Paul Holmes and Erebus

[ampan]

The crew weren't promised several redundant levels of safety. That was the promise made to the passengers, and one of the most important safety barriers was the crew itself.

The other members of the crew should have objected, but they were not given much a chance.

As for these NZALPA 'believers' who say that the same thing would have happened to them, this is what a C141 pilot did about 45 minutes after Capt Collins: "At the time we we navigating entirely by INS. We maintained 16000 feet until McMurdo picked as up on radar; as I remember, this was at about 38 miles."

It wasn't rocket science and it wasn't complicated, either today, or in 1979, or in 1949. When you're at 16000 feet, you won't hit a 13000 foot hill.

[framer]

It wasn't rocket science and it wasn't complicated, either today, or in 1979, or in 1949. When you're at 16000 feet, you won't hit a 13000 foot hill.

And right there, in one simple sentence, is a nugget you can tuck away inside your flight bag and carry with you your entire flying career.

If you are involved in a company SMS then there are plenty of more complicated lessons to take from this event. If you fly aircraft, an awareness that the systems always fail in some way or another and that you can never rely on

several redundant layers of safety

is all you need, because you are the last layer of safety. That is your job. That is why you get paid more than the First Officer.

[FGD135]

I just can't believe some of the silly statements that have been made recently in this thread.

When you're at 16000 feet, you won't hit a 13000 foot hill.

Silly, stupid statement that simplifies things so much that it ignores the important things.

... because you are the last layer of safety. That is your job. That is why you get paid more than the First Officer.

framer, this statement reveals that you do not understand the concept of safety levels. According to your statement, the other levels don't matter and you don't need them! You are saying the last level has the responsibility of preventing the accident.

If you fly aircraft, an awareness that the systems always fail in some way or another and that you can never rely on

So would you ever take off framer? With that view, you would never leave the ground! This is the third time I have had to put this to you. Here is the fourth: You MUST place a great deal of trust in others when you take flight. You have no other choice.

framer, you just make one motherhood statement after another.
I get the feeling that your statements are for the purpose of making *you* feel cozy that this accident wouldn't have happened to you. I think this is delusional and will contribute to a false sense of security.

I have been flying commercially in GA for 20 years but have no qualms in coming out and saying that I believe the result would have been exactly the same had I been the captain of that DC10 on that day.

Try saying this same thing to yourself. Say it with nobody within earshot if that helps. You will find that it doesn't hurt and will in fact set you free from those self-imposed ego constraints. You will then find yourself looking at the facts in a completely different light.

I think that for many pilots, there is a desperate belief that things would have been different had they been captain, and they will go to great lengths to support and strengthen that belief.

Bravo, Givelda, for the admission you made at post #221. Like Mahon, Vette and our Dozy, you can be trusted to look at the facts completely impartially. And thanks for that map.

After all, that was how the military pilots did it.

Another silly statement. Ampan, the military pilots weren't there for the sightseeing!

There is something quite striking about the arguments made by those that believe the crew had some responsibility for the accident: they never use the word "reasonable".

That word, reasonable, is a very important one.

To judge whether the crew were responsible, you first must ask whether they made any mistakes or broke any rules. They didn't, so next you must ask whether their actions were reasonable.

So what did they do that was not reasonable?

Putting a DC10 at 1,500' and 250 kts in a region you've never been to before? Not unreasonable when he believed he would be visual with a visibility greater than 20 Km. He had been told that the low cloudbase was only over Ross Island and that everywhere else was effectively CAVOK.

Descending where/when they did? Not unreasonable given that the real, if unwritten, SOP was for a flypast of Ross Island at 2-3,000 feet. Making this the real SOP was the fact that every other crew had done this, plus the reinforcement of the briefing, plus the reinforcement of the crewroom chatter. Bear in mind that Chippindale himself did not fault the pilots for doing this.

Not plotting their position on a topo map of suitable scale before continuing the approach towards Ross Island? Why should they have done this? Their problem was the low cloud, NOT any uncertainty of position.

Were they supposed to have somehow realised that, in fact, there was actually, uncertainty as to their position? Just where, reasonably, were those prompts/cues to have come from?

[framer]

framer, this statement reveals that you do not understand the concept of safety levels. According to your statement, the other levels don't matter and you don't need them!

