Tag Archives: Unstable Approaches

Asiana 214 Mishap, SFO, July 6, 2013: Stable Approaches & Go Around Procedures

Captain Paul Miller in cockpit

Captain Paul Miller

Stabilized Approaches must be part of an Standard Operating Procedure (SOP) and not just a criteria, policy or even best practice.  Procedure means that it is a written set of steps and explanatory notes. Procedures are trained by the airline, checked by the FAA and continually verified, reviewed and updated by the FAA and the airline. Part of a stabilized approach procedure must include a verbalized communication of the aircraft state and progress at several points in the approach and a verbal command to continue the approach or to Go Around as the aircraft passes these points, if the procedural steps are not achieved.

One of the most important things we learned in the recent June 18, 2013 Flight Safety Foundation (FSF) Go-Around Safety Conference, sponsored jointly by EuroControl, European Regions Airline  Association (ERAA) and the European Advisory Committee and the International Advisory Committee (EAC and IAC) of the FSF, in Brussels, addressed the idea of a criteria, policy or best practice as opposed to an SOP.

See the conference link at http://www.skybrary.aero/index.php/Portal:Go-Around_Safety  for more details.

It is one thing to have a 500ft or even a 1000 ft stabilized approach criteria.  It is quite another thing to have an SOP to which all flight crew members are trained, a procedure  that says if the crew does not accomplish a stabilized approach by 1000 ft or even 500 ft, the pilot monitoring, the flight crew member not flying, shall say verbally “GO AROUND,  APPROACH UNSTABLE.”  Then, as written in the SOP, the pilot flying (PF) procedurally initiates the Go Around (GA), as per trained, as briefed and published as well  as cleared by air traffic management procedures.
Even now, even with a stabilized approach criteria or policy at 500ft or 1000 ft and a even with a no fault GA policy, wherein the crews’ motive for going around is not called into question, air lines still need to have an SOP, a written operating procedure. The SOP must require crew members to verbalize, “1000 feet, APPROACH STABLE, CONTINUE,” or “1000 feet, APPROACH UNSTABLE, GO AROUND.” The SOP must define action, not just policy or criteria. It must require prescribed actions for the flight crew to do at that point of decision. This is not a talking point, a time for discussion or observation-this is a time for action.

Why is this important now? So, it has actually always been important.  However, now we need to understand what happened to Asiana 214 and learn from it. Having seen the video of Asiana 214 approaching San Francisco International (SFO) runway 28 Left, one thing is quite noticeable. The approach appears very flat and very low in altitude at a considerable distance from the stone wall at the waters edge. In lieu of the standard three degree glide slope, the aircraft appears to have essentially leveled off at 100 or 150 feet 1/4 to 1/2 mile from the seawall. This is not good, but also this did not happen in the blink of an eye. This was a developing situation, to which the crew did not report their state nor did they take action.

Keep in mind that the designated normal landing area on any runway is between 500 feet and 1500 feet down the runway. Add onto that, the stone wall at the waters edge appears to be approximately 1000 feet from the end of the runway.  So the aircraft landed approximately 2000 feet short of the intended point of landing, that is, the normal landing area on the runway. This means that the aircraft had descended approximately 50ft and 100ft  below the normal 3 degree glide slope. This deviation did not occur 4 or 7 seconds from landing, but possibly 1/2 mile or more from the end of the runway. This very large deviation was neither reported nor reacted to initially as it occurred, but only just a few seconds prior to impact. By apparently not having a reporting and Go Around procedurally ingrained by training, the crew was left to determine where they were by their own criteria or judgment and then decide on a course of action and then to take that action: too much to think about in too little time and space, in my opinion.

If and when the aircraft passed 1000 feet in altitude or even 500 feet in altitude during the final approach, this would have still been a considerable distance from the runway. In this mishap the aircraft had to have been well below glide slope and thus very unstable at that point, a factor observable from the flightdeck and measurable against the requirement for a stable approach.

So did the airline have a procedure (SOP) that required the crew to verbalized stable or unstable and continue or go around? If so, why did the crew continue the unstable approach? Did the crew notice the unstable approach? Why was there a hesitation to act? Was there a procedure in place to which the crew had been trained? Or rather was there merely a criteria and a policy only in place? Can you see the difference between having a a trained GA SOP and having only a policy or only a criteria? A policy or a criteria gives talking points.  An SOP gives action required.

Early reports from the NTSB identified that the airspeed decreased from a landing target of 137 knots down to about 109 knots. Remember that normal target speed is 130% of the stall speed at that weight or 1.3 times the stall speed. So this decrease of 28 knots is a considerable change. But remembering the laws of aerodynamics, the aircraft nose attitude had to slowly pitch upward as the speed decreased. This means that the plane went from a normal landing attitude to a very high nose up attitude, again something very noticeable from the flight deck, another clue that the speed is unstable.

