There's an argument that says the 9/11 attacks couldn't have succeeded if the FAA and NORAD had only followed standard procedures, as at least some of the planes would have been intercepted in time. The 1999 case of Payne Stewart is occasionally used as an example of just how fast intercepts can be.
It's certainly true that there were initial media reports suggesting Air Force jets intercepted the plane only around 20 minutes after contact was lost. But this isn't actually what happened.
The NTSB accident report available online reveals this timeline.
Read this carefully and you'll notice a change of time zone, from Eastern to Central time. CDT is one hour on from EDT, so the lack of contact was first noticed at around 09:34, accepted as a loss of contact at 9:44, and the fighter didn't get to within 2000 feet of Stewart’s jet until 10:54. That's well over an hour between the controllers realising there’s a problem, to intercept taking place.
There's further confirmation of this in the statement to the NTSB of intercept pilot Chris Hamilton:
Around 39 minutes elapsed between Hamilton checking in with Jacksonville ATC and making visual contact with the Lear Jet.
The detailed NTSB factual report provides the following timeline.
The total time elapsed between the problem being recognised and intercept made was more than an hour.
A NORAD summary document found in the 9/11 Commission files confirms this timeline.
This situation was also recognised by some press and other reports, which provided more details (our emphasis).
And the 9/11 Commission added some comments of their own:
Putting these together with the NTSB report suggests the following points.
First, it takes time before ATC consider they’ve lost contact with a plane. The absence of any radio response was first noted at 9:33, but the controller continued trying to make contact for another 3 or 4 minutes.
Second, Nafeez Ahmed told us above that "procedures also require controllers to immediately alert the military to scramble fighter craft, if a plane deviates from its flight path and communication between the plane and controllers is blocked". But the Stewart case is proof that does not always happen. Radio contact problems began at 9:33, ATC considered the situation an emergency at 9:36, yet according to the reports the FAA only notified the US Air Force Rescue Coordination Centre until 9:45, and NORAD weren't alerted until 9:55 at the earliest. That's 19 minutes after the plane was considered an emergency.
And third, there's the NORAD response time to add on top of that. Here 13 minutes elapsed between SEADS being informed of the situation, and their finding two jets to scramble, then another two minutes before those jets were airborne. (It's conceivable that there may have been some overlap here: SEADS may have asked Tyndall to get two fighters prepared while they tried to find someone closer, then came back and asked them to scramble. The end result remains the same, though - a 15 minute wait).
34 minutes have now passed between a problem first being noted and fighters getting airborne. If this were duplicated on 9/11 then there would be a further few minutes before the fighters could reach their target, and this is a problem when you look at the amount of time available. The 9/11 Commission reported the gap between the “likely takeover time” (the earliest time anyone would have known about the hijacking) and the point each plane reached its final target as follows: Flight 11 (8:14 to 8:47 - 33 minutes), Flight 175 (8:42 to 9:04 - 22 minutes), Flight 77 (8:51 to 9:38 - 47 minutes), Flight 93 (9:27 to 10:04 - 37 minutes). (Source)
And those are the maximum possible times that might have been available for intercept. The reality is people had to become aware that the plane was hijacked, first. In the case of Flight 77, for instance, if you start the clocking ticking when the transponder was turned off, then that gives you only 42 minutes. Start it when American Airlines HQ knows the plane is hijacked, and you're down to 33 minutes.
What the Stewart case tells us, then, is that the FAA response to an emergency is not always quick, and if they performed in the same way on 9/11 then there's little chance that any of the hijacked planes could have been intercepted. And of course even if they had been, there was no shoot-down order: the only difference is that there would have been one or two more witnesses to the final moments of the flight.
Of course this is far too damning a conclusion for some to accept, and so they try to find new ways to minimise its effects. In the 9/11 Commission Report: Omissions and Distortions, for instance, David Ray Griffin comments:
It seems Dr Griffin is aware of the time zone issue, but can't quite bring himself to explain to his readers that, if true, the intercept took an hour longer than some people claim. Instead he talks of "confusion" without quite spelling out any example of that, in an effort to persuade you that the official NTSB report can't be relied upon.
Well, the timeline doesn't seem confused to us.
There's EDT in the first time, CDT (an hour later) in the second. The "7" following it is a footnote which reads:
It's not a typo, the report specifically mentions the moment the plane changed time zones, and uses the new zone in its timeline. This is arguably confusing - the fact that some journalists didn't notice the time zone change proves that - but it's not confused, and there's no doubt whatsoever that it's talking about an intercept that occurred more than an hour after the problem was first noted. (But please, don't take our word for it. Follow the above link and find out for yourself.)
Others use a simpler strategy, conceding that the Stewart intercept may have taken a while, but claiming that hijacking reports would have received a higher priority, especially after the first impact.