Nah not really FGD. I was trying to highlight a practical lesson that you can use as a pilot. I made reference to all the other systemic lessons that are in this event, people pretty much accept all those lessons, there are just a few who refuse to accept one very important lesson and thats very concerning.

I get the feeling that your statements are for the purpose of making *you* feel cozy that this accident wouldn't have happened to you. I think this is delusional and will contribute to a false sense of security.

Not at all FGD, someone asked me a week or so ago on this thread what would have happened if I was flying it and I honestly answered that I didn't know....doesn't really tie in with your cozy theory now does it.

[framer]

I just had a look back through the thread and it was you FGD who asked me what would have happened if I was flying it and I answered that I don't know.
I'm pretty keen on taking lessons out of incidents and accidents and I've put forward what I believe to be a good lesson from this accident, it was

Minimum Safe Altitudes are your last safety net that will protect you from mistakes and errors that yourself and others make.

I feel like if we all put forward one lesson, the one we personally think is the biggest from this event, one that can be of use in our careers, there might be some really good thought provoking stuff. If you could join me in that FGD it would be appreciated and possibly useful to myself and others.
Framer

[ampan]

Another important lesson is to stick to the plan, unless there’s good reason not to. In this case, the original plan was to check the weather about 100 miles or 30 minutes out and then make a decision. When they got to that point, Capt Collins checked the weather at Mac Station and it was, basically, no good. So he would have been thinking about going to the clear areas coming up on his right (eg, Taylor Valley), then going back to Cape Hallett and on towards the South Magnetic Pole, keeping a careful eye on the fuel. But then he was offered a radar letdown, the radar having a range of about 40 miles. Although that wasn’t part of the briefing, the captain would have performed a radar letdown on numerous occasions in the past, and he responded with “* * that’s what we want * * *”. In Vette’s reconstruction, he fills in the gaps and confirms that an experienced DC10 captain would have welcomed a radar letdown in those circumstances: “Collins expressed his relief: ‘Crikey. That’s what we want to hear.’”(p130)

As to the rules, the specific rule for the sector, and the general rule re AINS, was to stay above 16000 feet. VMC flying was an exception to both, but we know, and this is an undisputable fact, that Capt Collins knew that VMC was not assured when under the cloud. But that would not matter once his position had been confirmed by the radar operator. We also know, again as an undisputable fact, that Capt Collins was told, at the briefing, that the waypoint was at Mac Station. But again, the position of the waypoint would not matter once the radar operator had confirmed his position. The point is that when he accepted the radar letdown, he could not have been intending to go below MSA before his position was confirmed, because that would defeat the whole purpose – and the rest of his crew would assumed the same.

Against that background, what he did next is difficult to understand. At about 45 miles out, before he had even got within range of the radar, he says “I’ll have to do an orbit here I think … Well actually it’s clear out here if we can get down.” That was the first indication that he was going to go down, and rather then getting anyone else’s input, he immediately started to descend, such a minute later, it’s: “I’ll come round there and set that down (at) ten thousand”. So in less than two minutes, he went from 18000, through the MSA of 16000, down below 13000, being the height of a known hazard, right down to 10000 – without any reason and contrary to his own plan for the approach. And when he doesn’t get confirmation of his position from the radar operator, his solution is to keep going down, all the way to 1500 feet, rather than returning to MSA.

[framer]

Good point ampan, I hadn't actually thought of that. It reminds me of several 'beat-up' accident reports I've read where the beat-up was done on a whim while disregarding the original flight plan. I would like to see a list of several lessons we can take from this, we don't all have to agree with them, but they will at the very least make us think and then we are getting something from this discussion greater than arguments. Thus far we have;

From Ampan

Another important lesson is to stick to the plan,

From Framer

Minimum Safe Altitudes are your last safety net that will protect you from mistakes and errors that yourself and others make.

From Dozy

Any safety-critical system is inevitably more complicated than the human brain will initially think it is - always assume that your best may not be good enough and make sure every contingency is covered".

[chris lz]

Originally Posted by FGD135

NOT any uncertainty of position.

