critical phase of flight. The actual text of Part 121.542, part of which is known as the Sterile Cockpit Rule, is as follows:
- No certificate holder shall require, nor may any flight crewmember perform, any duties during a critical phase of flight except those duties required for the safe operation of the aircraft. Duties such as company required calls made for such nonsafety related purposes as ordering galley supplies and confirming passenger connections, announcements made to passengers promoting the air carrier or pointing out sights of interest, and filling out company payroll and related records are not required for the safe operation of the aircraft.
- No flight crewmember may engage in, nor may any pilot in command permit, any activity during a critical phase of flight which could distract any flight crewmember from the performance of his or her duties or which could interfere in any way with the proper conduct of those duties. Activities such as eating meals, engaging in nonessential conversations within the cockpit and nonessential communications between the cabin and cockpit crews, and reading publications not related to the proper conduct of the flight are not required for the safe operation of the aircraft.
- For the purposes of this section, critical phases of flight includes all ground operations involving taxi, takeoff and landing, and all other flight operations conducted below 10,000 feet, except cruise flight.
The conversation was enough to keep the flight crewmembers from focusing on the tasks at hand and that is when the trouble really started to unfold. For one, the captain, who was the pilot flying the aircraft, was responsible for monitoring the flight instruments with the first officer serving as a backup. Because of the conversation, the captain was apparently not paying the correct amount of attention to his instruments and the fact that his autopilot was engaged (a normal procedure) further distanced the pilot from getting a feel for what the plane was doing. Icing had certainly been detected and some of the flight displays reflected that information (flight display below by NTSB):
Ice Detected Indication on Engine Displays.
The captain, as was customary in icing conditions, increased the flight display reference speeds which gave him an extra margin of error in the display accounting for the higher stall speed. Among the aids now available to him was a higher speed on the “low speed cue” which tells him when the plane is about to stall. The low speed cue is part of the primary flight display and looks like this (flight display below by NTSB, annotation by Velozia Air):
Location of slow speed cue.
The first officer, however, obtained landing speeds for the non-icing condition equivalents, which were fifteen knots below the speed at which the control stick would begin shaking to indicate an impending stall. If the first officer was not looking at the low speed cue in the primary flight display, she would be looking for indicated airspeeds that were below the speed at which the aircraft could still fly. This most probably helped in the confusion that resulted in her not noticing the speed creeping dangerously towards the stall range.
As the plane continued to slow down dangerously close to stall, both pilots failed to notice what was happening, even though the low speed cue was getting closer to the indicated airspeed. The captain was definitely not paying too much attention to the aircraft he was “flying” because as soon as the stick shaker activated and the auto-pilot disengaged, he failed to react as he had been trained and was completely surprised by what the aircraft was doing. Surprised to the point he did not even know what to do about it. As the airplane began to enter a stall, the stick pusher mechanism pushed forward on the control column in an attempt to recover the aircraft, but the captain, who seemed not to know a stall was going on, decided to fight it and pulled up. Three times the stick pusher tried to indicate the need for stall recovery and in all three instances the pilot opted to pull back instead. The captain did add power at least. Eventually the aircraft lost the ability to fly at such a high nose up position and crashed into the ground. There was definitely enough time to save the aircraft had the crew responded with the way they were trained. They could have noticed the speed reaching the low speed cue (maybe if they weren’t talking), reacted correctly to the stick shaker, or “agreed” with the stick pusher. Over twenty seconds transpired during this time.
Now, for those of you outside aviation, I need to explain something. When you learn to drive your car you spend most of the time learning what to do when your car is fine. However, when you learn to fly an airplane and as you progress through training, you spend a large portion of the time learning what to do when things are going wrong in the aircraft. For example, I spent maybe 70% of my initial multi-engine training flying around on one engine, practicing engine failure procedures. Stalls are one of the most commonly practiced flight conditions and you start doing them as early as the second flight of your flight training career. While you practice other things as well, stalls are always present in training. Recovering from them isn’t really that hard… in very short form: you lower your pitch until you regain speed and increase power (there are other things involved of course). You do the procedure quite a bit of times (in initial training more than later on) so that when a stall occurs you react to it as quickly as you would try to regain balance if you lost balance while walking. Colgan Air did have a twist in their training (since corrected) in which they told their pilots to lower the nose, but the stall recovery angle was still a bit higher than normal. This was done to reduce altitude loss, but could prove potentially dangerous in an accelerated stall condition such as the one the Dash 8 was in. However, Capt. Renslow did not react with his Colgan or primary training, but rather with panic and surprise. I don’t know, maybe his lack of rest had something to do with it. His panic and the first officer’s lack of correction ended up making a fatal accident out of an easily correctable situation. The stall recovery should have come natural, but, again, panic and confusion seemed to dominate in the cockpit. Even if it didn’t come natural, the flight control pushing forward should have tipped off the pilot.
Another thing that happened that night was that the first officer raised the flaps without being asked by the captain (at least verbally) soon after the stick shaker. This did not conform to company policy and would have increased the stall speed of the aircraft a bit, making the situation even worse. Normally flaps are raised farther down in the recovery process once a safe speed is established. So we can assume there was a panic free for all going on aboard the aircraft if the captain indeed did not request the flaps up (it is difficult to say if the pilot didn’t request the flaps up using signage).
There are other factors of concern in this accident and which I’ll mention briefly as this post has gotten crazy long. Captain Renslow apparently had unsatisfactory performance during previous trainings, although more recent training had been OK. Also, Colgan Air has an established history of being run by a hard-line management that requires people to fly extra hours and takes punitive action on those who point out the safety problems of this. While Colgan changed their policy after the accident to a “don’t ask pilots asking for rest because of fatigue” one, they then reversed this as the airline felt pilots were taking advantage of it. This was followed by a more recent reversal of which the terms I am unsure of.
The NTSB is expected to make up to 25 safety recommendations as a result of this accident, some of which are the same recommendations they have been making for almost twenty years. Among these recommendations are issues relating to strategies for flight crew monitoring failures, pilot professionalism, fatigue, remedial training, stall training, airspeed selection procedures, training on stick pushers (many airline pilots really don’t know how it feels), and more. Of course, some of these recommendations will get implemented, but it is likely more than a few will have to wait another twenty years before the Federal Aviation Administration does something about them. It’s just the way our government rolls, what can I say? In the meantime, let’s hope that pilots remember their training when it comes to stalls.
Want to stay informed of further developments on this story? You can keep track of this and all our posts by subscribing to Velozia Air. Simply fill out your e-mail on the “Get Velozia Air Delivered to Your Inbox” section on the right side of this page.
Related Posts:
-
WGRZ-TV Flight 3407 Third Anniversary WGRZ-TV CLARENCE, NY - Sunday night marked the third... Read Full Story » -
WGRZ-TV Vigil on anniversary of Buffalo-area plane crash Albany Times Union CLARENCE, NY — A... Read Full Story » -
Buffalo News Vigil on anniversary of Buffalo-area plane crash Wall Street Journal AP CLARENCE... Read Full Story » -
'All Of a Sudden We Have No Response' From Flight 3407 Federal investigators have recovered the... Read Full Story » -
We now know that 50 people died aboard Continental Flight 3407 from Newark to Buffalo late... Read Full Story »
