View Full Version : Post accident flight inspections
14th Feb 2012, 18:56
There have been a few threads over the years which have dwelt on accidents or incidents where radio navigation aids may have played a part. With the stability and improvement of modern aids and flight management systems, such events are mercifully very infrequent and that is good. That apart, there is little to be seen on the role of any subsequent related flight check.
Out of curiosity, what sort of an inspection is carried out these days and why? In the UK, who calls for it and who does it? Are any parameters specified? Is it a ‘redo’ of a routine inspection or is it required to replicate the situation at the time so as to get a performance reading under similar conditions? Is there any check now when GPS might have been involved??
My interest is as a retired pilot and now a bystander. Many thanks.
16th Feb 2012, 04:11
As far as verifying radio aids after an accident is concerned, if it involved an aircraft fitted with a modern Flight Data Recorder, it should be perfectly possible to verify the ILS beam structure without further flights. The general principle is to reconstruct the aircraft's flight path back up from the touchdown point or the position it ended up at using data from either the Inertial Reference System or the Triaxial Accelerometer. The FDR also records the Localiser & Glideslope at regular intervals so the overall beam can then also be reconstructed. For example, if the aircraft was 20 feet to the left of the centreline but the Localiser gave the equivalent of 40 feet left, it would be evidence that the beam was off by 20 feet for the accident flight.
FDRs also record Latitude & Longitude usually from the FMS (which is actually taken from the GPS) so if the received GPS signal was off that too would be evident to an investigator.
Can't help you with UK Flight Inspection - in the US it's the FAA and in Canada it used to be Transport Canada but may be NavCanada now.
18th Feb 2012, 17:51
Thank you for that. It is food for thought and discussion.
I was hoping to spark an open forum discussion on the post accident flight test of radio navigation aids. Hopefully it would bring in the users (pilots) of radio navigation aids, the providers and operators of said facilities, the engineering brethren, plus the flight test pilots and navigation aid inspectors to discuss how they each and collectively saw the purposes, obligations and liabilities of the process.
Why? In this increasingly litigious world, many aspects of aircraft operation are examined forensically in the event of a serious incident - and more so in the event of a fatal accident where damages may be awarded. I can not recall when there was last a post accident flight within the UK where the interface between the user, the operator, the service provider and the flight inspection organisation was last tested. Perhaps this is an appropriate time to open the door.. especially with the addition of SES Regs.
There seems to me to be a disconnect between the ICAO documentation, which has an All-weather Ops base, and CAP 670 - Air Traffic Services Safety Requirements which approves and regulates Flight Inspection Organisations and deals with the operation of, inter alia, radio navigation facilities. This document appears to be predominantly engineering requirement based with little specific operational exposure.
This theme seemed right for exploration. Elsewhere on this site analytical discussions take place on engines, airframes, systems, performance, operators, knowledge base and even public expectations and perceptions. If this Forum is wrong, please could the Moderators move it. If the topic is inappropriate please would they delete it.
Rather than develop my thesis any further right now, are there any takers for a safety audit type discussion?
Just a couple of immediate questions – in the event of a post fatal accident tomorrow where a radio navigation facility might have been involved. Who does the flight test? Who calls for it? What does the test involve explicitly? And what may the follow up liabilities be? All these things should be in place now – and maybe they are.
21st Feb 2012, 12:05
Just to look at things a little differently;
A; In 1978 Lord Justice Denning made an observation in a case (Alidair V Taylor) likening aviation to the nuclear industry because of the potential fatal consequences of the slightest slip in the high standards expected and demanded. (perhaps someone can provide quote.)
B. ICAO Annex 10 is the basic and highest internationally recognised statement of requirements for the use of radio navigation aids in all-weather operations.
C. ICAO Annex 10 stated that irrespective of the operational objective, that the rate of a fatal accident during landing due to failures in the whole system, comprising the ground equipment, the aircraft and the pilot should not exceed in a million.
D. In Cat l this responsibility is vested more or less in the pilot…..
E. In Cat ll or lll the requirement must be inherent in the whole system.
F. ICAO Annex 10. Course bends are unacceptable when they preclude an aircraft under normal operating conditions from reaching the decision height in a stable attitude within 2degrees of pitch and bank and at a position from which a safe landing can be made.
