Go Back  PPRuNe Forums > Flight Deck Forums > Rumours & News
Reload this Page >

Spanair accident at Madrid

Rumours & News Reporting Points that may affect our jobs or lives as professional pilots. Also, items that may be of interest to professional pilots.

Spanair accident at Madrid

Old 21st Oct 2008, 22:33
  #2241 (permalink)  
 
Join Date: Jan 2005
Location: uk
Posts: 269
The root cause of this accident was the failure of both the flight crew and the maintenance staff to fully consider why certain system abnormalities were indicated and what would be the consequential effects of the maintenance actions that were carried out. Flight crew, in particular, need to maintain a systems knowledge equivalent to that attained in the conversion course, in order to understand the potential for knock-on effects if systems are disabled or degraded by maintenance action. A few 'what if?' questions might have prevented this accident. The lack of lateral thinking left a big black hole, into which the flight crew fell under pressure of the abnormal circumstances. As I have previously posted, checklist discipline and a refusal to be hurried is of paramount importance in stressful abnormal circumstances, as this situation was.
777fly is offline  
Old 21st Oct 2008, 23:52
  #2242 (permalink)  
 
Join Date: Dec 2007
Location: devon
Age: 80
Posts: 366
The engineers involved with the initial snag that the RAT was indicating the incorrect temperature probably thought that the indicating system was unserviceable and carried out their actions in compliance with the MEL where flight not into icing conditions was OK. It is possible that the high temperature the RAT probe indicated was caused by the aircraft being "in flight" for what ever reason heating up the probe, to prevent icing, and the conduction of the heat from the probe heating element caused an overheat indication of the RAT. Was this ever covered by the manufacturers maintenance training course which I presume the engineers attended?
Would this be a case where Boeing have picked up a "poisoned chalice" from the original manufacturer?
Oldlae is offline  
Old 22nd Oct 2008, 00:41
  #2243 (permalink)  
 
Join Date: Jan 2005
Location: uk
Posts: 269
Oldlae

No, but some lateral thinking about why those indications were happening might have saved the day.
777fly is offline  
Old 22nd Oct 2008, 02:39
  #2244 (permalink)  
 
Join Date: Aug 2008
Location: Canary Islands, Spain
Posts: 240
Boeing, allegedly, told everybody after Detroit's Northwest accident that the TOWS couldn't be trusted 100% and they should be very frequently tested, specially shortly before each take off.

Unfortunately, such recommendation was never made mandatory by regulatory bodies anywhere in the world.

And worse yet, it seems Spanair was never told about it (or they weren't smart enough to find out, one of the two or a bit of both).

I guess Boeing figured on their own analysis that that was the best and most effective course of action (incidently, also the cheapest).

I concur that, with frequent tests before each take off, the likehood of an unnoticed TOWS failure at the same time as an unnoticed misconfiguration is very unlikely and reasonably safe for such an old airplane. Not "fool proof", but we've seen even worse designs being used for everyday critical operations in all industries.

At least it adds another layer to the cheese. Now 3 things would have to fail:
-The pilots missing a configuration error.
-The TOWS failing around that same time and not having been noticed on daily tests.
-The pilots missing the pre-takeoff TOWS test revealing the TOWS have become inop or the TOWS failing exactly in the few minutes (instead of up to 24h) between the check and the takeoff.

The chances of all three holes aligning are no longer that great as in the other case, where only two holes needed alignment.

Not perfect, but a pretty good, cheap and easy "patch" for the time being, until a better solution is found, if deemed necessary. At least, it would've (should've) worked on the only two known cases since Detroit: Spanair and Map. But in neither of those cases the crew performed a TOWS check right before takeoff. It wasn't even required in their SOPs.

Last edited by justme69; 22nd Oct 2008 at 18:49.
justme69 is offline  
Old 22nd Oct 2008, 03:00
  #2245 (permalink)  
 
Join Date: Mar 2006
Location: England
Posts: 860
777fly re 2271 “the failure of both the flight crew and the maintenance staff to fully consider why certain system abnormalities were indicated and what would be the consequential effects of the maintenance actions that were carried out.”

