Torqued: 747 crew followed the procedures but didn’t make the best choice

 - December 11, 2006, 4:21 AM

In the June 2005 issue of AIN I wrote about a fully loaded British Airways 747 that was taking off from Los Angeles bound for London when one of its engines emitted a large fireball just after liftoff. I commented on the wisdom (or lack of wisdom) of proceeding on a 16-hour flight with one engine out and uncertainty about the extent of damage to the engine or the airframe.

A flight crew is at a real disadvantage when an event like this occurs because the information on which to base an intelligent decision is limited to the instruments and possibly whatever can be collected from the onboard reporting systems that transmit back to some company-designated location, such as a flight-operation center. A major concern in this type of event is that the rotating internal parts stay contained within the engine. These pieces contain a considerable amount of energy, and they have caused major damage in other events.

The tower controller saw the fireball that signaled the start of this event, and he started the internal process to provide the crew of the aircraft with all the help ATC could muster.

ATC began making preparations for the aircraft to return to LAX almost as soon as the controller saw the fireball. There was no doubt in his mind that this aircraft was returning to LAX, but that was not the case.

For this article I thought I would walk through the tower communication tapes as if I were conducting an investigation and give you my view of the words spoken and the actions taken that day. I will also use some of the investigation material developed by the UK Air Accidents Investigation Branch (AAIB).

The copy of the tower communication with BA Flight 268 that I have starts when the aircraft is number one in line to access the takeoff runway and the tower controller asks 268 Heavy to “advise ready to go.” The flight crew responds, “I will…”
The tower controller then handles other traffic without any further communication with 268 Heavy for almost five minutes. In any probe, investigators would ask the surviving flight crew about the events that occurred during this period. There are many reasons why an aircraft could taxi to this point in the takeoff process and not be ready, but any investigator would want to understand what was going on at this particular time.

At approximately the five-minute mark the tower controller both asked/advised the crew of 268 Heavy, “Are you ready? I have a couple of departures behind you. Are you almost ready?” “Yes, just one minute,” replied 268 Heavy. Again, any investigator would want an explanation of events occurring in the cockpit at this time. Although we all understand that this might seem like a long time to prepare for takeoff, it could easily be justified.

Approximately one more minute passed and the tower controller advised 268 Heavy to “taxi to Runway 24 left and hold, and if not ready to go in a minute I have to put you back in line.” The BA crew responded, “In position and hold. We are ready.”
Approximately another minute-and-a-half passed before the tower controller cleared 268 Heavy for takeoff, giving the wind direction and speed as well as the heading to fly after liftoff. The BA crew read back the clearance correctly.

There is nothing out of the ordinary in this sequence of events except what would appear to be some delays that could perhaps be easily explained. After all, we do not want our flight crews rushed or to take off unless they are properly prepared for what is to follow.

Engine Troubles
Approximately one minute and 20 seconds after giving the takeoff clearance, the tower controller stated, “268 Heavy, flames out of either number-one or number-two engine.” The crew of 268 Heavy responded, “We’re shutting it down.”
The tower controller then advised “268 Heavy to contact Southern Cal on 124.3. He knows about your issue.” The 747 crew acknowledged that transmission.

After this transmission, the tower controller followed the procedures for dealing with aircraft in trouble by making the required notifications. In fact, by the time 268 Heavy’s crew contacted departure radar, their situation was known (although the procedures require the radar controller to verify the aircraft condition again). Also, other controllers were making contingency plans for this event, following procedures developed for many types of emergency.

When 268 Heavy came up on departure radar, the crew was advised “to climb and maintain 5,000 if able and to advise intentions.”

The BA crew responded, “Roger, climb to 5,000 we are able…We had a surge on takeoff and we are doing our checks.”

The departure radar controller then stated, “Tower advises you had flames coming out of an engine and it was shut down.”

The crew of 268 Heavy responded, “We haven’t shut it down. We throttled it back and are doing our checklist.” The departure radar controller responded simply “understood.”

The next transmission from the radar controller vectored the 747 to a new heading; 268 Heavy acknowledged. Then the departure radar controller asked, “268 Heavy, can you tell me the number of people on board?” The crew responded with 351 plus one plus 18 crew. The departure controller then asked, “Which engine was it?”
The crew responded that it was the number-two engine. All of this follows established procedures. While the pilots were going through their procedures in isolating and evaluating the condition of their aircraft, the air traffic controllers were quite occupied. The airspace around LAX is busy and controllers must stay focused on the effect an aircraft in trouble can have on all the other operations.

After a few minutes, 268 Heavy called the departure controller to advise, “We have now shut down the number-two engine and will consult company to see what they want us to do.” The controller then asked to keep departure control advised. About two minutes later, while receiving another vector, 268 Heavy advised the departure controller, “For information we need about 10 minutes to contact company to see if we are coming back to LA or to continue.”