G. ICAO Annex 10. For Cat ll and lll performance in the approach phase it is desirable to restrict course bends to displacements of less than fifteen feet either side of the centreline and vertical displacements of less than 4 feet at 50 feet – this allows approx. 30 feet of wheel height at the threshold…
H. ICAO Annex 10. Automatic and semi-automatic coupling is effected to a greater degree by the presence of bends.
I. Excessive control activity after the aircraft has settled on the an approach path may preclude it from successfully completing an approach or landing.
J. Doc 8071 stated; the purpose of flight testing should be to confirm the orrectness of the setting of essential signal-in-space parameters and determine the operational safety and acceptability of the ILS installation.
K. Doc 8071.Flight testing encompasses those tests carried out in the air by a trained crew in a specially equipped aircraft. These tests re required to examine the resulting signals in space as they are presented to an aircraft receiving system after being influenced by factors external to the installation;
L. Doc 9071 The flight test crew must be experts in their individual fields, have sound knowledge and experience in flight testing procedures and requirements. Their competency must be of a high order.
M. Doc 8071. Flight tests, because they represent in-flight evaluation and sampling of the radiated signal in the operating environment will continue to be an important final requirement in the proof of performance of a facility.
All italics are mine. All references were valid but may no longer be so. Perhaps someone could verify their validity for today.
ICAO documentation covers the complex significant topic of bends and gives some possible causes.
A bend may be permanent or transitory. A bend is a displacement of the received signal. An aircraft following a bend will appear to be on course. An aircraft unable to follow a bend sufficiently accurately or quickly enough will appear off course. To a pilot a bend can seem very similar to wake vortex, windshear or a sloppy autopilot/FMS.
Are these ICAO observations etc still valid and reasonable? A Lord Denning type view would be significant.
These ‘requirements’ should be referenced or enshrined in UK material. ICAO sought that wherever possible ICAO terminology should be carried into national material. Can terminology differences cause scope for confusion if the specific purposes of the use of each is not defined?
Is CAP 670 to radio navigation aids as airworthiness requirements are to the acceptability of aircraft and systems. Or similar Flight OPs requirements.
(For the non pilot community, Cat lll operations allow operations where no minimum cloud base is required and forward visibility against a very powerful light is not much further than the length of a large passenger aircraft. You ould not want to drive in it or even walk in it on a dark night. Try and visualise 15 feet, then four feet then 30 feet and reference it to your own height.)
There have been two recent fatal accidents where radio navigation aids have been involved and the ICAO system (the ground installation, the aircraft and pilot) has failed. Many years ago there was a very serious incident at Heathrow when a 747 on A Catlll approach supposedly came within 70 feet of the Penta Hotel. To my knowledge, that incident was never flight checked at all, let alone under similar conditions. (The subsequent jury in court were, I think, stated as having an average age of 26 and no knowledge of aviation other than as fare paying passengers.) The industry lost a significant opportunity and obligation to learn. Lets not wait for the accident to have a critical look at the system.
In the event of a failure of the ICAO system would you expect there to be a benign request for damages by the victims representatives? Look at some of the punative damages awarded. Look at the efforts made by the agencies elsewhere to trace the accountability of factors and responsible operators. Cork accident or the recent Schipol accident come to mind.
It is possible that the previous posts have provoked consideration. It is not necessary for a response to indicate reaction. If someone in qualified authority is prepared to say that all is well, that would be good. Probably so too would be a statement that this is all rubbish and the product of the obsessed mind of SLF. - who owes me a Duty of Care and is it met?
Genghis the Engineer
21st Feb 2012, 16:42
Do you mean these people?
Flight Calibration Services ltd | The Alternative Flight Inspection Solution! (http://www.flightcalibration.co.uk/)
Cobham plc :: Mission Systems, Aviation Services, Darlington DTV, Flight Inspection Services (http://www.cobham.com/about-cobham/mission-systems/about-us/aviation-services/darlington-dtv/products-and-services/flight-inspection-services.aspx)
ILS calibrations are required very regularly.
I've not on the other hand heard of calibrations being required following an accident. Perhaps somebody who works in airport operations can advise on that?
21st Feb 2012, 21:31
Blurring the distinction between fact and fiction -
Ernest K. Gann's "Band Of Brothers" is a story of airline pilots seeking to exonerate one of their own after an approach accident. They form their own navaids check team, fighting a government straitjacket, and eventually find what led their colleague astray.
