Gulfstream IV jet ran off a runway while taking off

[Gabriel]

Note what you say Gabriel. The run off is 550m/1811ft from the threshold to the hedge line with a further 94metre/288ft metres to the edge of the ditch. Sounds a lot but an aircraft travelling at speed with difficult retardation on grass for at least half those total distances is going to eat that up in seconds. From the marked end of the runway to the end of the tarmac is 320m/1055ft, from there on its grass and hedges.

Not that I disagree (I don't), but...

The Federal Aviation Administration (FAA) requires that commercial service airports, regulated under Part 139 safety rules and federally obligated, have a standard Runway Safety Area (RSA) where possible. The RSA is typically 500 feet wide and extends 1,000 feet beyond each end of the runway. The FAA has this requirement in the event that an aircraft overruns, undershoots, or veers off the side of the runway.

The EMAS technology improves safety benefits in cases where land is not available, or not possible to have the standard 1,000-foot overrun.

That comes from the very document you linked to back your point for the EMAS.

[elaw]

Is anyone here a fan of the show "Futurama"?

This is starting to sound like the episode where they crash a tanker (spacecraft) that's supposed to be "leakproof" because it has 6,000 hulls, it leaks, and everyone is wondering when they're going to smarten up and start building tankers with 6,001 hulls...

[brianw999]

That comes from the very document you linked to back your point for the EMAS.

Yes Gabriel, I'm aware of that and the 1000 feet requirement.....I say its not enough in this circumstance given that two thirds of the run off area is poor quality tarmac and grass.

By the way, I've added to my my post since I first put it up.

[elaw]

They say a picture speaks a thousand words. This could help illustrate what the pilots had to contend with. The yellow line is the distance that a Gulfstream 4 needs to get airborne at MTOW. The word DIE is where they did exactly that.

But... is there any reason to believe the aircraft was at, or even anywhere near, MTOW? We're talking here about a plane with a max. range of over 4,500 miles, but from Hanscom to the most distant point in NJ is only ~275 miles.

On a related note, that thing the media keeps calling a "ditch" is actually a streambed. So it does have a purpose... it's not like some malicious party just went and dug a trench there to ensure that any a/c that ran off the runway would be destroyed.

[elaw]

two thirds of the run off area is poor quality tarmac and grass.

Hey you better not complain too much about the "grass" thing... it was you Brits who got us started on that expensive, high-maintenance, weed-infested, ANNOYING stuff...

[brianw999]

But... is there any reason to believe the aircraft was at, or even anywhere near, MTOW? We're talking here about a plane with a max. range of over 4,500 miles, but from Hanscom to the most distant point in NJ is only ~275 miles.

On a related note, that thing the media keeps calling a "ditch" is actually a streambed. So it does have a purpose... it's not like some malicious party just went and dug a trench there to ensure that any a/c that ran off the runway would be destroyed.

I was using the maximum possible scenario.

[Gabriel]

Airworthiness = the ability of the plane to safely fly and land. But call it 'the ability of the plane to fly' if you wish.

I concede that this is the English definition of "airworthiness".

However, within the aviation industry, it's uses differently to state the legal condition of the airplane to fly. The manual says that cracks up to 2 inches are acceptable in this zone. There is a crack of 1.9 inches? Airworthy. The crack grew to 2.1 inches? Not airworthy. There is an AD to replace the concentric rudder actuators before 2020. December 31, 2019? Airworthy. January 1, 2020? Not airworthy. The annual inspection is due in October. October 31? Airworthy. November 1? Not airworthy.

So we are in the take off run nearly at Vr and the FO says "Today is November 1! The annual inspection is overdue!" and the PIC replies "Airplane not airworthy, reject the take off!!!"

Doubt = a feeling that something is preventing liftoff.

Rather, a feeling that a safe lift off WILL not be achievable. Something can be preventing the lift-off NOW (overweight, wrongly claculated Vee speeds, CG or trim in a different position that supposed...) but not prevent the ability of the plane to fly in a couple of seconds.

Serious Doubt = a strong feeling based on professional experience that something is preventing liftoff.

A strong feeling of what I have just said. Judgement. That's why the abort in Teterboro was a reasonable decision even if a decision to continue the take-off would ave prevented the accident altogether, and would have been a reasonable decision too.

It is always best to take-off. That doesn't mean anything if it won't take off.