We might argue in return that systems under stress don't necessarily produce improved results: you might hope that individuals faced with an unparalleled national emergency might perform better than in the relatively quiet situation surrounding the Payne Stewart flight, but it wouldn't necessarily be surprising if the opposite were true.
We also suspect that there was nothing of a higher priority going on at SEADS or the FAA during the Stewart incident, and if so it would surely have received their full attention.
However, the key point here is the whole "but Stewart's situation was different and on 9/11 it should have been much faster" argument is speculation and conjecture. And people can put that forward just as much as they like, but it won't change the underlying facts: the Payne Stewart timeline does not support the idea that the FAA immediately informed the military whenever they had a problem, or that intercepts typically took place in 10 or 20 minutes.
Is the intercept time in the Payne Stewart case really typical? Perhaps looking at other cases can tell us more.
On January 10, 1980, football coach Bo Rein was travelling from Shreveport to Houston on board Cessna Conquest N441NC. At around 03:40 GMT Fort Worth Centre made repeated attempts to contact the aircraft, but without success. They asked a nearby Pan American flight to contact the plane instead. The Pan American crew did so at approximately 03:45 GMT, telling the pilot to contact Fort Worth on the assigned en route frequency. The Cessna pilot acknowledged this, and heard him attempt to check in, but reported his transmission as "very weak and barely readable". Fort Worth did not receive the transmission at all.
N441NC had climbed through his assigned altitude by the time the Pan Am crew reached him, and continued to climb until he reached his absolute altitude of around 40,000 feet. No further communications were reported with or from the pilot after his attempt to contact Fort Worth.
The NTSB case documents don't state when a scramble was ordered, unfortunately, reporting only that "F-4 aircraft from Seymour Johnson Air Force Base were scrambled to intercept the aircraft as it flew into the North Carolina aircraft". We do have a better record of the scramble times, though.
At approximately 5:03 GMT Aersospace Defence Command advised that they were scrambling two aircraft from Goldsboro (the Seymour Johnson flights).
"A few minutes later" the scramble was initiated.
Washington ARTCC began working the two fighters at approximately 05:22 GMT, vectoring them towards the Cessna, and they intercepted the flight at around 05:29 GMT.
The last time anyone heard from N441NC would have been around 03:47 GMT, then. We don't know when the military were alerted, but it seems there was no decision to scramble until about 05:00, 73 minutes later. ATC heard about this at 5:03, began assisting the fighters at 5:22, and the intercept was made at 5:29, 102 minutes minutes after the pilot's last communication and 109 minutes after contacting anyone on the ground.
Helios Airways Flight 522
Helios Airways Flight 522 (HCY 522 or ZU522) was a Helios Airways Boeing 737-31S flight that crashed on 14 August 2005 at 12:04 EEST into a mountain north of Marathon and Varnavas, Greece. The timeline for the incident is as follows (AAll times in EEST - UTC + 3h)
- 9am (0600 GMT) Helios Airlines flight ZU522, a Boeing 737, with 115 passengers and six crew members on board takes off from Larnaca International Airport in Cyprus, heading for Athens, Greece, and then Prague, Czech Republic.
- 9:37am (0637 GMT) The jet enters Greek airspace and is identified by Greece's Civil Aviation Authority.
- 10:07am (0707 GMT) Control tower at Athens International Airport is unable to establish communication with the plane.
- 10:20am (0720 GMT) Air traffic controllers notify their counterparts at Larnaca who say the plane reported a problem with its air-conditioning system before entering Greek air space.
- 10:25am (0725 GMT) Greece's Civil Aviation Authority notifies Defence Ministry's national search and rescue centre.
- 10:30am (0730 GMT) The Greek Defence Ministry issues a Renegade alert, a standard aviation procedure when a plane fails to respond to the control tower.
- 10:55am (0755 GMT) Two F-16 fighter jets scramble to locate airliner.
- 11:20am (0820 GMT) Fighter jets spot the Cypriot jet over Aegean island of Kea, but are unable to communicate with pilots.
- 11:25am (0825 GMT) Fighter jets approach the plane and report that the co-pilot of the Cypriot plane appears unconscious in the cockpit, while the other pilot was absent. Oxygen masks also seem to have been activated in the plane.
- 12:05pm (0905 GMT) Airliner crashes near coastal town of Grammatiko, about 40 kilometres north of Athens.
We need an official source for all these timeline details before firm conclusions can be drawn. However, as it appears at the moment, the lack of contact was noticed at 10:07 am. It took 23 minutes for this to produce a Renegade alert. It was then a further 25 minutes before jets were scrambled, and another 25 minutes before those jets spotted the missing plane. The total time from noticing the issue to a jet arriving in the vicinity was therefore 73 minutes.