The story is a fictional appendix to a real accident - Accident Database: Accident Synopsis 02161968 (http://www.airdisaster.com/cgi-bin/view_details.cgi?date=02161968®=B-1018&airline=Civil+Air+Transport)
... and lends some insight into the investigation process.
21st Feb 2012, 21:59
I've not on the other hand heard of calibrations being required following an accident....... From CAP493 Appendix C:-
Flight Inspection of Radio Navigation Aids and Radar
2.1 Types of Flight Inspection
The flight inspection of radio navigation aids and radar can be divided into four basic
a) A commissioning flight inspection to determine the state of a facility before it is
brought into operational service;
b) Routine flight inspection at specific intervals to check that facilities are operating within tolerances;
c) Special flight inspection in addition to routine flight inspection whenever the
performance of a facility is suspect and airborne measurement is required. This
may affect all or part of the facility;
d) Accident/Incident flight inspections to determine the performance of all
equipment, used or considered to have been used by the aircraft concerned, which
could have contributed to the accident or incident.
Genghis the Engineer
22nd Feb 2012, 11:42
Very sensible, but does anybody know how often (c) or (d) of that list actually happen?
1st Mar 2012, 11:06
I've been working for Cobham since 1999 and flying King Airs on flight inspection as a Captain for the last 6 years.
The routine inspections are generally every 180 days, I can't remember ever having done a post accident inspection myself but the company have been involved in them at Amsterdam, Heathrow, Birmingham and Cork (i think).
Any accident on a approach will probably result in a post-accident check to ensure that the ILS wasn't to blame. It's rare that we find major issues with ILS's but they do quite often require "tweaking" to keep them in the required, very tight, tolerances.
Currently we look after systems for the MoD, UK civilian customers, a lot of stuff in Spain, Portugal, Belgium, the Netherlands, Germany, lots of places in the Middle East and the RAF's overseas stuff in Cyprus, the Falklands and at Camp Bastion.
Need to know anything else just ask ;)
8th Mar 2012, 14:52
Thank you for your replies and forbearance reading my poorly previewed and print-dense posts.
Is it adequate that a flight inspection aircraft is not required to be capable of flying the system, even in good weather?
In order to operate an airport in bad weather an airport operator needs to provide an adequate instrument landing system. This safety critical system has an ICAO safety objective which must be met across the ground installation, the aircraft and the pilot. The airport operator can not offer his equipment for use without a certificate. A certificate can only be issued by an approved organisation.There are recognised anomalies and factors which influence the radiated signal and its effect on the aircraft using the facility.
From CAP 670 ATC Safety Services Requirements. This certificate is issued on the strength of a prearranged flight check at 6 month intervals under ‘calibration’ conditions of weather and traffic. The purpose of this check is to confirm the accuracy of the radiated signal in space meets engineering standards. Reasonable because the base signal radiation pattern has to be correct or the resultant signal in space can not be.
Can this check reasonably be interpreted as meeting the complete industry expectations of safety of performance, to meet the ICAO Safety Objective, under all acceptable non-calibration conditions; be that airport owner and operator, or the users within the airline industry?
I guess it must be, as CAP 670 permits this inspection by an aircraft which is not required to be capable of using the ILS to Catll or Catlll standards even in good visibility - let alone have an autopilot.
CAP 670 (My italics, bold and underline.)
Annex 1 - Instrument Landing System
5.1 Manual flight control using only the mandatory navigation instruments is not considered sufficiently accurate for inspection of the following types of ILS:
• Category III systems.
• Category II systems.
• Category I systems, which the operator wishes to use for autoland in good visibility.
5.2 For inspection ofthe above systems the aircraft shall be fitted with equipment which will providerepeatable following of the required path (1549). Systems considered suitable to this purpose include telemetry of the ground based tracking system’s output to a separate instrument in the aircraft, OR anautopilot. If an autopilot is used the CAA shall be satisfied that it is capable of safe operation down to 50 feet above the threshold elevation (1550).
I have no concerns about the measurement of transmitted signal accuracy (other than post-accident) but not sure how matters such aircraft stability and attitude at the threshold, acceptability of bends and control motion noise etc. are determined. All are cited as factors in the chain between transmitter and an acceptable installation. Maybe this is met by some other agency and some other way but if so where does it fit into meeting the safety objective?