Rounding off for argument sake, a stall speed of 100 knots, with a 30% buffer gives a landing speed of 130 knots. The stall warning stick shaker come on 5-8 knots above stall speed (or more correctly stall angle of attack).  So the crew was not flying the aircraft just a little slow; the crew was flying substantially slow, dangerously slow and into the region of rapidly rising induced drag.  The crew appeared to be unaware of the rapid decay from 137 knots to 109 knots right into the stall. I wonder if the subject of the region of reverse command in swept winged aircraft was part of the Asiana B777 training program? Again, there may be a reliance on subjective values such as airmanship, in lieu of a set of written procedures, approved, trained and checked. But all flight crew of swept wing aircraft must have a very good understanding of the region of reverse command and the grave consequences of allowing the aircraft to enter into that part of the flight envelope. Yet, how many commercial pilots can recount the danger of the region of reverse command? Does the FAA require it?

Remember it is not uncommon for many crew to add 5-7 knots to give a margin above the minimum target speed. So that would mean many pilots would be flying at about 142 knots. The aircraft landed 30 or more knots slow. This would require a significant increase in angle of attack and nose up position, which was seen in the video at the last moments. This high pitch attitude rapid change is another major deviation from a stabilized approach procedure.  This deceleration should have been recognized immediately as it occurred initially below 137 knots and the nose high attitude should have been corrected by the crew or the crew should have commanded a Go Around. The GA SOP should have been initiated as soon as the airspeed fell below the target speed. The correction would have been to add substantial power and lower the nose slightly.  The value of a written stabilized procedure is that this procedure is trained over and over again until the crew is exceptionally skilled in close to the hand eye coordination needed when a lot is going on.

As it happened the aircraft was well below glide slope and very slow below landing target speed. Both of these serious deviations from the stabilized approach requirements occurred well before the aircraft arrived at the sea wall. The deviations began more than a half mile out and increased as the flight approached the sea wall. The crew began a discussion during this period, but did not take action during this period. This is the danger of having a criteria and/or a policy in lieu of a well trained GA SOP.

Not having a well trained Go Around SOP based on deviations from Stabilized Approach is a severe risk to safety. Not having a training program that covers the aerodynamic principle of the region of rising induced drag and does not address corrections is a severe risk to safety as well.

Concerning the Instrument Landing System (ILS) electronic glide slope for runway 28 left being inoperative, I would ask very serious questions of the FAA and the airport authority as to why they chose to temporarily decommission this system and why the decommissioning lasted for such a long period of time. This puts a burden on the crew members in the cockpit. This is what many people call the tail wagging the dog. Why have all this expensive ILS technology, engineering, equipment and training, if in the end all of it is deferred to a construction company that is laying concrete? Whose decision is that? Either the ILS is important and we had better damn well keep the system up and operating, or it is not important and we should train pilots to do visual landings. But the FAA and airport authorities want the argument to go both ways and always be in their favor. I do not agree with this tail wagging the dog theory. If the runway is “under construction,” then close the runway and allow the construction crews to do what has to be done. Do not close all of the vital instrumentation and just keep the concrete portion of the runway open for revenue generating purposes. This is not fair to the flight crew and this is most certainly not fair nor safe to the flying public, which by the way is also the paying public. Again the least of all parties in this case is the concrete construction company. Therefore it ought to be they who accommodates the operation, not the operation that accommodates the concrete company.

Decommissioning the ILS  requires the crew to build an artificial glide slope in the flight management system, using runway data and alternate procedures. This would have been a good training opportunity for the crew to build that approach within the aircraft flight management system. This could have been done as part of the preparation for the approach. Again though, was this Asiana SOP?

This is a standard and written procedure in the B777 flight manual. If that procedure was not done, the next guide to a usable runway glide slope would have been the precision approach path indicators known as PAPI mounted along side the runway. The PAPI  will indicate to the crew when they are below the glide slope, an illumination of four red lights to the left side of the landing area of the runway. We are not sure from reports that the crew was using the PAPI to assist them in the landing.

One last thing, as the initial operating experience instructor pilot sitting in the right seat should have said when things got far out of hand, “Okay, I’ve got it,” and flown the aircraft back into safety, again as a procedure for being the person in command. But the actions taken were to order a Go Around to the pilot flying, but too late for any reasonable reaction from the pilot flying. He should have taken the plane and done the Go Around when he saw that things were out of hand.

A lot going on here to discuss, no? I think so. We all would be better off tomorrow if we all tried to understand what happened to Asiana 214 today. We should see many ways on how our own airline could be made safer from the discussion of the mishap of Asiana 214.

Stabilized Approach Procedures, Go Around Procedures and substitutes for ILS procedures all must be part of the skills sets of B777 crew members and for that matter all commercial flight crew members no matter what aircraft they are operating.