G-BNLL report published
 

The original thread about this BA 744 colliding with a building at JNB was moved from here to the African Aviation forum after it eventually decended into triviality.

The accident report has now been published by SACAA and it contains lessons for all of us, so maybe the moderators will allow discussion to return here rather than remain quietly hidden in the Africa forum.

http://www.caa.co.za/Accidents%20and...ports/9257.pdf
It has taken a long time but this is an excellent report well worth reading.

My quick summary : the crew were suckered into a mistake by faulty taxiway lighting, as usual there were many contributing factors.

Very interesting reading. Two points.

To those who mutter about the time taken to see the official reports - take one hour to read this report- all 130 pages- and see the detail involved for a relatively minor incident. Well done SACAA.:ok:

Secondly, as expected, a chain of errors,
Personally I find the signage/lighting deficit slightly more significant than the briefing error. (Not well done Acasa.) :=
But my experience of ORT is now more than 15 years old. I recollect it being quite hard work at night.

Flying Bean is online now Report Post

Couple quibbles with the report
 

Engine position mis-identification on Page 23
If you don't know how 744 engines are numbered, ask someone.

Communications description on Page 27

And my browser shows a date at the bottom of each page which is 5 months and 12 days prior to the accident.

Something's lost in translation here.

That's simply the date corresponding to the latest version of the proforma, nothing to do with the contents.

Here's another example with exactly the same footer: http://www.caa.co.za/Accidents%20and...ports/9315.pdf

The PDF containing the BA report is datestamped 19 June 2015.

Failure to NOTAM maintenance work that resulted in lighting being unserviceable (report para 2.1.21.4) := - that's gotta be high up the list of causal factors, hasn't it? :ugh:

I have ploughed through it. The oddly written introduction/Executive summary was clearly written by somebody different to the main report which is by and large well written and incredibly thorough. It would have benefited from some additional proofreading and tech review but the typos don't in any way alter the conclusions which are as succinctly summarized by the Ancient Geek.

The tensions between ACSA and SACAA are evident and there is some finger pointing in a way you wouldn't find in an AAIB report. Its slightly odd that the dual role of SACAA involves it mentioning itself in the context of the incident which it is investigating but you'd be hard pushed to find fault with the particular instances where that occurs.

The conclusions of OverRun in the PPrune thread at the time that "M" was constructed as a Taxiway were vindicated, as was the decision by BA to return the flight crew to flying duties soon after the incident.. the picture indicating the failed centerline lighting and signage onto Bravo was pretty damning. Jury is out in my mind on whose fault it was that the lessons were not learned or engineered out from the carbon copy incident of a previous BA 744 incident in the same location.

Worth the wait.

I am not saying that it would have ended differently with me there, but one has to admit that it doesn't look to good to have written off a 747 due to a couple of burned out taxiway lights and a sign not lit. This is in Africa.

First 45 degree turn to the left. There were further green lights on that taxiway and a yellow line. What about the blue lights.

Apparently there was a briefing page 10-P6 with a caution about this area being confusing that was not read. That doesn't help.

Frequently there tends to be a hint that something is going wrong prior to the actual damage happening. It was noted that the taxiway was narrower than expected all of a sudden and a nearby building that looked too close. You have to do something at that point. it just shows how you really do have to be ready as PNF to jump on the brakes even at a major airport.

After this accident and reading about others, I like to look at the taxi chart for the briefing and mention if the taxiway goes beyond the end of the runway or similar and brief it.

An airport moving map(AMM) installation would have given the crew the tools to overcome any situational awareness issues. But that involves money and a commitment to spending money to prevent incidents like this. An example of an accident that likely would have been avoided if this company had that commitment.

My hat off to Accident and Incident Investigations Division (of SACAA). Well done, and as Pinkman says - incredibly thorough.

The delay in the release of the report speaks volumes about the pressures that must have been applied. It is a pity that the summary forgot to adequately mention the missing taxiway lighting and darkened sign, and still gets confused about the difference between a taxilane and a taxiway. But the body of the report does a good job and has lessons for all of us to ponder.

Exec summary / Briefing
 

OverRun makes a good point: the executive summary is at odds with the main report. I was being kind in my comments.

JammedStab: your point about the briefing is well made; the crew had carried out a thorough briefing and hazard analysis for the route that they had expected to take and were clearly surprised by the initial instruction to "start, push, and face South". Would the expectation be that they then stopped and re-briefed for a new route (which had not been given to them in its entirety at that point)? What happens if the route is given piecemeal? Not being defensive, just trying to understand what is considered "practicable". The crew thought they were on a taxiway and it would be interesting to do some human factors analysis.. when the report says "lack of situational awareness" what it really means is that a combination of high workload, visual miscues (lights and taxiway construction), and probably confirmation bias, led them to continue.

Which way does Bravo turn?

