Cross Air 3597 24/11/2001

[PPRuNeUser0163]

Hi guys,

saw this crash involving a BAE146/RJ100 on nat geos Air Crash Investigations- and it got me thinking (something which was NOT mentioned at all during the program)..

a.) does the BAE146 have EGPWS/TAWS? I assume it would being an rpt category aircraft?

b.) why didnt either of these systems warn the crew of an excessive terrain closure rate with the mountain (via one of the modes of even a basic gpws) or give a 'pull up' type warning

It was very mystifying this crash- so I was just wondering about the above questions.. I can't help to think of how easily it could have been avoided (like with so many accidents!) but speaking from an armchair position id rather not go too much into this..

Also- interestingly given they got taken out by Swiss- would anyone know if excessive cockpit gradients exist within the above mentioned airline- (if crew which were possibly similar to the accident crew were retained and still fly for Swiss)..

Cheers,

N

[IGh]

A final rpt is available (pdf) on the web-site for Swiss AAIB, use the registration & date
BFU - BEAA - UIIA - AAIB (http://www.bfu.admin.ch/en/dokumentation_berichte_suchen.htm)
24.11.2001
6 km East Zurich Kloten, ZH
Switzerland HB-IXM

From Swiss AAIB report, pg 11:At 21:05:21 UTC flight CRX 3597 reported on the aerodrome control frequency. When the aircraft reached the minimum descent altitude (MDA) of 2390 ft QNH at 21:06:10, the commander mentioned to the copilot that he had certain visual ground contact and continued the descent.

At 21:06:36 UTC the aircraft collided with treetops and subsequently crashed into the ground. The aircraft caught fire on impact. Twenty-one passengers and three crew members died from their injuries at the site of the accident; seven passengers and two crew members survived the accident.

Investigation
The AAIB set up an investigation team designated to investigate an aircraft accident ...
The Analysis Sections do address several of the "human factors" you questioned. Regarding the "ground contact" observed by the pilots, from pg 111:"... the commander was provably used ... to descend below the minimum descent height, even when he had sight only of the ground and not of the runway. As the incident of December 1995 shows, the commander obviously believed he was capable of carrying out such a procedure even at night and under instrument conditions.
In view of the meteorological conditions, visual contact with the runway was highly improbable, since when it flew through the MDA the aircraft was still at an oblique distance of approximately 4.4 NM (8.1 km) from VOR/DME Kloten and 3.5 NM (6.5 km) from the runway threshold.
At 21:06:22 UTC a radar altitude of 500 ft RA was reached and the ground proximity warning system reported: “Five hundred”. This call-out may have caused the commander some displeasure ..."
From pg 120+ of the Swiss AAIB Investigation Report:3 Conclusions
...
• The ground proximity warning system (GPWS) issued no warnings because throughout its flight path the aircraft was outside the envelopes of
mode 1 – excessive descent rate and
mode 2B – excessive terrain closure rate.
...
• The recordings of the CVR and the radio transcriptions prove that immediately before reaching the minimum descent altitude (MDA) the copilot was occupied by tasks.
• The operator’s procedures specified a clear division of tasks between pilot flying and pilot not flying for this flight phase. The flight crew did not comply with these specifications.
• The commander deliberately violated the minimum descent altitude (MDA) for the standard VOR/DME approach 28.
• The copilot made no attempt to prevent the continuation of the flight below the minimum descent altitude.
• None of the crew members had visual contact with the runway or with the approach lights. Therefore the conditions for going below the minimum descent altitude (MDA) and continuing the final approach visually were not met.

[PEI_3721]

nkand, the accident occurred before EGPWS was mandated.
It is a sorry fact that the aircraft concerned was the last one to be delivered before EGPWS was the standard factory installation. The operator had plans for retrofitting the equipment in all other aircraft.
With the aircraft correctly configured for landing, no Pull Up warning would be given by GPWS in the accident situation.

With hindsight (even foresight) the accident was avoidable – why should an environmentally stipulated noise procedure lacking precision approach aids, and with visibility limits which did not match the altitude profile, be mandated instead of using a proven ILS approach.
Local and national politics!

As reported the cross-crew gradient was steep, but also, and perhaps a factor, the Captain had been the First Officers basic flying instructor, which could result in an instructor/student relationship in the difficult conditions. In particular, this related to the erroneous use of the APU to feed the airframe antiicing system, which would give low pressure alerts. These warnings were not widely reported as they occurred early in the descent, but would have been more prevalent (relevant – in icing conditions) during the approach and probably tied in with a comment from the Captain to ‘ignore the alerts’ – hence promoting an attitude that the instructor knows best; yet the First Officer (in learning mode) still attempting to understand the situation.
The flight deck instruments were configured in an unusual manner suggesting although the Captain was flying a VOR/DME, auto/FD approach, the EFIS ND was not displaying the HSI beam bar nor the inbound track in map mode – only the then ‘unauthorized’ FMS track.