But that's debatable. Their certainty was based on trusting AINS. It probably should not have been. The logic makes more sense after reading framer's response to my AINS question: he stated he's never been aware of any SOPs that allow AINS to be used as a substitute for ground based aids below MSA, primarily because of the drift factor. Wasn't that what the crew were in effect doing? Whether that was reasonable I guess depends on how high one thinks taking that risk would be. Would framer or anyone like to take a stab at this? How risky, in general, would you view Collins' use of AINS that day? Recklessly negligent? Understandable if not optimal? Or somewhere else on the danger curve?

[FGD135]

framer,

I find it a bit of a worry that you think the main lesson to be taken from this accident is:

Minimum Safe Altitudes are your last safety net that will protect you from mistakes and errors that yourself and others make.

Neither Mahon nor Chippindale considered the where/when of their descent to be an error/mistake. You seem to be the only one with this opinion.

And, this is NOT a lesson from the Erebus crash. This was a lesson that was presented - and learned from - way earlier in the history of powered flight.

And, the premise is false. MSA is just one of several safety layers. It will NOT necessarily protect you from the mistakes of yours or others.

Finally, I will suggest that by proposing this, you are looking more at the "whats", rather than the "whys".

framer, I asked you to summarise the opinions of those ANZ Captains who stated that the accident wouldn't have happened had they been at the controls. Please do so.

Another important lesson is to stick to the plan,

Another non-lesson! "Sticking to the plan" might have been a lesson in the first few years of powered flight. I can imagine a commander of the Royal Flying Corps finding that his charge crashed his Sopwith Camel into a tree because he didn't stick to the plan, but nowadays, we know that sticking to some plan can be just as dangerous as not.

Take fuel exhaustion and "VFR into IMC" accident categories, for example. Both of these frequently involve situations where a change of plan would probably have saved the day.

And besides, Collins WAS sticking to the plan! The plan was to give the passengers some lovely sightseeing - the first stage of which being a flypast of Ross Island at 2-3,000'.

To me, the lesson from this accident is that sometimes, those outside aviation will do a far better job of identifying the failures that led to the accident than those inside!

Collins' desire to get visual was perfectly normal and understandable and was in accord with the commercial obligations ANZ had towards the passengers. He broke no laws and contravened no established wisdom by doing this.

He did what he had to do! 99.9% of commercial flights terminate with a visual stage, but just how one gets visual depends on the environment and circumstances.

I had been going to list "organisational entropy" (OE) as the main lesson from this accident, but I'm sure there would have been accidents before this where that factor played a major part.

OE is an extremely difficult failing to rectify. The aviation company I work for at the moment - and all the previous ones - all have OE to the same extent as that which prevailed within Air New Zealand in 1979.

chris lz, my apologies, but I have run out of time for the moment. Next post I will address some of the questions and issues you have patiently been raising.

[prospector]

And besides, Collins WAS sticking to the plan! The plan was to give the passengers some lovely sightseeing

If the weather was suitable, and it obviously was not.

The more important plan was to get them home again.

This was a lesson that was presented - and learned from - way earlier in the history of powered flight.

And the first commandment applied then, it does now, and always will,

Thou shalt not make a stuff up or the ground will arise and smite thee


.

Neither Mahon nor Chippindale considered the where/when of their descent to be an error/mistake. You seem to be the only one with this opinion.

Mahon obviously did not, it was the backbone of the Chippindale report.

Initiated a descent below both the IMC 16.000ft and VMC 6,000ft minima for the area in a cloud free area

And I can assure you many airline experienced pilots agree with that opinion.

.

The aviation company I work for at the moment - and all the previous ones - all have OE to the same extent as that which prevailed within Air New Zealand in 1979

.

So it would be a reasonable assumption that if we hear you have flown into a mountain, it would not be your fault but Company "OE"???

[framer]

this is NOT a lesson from the Erebus crash. This was a lesson that was presented - and learned from - way earlier in the history of powered flight.

I agree whole heartedly that it is a lesson that has been learnt before, and then forgotten. What I am getting at is , how many times do we have to learn it? I can think of three New Zealand crashes off the top of my head that would not have ocurred without a willful violation of a safety altitude.

MSA is just one of several safety layers. It will NOT necessarily protect you from the mistakes of yours or others.

True but if we are waiting for one lesson that will prevent all crashes we will be waiting a long time. Respecting safety altitudes will definately protect you from some crashes. That is worth while isn't it?