That's exactly what I said in the previous post. The pilot has to judge, with the (likely incomplete) information that he has available, his experience, the procedures and guidelines, and his intelligence, if the plane will be able to lift off or not. If all the engines fail or a wing is left behind on the tarmac, the only reasonable judgement is that it won't. Other cases are not so clear and the pilot must assume the risk of whatever decision he takes. History shows a bias towards continuing the take-off being safer. But there is no guarantee.

V1 is not a judgement call because it can be calculated. Can we not also use computational power to calculate a speed beyond V1 where the a/c will depart the runway during RTO at or below [x] kts, [x] being the allowable limit for a controlled crash beyond the runway? Call it Vrto or something...

V1 is also calculated beforehand so the pilot does not have to make spontaneous calculated estimations. A Vrto would have the same value.

I don't see any absolute limitation to doing that. But we are VERY far away from that day.
So far away, that today a reject at V1 can leave the plane stopped with a lot of runway ahead. Remember that V1 is a speed selected BY THE OPERATOR within a range. The minimum V1 is the one that is greater than Vmcg (minimum control speed on the ground with the critical engine failed) and than the speed that, if the take-off is continued after an engine failure, will put the plane 35ft above the runway by the end of the runway or of the clearway. The maximum V1 is one that is less than Vmbe (maximum brake energy speed) and than the one that, if the take-off is rejected, will make the plane stop at the edge of the runway or stopway. So unless the take-off weight (or thrust reduction) is limited by runway length, there will be more runway than necessary to either continue after an engine failure at V1, or reject at V1, or both.

And yet, pilots don't have this information for each take-off. So let's begin here before moving to your level, which would involve taking into account not only the runway lengths (TORA, TODA and ASDA) but also what's beyond the runway.

The problem with leaving this to pilot judgement involves the same human factors as leaving the decision to continue a landing without the runway in sight (go-itis, blind confidence, compromised situational awareness, risk taking).

Yes, humans are problematic. We have to deal with that. A wrong go/no-go decision past Vr is one of the things that I fear less about the human-factor risks in planes.

I'm going to make a judgement call that rto should definitely occur before flap retraction speed...

Unless the flaps are already retracted (like in many of the accidents that I cited in the previous post).

Side note: I had the feeling that if you had been the captain of the EK A340 flight, we would still be counting dead bodies (including yours).

[MCM]

The point is that V1 is that point.

The concept, as we all know, is that a reject at V1 MAY use up all the available stop distance. So, ANY time you reject above V1, you're working on the philosophy that, yes, you may die - but you are LESS likely to die than if you continue the takeoff.

Please don't try and complicate this more than it needs to be. The RTO decision is already a very difficult one, and stopping above it is known to be a very serious decision. To then say 'oh, but there's another point that you can reject and go off the end of the runway and hurt only a few people', and expect that speed to be monitored and considered is simply not practical.

Safety arresting areas (EMAS) only work to an extent, and to stop an aircraft in the same area that a normal RESA could. They are expensive to install, require maintenance, and are a secondary option compared to a proper, normal RESA. They are a great option when there is no other practical solution, but they are not the preferred answer by anyone, including the international pilot association who push for all sorts of safety improvements.

We have to be practical and economic - we cannot expect every airport to put in an EMAS on the possible opportunity that an aircraft goes off the end - the cost would be prohibitive to the industry.

If you use more than your 1000ft RESA, current design EMAS systems wouldn't have stopped you either.

[3WE]

Is anyone here a fan of the show "Futurama"?

This is starting to sound like the episode where they crash a tanker (spacecraft) that's supposed to be "leakproof" because it has 6,000 hulls, it leaks, and everyone is wondering when they're going to smarten up and start building tankers with 6,001 hulls...

Don't watch the show.

But your concept is huge.

Should southwest cease all operations at midway?

Compare dfw runways to dal....

Ban all runways shorter than 14k!

[vaztr]

Would all of the PICs thinking change if they were on a runway with little runoff

I'm thinking of the one in Brazil? where the 737? landed hot and ran into the gas station after dropping 20 metres at the end of the runway.

I'm guessing that given that sort of runway would cause the pilot to be very cautious very soon after V1.

I doubt he'd be thinking 'if I gun this a little more I could be airbourne with a few more knots of speed'

VAZ

[Evan]

Please don't try and complicate this more than it needs to be.

I concede that situations in which a plane cannot take off within the runway length are extremely rare and that most that fail to lift off after Vr will still lift off before reaching the end of runway. I'm only commenting on this particular accident because 168kts seems pretty high runway speed for an airworthy G IV. I can only assume that something was preventing it from ever getting airborne.

