Rejecting A Takeoff After V1…why Does It (still) Happen?
Aborts at a high speed (above V1) are rare. However when they occur the outcome can be a serious accident. The following paper discusses why this still occurs despite efforts from the industry to stop this.
The analysis in the paper showed that there is still plenty of room to improve takeoff safety and reduce the number of unwarranted rejected takeoffs above V1. What do pilots and safety specialists think about this? All views are welcome but please read the paper first!
You are correct. Pilot simulator training often presents RTOs as engine-related events while the Takeoff Safety Training Aid (1993!) gives recommendations about other failure conditions to consider. The majority of all RTO accidents were not related to engine problems. In these cases it is possible that the pilots were not fully prepared to recognise cues of other anomalies during takeoff. Pilots often interpret these other anomalies (like a tire burst) as events that threaten the safety of flight and decide to reject the takeoff at any speed. Indeed it is not easy to put these other than engine failure anomalies in the simulator.
Have long since retired; but during the 50 plus years of flying experience which included the careful study of accident reports and countless papers on the dangers of high speed aborts, I have always conditioned my thinking to continuing the take off from 15 knots below V1 unless I was absolutely certain the aircraft would not fly.
One could speculate on how to define how you would know for certain if the aircraft would not fly. It all gets too hard. While the 15 knots below V1 figure was always in my mind it was a personal private choice and I never told the first officer. My point being to avoid a legal or company SOP argument recorded on the CVR. There have been many instances of fatal overuns caused by stuffed up aborts near V1 but very few of clipping an obstacle in the take off splay in the go situation.
However there is at least one case when a T/O rejection above V1 maybe saved lifes. About 10 years ago or so, an Air Liberté MD83 stroke a Shorts 330 at Roissy CDG, France. Basically the MD was performing its T/O roll after having had his T/O clearance when ATC did a tremendous mistake, allowing a Shorts 330 to line up from a mid position taxiway on that same runway. Nightly operation. Visibility less than average. Language misunderstanding (french language ATC <-> MD83 vs english language ATC <-> Shorts). The MD's left wingtip stroke the Short's cockpit, instantly killing the unfortunate F/O. Although the MD was a few knots above V1 its crew chose to abort their T/O and I think they took the fair decision as 1/3 to 1/2 of their left hand wing was torn off. And as they were light on a pretty long runway they managed to come to a full stop before the runway end.
Had they chosen to continue their T/O, none would have been able to say the airplane would have flown safely.
Final report on the high speed RTO overrun with a PSA Airlines CRJ2
The final report on the high speed RTO overrun with a PSA Airlines CRJ2 at Charleston on Jan 19th 2010, has been released: DCA10IA022
From the report:
The Capt. stated that he did not think the first officer had called out V1 before he initiated the RTO. The F/O stated that he thought that the RTO was initiated before reaching V1 and that he knew that an RTO should not occur after reaching V1.
NTSB Report: “The captain initiated a rejected takeoff (RTO) about 5 seconds after he started moving the flaps and when the airplane was at an airspeed of about 140 knots, which was 13 knots above V1.”
Then there was the TWA L1011 at KJFK (13R) that "aborted" just after liftoff. That was a totally asinine decision. The runway length and the benign overrun area (at least that day) saved their bacon in spite of themselves. It destroyed the aircraft, though.
I was astounded that both TWA management and the FAA supported this utterly insane act by the captain. The NTSB, though, saw it through untinted glasses, if I recall correctly.
Aterpster. I think you are referring to the following case:
IMMEDIATELY AFTER LIFTOFF THE STICK SHAKER ACTIVATED, AND THE FIRST OFFICER, WHO WAS MAKING THE TAKEOFF, SAID 'YOU GOT IT.' THE CAPTAIN TOOK CONTROL, CLOSED THE THRUST LEVERS, AND LANDED. HE APPLIED FULL REVERSE THRUST AND MAXIMUM BRAKING, AND TURNED THE AIRPLANE OFF THE RUNWAY TO AVOID A BARRIER AT THE END. A SYSTEM DESIGN DEFICIENCY PERMITTED A MALFUNCTIONING AOA SENSOR TO CAUSE A FALSE STALL WARNING. THE SENSOR HAD EXPERIENCED 9 PREVIOUS MALFUNCTIONS, AND WAS INSPECTED AND RETURNED TO SERVICE WITHOUT A DETERMINATION ON THE REASON FOR THE INTERMITTENT MALFUNCTION. THE FIRST OFFICER HAD INCORRECTLY PERCEIVED THAT THE AIRPLANE WAS STALLING AND GAVE CONTROL TO THE CAPTAIN WITHOUT PROPER COORDINATION OF THE TRANSFER OF CONTROL.
NTSB REPORT AAR-93/04
I don’t think it is a true RTO as the aircraft already left the runway. Still it is interesting to notice the reaction of the co-pilot on the stall warning in the light of many (also recent) occurrences involving inappropriate reaction to stalls.
I don’t think it is a true RTO as the aircraft already left the runway.
Having taken off the same aircraft (probably the same tail number) from that runway many times, my subjective conclusion would be a rejected takeoff. My conclusion is supported by the NTSB investigation into the accident:
The NTSB title of its investigation report is:
Aborted Takeoff Shortly After Liftoff
Finding Number 12 of the report:
12. The captain made a "split second" decision to reject the takeoff by reducing the engine thrust. His decision was very likely based, in part, on the perception of available runway to stop the airplane.
BBf, BOAC; Re HS748 Stanstead. Read the AAIB report carefully, re decisions and actions (3/2001 (EW/C98/03/7)). With hindsight, the decision was ‘incorrect’ (yet all decisions at the time are correct from the perspective of the decision maker), however the resultant fire from which a correct decision might be concluded, was perhaps exacerbated by other ‘erroneous behaviors’ in procedure.
