MK Airlines B747 crash at Halifax

[Mode7]

Hangin' On, MK was audited by tyhe FAA (US and Canadian) and the CAA during the inquiry into the Halifax crash.

[ShotOne]

We're still blaming the berm. This seems a perverse and unhelpful twist of logic. Presumably if this accident had happened at Ostend or some other MK destination we'd be blaming the shopping centre/hill/apartment block/tree unfortunate enough to get in its way. Many airports ARE in built up areas. That's not because of greed. It's because our passengers pay us to fly them to and from where they live.

[MungoP]

ShotOne.......

Exactly my point....In this case the accident would appear to be the result of a mistake by the crew...the cause behind the cause was almost certainly fatigue...but whatever the cause of this event, and there will be others, safety concerns of all aspects of aviation need to be upgraded to allow for developments in the industry.....that includes the airport environment as well as flight duty rosters....out of the two I believe that rostering concerns should be coming under the microsope rather more just now......only changes forced on the companies involved, whether operators of aircraft, airfields or whatever by the regulatory bodies will bring about the improvements....the commercial concerns themselves have to think of the bottom line...changes in regulations at least will ensure that all commercial concerns will be playing on a level field...ultimately the costs will be borne by the clients, and that is how it should be...........

[PickyPerkins]

Well it looks from these photos on the TSB web site that that 9G-MKJ did indeed did make a “ski-jump” take off over the trees to land beyond the road at the left in the second picture below, although by then the aircraft was minus the entire tailplane.
http://www.tsb.gc.ca/en/media/photo_...4/photo_02.asp
http://www.tsb.gc.ca/en/media/photo_...4/photo_04.asp
(Sorry to post these photos as links rather than as images but, as someone else also commented, PPRuNe does not seem to like asp files.)

In both pictures the horizontal stabilizer is visible at the start boundary of the trees.
The larger white object in the center of the first photo is the fin, also shown in this picture with the “MK” visible.
http://www.tsb.gc.ca/en/media/photo_...4/photo_10.asp

The T-junction in the second picture is directly in line with the runway. Click on the caption of the second photo on this page to get an enlarged view of the runway from space.
http://www.cbc.ca/story/canada/natio...fax041014.html

These picture seem to me to show that the trees between the horizontal stabilizer and the road, and the road itself, are entirely undamaged. So the plane must have passed over them.

So in summary, the plane:
left its wheel marks in the berm,
left the fin between the berm and the trees,
left the horizontal stabilizer at the start boundary of the trees,
the rest of the aircraft must have passed over the trees and the road to crash and burn beyond.

Does anyone else see a different interpretation?

Cheers,

[broadreach]

PP, I expect the accident investigation will confirm your analysis. Ski-jump is a good analogy. Had the gradient up the berm really been that of a ski-jump the aircraft might have actually become airborne without breaking its back. Which is what one suspects happened when the main gear hit the berm. The centre of the fuselage would have been projected upwards and the after half of it would have hit the berm at an even greater angle than the aoa during the last few hundred metres of the takeoff run. Thus knocking the heavier tail section off and bouncing the aircraft over the trees and the road, rear hull flapping and engines still at full power, into the quarry beyond.

The whole thing is one of those nightmare scenarios where cause, sequence and outcome seem inevitable. Maybe, because of its slightly higher profile than accidents in Nigeria, Equador or other places that don't attract so much attention, it will nudge operators, regulators and airport authorities.

[targaman]

Aircraft loadin error kills!

http://www.emediawire.com/releases/2005/4/emw233476.htm

Comments?

The investigation referred to the Cessna accident at Pelee Island, Ontario is ongoing and TSB Canada has recomended standard passenger weight tables NOT be used on aircraft with 9 or fewer passenger seats.

What about MK Airlines B747, Halifax, Noca Scotia, 14 October 2004.

Loading error? Will the load sheet be used as evidence or a full reconstruction of weight & balance carried out?

As a pilot how do max weight, balanced field, ATM derated field limit take offs on a hot night and a long duty grab you?

I know what I grab!

I would like feedback for my research into weight & balance error as a causal factor in many large aircraft accidents. Email me through www.avibridge.com or www.oacs.homestead.com please.

[barit1]

a = F/M

INS-equipped aircraft have all the raw data necessary to compare ACTUAL aircraft acceleration on TO vs the rate IMPLIED by AFM. Why this has not become a standard go/nogo check is beyond me.