The conversations I had with two Air NZ Captains about this contributed to the opinion I have formed about this. I'm not speaking on their behalf, just mine so I won't 'summarise' any more than that.

Finally, I will suggest that by proposing this, you are looking more at the "whats", rather than the "whys".

I can see why you feel like that. I am focussing in on one element because it is disputed. The other reasons (and there are many in this case) are not. It is suggested by Paul Holmes, Mahon, Dozy, yourself, and obviously many others that the crew bear no responsibility. I personaly think that is a dangerous position to take so I put my view forward in the discussion.

[framer]

Another important lesson is to stick to the plan,

Minimum Safe Altitudes are your last safety net that will protect you from mistakes and errors that yourself and others make.

Any safety-critical system is inevitably more complicated than the human brain will initially think it is - always assume that your best may not be good enough and make sure every contingency is covered".

To me, the lesson from this accident is that sometimes, those outside aviation will do a far better job of identifying the failures that led to the accident than those inside!

..............................

[ampan]

FDG135 #290: Of course you change the plan if you have to, but he didn't have to. If he was so keen to see McMurdo Station, he already had the best method, which he had announced to the passengers, which included the two extra crew members. Look at the situation from F/O Lucas’ perspective: The radar letdown was announced over the PA, which would have made perfect sense to him. He would then have noticed the first descending orbit, followed shortly thereafter by another announcement over the PA: "Captain again ladies and gentlemen we're carrying out an orbit and circling our present position and will be descending to an altitude below cloud so that we can proceed to McMurdo Sound ... thank you". F/O Lucas, as well as the spare F/E, would have assumed that the radar operator had confirmed the aircraft's position. This could explain the some of the garbled background comments, in that one of the spare crewmembers might have gone up to the cockpit and discovered that they were still trying to contact the radar operator, when they were already below MSA. Whatever the situation, Capt Collins’ quickly-hatched plan was a bad one, and the reason came out of his own mouth.

The other obvious reason for waiting (for the radar operator to confirm his position) was other aircraft. There were helicopters flying on the day and there were other fixed-wing aircraft around – and he was flying without flaps at 260 knots, and went from 18000 all the way down to 2000. Putting the terrain to one side, I would suggest that most DC10 captains of the time would have wanted to ensure that they were on the radar screen before going down through 16000 feet of airspace, whatever the radar operator had to say about it. (In any event, he was only cleared to 2000 feet, yet dropped another 500 feet without any clearance from the tower, right into the band that the helicopters were probably operating in.)

[FGD135]

Mahon obviously did not, it was the backbone of the Chippindale report.

Ah yes, looking through Chippindale's report, I see that he did indeed have an objection to the descent occurring when/where it did. The contrary had been stated earlier in this thread by someone else and I had just been repeating that.

But this does not change my views as to whether that was a reasonable act on the part of the flight crew. I still believe it was, given all the circumstances.

Those (including Chippindale) that insist he "broke an SOP" by descending where/when he did need to realise that that "SOP" was not really an SOP.

An "SOP" is not an SOP when nobody observes it. Not only was nobody observing it, its disregard was condoned by the CAA. And as the RC revealed, ANZ had distributed a pamphlet to every postal address in NZ, advertising the fact that the flights were passing Ross Island at low level.

Some SOP!

... if we hear you have flown into a mountain, it would not be your fault but Company "OE"???

It would take an investigation to determine whether it was my fault or the company's OE. I just hope it would be somebody as open-minded as Mahon doing the investigation.

I think it would be fair to say that ALL pilots hope somebody like Mahon will investigate their crash.

chris lz,

Re "certainty of position". This subject is largely philosophical.

No pilot is ever 100% certain of their absolute position - and the word "position" itself has aspects whose discussion are more suited to philosophy!

The degree of certainty varies, and the criteria for "certainty" varies between flight phases. What would pass for certainty in the middle of a long overwater crossing would not suffice during an instrument approach, for example.

The question of position certainty is really a question of degree. There is a dividing line between "sufficiently certain" and "insufficiently certain" and precisely where that line is drawn is a matter of opinion.

For a pilot to be "certain" of his position, he will usually have built a case based on several independent pieces of evidence (inputs). (E.g. INS position roughly agrees with VOR/DME; Distance to run agrees with time since last positive fix, etc - there are dozens of combinations). It would be a rare pilot to use only one input.