I believe there used to be random checks with inspection aircraft just fitting unannounced into the traffic system and sampling the facility. This concept seems to have gone - probably for commercial reasons - but it would offer a measure of reasurance.
How are the requirements of a post accident or incident to be met as specified – let alone with the added disruption possibly caused by the accident itself - and how and by what standard is the ILS deemed safe for operational use? One fatal accident in 10,000,000 landings is the required safety objective.
I am just SLF and curious. Reassure me please.
8th Mar 2012, 18:18
Non precision approaches in EU are tested to 100 feet below system minima eg: NDB, VOR, SRAs
ILS CAT 3 has Roll Out Guidance check for auto-land which is flown at 50 feet manually (180KIAS gear up in Multi crewed B200/350s) as there are no Calibration aircraft with an A/P that will respond accurately enough.
ILS signal in space is checked both above and below glide path, as well as left and right of localiser based centreline.
Cobham and Aerodata GmbH use a Flight Inspection system that utilises a ground based laser- this requires suitable MET visibility to lock onto the aircraft's reflector - it gives extremely good accuracy and acts as a datum line for the ILS calibration.
They also test and assist with the system design of SIDS/STARS/PARs/MLS/ and GPS based approaches.
Post Accident: I recall LHR was tested after the BA B777 glider incident.
10th Mar 2012, 12:30
Perhaps looking at a door from different sides. The purposes of a Post Accident are unclear.
Is it to check the accuracy of the signals in space before the aid can be brought back into service?
Is it to have any significance in studying how the system failed?
My understanding is the latter case.
From Post 7; d) Accident/Incident flight inspections to determine the performance of all equipment, used or considered to have been used by the aircraft concerned, which could have contributed to the accident or incident. (No restriction to accuracy of signals in space.)
If it were the former then it could have been more cleanly stated.
To put bones on the question. Take the case of the NO incident in 1989 when a BA747 came very close to the Penta Hotel at Heathrow. Would a PAFI have simply looked at the signals at say 0300the next morning and said OK. Or would the PAFI perhaps have looked at the effect of a 747 moving slowly on the runway, looking for a turn-off in fog, and a subsequent delayed clearance to land? Perhaps someone in the FI community can say with significant certainty that the aircraft on the runway could not have been a factor.
A PAFI seems a process outwith CAP 670.
What is the purpose of a Post Accident flight and how is it accomplished? What is the required standard?
10th Mar 2012, 14:05
Herewith a statement of the purpose of a PAFI from the daysof CAAFU.
A flight inspection having top priority carried out on an ILS which was used, or might have been used by an aircraft involved in an accident and which might have contributed to the accident. The post Accident flight inspection is to be completed as soon as possible after the accident and should normally be completed within 12 hours of the accident.
That is my baseline. It appears to have been diluted.
Is it significant that a PAFI still takes precedence overflight by Heads of State? That would seem to endorse the urgency element.
To Angel Orange. With the greatest respect, how was that CAT lll requirement achieved previously? What was the 748 capability, both in developing the system at the BLEU and in performance testing under CAAFU?
I also note the need for the aircraft to respond accurately enough. Why should it need to respond here especially when it is not needed to respond elsewhere?
Eric T Cartman
11th Mar 2012, 11:01
Very sensible, but does anybody know how often (c) or (d) of that list actually happen?
In 42 years of ATC, I've only come across this once. I guess the angel on my headset has been looking after me ! ;)
In the early 80's someone landed safely at Liverpool on the closed runway 08 instead of 09. I believe he said he had been misled by a false back beam from the 27 ILS. There was no ILS for 09 - it was SRA's to 1/2 or 2 miles, or a visual. Anyway, CAAFU promptly sent a 748 from Teesside & carried out item d) = "Accident/Incident flight inspections" in spekesoftly's list. Needless to say, there was no fault with the ILS.
12th Mar 2012, 03:51
Assuming we're talking Air Transport rather than GA, I would still regard FDR data from the accident aircraft more meaningful than surveys even within 12 hours. To my knowledge there have been a couple of reports of bad Glideslope that did not show any evidence to a later survey aircraft. My suspicion is that it was interference caused by a cell phone tower adjacent to the beam which only was apparent during normal daytime operations but non-existent after hours when the check flight was performed.