It was not so much the lack of lateral thinking as the lack of direct thinking – consideration of the consequences of work on any system that could result in the combination of errors.
If the combination of errors has to be considered, then why isn’t this done by someone higher up the management chain? The certification requirements shave the subject but fall short of hard defences. The manufacturer or FOEB (predominantly operators) who assemble the MMEL could have considered the possible errors and required a TOCW test after any work on adjacent systems. Are the FOEB qualified to think about the likelihood of error (human factors) or the consequences of error?
But this is all in hindsight, what we require is the foresight to avoid the next major accident, which most probably will not involve TOCW.
PEI_3721 is offline  
Old 22nd Oct 2008, 03:33
  #2246 (permalink)  
 
Join Date: Jan 2005
Location: uk
Posts: 269
PEI 3721

Sorry, but that is exactly what I was saying. Lateral, direct or 'out of the box' thinking is required. In my experience, MEL rectifications and alleviations follow simple consequential paths and usually, but do not always, anticipate knock-on effects, particularly when multiple failures are involved.. The resultant effects can be dire if the rectification process addresses an apparently simple fault which is, in fact, just a symptom of a far larger problem. A classic case of treating the symptoms rather than the disease itself. This appears to be what happened in this incident.
777fly is offline  
Old 22nd Oct 2008, 09:20
  #2247 (permalink)  
 
Join Date: Oct 2001
Location: Gatwick
Posts: 1,964
777fly

Lateral or Outside the `box thinking is what regulators have tried to move away from in recent times and I agree with this. It is the quality of the procedures that you follow that need to be changed. The MEL should be doing the thinking outside of the box, not the engineer.

If you want a change in approach, then you need to remove the MEL as the first go to book, if an aircraft has a fault on prior to departure. If you had to quote a maintenance/schematic manual or wiring diagram reference along with a MEL reference, then you would be forcing the engineer to be giving him/herself a quick refresher on the system to be deferred.

But this approach lengthens the time it takes to dispatch the aircraft and so costs money. So write better MEL's.

How do you prove that the engineer had an understanding of the air/gnd system to the type of aircraft. He/she may have sat the course and passed the exams, but in the modules dealing with air/gnd sensing and anti ice protection, only got 75%, meaning that 25% of the syllabus was not understood. That is why you don't think outside the box. You follow procedures, i.e. what the MEL tells you to do.
Litebulbs is offline  
Old 22nd Oct 2008, 11:21
  #2248 (permalink)  
 
Join Date: Apr 2002
Location: FUBAR
Posts: 3,352
I have never flown for a company ( & I have flown for a few ) where engineers were not under real/perceived pressure to get the aircraft back on line ASAP.
Asking in this situation that the person concerned thinks laterally could in fact merely distract him from carrying out what may be a complicated task in itself. Theoretically lateral thinking should be great, but realistically a robust and comprehensive MEL procedure which does that for him would achieve much the same result whilst also covering the poor blokes back when he has to explain why the troubleshooting took so long.
Of course persuading Airlines & Authorities that the MEL should be more comprehensive won't be so easy, as someone has to actually produce & authorise it, and future defects will be more time consuming.
Blinkers tend to be worn regarding safety, and the lessons quickly forgotten at the temple of the great God of commercial expediency, don't know how many accidents it would take to change that unfortunate mantra.
captplaystation is offline  
Old 22nd Oct 2008, 13:32
  #2249 (permalink)  
 
Join Date: Sep 1999
Posts: 542
Going after the engineer is not the answer.You go after the system.If you have to play the blame-game,attack from the top-down.The flight crew will carry the can but if you want the big picture,you have to delve deeper.Why did the Spanair CP not collate and disseminate the lessons learnt from Detroit and Lanzarote,why did he not emphasize the frailties of the MD-80 air-ground system to his crews,and why did he not instigate Boeing's recomendation?
Encourage a training culture where system knowledge is taught at a much deeper level.Dont scratch the surface.Very often a pilot only knows that if X happens he must do Y.He may or may not know why X has happened and why Y is remedial.And you test this level of in-depth understanding orally in a classroom with visual props/aids with engineers as instructors .This CBT is okay but its superficial and promotes rote memorization over lateral thinking borne out of a thorough understanding of the systems.Of course its cheaper and less time-consuming which is why they do it.
They spend time and money on CRM classes telling us that the flight deck is a democracy(which it isnt) and what they should be doing instead is devoting those resources on a return to the fundamentals.Return the role of the CP to its original glory;he sits at board level,is divorced from economics totally,and fights for his one and only mandate;safety.
Rananim is offline  
Old 22nd Oct 2008, 13:41
  #2250 (permalink)  
 
Join Date: Apr 2002
Location: FUBAR
Posts: 3,352
Totally with you on that score.
It is very easy to blame the last person to handle the "component" particularly when they are dead, the blame then passes to the next in line, the poor engineer.
Much more relevant to ask those difficult questions you identified of the post holder/ the regulators/ the manufacturers. . . . . . but SO much easier to blame the pilots & engineers thereby conveniently ignoring the dumbing down of knowledge/respect that has for fiscal reasons been encouraged in this profession for too long now.
All of the responsibility/ none of the authority that is the dream scenario for bean counters, and is very close to the current status "enjoyed" by those on the front line.
captplaystation is offline  
Old 22nd Oct 2008, 15:45
  #2251 (permalink)  
 