It turns Left

Never been able to turn up the airport and board the aircraft in 5 minutes at JNB.

I would think that while at the gate during the briefing, one can brief the expected route and when there are parallel taxiways, look to see if there are any unusual features. If not then specific detailed mention may not be necessary. If there are unusual features such as this one and then brief...." if we happen to get a taxi clearance on taxiway B, we have to be careful to make sure that we make the 45 degree left turn which is located not very far after exiting the ramp. Otherwise we will be in an area that we can't exit. Bravo is the first left turn after the end of the ramp". Takes 10 seconds. If something comes up as you stated when cleared to push back, a convenient time to re-brief can be during the push prior to start. Or stop and brief. I find it difficult to taxi properly and try to figure stuff out, especially at night in unfamiliar places.

from a practical perspective, sometimes having the info for an aerodrome printed on many pages doesn't help. here mention is made of a warning note on pg 6 about this.

Let's be realistic, time constraints, workload issues and cockpit ergonomics, especially older airliners, are not conducive to a making a comprehensive briefing covering reams and reams of paperwork.

Aerodromes have become waaay too complicated, my base is Narita, have a look at the taxiway layout there, it's a set up already.

Singapore has a brilliant system at night, "follow the greens". This should be standard worldwide. it would prevent a lot of problems.

Copied, I believe, from the system pioneered at Heathrow (and subsequently Gatwick).

Or convince CAAs that iPad with GPS receiver in connection with JeppFD displaying precise aircraft location on taxiway chart present greater benefit than risk of any possible EM interference...

Much cheaper alternative that works everywhere where airport chart is georeferenced.

Correct, but it is not really a valid excuse as we have to just deal with the way things are. In order to mitigate this, I have notes for every airport I fly to commercially. They are on my laptop and I have them printed out on paper in a small pile in my room. When I am going somewhere, the paper copy gets brought with me for quick reference. It is just a single page of notes divided up into arrival, arrival taxi, departure taxi, and departure and whatever other random notes I might feel is important. Some airports have only a few notes, others are quite extensive.

Much of the information has come from pre-reading the Jeppesens and the many briefing pages, in advance of going into the airport the first time. That way, I don't have to worry about sorting through reams of Jeppesen pages on the ground while in a rush or in flight, especially a short flight. I can just look at my cheat sheet for quick reference. It gets updated on occasion as required with computer updated and a new page printout that goes into the pile.

Bottom line, I am just not nearly as sharp as some of those smart guys who really seem to know their stuff, so I write it down and review.

It can have out of date info and it is my responsibility to check and see if the Jepp pages have been updated since I was last at an airport.

Don't ever be afraid to slow down and ask the GMC ATCO for help folks.
It always amazed me, on the cockpit trips I did, (after sitting 9 floors up in the tower for many years), how different it all looks at night, in the rain, at pilot eye-level, especially taxying towards the terminal buildings. It seemed to become a 2 dimensional mass of lights. This is a classic example of lots of little cheese-holes lining-up.
Safe taxying y'all.

"Correct, but it is not really a valid excuse as we have to just deal with the way things are. In order to mitigate this, I have notes for every airport I fly to commercially. They are on my laptop and I have them printed out on paper in a small pile in my room. When I am going somewhere, the paper copy gets brought with me for quick reference. It is just a single page of notes divided up into arrival, arrival taxi, departure taxi, and departure and whatever other random notes I might feel is important."

Proof positive, if ever it was needed, that the very expensive aero info supplied to our profession is not and never has been up to the required standard.

Neither the content nor the delivery, in graphical design terms, is user friendly and the same trap awaits in other locations for those in our profession.

I recall in the last decade complaining to my airline management and training departments about the standard of the notams and their presentation but alas in the absence of a "Tombstone Imperative" neither interest nor action was taken.

There will be more BNLL type accidents/incidents before the nettle is grasped by the xAAs, if ever. And certainly any solution which costs (more) money will have a hard time being implemented.

"To those who mutter about the time taken to see the official reports...."
Rest assured I have not and will not mutter about delayed reports, my highlighting the delay to the LHR Airbus cowlings accident will be continued at a higher volume until we are all able to learn the lessons currently being obscured.

Or do you want aircraft accident reports to follow the example of the UKs much delayed Chilcott Enquiry?!

Airport qualifications for air carrier operators are important, but local familiarity with the airport makes or breaks the deal, so to speak.

I cannot download the report but I hope the station manager was held accountable.

Zooker, very true. And - in the US particularly - the use of only blue sideline taxiway lights, and only rare use of green centreline lights, is awful. You're faced with a sea of blue - blackness - blue - blackness - blue.... Some of the blackness is taxiway but most is not. But try distinguishing which bit of blackness is the taxi bit & which is not at a strange airport bloody difficult.