Learning points from this accident are the power of startling or extreme graphic events to bias decision making, and that on balance a Go-Minded attitude (with correct engine shut down procedure) should provided a similar or higher level of safety than a decision to land straight ahead on a long runway.
In similar circumstances – a surprising event with a long runway ahead – another Captain related that he did not believe that the cues (noise and vibration from a thrown tyre tread) could be generated by a normal aircraft – hence the decision to a land ahead – surprise and incomprehension. However, on a shorter runway the takeoff would have been continued, but no reasoning for the change could be offered. This indicates the complexity of the decision making process, highly dependent on the perception of the situation in a time pressured environment, and bias – our predispositions towards particular actions (knowledge and understanding of risk – see the Training Aid).
I doubt that extensive simulator training would ever cover the range of scenarios, nor generate the requisite surprise; thus it might be more effective to practice the drills associated with stop or go in relation to V1, and as the Training Aid suggests, be Go minded – education and knowledge to dispel inappropriate bias. In addition, from the report (#1), keep the decision choice to a simple ‘if – then’ option, which entails considering the reality and probability of operational scenarios – keep the SOPs practical. Human judgment should still the best solution in ‘unforeseeable’ circumstances, and although we are ‘ground dwellers’, flying – like the birds – offers another dimension for safety, as well as more time (and longer runways) for the subsequent choice of action.
I don't want to drag out an old accident here, sp, but my copy tells me that the action to re-land was considered 'sensible in the circumstances' by AAIB and I recall the Captain telling me that it had been suggested to him unofficially that the wing may well have burnt through during the positioning to land after continuing. I judge the decision a good one.
It is, after all, what we are paid to do as Captains - to take decisions at the time for which we stand responsible and sometimes 'normal' rules need to be abandoned. 'Never say never'?
BOAC, re HS 748 accident, which the AAIB investigated and reported circumspectly – avoiding ‘error’ and blame. I don’t have a problem with the Captains decision in the particular situation, but what about a shorter runway etc, etc. Nor do I have a problem with discarding standard procedures in ‘non standard’ conditions. The HS748 accident suggests that the cues invoking surprise might not have been considered or realized in training – and perhaps not even possible for any aircraft type.
These are the areas that require improved RTO training. The Training Aid provides a range of situations and considerations but perhaps fails to provide adequate guidance on probabilities of occurrence and human behavior aspects.
The introduction/use of ‘if unsafe or unable to fly’ IMHO is detrimental to safety. GlueBall touched on this (#4); it is unlikely that sufficiently reliable cues or time will be available for a decision. Environmental issues, i.e windshear, might be judged a failure in the decision to commence the take off, if so then we need to improve awareness and risk assessment. Similarly a truck or even an elephant on the runway (I’ve experienced one of those) is a real but low probability event, but again aspects which might be determined before V1 thus avoiding the possibility of rejecting ‘above V1’.
The report (#1) appears to identify problems with pilot’s inability/reluctance to change the decision/action process at V1. Perhaps because of poor speed awareness, or where correctly deciding to reject before V1, slow actions or incorrect technique result in an overrun; again, these are basic training issues. However, IMHO the dominant issue is dealing with surprise – shock, even fear, an aspect which can be in most abnormal / emergency situations. Thus, how do we train pilots to manage surprise?
"...decision was very likely based, in part, on the perception of available runway to stop the airplane."
Be careful with that illusion, because past V1, especially at max gross weight, [eg:B747] the available brake energy is in La-La-Land. The brakes will go on vacation before the pavement ends...and then you just keep on rolling and rolling!
There is a tremendous amount of outcome bias in this thread.
What's a good decision by a captain? One that saved lives. What's a bad decision by a captain? One that cost lives.
The rightness or a wrongness of a decision in the minds of too many people is simply a function of the results that followed from it. But hindsight is always 20-20. That has absolutely nothing to do with safety as a process.
The key question is does any improvement in procedures or training make a difference in safety outcomes when the outcome of the relevant decision making is unpredictable in advance. Faced with a whole host of unknown factors successful outcomes often don't come down to training or procedures but old fashioned luck. Honest captains admit this.
Not that here are many honest people around, captains or no.
I don’t have a problem with the Captains decision in the particular situation, but what about a shorter runway etc,
- nor did the AAIB!
SP - mountainbear has summed it up neatly in his para 2. If we are to talk forensic 'what-ifs' years after ('shorter runway etc'), 'whatif' the wing had fallen off downwind, having had 3000 ft or so of tarmac to land on?
As I said in my previous last para - when the chips are down you make the decision and mountainbear's para then analyses it. Short of 'training' by popping a large paper bag behind PF in the sim, that sort of 'surprise' is what the human brain copes with, either successfully or not. (Interesting to see a scientist write the software for that decision.)
Losing 2 motors on a heavy 74 out of Brussels and stopping at or above V1 brings the probably inevitable crash scene inside the airfield boundary - right or wrong? I do not wish to judge.
V1 is a threshold where below this you have better chances to survive when you abort, above this you have better chances to survive when you continue. it is not a guarantee that all will be nice when you continue above v1. when e.g you have an all engines fail , wing seperation etc. you need not do discuss the V1 - but when this happens there is no procedure anymore that gives you a chance to handle this situation, you may simply crash and burn.
we have situations where the pilots ( e.g concorde crash) did all right regarding V1 and the outcome was killing all the people.
nevertheless- there is no time to think about this on takeoff roll and in the vast majorities of failures continueing above v1 gives you a better chance to make it.