Join Date: Mar 2002
Location: Florida
Posts: 5,219
Litebulbs

Lateral or Outside the `box thinking is what regulators have tried to move away from in recent times and I agree with this. It is the quality of the procedures that you follow that need to be changed. The MEL should be doing the thinking outside of the box, not the engineer.
Right on mate

Please paste this into all threads following an accident
lomapaseo is offline  
Old 22nd Oct 2008, 16:25
  #2252 (permalink)  
 
Join Date: Dec 2007
Location: france
Age: 71
Posts: 74
SPA83 is offline  
Old 23rd Oct 2008, 00:37
  #2253 (permalink)  
 
Join Date: Oct 2006
Location: Spain
Posts: 82
Rananim and Spa83:

Agree 100%
agusaleale is offline  
Old 23rd Oct 2008, 03:38
  #2254 (permalink)  
 
Join Date: Aug 2005
Location: fl
Posts: 2,561
Lateral or Outside the `box thinking is what regulators have tried to move away from in recent times and I agree with this. It is the quality of the procedures that you follow that need to be changed. The MEL should be doing the thinking outside of the box, not the engineer

Then we need to do a lot of work on MEL's because it didn't work this time. Smart engineers and pilots would have prevented this disaster. Systems knowledge by either would have made them realize that the only time the RAT heater works is in the air. Disconnecting the RAT heater does not fix the problem, only the symptom.
bubbers44 is offline  
Old 23rd Oct 2008, 08:21
  #2255 (permalink)  
 
Join Date: Aug 2008
Location: Canary Islands, Spain
Posts: 240
An interesting article published yesterday:

Safety slip in Madrid crash also seen in U.S. - USATODAY.com

It talks of 55 voluntarily reported cases of bad takeoff configurations in the past 7 years or so in the USA alone. That's a lot more than I had found, that were around a dozen or so cases, but once I found it was pretty common, I didn't continue looking that closely. Most of them, of course, were catched in-extremis by the Take Off Configuration Warning Systems.

The Spanair pilots were of the few unlucky ones (together with Lanzarote and Reagan's cases, i.e.) that had an unnoticed TOWS failure shortly before they needed it the most. After so many years and so many million flights, I guess it was due to happen.

It comes to point to that, even with sufficient training, experience, safety culture, management, maintenance, etc, human error is still a piece of the puzzle that just can not be avoided and therefore needs as much help as possible from technology, etc.

But I don't think I'm saying anything new here. Traffic accidents, even by professional and experienced taxi/bus/truck drivers, happen every single day when humans in charge of vehicles make bad choices against everything they have been trained for. And no, most of them are not "careless" or "suicidal", they are just humans carrying their children to school everyday but don't even bother doing a basic visual check of all four wheels before entering the car. And constanly, driving schools, TV safety campains, police controls, improved vehicle designs, better roads and signage, etc, etc are reminding us to watch our speed, buckle up, not forget to turn on lights at night, etc, etc, etc.

And yet, we all do make those basic mistakes at times, putting our own lifes at risk against our better judgment for "unkown reasons".

BTW, does anybody know if there was ever an investigation report for the Indian Airways accident of Dic 17 1978 of B737 VT-EAL that explained why it tried to take off w/o slats and it crashed?

Last edited by justme69; 24th Oct 2008 at 00:34.
justme69 is offline  
Old 23rd Oct 2008, 08:57
  #2256 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,584
I have no knowledge of the MD and the effect of the A/G sensor, but it appears that the a/c was 'flyable' when stuck in the air mode. It appears the possible incorrect A/G state is not obvious on his a/c. IF the crash was caused by failure to configure correctly for take-off, then in my opinion it was not the MEL, the engineers or the TOWS failure that CAUSED the crash. That is a pilot's view. I know there will be howls of protest from the safety psychologists.

Certainly, having had an A/G stuck in 'Air' mode on landing in a 737-400 (Classic), it would be instantly obvious to crew that all was not right on that type.

I fully support the call for a 'look' at the MEL philosophy to ensure that possible causes for symptoms are checked/trapped in the 'comfort' of the manufacturer's tech office to take need to diagnose complicated systems away from the LAE, and I can easily see now how LAEs will be MOST reluctant to 'quick fix/dispatch' in future which will impact on airline programmes. I cannot, however, see how the possible dispatch of this a/c in 'Air mode' was a direct cause of the accident.

As we have seen, and know, it is by no means impossible for a fully serviceable and sounding TOWS to be ignored by crews, so we need to add this to Justme's list, making 4?

"At least it adds another layer to the cheese. Now 3 things would have to fail:
-The pilots missing a configuration error.
-The TOWS failing around that same time and not having been noticed on daily tests.
-The pilots missing the pre-takeoff TOWS test revealing the TOWS have become inop or the TOWS failing exactly in the few minutes (instead of up to 24h) between the check and the takeoff."
BOAC is offline  
Old 23rd Oct 2008, 09:05
  #2257 (permalink)  
 
Join Date: Aug 2008
Location: Canary Islands, Spain
Posts: 240
As we have seen, and know, it is by no means impossible for a fully serviceable and sounding TOWS to be ignored by crews, so we need to add this to Justme's list, making 4?
Well, in that case, it would only make it a "one hole" cheese.