Centerline greens, however, are a much better way of marking the way. You have a clear line of breadcrumbs to follow.

One lesson to come out from this for airports is that runway / movement area inspections by ARFF are no good for electrical or pavement inspections - IMHO.

This is not to detract from the valuable job that ARFF do in emergency responses, nor the value of their routine inspections. But this accident has shown, like other accidents I know of, that such "drive-by" inspections are limited. Their frame of reference is to detect gross deficiencies such as wheels, flaps, cargo doors, etc., lying on the runway, or other obstacles such as dead animals. That is as far as the ARFF are trained to do, and as much as can be expected from their inspection.

They cannot detect pavement problems such as ruts, failures or slab breakage, and nor do they detect electrical/lighting problems. The report makes it clear that at JNB they missed the unserviceable sign and the many non-functioning taxiway lights – in section 1.11.5.2. I contend that for an airport to rely on ARFF to do the inspection tasks of pavement engineers or airport electricians is delusional, because ARFF are not trained in those tasks. What an airport must have is a routine inspection programme by the senior groundsman / technical officer for pavements, and by the electrician for lighting, signs, PAPIs etc. That is the ONLY way that pavement and lighting deficiencies can be assured of being detected.

Section 1.17.3.4 of the report outlines the ACSA Runway and Taxiway Inspections Procedures, which looks good on paper (and I am not being sarcastic). This includes a “Maintenance or Engineering representative to be in the inspection team”; and at ACSA International Airports “the electrician on duty is required to complete an inspection focusing on the serviceability of lights during the course of the day or night”. But it didn’t work; report section 2.1.21.4 refers to “daily runway and taxiway daily inspection sheets and maintenance checklists as proof of actions taken”. It looks good. But the reality is all that got generated were bits of paper with ticks on them. The defective lighting and signage still existed and they were a significant factor in this accident.

I don’t want to single out JNB or ACSA here; I have seen exactly the same problem at other airports in South Africa and I have seen it in other countries. It is a system-wide problem, and the only solution is separate out the routine “drive the runway” inspection from the pavement + electrical technical inspections.

After several tries I managed to download the report and read it. An extremely thorough report but perhaps overly detailed in places.

I know this has been mentioned before, but what I found staggering was that this was an exact repeat, with a more unfortunate outcome, of an earlier mistake at the Bravo/Mike intersection, by the same airline!

How come that one fell through the cracks at BA?

No problems with that in my airline... been doing it for a few years now. Works very nicely.

Can I ask whether I've misunderstood a point in the body of the report?

I thought that what it was trying to say was that Mike had been described on paper as a taxilane. But physically it didn't look like a taxilane, and had all the visual characteristics of a taxiway. Therefore, even if you were aware from your paperwork that Mike was approaching, you wouldn't be expecting to see it look like a taxiway; and therefore, if you saw a taxiway ahead of you, you might mistakenly think that it was still Bravo and continue along it.

If I've got that right, it's a relatively subtle point, and I wouldn't be surprised if the summary writer didn't manage to grasp or convey that.

I24, who is to say that ONLY BA had had the problem previously?

You might find that it was BECAUSE of the comprehensive nature of BA's ASR system that they were able to search for previous issues at JNB and found it.

iirc from reading report, the JNB authorities had no trace of it.

If the local authorities had no trace of the previous BA incident, how many other previous incidents at that intersection have they 'conveniently' 'no recollection' of from other airlines?

Remember also that local operators in and out all the time have much better local knowledge, the true test of an airport layout and procedures and deficiencies is gained from irregular users.

Just a countering view ....

"Just a countering view ...."

Which has conveniently ignored the point of the question.........any interest declared, perchance (when not on the golf course)?

The fact that the SA CAA publish a clear warning on their AIP Ground Movement Chart for JNB, indicates that taxiing down Mike in error, was and is a known threat. Normally published 'Hot Spots' and warnings for taxying are normally derived from past errors that have been made. I would be very surprised if BA was the only operator that has had aircraft involved in entering Mike in error.

The SA CAA obviously felt that the threat was worthy of mention on the AIP taxi chart itself and not confined to the text portion of the aerodrome entry. If Navtech had done the same and published the warning on the chart itself, then perhaps there would have been more likelihood of the crew spotting this and perhaps even briefing it before hand?

Exactly my point. I doubted that BA were the only ones EVER to make that mistake and hence questioned HOW and IF any previous instances were reviewed for severity and repeated occurrences by the authorities.

My interest is in finding out the facts and them being accurately and impartially analysed to prevent any future event. If (at least 2) crews have made the same mistake at the same point in recent years (I suspect more), shouldn't that sound a warning bell saying that more diligence and less sweeping under the carpet is required?

For completeness, I am no longer employed by BA but was until retirement a BA 747-400 Captain who operated in/out of Jo'berg between 1997 and 2009 and think I know the place fairly well (for a foreign operator - about 20 times between 2006-2009).

