Sunwing Airlines 737-800 take-off at Belfast

[Bergerie1]

The crew entered the wrong temperature into the FMC before departure resulting in too low a thrust setting. When they noticed that the aircraft was 'not accelerating normally' they continued and did nothing. The aircraft was so low that it struck a runway light situated 29 metres beyond the take-off run. Surely most of us would immediately have applied full power. I know acceleration can be difficult to judge - but - but what do readers think?

https://www.flightglobal.com/news/articles/probe-into-737-near-overrun-details-temperature-blun-453814/

[Dave Gittins]

Haven't I seen this before ? .. it's either an old report or some folk never learn.

[Bergerie1]

Probably, as it happened in July last year. But I posted it because I am amazed the crew did not apply full power, either during the take-off run when they became aware of the incorrect acceleration, or immediately after leaving the ground when presumably the climb was rather shallow. Is it that crews place too much reliance on the auto-throttle system without maintaining a common sense overview? Or are there other factors at play here?

[DaveReidUK]

https://www.gov.uk/government/news/aircraft-accident-report-aar-22018-c-fwgh-21july-2017

[Hotel Tango]

They are of course not the first, but it's still frightening to think that 2 pilots could be so incredibly incompetent as to not react in a more timely fashion!

[Geriaviator]

This was discussed last summer, but the Daily Mail has resurrected it this morning so it must be news again :rolleyes:

[CAEBr]

This was discussed last summer, but the Daily Mail has resurrected it this morning so it must be news again https://www.pprune.org/images/smilies2/icon_rolleyes.gif

Obviously debated at length here last year, but the official AAIB report was issued yesterday as posted by Dave, so the Daily Fail have now given it to the masses to justify why flying is such a risky business........ :oh:

[Bergerie1]

Clearly I missed it last year but only found the FlightGlobal article today - my apologies to one and all!

However, (1) this type of incident is, unfortunately, fairly common. (2) Mistakes when entering data are easy to make, and I realise, with the use of derated power, neither is it always so easy to recognise what is a normal power setting - it has become more complicated since my day. (3) I also realise that a pilot's subjective appreciation of acceleration is not reliable, and take-off monitoring systems are difficult to engineer. But (4) I find it very hard to understand why the pilots did not increase power once a problem had been detected - and that they did not do so until 4kms from the end of the runway and at 800ft aal.

With both engines working, it is not as though they needed to beware of VMCG and VMCA, both of which can be a factor following an engine failure on take-off. Is it yet another case of lack of basic stick and rudder skills?

[Starbear]

Obviously debated at length here last year, but the official AAIB report was issued yesterday as posted by Dave, so the Daily Fail have now given it to the masses to justify why flying is such a risky business........ :oh:

But they are not entirely wrong are they? When incidents like this keep happening. At least since1982 with, oddly enough another B737 -Air Florida Flight 90, into the Potomac. Yes I know the lead up was different but once on the runway, pretty much same effect if not results...luckily. Many more examples out there.

[Timmy Tomkins]

Some years ago there was a suggestion of a blue light speed. Obvious blue light by the runway, your perf calcs say you should be doing, say 80kts, passing that; if you are not then you need to do something.

Old heavy turboprop with crap performance, we used 30sec to 80 kts as a ball park, just as a check.

[ManaAdaSystem]

Some years ago there was a suggestion of a blue light speed. Obvious blue light by the runway, your perf calcs say you should be doing, say 80kts, passing that; if you are not then you need to do something.

Old heavy turboprop with crap performance, we used 30sec to 80 kts as a ball park, just as a check.

A check of computed N1 vs FMC N1 is all you need to avoid this.

[nicolai]

(5) reporting of this incident was quite poor to say the least. Bluntly the crew tried to evade reporting it - they basically scuttled off hoping noone would notice. Then the airport dropped the ball by not calling the AAIB hotline as the AAIB intends they should, and eventually someone at Sunwing sent their regulator (Canadian TSA) some paperwork probably hoping it would gather dust in an in-tray. Unfortunately for them and fortunately for us, the TSA is not incompetent, the AAIB does check its messages in office hours, and then the investigation started.

[Dave Gittins]

Sadly this occasionally happens with disastrous consequences, MK Airlines 747 at Halifax. https://en.wikipedia.org/wiki/MK_Airlines_Flight_1602

[Doors to Automatic]

As a non-pilot, albeit a reasonably well-informed one I would like to think, I can't understand how situations like this arise.

We are talking about two experienced pilots (or at least one very experienced one). They will have taken off many times and know their machine backwards. They should know weights, thrust settings and speeds like the backs of their hands.

An error of input is understandable but surely when it throws up a thrust setting of 81% for a four-hour flight with a full load, when previously it came up say with 90+ % for a Palma or Ibiza, alarm bells should start ringing in their heads? Especially as there are two of them!

Or am I missing something?

[paradoxbox]

Automation Engineering Induced Ineptitude is the cause of this 'un
Click click, when the autothrottles ain't workin like you thought they were going to, push those bad boys to the wall. Works every time.

81% N1, you have got to be kidding me. That is gross incompetence of the worst kind. Does that airline have checklists?

[admiral ackbar]

Not excusing what they did but apparently those types of N1 values (under 90%) are not uncommon on cold winter days in Canada. Does not mitigate the fact that they should have slammed the throttles at some point but perhaps that is part of the reasons they missed it initially.

[edmundronald]

Not excusing what they did but apparently those types of N1 values (under 90%) are not uncommon on cold winter days in Canada. Does not mitigate the fact that they should have slammed the throttles at some point but perhaps that is part of the reasons they missed it initially.

Apparently they did select TOGA thrust at some point but the logic disabled any effect if the selection.
They should have moved the throttles.


As an SLF I find it incomprehensible that a pilot who sees the runway threshold below him at an altitude < 1ft -that must be hard to miss even for a pilot used to automation- does not slam the throttles forward while *trying* to climb, and waits a minute or so to a distance of 4km. But of course my lack of understanding presumably stems from SLF stupidity. I am also amazed that the plane did not descend again when ground effect was lost.

Edmund

[Denti]

Not excusing what they did but apparently those types of N1 values (under 90%) are not uncommon on cold winter days in Canada. Does not mitigate the fact that they should have slammed the throttles at some point but perhaps that is part of the reasons they missed it initially.

Not even uncommon, depending on weight, in Europe during the summer. Lowest N1 i had in a 737 was 75% for take off, and that was well within the limits with a huge stop margin. Yes, derate plus assumed temperature, a 4000m runway and a short sector in a 737-700 helped there, but if you fly different variants, it is pretty much impossible to do a rough number crosscheck as you can see wildly different performance figures depending on options during the performance calculation. Even the V1 for the same conditions can vary by over 30kts depending on improved climb speed schedule or not.

[DaveReidUK]

A check of computed N1 vs FMC N1 is all you need to avoid this.

They didn't do that because they couldn't.

As the report makes clear, Sunwing's EFB performance app doesn't display the computed N1.

[Bergerie1]

Another incorrect FMC entry. This one is different but with a similar result.

https://www.flightglobal.com/news/articles/el-al-787-struggled-to-lift-off-after-40t-weight-dat-453942/