Just one of many examples: Leased Airbus on a Hadj flight, OEI at rotation, barely made it around the circuit to land safely. Aircraft was impounded & everything weighted. Bottom line - the dispatcher was jailed for clandestine shipment of several tonnes of goodies.

There's a simple way to overcome this - measure acceleration. Can't beat hard data.

[Earl]

Some one correct me if I am wrong.
This was a freighter not a passenger plane.
I have done many overweight Haj flights.( passenger)
And many cargo freightor flights.
All were done at max takeoff weights.
Even 8 tons overweight on the 747 classic we got airborne,
The aircraft would not climb or maintain flight planned alltitude even with MCT.
Running the charts in reverse showed us 8 tons over weight.
At destination after dropping our passengers and cargo the aircraft flew right on the charts.
Onthis flight out of Halifax.
The load sheet was done by the loadmaster and checked by the Captain.
He relied on information provided to him by the load planners and loadmaster.
The aircraft load was in the proper configiration as the 747 would show this through the nose gear sensor.
I understand that the cause of this was not the proper EPR setting.
Which anyone can see was way too low.
Maybe I am wrong here but if the proper EPR is not set then this new and improved system wont work.
Back to the basics here.
You cant re-invent the wheel.
I feel for the crew as they were probably tired and did not notice this.
Could happen to any of us!

[BOAC]

but if the proper EPR is not set then this new and improved system wont work

- this idea has been aired on PPrune before, and the thing is that it WILL work as the acceleration will not match the requirement for that runway - be it EPR/wt/binding brakes whatever. The only thing it will NOT show, I guess, is that you lined up at the wrong point and are running out of tarmac.

[Earl]

So what is suggested here is some type of runway cross acceleration check?
How about an EPR cross check prior to block.
This crew was obviously tired and missed some things.
EPR was one of them.
Could have happened to any of us.
Hind sight is always 20/20.
As far as an acceleration check I dont think this will come to pass.
Unless you have unreliable airspeed instruments that the MEL requires you to do.
All though if this was added as a techniqe then it would help.
But how many pilots would be looking down at this point instead of the airspeed indicator and the engine instruments and other scans.
Much easier to check the EPR?
But then again you can never have enough information.

[Joetom]

GPS/DME v's EGG timer wired to a go/no go bit of kit and this subject can be history....did they/we put men on the moon.?

[Earl]

No one is ever proud of doing overweight flights.
But when the company you are contracted to constantly lies about the cargo weight it happens.
Things are far from perfect in this part of the world with little or no accountability.

[OverRun]

Earl,

Good words man.

Maybe I'm feeling guilty about the berm. Not mine. But there is one that is. Guess I'd better go get it levelled.

[targaman]

Weight & Balance Check

We in the industry (Pilots-Dispatchers) have been using an 85 year old paper based system of weight & balance measurement that is ...........

Read my press release http://www.emediawire.com/releases/2005/4/emw233476.htm

We still (2005) rely on the presumed accuracy of the load sheet when in fact it has been shown many times to be in error.

What is needed is an external aircraft weight & balance system that provides a final check & warning of an overweight and or out of balance condition.

I have patented such as system and am looking for dvelopment funding for a computer model and a prototype for trials in NZ.

Any support from this forum or any other source would be gratefully accepted.

In the MK Airlines case I suspect that once again the load sheet will show a correct weight & balance data. I have made it known to TSB Canada that I am a researcher into what I call the Ice Berg Effect of load sheet error risk to large aircraft and have the support of many accident investigation bodies such as NTSB and TSB. They agree with my concerns but are unable to do anything but make recomendations.

The lack of alternative weight & balance evidence in cases where reconstruction of the aircraft is not possible is a serious concern for all of us.

For those of you interested in more info on this topic visit www.avibridge.com

Keep the discussion going and make your MP aware that the load sheet is still the only evidence available at the scene of the accident in many cases.

Does anyone know if the FDR in your company records weight & balance from the onboard weight & balance system?

I need hard evidence from many sources of the status of your aircraft's OBWBS-active or deactive?

All info will be treated confidentially and is only used in research.

I am free of direct pressure from operators as I am now retired. I know the very real threat for whistle blowers both from a union perspective and from a sole contractors view.

Sorry for the long winded post but I believe it is a serious problem that needs urgent attention. I am doing my bit.