Remote monitoring of physical surfaces is difficult, but checking individual lights is within the compass of existing technology. If my car can check the functioning of each of its light bulbs, or LEDs, every time I swich on the ignition, surely a similar system can be devised for all functional lights at an airport.

A simple-minded scheme based on current drawn by a long string of lights may not be sufficiently sensitive, but including sufficient intelligence at the socket, with reporting back along the ower cables, should not cost more than a few pence per light. In total, less than a superannuable B747.

Too late for this particular horse, but such a scheme might be an effective bolt on future stable doors. If it worked, of course, we would never knoe because there would be no unknown missing lights to lead flight crew astray.

I have my doubts about the fixes for this problem.
Lights will fail, inspections will at times be less than diligent, crews will get distracted etc etc and it is likely to happen again.

A better fix would be to put a dogleg in Mike so that it joins Bravo at a right angle after the bend. The current straight on junction is an accident waiting to happen.

A distinction has to be drawn between the regulator and the airport operator. There is no denying that there were shortcomings that were attributed to the airport operator. The regulator on the other hand obviously accepted that there was a threat and highlighted this in a unambiguous way in the AIP for the benefit of the end user i.e. the pilot. For this to happen then I suspect that previous occurrences were not hidden from a wider audience, otherwise nothing would be published.

Another interesting question is how many BA pilots would have been aware of the previous incident at JNB? Not that one expects a pilot to trawl through years of ASR information prior to each flight. However previous errors made by colleagues can prove to be truly invaluable in preventing a similar occurrence. What is key is making such relevant information accessible and not hidden in amongst safety databases.

Doesn't BA have an airport briefing package for each destination giving general info and cautions.

Sorry for being stupid, but could someone please explain to a simpleton what the picture on page 4 is supposed to convey?

There are two markings, one long red dotted line and one shorter white dotted line. From what I know the crew followed the shorter, white, dotted line and ended up in a building. The picture description appears to say that the crew expected to taxi along the red dotted line, completely opposite to the white one.

But reading the following pages it is, to me, somewhat vague if they really expected that. Anyone care to help me out here?

The crews' expectations are explained BEFORE the photograph - on page 3

Section 1.1.4.1, item (ii)

"The expected taxi route, which was based on previous experience, was discussed and agreed on by the crew. They expected the taxi route to be facing tail south, with the intention to use taxiway Echo, turning onto taxiway Alfa for a "full length taxi" and "to keep going the extra 200m straight to the end of Runway 03L". i.e. - the red path.

Items (iii) and (iv) cover further expectations about problems along the red expected route.

To me the critical evidence is the picture of theintersection with the missing lights - unless you were really on the ball the curve of the Bravo taxiway beyond the intersection is very hard to pick up

Mike isn't brilliant either but when you have pre-briefed "straight on to the end" you can see how it happened

The only other thing that concerns me is that even though the CC reported dripping fuel the flight crew kept the engines running and then started the APU - that could have turned an incident into a full on disaster

HH,

You weren't there, you didn't see the leak size or location. After an incident like this it would be so easy for a crew succumb to the 'oh **** what have we done' aspect and lose control of the situation. They remained calm and rational and made sensible decisions with the information they had and are to be commended for that.

Out of interest have you ever tried to set fire to Jet A1? It's harder than you think. People get very badly hurt evacuating from a 747, it's not a decision to take lightly.

Lad

Funily enough I HAVE tried to set fire to JetA and you are correct - its much better to use Vodka as a fire-starter

I just think that knowing you've hit something and then getting a report of a leak you'd shut down - no reason to evacuate but you aren't going anywhere until someone outside has had a look. They had no idea how much they'd lost, or where it was leaking from. It was pretty dark out there.............

Thankyou very much, pattern_is_full!

I guess I didn't grasp the subtleties of "facing tail south" and "facing south" (as in "facing face south"), to me it was unclear, but I guess for you natives it is piece of cake.

I would agree with the person who noted the marked difference in the summary and the report conclusions. The summary looks hurried, as if they ran out of paper or something...

"didn,t discuss.." I may suggest anyone to try to stop an airliner after pushback,to discuss the taxi routing. Try it at CDG or AMS and you will be missing much more than your slot. The whole thing is based on that You know every airport by heart,so every instruction of taxi clearance is clear in your mind,including every work in progress. Loophole lies in there,because taxiing is always full with preparations,so everyone is supposed to know the routing. Understanding those instructions is left for the person who hears them,so... We were once the first to taxy at BRU morning cargo flight , all departing at the same time... Barely visible taxiing,i missed one sign for my captain,so we turned to the inner one too soon. The whole convoy of about twelve of us,followed wthout a question asked... Follow the one who seems know-syndrome.