The departure controller advised he would keep vectoring him until they decided. He also stated that he would keep him clear of all Level 2 and above precipitation that he could see on his scope. It is in events like this that we see just how good the ATC system we have in the U.S. really is.

Aircraft Command
About seven minutes later the departure controller asked if 268 Heavy had any update. The response was “Yeah, 268 Heavy just decided to set off on our flight plan route and get as far as we can so we would like to continue our flight plan.”

At this point some of the previous transmissions would be of concern. First, the crew’s comment that it would “contact company to see if we are coming back
to LA or to continue” would raise the question of who is in command of this aircraft. This is a dilemma spawned by technology that allows an aircraft to remain in communication with company headquarters from anywhere in the world.

Many readers might recall the Alaska Airlines accident into the ocean near Los Angeles. In that accident the flight crew gave up time and the opportunity to land at other airports by following communications from maintenance control, which was interested in troubleshooting the system instead of getting the aircraft on the ground and then troubleshooting the problem. There is nothing wrong with the flight crew consulting with other experts who can advise, but the responsibility and decision-making rests with the pilot-in-command.

About all the FAA has on the circumstances of this flight is in the ATC recordings as the 747 continued onwards and out of U.S. airspace. When the aircraft made an uneventful emergency landing in Manchester, England, the AAIB conducted a field investigation into the facts and circumstances of this flight, and what follows is taken from that report.

The AAIB investigator debriefed the flight crew to understand their view of the facts and circumstances and to develop a timeline for the event as well as their actions. The aircraft’s flight data recorder and FOQA quick-access recorder helped investigators gather data.

Unfortunately, the flight data recorder had a malfunction and did not record the first one hour and 14 minutes; fortunately, information contained on the quick-access recorder filled this void. The crew interview and the recorders helped the investigators develop a good picture of the facts. A short time later the engine was torn down and inspected to understand what had malfunctioned.

Uncontained Engine Failure?
The crew statements contained a fact that we could not glean from the ATC recordings. While in the decision-making stage, one of the pilots went back into the cabin to look at the number-two engine for damage, which he was only partially able to accomplish. Could the crew have been looking for signs of an uncontained engine failure?

We might never know since the AAIB investigators did not ask that question, but the engineering staff would surely recognize that possibility, especially since the crew reported an engine over-temperature.

In addition, the pilots stated that they tried to advance the number-two throttle from idle but were unable due to “audible surge noise.” Attempts at a higher airspeed had the same results. Again, these facts would raise concerns from any knowledgeable engine specialist about this being more than solely an engine-surge problem.

I have another concern: I wonder why only an engine surge was considered as the cause of this problem when there are clearly other possibilities, such as a contained or uncontained engine failure. Uncontained engine failures can and often do cause additional damage to the aircraft or systems, and sometimes this damage is not readily apparent from the cockpit instruments or from the limited view afforded by a cabin window.

In addition, the engine damage described in the AAIB report indicates that the temperatures inside the engine were high enough to damage (melt) some blades, and I believe that the over-temperature likely occurred during the reported fireball since the engine was throttled back and the crew was unable to advance the power lever past idle due to “surge sounds.” The AAIB report states only that the blades were damaged as a result of high temperatures.

The AAIB also reported that the captain “had spoken with the operator’s base at Heathrow by radio and had been advised that it would be preferable to continue the flight but that the course of action was the commander’s decision.” The fact that this communication occurred raises the question of who was in command of this flight. The AAIB report does not address this issue.

The AAIB report contains considerable information about the fuel-handling training provided to the airline’s pilots, explaining that there was enough fuel remaining on the aircraft to continue on to London. The report notes, in a tortured way, that there really wasn’t an emergency when the crew decided to land in Manchester.

British Airways has a procedure in the flight operations handbook that allows a flight to continue on to destination with an engine out.

The AAIB report also states that British Airways is required to comply only with the UK’s aviation rules, a position the U.S. disagrees with. The FAA filed an enforcement action against British Airways for violating U.S. rules (FAR 129.11a and 91.7a) and sought a $25,000 fine. After some high-level discussions, the FAA settled this case with a commitment from BA that it would not use the disputed engine-out procedures while flying in the U.S.

Some have criticized the FAA for this decision, but I take a different view. The amount of the fine, $25,000, is nothing to a company that has a cash flow as large as British Airways’, and the traveling public is better served by the elimination of this procedure than by the collection of the fine.

One of the purposes of conducting accident or incident investigations is to prevent similar events in the future. To accomplish that, every investigation must explore every facet of the event, follow it to its conclusion and uncover the little things that can make a difference in the future. I don’t think that happened in this case. Also, there is a requirement for all code-sharing airlines to review the procedures of their code-sharing partners. It will be interesting to see whether the next review of British Airways by its U.S. code-sharing partners makes any mention of this procedure.