Midway revisited
 

With the recent NTSB report placing a great deal of blame on the pilot for the SW Midway accident – I thought I would bring up a confusing subject in the hopes that experts from the technical forum could shed some light on it.
As this accident took place in the US I would appreciate limiting this discussion to FARs and/or US Ops Specs.
It is clear in FAR 121 (121.195) that an aircraft may not depart so as that it will not be able to land within 70% of the available destination (or alternate) runway length. I can find no FAR that refers to actually landing on a runway within 70% of the available length -once you have departed. In fact many of the “inflight” performance #’s provided to the pilot refer to distance to 50’ over the threshold “to stop.”
There has been a long standing Ops Specs referencing visibilities below 4000’ – which muddy the waters even further – they say
“A pilot-in-command of a turbojet airplane shall not begin an instrument approach procedure when the visibility conditions are reported to be less than ¾ statute mile or RVR 4000, unless the following conditions exist:
(a) Fifteen percent additional runway length is available over the landing field length specified for the destination airport by the appropriate Sections of the CFR.”
What do they mean by “the appropriate Sections of the CFR.” Are they referring to the landing distance specified for the destination airport before departure?
If this is true (and why do they not just specify the FAR here) then any computations of a performance computer would need to also provide the additional 30 % buffer + an additional 15% for WET and an additional 15% for the reduced Vis and from my understanding of these computers – an additional additive for braking action !!!!
Yet I have seen educated and competent Airline Pilots unequivially state that once you are inflight – this 70% requirement no longer applies.
So my question is:
What did the computer tell the pilot? Given the conditions – was he capable – of stopping only on the surface provided? Or was he capable of stopping on the surface provided, plus all of the above mentioned additives – which means that he should have been able to stop in about 50% of the runway available?
I have read many stories about this accident and have seen numerous statements that the A/C should have been able to stop on the runway surface – even with the reduced braking conditions – yet when I look at some of the performance #s it appears that they mean with no additives beyond those additives – not considering an additional 30% or the additional 15% as specified in the Ops Specs.
I used to instruct in a 737NG and as an enthusiastic Instructor would pester the Engineers with questions about the auto-magic. When they were tired of listening to me they would say “it’s in the algorithms.” It used to piss me off – but I must confess – later in my instructing career – I used to respond with that answer often. But did this pilot get led down the path with these algorithms. What is the computer really saying – that with the given conditions – he will be able to stop with his nose wheel on the end of the runway – if he did everything perfectly? Granted – this is what we do when we depart on a field length limited runway and we have an engine fail at V1 – but that is for a once in a million chances with no time to make a decision – this is a situation where there is time to make a decision – but I’m afraid most – including me do not have a clear understanding of the regulations – nor what this computer is saying.
Please specify a specific regulation when replying. Interpretations are plentiful in this area – I’d like to stick to facts. I was surprised by the harshness of the NTSB’s report of the pilot’s blame in this – if it is how I see it as far as ambiguity – I think he got a raw deal.
LK

You are asking a bunch of different questions in your post. I am not familiar yet with all details of the report, but I did read the preliminary safety recommendation (January 2006) and commented on it in Risks:
http://catless.ncl.ac.uk/Risks/24.16.html#subj9.1
Some of your questions are already answered by that, as follows.

The computer was calculating stopping distance based on immediate deployment of reversers. Stopping distances as calculated in the operations manual do not take into account reverser deployment. So there is a mismatch of information there which was not explicitly stated as the pilots were using the tool. Also, the "WET-FAIR" runway-conditions parameter that the pilots entered into the calculation obviously did not accurately represent the actual state of the runway with regard to available friction.

With these two assumptions, one mistaken and one not fulfilled during the landing, the calculation said the aircraft would stop with 560 feet of runway remaining.

The NTSB pointed out that had the reverse-thrust "credit" not been used, the calculation would have indicated that they did not have enough distance available to land, with either runway condition of "FAIR" or especially of "POOR", which it was.

That all does not have to do with regulations. That has to do with a calculating tool using assumptions which do not match the pertaining conditions (including delayed reverse thrust). One is going to have problems there no matter what the regs say.

PBL

- they do in the Boeing QRP for the NG. I do not know what Ops Manual figures SW use.

Loke - looking at the NTSB Pdf files, it appears this crew 'expected improved stopping using reversers' which does indicate a training (or even Ops Manual) issue since the Boeing figures I use assume use of No 2 detent, but it also states that 'other crews are aware' of this need. The NTSB safety recs are interesting in particular 4) and 5) as you point out

4) Require all 14 Code of Federal Regulations Part 121, Part 135 and Part 91 subpart K operators to accomplish arrival landing distance assessments before every landing based on a standardized methodology involving approved performance data, actual arrival conditions, a means of correlating the airplane’s braking ability with runway surface conditions using the most conservative interpretation available, and including a minimum safety margin of 15 percent.