Giving Collins "certainty" of present, recent and near-future position would have been the following three inputs (or "facts"):

1. The AINS position was visually confirmed over Cape Hallett;
2. The AINS would accumulate no more than 1 mile of error between Cape Hallett and McMurdo Station;
3. The route had been plotted on a map the night before;

In Chippindale's opinion, the pilots of TE901 were not sufficiently sure of their position. What he really meant was that, within the context of the environment and what they were intending to do, they should have possessed a higher degree of certainty.


And this is all very obviously true, but is something that can only be said in hindsight (Mahon's opinion).

Their certainty was based on trusting AINS.

Yes, but this was perfectly reasonable. I cannot conceive of a navigator that would not have had a similar degree of "trust" for the AINS at that point.

The AINS had got them to the point where they could transition to visual navigation. The accident occurred whilst the aircraft was being navigated visually.

[Desert Dingo]

Ampan: It appears you have a problem with reading comprehension regarding the CVR record.

Mac centre never cleared TE901 to descend to 2000 feet.

What Mac center did do was to clear TE901 to descend maintaining VMC and keep them advised of what level was being maintained.

2000 feet was a level that F/O Cassin said they were descending to at a time when he had radio contact on HF. Some time later they descended further to 1500 feet (which seems appropriate as McMurdo had reported overcast at 2000 feet). F/O Cassin was never able to advise the final level being maintained because of the radio problems.

I’ll refresh your memory and even use Chippindale’s version of the CVR.

0032:07 RDO-2 (HF)
We'd like further descent and we could orbit in our present position which is approximately 43 miles north, descending in VMC.

0032:08 CTR (HF)
Roger Kiwi New Zealand 901, VMC descent is approved and keep Mac Center advised of your altitude.

0032:10 RDO-2 (HF)
Roger, New Zealand 901, we're vacating one eight zero. We'll advise level.

And later when Mac centre requests a cloud layer report

0044:57 CTR (HF)
Understand the bases are at ten thousand?

0045:00 RDO-2 (HF)
Affirmative. We are now at six thousand descending to two thousand and we’re VMC.

[Do a text search and that is the only time “two thousand” appears in the radio transcript.]

So when you write

(In any event, he was only cleared to 2000 feet, yet dropped another 500 feet without any clearance from the tower, right into the band that the helicopters were probably operating in.)

I call BS. You are making false assertions without a shred of evidence to back it up.

[prospector]

Affirmative. We are now at six thousand descending to two thousand and we’re VMC.

Would appear to be rather unusual adding "we're VMC". It is the only way they could be at that altitude without getting a descent clearance, and they would not have got a descent clearance as such before they were identified by radar'.

There was only, as standard practice for VMC descent, advisories given to the controller as to their altitude and intentions.

In a VMC descent, responsible for their own terrain and traffic separation as they had requested, all the controller had to provide was traffic information.

There was a danger operating at that altitude, as has been mentioned before on this thread. One of the reasons for the minimum descent altitude, 6,000ft, was that at that altitude there was no danger of bird strike with the numerous skua gulls, big birds capable of causing a lot of damage.

Wonder what would have been the outcome if one of the flights that did descend below 6,000ft did have such a birdstrike that caused substantial damage, which at 260kts plus it certainly would have, and the reason given was that "The controller invited me to do a low level run". Would not hold much water as areason..

The contrary had been stated earlier in this thread by someone else and I had just been repeating that

.

Would not worry to much about that, would appear to happen quite often on this thread.



.

[framer]

[chris lz]

FGD135,

Thank you for your reply. I am not one of those who say this was just another ordinary CFIT. So whatever Collins misjudgements regarding AINS may have been, I cut him a lot of slack. The coincidence factor was pretty high on that day. Whether Collins' confidence was reasonable is still an open question in my mind. But I doubt he could be accused of gross negligence. The airline itself described him as "meticulous."

The question marks for me remain as follows:


-"certainty" - of course, it's a question of degree. But acknowledging this doesn't itself show the crew had "reasonable" certainty.


-Cape Hallet - If there had already been a drift of 3 or 4 miles, would they have noticed it at that point? So we have the possibility of 4-5 miles drift by the time they were in Lewis Bay.