It mayyet prove to be a causal factor in the MK Airlines case.

[Belgique]

Halifax MK747F Bottom Lines

Built in Software Traps and Pilot Fatigue
Main features of the Halifax MK Airlines 747F Crash Report

In their report on the Boeing 747 crash at Halifax that killed seven people, investigators say a lack of crew training on new takeoff software, combined with fatigue from working long hours, were likely factors in the accident.

Last year, following the 14 Oct 2004 crash of the MK Airlines cargo jet at the Halifax airport, investigators at the Canadian Transportation Safety Board said the crash was caused by an improper setting of the throttles by a crew member.

As a result, the heavily laden cargo plane, weighing more than 350,000 kilograms, didn't have enough speed or power to make it off the runway.

The tail of the aircraft broke off after striking a berm at the end of the Halifax runway and the plane plunged into nearby woodlands, where it exploded into flames.

But further investigation has turned up problems beyond the throttle settings; specifically the crew's training on new cockpit software, said Bill Fowler, the board's lead investigator.

The software known as the Boeing Laptop Tool, or BLT, calculates the proper speed and thrust of the plane - based on information the crew inputs on weight, the length of the runway and the weather conditions.

Fowler said the software was set for the plane's weight at its previous takeoff in Hartford, Conn. - almost 110,000 kilograms lighter than the Halifax load.

After interviewing MK airline staff, the board concluded the training process on the software was incomplete.

"Did all of the training get done? We didn't feel it was adequate," said Fowler.

"The company undertook to implement this (software) package following guidance material. The question arose: did they do it adequately? In our view, not all of it was."

Fowler said the software could allow the crew to use the previous takeoff's weight data, if they were not familiar with program.

The investigator declined to elaborate further, saying the full analysis on the role training played will become public when the board's final report is published in several months.

When asked about the crew's training in an interview, Steve Anderson, a company spokesman, said the training was adequate.

"We've been using the program for many months prior to the accident. We'd never had any information of any problem from any of the air crew previously," he said from the company's headquarters in Hartfield, East Sussex.

Anderson also disagreed with the investigator's concerns about pilot fatique.

"They'd had sufficient rest at different stages of the journey. It's a simple as that. We don't believe it played a role in the accident."

At the time of the crash, the crew had been on duty for 19 hours. Had they completed their trip, Fowler says they would have been on duty for 30 hours.

In Canada, regulations allow a similar-sized crew a maximum duty period of 20 hours. In the United States and England, the maximum is 18 hours.

As the cargo plane was registered in Ghana, however, MK Airlines wasn't required to follow the North American or British standards.

Anderson said the airline has since increased staff training on the Boeing software and reduced the number of on-duty hours for crew.

Fowler said he can't be absolutely certain a lack of knowledge about the software was the main cause of the error that led to the crash, because the cockpit voice recording of the pilots' final conversations was destroyed by fire.

But he adds, "all of these factors come together, the level of awareness, the training and procedures . . . The best we can do is we can look at it and say these were the factors that were present at the time and we feel they would likely have contributed."

"The BLT and fatigue were significant elements in this."

Anderson said the airline is currently attempting to register its fleet of jumbo jets in the United Kingdom, rather than in Ghana.

"Part of the process of going through that is to have the Boeing laptop tool certified and approved by the U.K. aviation authorities, which has just been completed," said Anderson.

This Friday, the company plans a memorial service in the United Kingdom for the seven dead crew members.

A company official also plans to return to Halifax, where a small memorial will be held.

The dead Britons were Capt. Michael Thornycroft, a resident of South Africa; as well as Capt. David Lamb and flight engineers Pete Launder and Steve Hooper, all residents of Zimbabwe.

Also killed were ground engineer Mario Zahn, a German who lived in South Africa; and loadmaster Chris Strydom and first officer Gary Keough, both of Zimbabwe.