If so, the best practical solution I see for that scenario is to strengthen the things that will prevent such a scenario, such as loading procedures and restrictions, regular maintenance inspections, vigilant walk arounds and flawless adherance to pre-flight checklists. It's one of those accidents you have to prevent before you ever reach V1.

[3WE]

I'm only commenting on this particular accident because 168kts seems pretty high runway speed for an airworthy G IV.

Pretty high, kind of high, somewhat high, a little high...

I'm with Gabe- the speed is kind of reasonable given normal parameters plus training plus reaction times.

[elaw]

It is always best to take-off. That doesn't mean anything if it won't take off.

That's exactly what I said in the previous post. The pilot has to judge, with the (likely incomplete) information that he has available, his experience, the procedures and guidelines, and his intelligence, if the plane will be able to lift off or not.

You forgot one key thing: in situations like the one we're discussing here, the pilot has to decide very quickly what to do.

Consider this just-slightly-fictional example: you're sitting in an office at work, and the fire alarm goes off. Because of some bizarre circumstances, you have 3 seconds to decide whether to evacuate the building or not. And no matter which decision you make (stay or leave), you will die if the decision is not correct. Other than those limitations, you can use any & all resources available to you to make your choice. What's the chance you'd choose correctly, regardless of your training?

Gathering and processing information, if it's available at all, takes time - and sometimes the necessary amount of time is simply not available. In that case, luck is going to determine the outcome.

[Gabriel]

Would all of the PICs thinking change if they were on a runway with little runoff

I'm thinking of the one in Brazil? where the 737? landed hot and ran into the gas station after dropping 20 metres at the end of the runway.

Off topic, but:

It didn't land hot. It landed at the right speed and right spot. The probelm is that the pilot forgot to close one of the throttles so the airplane didn't "interpret" that the intention was to stay in the ground and complete the landing (both throttles must be at idle of this logic to apply, let's say that the plane thought "maybe the pilot wants to go around) so the autospoilers didn't activate and the autothrust didn't disengage and, as the plane slowed down, kept increasing the thrust on the throttle that had not been idled in an attempt to hold the speed it was programmed to hold.

I'm guessing that given that sort of runway would cause the pilot to be very cautious very soon after V1.

I doubt he'd be thinking 'if I gun this a little more I could be airbourne with a few more knots of speed'

I'd say that it's exactly the opposite. If you have a minimum acceptable runway and cliff immediately after it with a mined field at the bottom, you know that an overrun is a death sentence. So after V1 the only choice is to go and hope that the plane will eventually lift off. An overrun, no matter how smooth or slow, will not save you.

[Gabriel]

You forgot one key thing: in situations like the one we're discussing here, the pilot has to decide very quickly what to do.

Consider this just-slightly-fictional example: you're sitting in an office at work, and the fire alarm goes off. Because of some bizarre circumstances, you have 3 seconds to decide whether to evacuate the building or not. And no matter which decision you make (stay or leave), you will die if the decision is not correct. Other than those limitations, you can use any & all resources available to you to make your choice. What's the chance you'd choose correctly, regardless of your training?

Gathering and processing information, if it's available at all, takes time - and sometimes the necessary amount of time is simply not available. In that case, luck is going to determine the outcome.

I fully agree with you except for one minor detail: I didn't forgot that thing.

That's why the decision is after V1 you GO, and that decision is taken before advancing the throttles, and the only good reason to change that decision is that you have enough complelling evidence to judge, in that very little time, that the plane will not fly. If you have doubts, then don't decide anything, just keep up with the original plan and GO.

Why? Because history has shown that, in most cases, the airplane WILL fly before reaching the end of the runway (and sometimes even after, like the EK A340 case). So if you can't make a rational choice based on current information, you will have to bet, and you know (because of your training) that in most cases the plane WILL fly, especially in those cases where you could't convince yourself that the plane will not do it. And because of the limitations that you've mentioned, to convince yourself takes a very strong evidence like a second engine failure (after the first one failed 5 seconds ago), that you have just crashed against a truck that has just crossed the runway and now the plane (or its remains) are dragging on the tarmac.

There is a saying in aviation: If in doubt, don't. So you had already decided that you would not abort the take-off after V1 except if you decide that the plane will not fly. Have you decided it? No enough information or time to process it? In doubt? Then don't abort.