5) Immediately require all 14 Code of Federal Regulations Part 121, part 135 and Part 91 subpart K operators to conduct arrival landing distance assessments before every landing based on existing performance data, actual conditions, and incorporating a minimum safety margin of 15 percent. (Classified “Open—Unacceptable Response” and “Urgent.” Supercedes Safety Recommendation A-06-16.)

My bold - and no mention of visibilty

A-06-16 (Previously issued)
Immediately prohibit all 14 Code of Federal Regulations Part 121 operators from using reverse thrust credit in landing performance calculations. (A-06-16)

This recommendation (previously classified “Open—Unacceptable Response” on May 8, 2007) is classified “Closed—Unacceptable Action/Superceded” by Safety Recommendation [5] in section 2.3 of this report.

That is new to me; thanks for the heads-up. What I wrote is what I had understood back in January 2006. I find implicit, but not explicit, confirmation in the following two findings:

It sounds to me as though "Boeing's recommended airplane performance data" do not include credit for thrust reverse.

Can we confirm this?

PBL

- it seems pretty clear to me and to the NTSB - 'other crews' knew this, this crew is thought (either not to have planned for it or) not to have known it according to the report.

- not sure what you mean by that. As far as we are concerned, these figures are in the Boeing QRP.

A check of the NG Flight Planning and Performance Manual (www.myboeingfleet.com) shows normal configuration stopping charts that calculate reference stopping distance (130,000 lb) that includes credit for two thrust reversers. Adustments are then applied (as required) for one reverser only or no reverse thrust.

At my company, the manual (FOM) shows only the distances without credit for reverse thrust.

JD - I think that is an excellent way to do it. Otherwise, flying NG and Classic there is always the big 'Gotcha' waiting to bite.

Lets face facts here.
The Southwest crew concerned landed on a limiting runway, the condition of which was poorly reported.

Has happened before, more than likely will again.

Beware.
You had better be darn sure of the landing distance required, otherwise BIG problems result.

Sorry folks, them's the facts.

The crew, the company, the FAA...but yeah, the crew.

Revisted - again
 

411A, et al,
I misspoke when I cited FAR121.195 – I said that the requirement for Takeoff was to be able to land within 70% of the Arrival Runway – it is of course 60% for a Turbojet. This, understanding that, in all likelihood, an additional 15% had been added to that for forecast wet conditions.
You stated, “You had better be darn sure of the landing distance required” – and I agree with you.
My question – and I have an entire bottle of wine waged on this, so you can imagine how important it is:
Is there any regulation, Ops Spec, or “legal” interpretation of the FARs that mandates the additional runway required for the specific conditions at the time of landing. The NTSB says that the computer advised the pilots that they should have had an additional 560 feet before the end of the runway to stop the A/C. This is nowhere close to the % increase required by 121.195. The NTSB does not fault the pilot on not leaving enough margin – they fault him for not calculating the information correctly and for the problems with the thrust reverser.
So – again – is there a specific FAA document, CFR or anything from the FAA, that requires a pilot – or an operator – to use pre-takeoff margins (or any other specifically prescribed margin), to the actual conditions upon arrival on that runway.
I understand the 91.13 and the 121 regulation, saying that the pilot cannot continue into a dangerous situations Regs – not talking about that here.
Just to be clear – if the pilot had calculated everything correctly on his approved computer and had been advised that he could land on the runway provided with no margin, would he have been legal?
Thanks for any comments or information,
LK

Once airborne you are only required to calculate and verify that you can stop within 100% of the available runway, the other figures quoted are for dispatching purposes. The NTSB proposes the FAA require a 15% margin be added to that in-flight requirement going forward. At SWA that has since been adopted. The onboard computer that night showed 560 feet stopping margin based on FAIR braking, and 30 feet margin based on POOR. The number included credit for thrust reverse as per Boeing data on the NG series of 737 aircraft. Braking action and tailwind proved worse than expected and when combined with lack of thrust reverse application in the initial landing rollout made the calculations invalid and led to the overrun.

Once airborne you are only required to calculate and verify that you can stop within 100% of the available runway, the other figures quoted are for dispatching purposes.

As with any decision process, consider

(a) if the situation is a real emergency and action is required right NOW, then one needs to know where the worst case boundary fence is located and do the best one can ... ie what is the 100 percent figure and one expects to overrun the end a little ...

(b) if one has an emergency but resolution is not required right NOW, and there are several options available to the resolution process ... would it not be sensible to consider the certification requirements .. ie, in this case, if one had the choice of, say, three runways to recover to .. then one might prefer one of the runways which had the certification factor available ... just one less argument to face at the enquiry .. ?