1) The pilots fail to set/verify the right takeoff configuration.

--- It doesn't matter if the TOWS fail or not, the pilots are going to ignore them.

They take off, and, if the conditions are not very favourable for a recovery (wind/weight/response time/engine thrust available/runaway length/pure random luck of the stall behavior/etc) they crash. That's it.

Last edited by justme69; 23rd Oct 2008 at 09:20.
justme69 is offline  
Old 23rd Oct 2008, 09:19
  #2258 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,584
My apologies - I agree - I was trying to look at alternative routes through the holes to the same end, not related to this crash, and should have had 2.

1) The pilots fail to set/verify the right take-off configuration.
2) The pilots ignore a warning.

That is with a fully serviceable a/c, taking all questions of LAE actions and MELs out of the equation. This is where the last firewall needs to be.
BOAC is offline  
Old 23rd Oct 2008, 10:03
  #2259 (permalink)  
 
Join Date: Aug 2008
Location: Canary Islands, Spain
Posts: 240
We indeed know of at least one case (LAPA), but possibly one or two more that weren't properly investigated, where that happened.

Also, please note that the MD-82 that suffered the accident wasn't in "air mode" strictly speaking.

Most of the indications for "air mode" come from the front wheel sensor, others from the back. But even the ones in the front are divided between the left and right "wheel". Each of those sensors feed between half-a-dozen and a dozen relays, which in turn feed the "correct state signal" to between 1 and 4 devices each.

The Spanair MD-82 probably had only one of those almost 20 front wheel logic electrical relays fail, one that ultimately only affected two devices: the RAT probe heater and the take off configuration alarm (the other two devices connected were redundantly serviced by the right wheel circuit). So only the TOWS and the RAT probe heater thought they were "on the air". The rest of the airplane thought it was "on the ground". The absolute only signs that the airplane had any failures whatsoever would've been: a RAT probe heater turned-on while on the ground and a TOWS that wouldn't have responded with sound to a bad configuration takeoff check. If none of those symptoms were noticed, nothing would've seemed wrong at all with the airplane, which otherwise was, of course, perfectly fit to fly.

It was almost out of luck that the pilots noticed the RAT heater on while on the ground, giving them and maintenance the only chance to figure something was wrong at all. But they missed that chance. Also, if a mandatory TOWS check would've been included in the checklists (and actually performed by the crew), the failure would've also been noticed. But unfortunately the whole RAT issue was pbbly also the "spark" that made them distracted and rushed and finally triggered the oversight on setting flaps and the tragedy.

The MAP MD-83 in Lanzarote had the whole set of relays from the left front wheel "disconnected" by a pulled circuit breaker. And still, because most of those devices are actually ventilation/air conditioning/etc and a good number are redundantly serviced by the right wheel circuit, the pilots still didn't notice the "air mode" logic that affected probably a dozen or so non-redundant devices. They forgot to set the flaps/slats as well and took-off with an effectively disconnected (but perfectly serviceable) configuration warning alarm, so they didn't realize anything was wrong until the airplane started to stall.

Last edited by justme69; 23rd Oct 2008 at 13:06.
justme69 is offline  
Old 23rd Oct 2008, 12:15
  #2260 (permalink)  
Per Ardua ad Astraeus
 
Join Date: Mar 2000
Location: UK
Posts: 18,584
I'm with you, justme.

I am basically getting uneasy about posts like 777fly's, bubbers and others with suggestions that those involved (the crew and engineer) should have 'correctly diagnosed' the problem. (Leaving ?possibly? AB aside), an aeroplane - if correctly set up - will fly without any of these 'failsafes'. I just cannot see how blame can be attached to these poor folk for not understanding a complex logic system of switches and relays. The MEL system NEEDS to have been constructed with all this done in advance, surely? Certainly the limited tech info available to me in my seat on the ramp would in no way have allowed me to follow all the relay switching involved nor to be CERTAIN there was no other failure in the system that could heat the RAT sensor - and it really is not the pilots' job to do so either. The logical progression here is for me to insist on full access to maintenance manuals before I accept a signed off defect. NO - I am trusted to fly the a/c properly and I trust the engineers to fix it likewise.
BOAC is offline  

Thread Tools
Search this Thread

Contact Us - Archive - Advertising - Cookie Policy - Privacy Statement - Terms of Service - Do Not Sell My Personal Information

Copyright © 2018 MH Sub I, LLC dba Internet Brands. All rights reserved. Use of this site indicates your consent to the Terms of Use.