[Rockhound]

MK 747 crash in Halifax - new findings

The Oct 12 Globe and Mail newspaper reports preliminary findings, just released by the Canadian Transportation Safety Board, as to the cause of the MK 747 freighter crash in Halifax on Oct 14, 2004.
----------------------------------------------------------------------------------
POOR TRAINING TIED TO JET CRASH, PROBE FINDS
By Michael Tutton
HALIFAX -- A year after a fiery crash of a Boeing 747 killed seven people, investigators say a lack of crew training on new takeoff software, combined with fatigue from working long hours, were likely factors in the accident.
Last year, after the Oct. 14 crash of the MK Airlines cargo jet at the Halifax airport, investigators at the Canadian Transportation Safety Board said the crash was caused by the throttle being set improperly by a crew member.
As a result, the heavily laden cargo plane, weighing more than 350,000 kilograms, didn't have enough speed or power to make it off the runway.
The tail of the aircraft broke off after striking a berm at the end of the Halifax runway and the plane plunged into a nearby woodland, where it exploded into flames.
But further investigation has turned up problems beyond the throttle setting; specifically, the crew's training on new cockpit software, said Bill Fowler, the board's lead investigator.
The software known as the Boeing Laptop Tool, or BLT, calculates the proper speed and thrust of the plane based on information the crew inputs on weight, runway length and weather conditions.
Mr. Fowler said the software was set for the plane's weight at its previous takeoff in Hartford, Conn., almost 110,000 kilograms lighter than the Halifax load.
After interviewing airline staff, the board concluded the software training process was incomplete.
"The company undertook to implement this [software] package following guidance material," Mr. Fowler said. "The question arose: Did they do it adequately? In our view, not all of it."
Mr. Fowler said the software could allow crew who were unfamiliar with the program to use the previous takeoff's weight data.
The investigator declined to elaborate further, saying the full analysis will become public when the board's final report is published in several months.
When asked about the crew's training in an interview, company spokesman Steve Anderson said it was adequate. He also disagreed with the investigators' concerns about pilot fatigue.
"They'd had sufficient rest at different stages of the journey. It's as simple as that. We don't believe it played a role in the accident."
At the time of the crash, the crew had been on duty for 19 hours. The full trip would have been 30 hours, Mr. Fowler said.
In Canada, regulations allow a similar-sized crew a maximum duty period of 20 hours. In the United States and England, the maximum is 18 hours. But as the plane was registered in Ghana, it wasn't required to follow those standards.
-----------------------------------------------------------------------------------
Rockhound

[filejw]

19 hrs on duty and not fatigued? People like Mr Anderson shouldn't be allowed to run an airline.

[barit1]

Operational redundancy

The world of commercial aviation is built on operational redundancy: two pilots, two donks (or more), multi-redundant nav, pressurization, everything essential to safety.

Except:

Takeoff planning includes a choice of thrust settings (fixed derate, FLEX, full rated...) predicated on runway, ambient, and TOGW variables. Everythings fine if the raw data are correct and the sums are worked properly and the brakes aren't dragging and you don't have a flat...

My lunchbunch colleague B-52 driver told me they always used (in the 50s-60s) an acceleration check - a time limit from 70 to 120kt I believe. If the time limit is reached before 120 is seen, it's a reject.

There are smarter ways to do this in 2005, but the principle is the same. It would provide a measure of operational redundancy where now there is none.

[FlexibleResponse]

When asked about the crew's training in an interview, company spokesman Steve Anderson said it was adequate. He also disagreed with the investigators' concerns about pilot fatigue.
"They'd had sufficient rest at different stages of the journey. It's as simple as that. We don't believe it played a role in the accident."
At the time of the crash, the crew had been on duty for 19 hours. The full trip would have been 30 hours, Mr. Fowler said.

Anderson states that as his pilots killed themselves after only 19 hours into a rostered 30 duty, it doesn’t count as fatigue-related.

One is led to speculate that if they had instead killed themselves on landing after 30 hours of duty that Anderson would also claim likewise that it would not qualify as been fatigue-related?

It would be a great shame for the aviation community not to learn the salient lessons from this accident and the tragic waste of human life in just attributing it to “Poor Training” and which might merely be corrected by the operator entering the following amendment into the Ops Manual:

“Don’t forget to input the new takeoff data into the computer; otherwise it will default to the last entered data for the previous takeoff”.

If you scratch the surface of any airline that rosters 30-hour duty periods you will find a lot of dirt under your fingernails.

[DOVES]

barit1 wrote:

--------------------------------------------------------------------------------
My lunchbunch colleague B-52 driver told me they always used (in the 50s-60s) an acceleration check - a time limit from 70 to 120kt I believe. If the time limit is reached before 120 is seen, it's a reject.
--------------------------------------------------------------------------------
Why don't you give a look to this:

http://www.pprune.com/forums/showthr...75#post1597375


Fly safe
DOVES