B744F off the runway in YHZ

[punkalouver]

This was definitely a landing where margins were tight. According to the report, despite all the notams, the longer runway(runway 23), which was into wind, actually did have an available approach for landing(NDB and LNAV). Being nighttime with the weather at non-precision minima and a wet surface, this would seem ideal except for one thing, the threshold was displaced by 1767 feet, something that would seem unusual for any large aircraft to land on, especially a 747. That being said, a 757 did land on the long runway earlier in the night.

With the decision to land on the 7700’ runway 14 with a 20 knot crosswind having been made, there was little room for error. But an accumulation of errors was made, starting with a belief that there was a reasonable margin for this landing, a landing that was not at a particularly light weight. It seems surprising that the captain would choose a flaps 25 landing versus flaps 30. While new on type, the report states that he “had more than 1000 hours PIC experience flying large aircraft worldwide” leading one to think that he would recognize that this was a short runway for an aircraft like the 747. The choice of flaps 25 instead of flaps 30 added 6 knots to the approach speed.

Unlike the other two 747’s in the fleet, this aircraft had performance charts that showed unfactored landing distances, meaning that there was no 15% safety margin added to the actual landing distance values. This information was mentioned on the performance chart but only in the fine print below the calculation table. In addition, the fine print mentions an air time of only 4.22 seconds from crossing the threshold to touchdown, something that’s unlikely to happen in the real world and is for Boeing test pilots(I believe this allowance was removed when TALPA came into effect). It is a good example of how valuable it can be to read the fine print which frequently has critical information that is not obvious.

The crew calculated a landing distance of 6000’ while the TSB calculated 6375’. What adjustment that was missed by the crew is not entirely known. There was an error that catches some pilots, an unnecessary wind additive to the approach speed. There was no headwind component, only a crosswind at 90° when the landing performance calculation was made. Yet the crew added another 5 extra knots to the approach making the Unfactored landing distance 6735’. Even when a 5 knot gust was reported later on, it need not have been considered as the crosswind had become a tailwind and gusts do not affect approach speed calculation with a tailwind. Therefore, with the flap 25 choice and the adjustment for the winds, there were eleven extra, unnecessary knots on the calculated approach speed.

Upon reaching the final approach, the wind had changed direction to become a quartering tailwind, adding close to another 500’ to the landing distance. In reality, they had an approach that had almost no room for error for what were certification test performance landing numbers. They confirmed that they were unaware of the small performance margin as they did discuss the new tailwind but appeared satisfied to continue on.

According to the report, there was ‘heightened anxiety’ in the comments made by the crew on short final but that was related to the crosswind. Unfortunately, their indicated speed was 9 knots faster than even the planned, unnecessarily high approach speed and also were slightly high by the time they crossed the threshold. That being said, they were able to touch down at 1350’ past the threshold which maybe could have somehow allowed a stop just by the end of the runway.

However, despite only a 15 knot crosswind(perhaps the wind did pick up with reports averaging it out), the landing was obviously not handled well and they touched down ‘firmly’, crabbed almost 5° and likely struck one of the nacelles. Although Boeing allows a crabbed touchdown, there appears to have been enough of a lateral displacement after touchdown that the report states ‘preventing a runway side excursion became their priority’. Perhaps the aircraft was drifting to the right upon touchdown due to a corrective action after being blown downwind of the centerline.

The control difficulties on landing led to other events on the rollout resulting in decelerating devices not being activated promptly. The No. 1 thrust lever which had been closed, was advanced forward of idle. One wonders if this was due to the pilot accidentally knocking with his hand in a rush for reverse while trying to control the aircraft. Because of this, no reverse was initially activated on the No. 1 engine and the speedbrakes retracted. The air-ground logic switches in the landing gear changed momentarily to Air Mode during this time period. The No. 1 thrust lever was brought to idle which allowed the speedbrakes to deploy and its reverser to be deployed but then the autobrakes disengaged for reasons unexplained, although large rudder pedal displacements were being used. Maximum reverse was selected but the captain was initially unaware of the autobrake disengagement. Manual brake application began 8 seconds after touchdown, but maximum braking effort did not occur until 15 seconds later, when the aircraft was 800 feet from the end of the runway.

Perhaps in the end, one needs to recognize when an approach has little margin for error, which may only be detailed in the fine print. Once that is recognized, one needs to ensure that in such situations, if any factor moves to a negative performance result, the approach is abandoned. In this case, a new tailwind with the approach speed faster than planned while slightly high on approach should have been enough to trigger a go-around. As for the touchdown and rollout, that is a pilot skill requirement. In this case, it needed a better level of proficiency, something that 110 hours of long haul flying on type may not give you.