-However acurate AINS was, we have it from prospector that it was never intended as a substitute for ground aids. That implies they used it improperly.


-They were not picking up signals from the TACAN.

-They had no VHF contact.

-Regardless of what one thinks of the above, I suggest it's common sense that before beginning a VMC descent in a mountainous area, without any ground aids available, you first visually identifiy the mountain(s) itself. To me that just seems intuitively basic to any philosophy espousing "fail safe flying."


When you combine all this together, I can certainly believe that many pilots would not have did what Collins did - and that's in fact what many have stated. Why should we be out to condemn them, especially if they are telling the truth? Perhaps we should at least say that maybe this is an irresolvable grey area. Possible?

Cheers

[FGD135]

... your thoughts on what Nigel Roberts says 47 seconds into the youtube video

He describes a flight that appears to have been observing the "16,000' descent limit". He goes on to describe the next flight as one that clearly wasn't.

Yes, the odd flight does appear to have observed that "limit", but in Mahon's opinion, most didn't.

I suggest it's common sense that before beginning a VMC descent in a mountainous area, without any ground aids available, you first visually identifiy the mountain(s) itself. To me that just seems intuitively basic to any philosophy espousing "fail safe flying."

It is not necessary to see and identify ground obstacles before switching to visual flight, but you must be in VMC at that point (in order to see that you can continue in VMC). You don't even need to know your position - you can be totally lost!

The only requirement re visual flight is that the necessary meteorological conditions exist.

Terrain and other obstacles beside the flight path don't matter. The only thing that matters is that you can see a sufficient distance along the path the aircraft will take. Wise aviators will have an escape plan up their sleeves for some scenarios involving visual flight, however.

At the point where Collins realised he didn't have adequate visual reference it was too late for any escape.

[ampan]

Chris Iz - They checked the black box to see what happened at Cape Hallett. This is from MB Wylie, an NZ government accident inspector:

1. "The aircraft was overhead the Cape Hallett waypoint at 2351 hours GMT at flight level 330. This altitude, plus or minus 45 feet, was maintained throughout.

2. During this period the aircraft was flown continuously by the number two autopilot engaged in the Command (normal on) mode.

3. Prior to turning into the Hallett – McMurdo leg, the role mode of the Flight Guidance System was in ‘NAV TRK’ and the aircraft heading was 327.5 degrees plus or minus 0.5 degrees to maintain this track.The Grid track for this leg was 324 degrees, the difference being due largely to drift.

4. The aircraft overshot this turn requiring it to make an intercept on the out-bound track. This is unusual, as the flight guidance system would normally compute its rate of turn based on the selected bank angle and commence the turn at a point that would allow the in-bound and out-bound tracks to be tangential to the radius of the turn; ie, the aircraft wouldn’t actually reach the turning point. Bank angle in this form was limited to five degrees. One possible explanation for this overshoot is that as the aircraft started to turn, the crew reduced the bank angle from a normal position of 20-25 degrees to the five degrees setting. This would allow better viewing from both sides of the aircraft, but because of the small bank angle, the aircraft would overshoot the next track.

5. At 2356:43 hours the NAV TRK mode was again displayed as the aircraft locked onto the Hallett-McMurdo track. The initial Grid Heading was 360 degrees on this leg. This backed one degree as the aircraft corrected for changes in drift. The Grid track on this leg was 358 degrees.

6. At no stage was the “Heading Select” mode engaged. This would occur if the aircraft was manually steered through the autopilot.

7. If the crew had manually updated the system at any time after leaving the Christchurch area the words “Manually Updated” would be displayed above the ‘present position’ shown on the performance page of the CDU. These words were not displayed.

8. Based on the known error in the AINS at impact; the error in the system one hour earlier when the aircraft was over Cape Hallett would be less than three nautical miles. With the aircraft at altitude it is doubtful if the crew could visually improve the system accuracy by making a manual update.

9. The crew of NZP made a number of radio calls (4 on VHF2, 6 on HF1, 11 on HF2) during the period. The voice recorder record does no extend to cover this period so the context of these messages is not known."



One slightly interesting thing about the waypoint is that it's right on Cape Hallett, which, under the Antarctic Treaty, is a 'no overfly' area. That might have been different in 1979, but it's more likely to be yet another SNAFU by AirNZ.