The report concluded an act of deliberation by the Captain – descending below MSA, but without providing any substantiating evidence. How can an investigation determine thoughts and intent without verbal evidence? How can a report decide what the crew saw or thought they saw?
Simulations of the event identified a significant risk of misidentifying village lights at MDA as being the airfield.
It was demonstrated that the preceding aircraft flew level at MDA to the MAP (at minima visibility) and then landed off a 6 deg approach – poor procedure design and questionable crew procedures.

The report chose not to discuss inter-crew coordination at and after MDA, particularly where the PF calls contact and the PM remains head down and monitors the progress of the approach. The PM has no way of determining if the ‘contact’ is appropriate or otherwise (a {insolvable} hazard of non precision approaches and head down monitoring?), thus if the flight path appears reasonable (to an inexperienced pilot) then only the late and rapid reduction in Rad Alt could cue an unsafe situation.

Many lessons can be learnt from this accident. Unfortunately the report chose to focus on the Captain, the organization, and unfairly (IMHO) some key airline managers. The responsibilities of regulatory authority, airport, and local area management were not highlighted.

[flydive1]

Local and national politics!

International, actually, problems with Germany.

Also- interestingly given they got taken out by Swiss- would anyone know if excessive cockpit gradients exist within the above mentioned airline- (if crew which were possibly similar to the accident crew were retained and still fly for Swiss)..Kind of the opposite actually, Swissair was taken over by Crossair to form Swiss.(More or Less)
Yes there was lots of problem at the beginning between the different source crews(difference in salary, treatment, aircrafts, etc)
Do not know how's the situation at present.

[Mercenary Pilot]

With hindsight (even foresight) the accident was avoidable – why should an environmentally stipulated noise procedure lacking precision approach aids, and with visibility limits which did not match the altitude profile, be mandated instead of using a proven ILS approach.
Local and national politics!

Sounds like JFK!

[IGh]

PEI mentioned that important technical improvement not available in this CrossAir ALA-CFIT case: the newer TCF (Terrain Clearance Floor) included in the later Enhanced GPWS.

PEI also mentioned "... the accident was avoidable ... lacking precision approach ... instead of using a proven ILS approach...."

Before Enhanced GPWS, there were several cases of pilots flying an ILS approach, using a False Glideslope INDICATION, with NO Flags. One of these case happened at Zurich [sorry that Swiss Report must be in German]:
"... über den Unfall des Flugzeuges DC-9-32, ALITALIA, Flugnr. AZ 404, I-ATJA am Stadlerberg, Weiach/ZH, vom 14 November 1990 ..."
Alitalia 404 / 14Nov90 DC-9-32 I-ATJA , en route from Milan ... descended on ILS-approach to Zurich Rwy 14, impacted about 9.6 km from airport at night. All fatal, 40 passengers and six crew.

Even an ILS-approach (glideslope) proved unreliable sometimes. There were several presentations from investigators describing such descents on false glideslope, here's that Air New Zealand B767 /29July2000 at Apia:
Aero 21 - Erroneous Glidescope (http://www.boeing.com/commercial/aeromagazine/aero_21/glidescope_story.html), and NZ's CAA offers their report "CAA Occurrence No. 00/2518"

About ten years ago BEA-guys made a presentation describing an MD80 CatIII coupled-approach to Paris, the crew flew final down to DH, then did a Missed' -- but the controllers noticed their position as still several miles out on final [good AAR, in English, but I can't locate the report].

Enhanced Ground Proximity Warning Systems Near Miss

In February 1999, an Air France Boeing 777 was approaching runway 10 at Rio de Ja-neiro’s Galeao International Airport. The flight was headed in an easterly direction thus blinding the pilots by the early morning sun. In addition, the runway itself was shrouded in mist. The pilots however continued their approach; relying on the Instrument Landing System (ILS), that provides the proper glide slope down to the runway.

Unknown to the flight crew, the Brazilian technicians were servicing the ILS transmitter. Following a false signal sent by the airport transmitter, the jet was descending toward a hill 4.8 kilometers short of the runway. Bertrand de Courville, Air France’s manager of flight safety, said that at the last moment the descending jet broke out of the mists.

The pilots, expecting to see the runway, were startled to see the hill. At the same moment, the jet’s Enhanced GPWS displayed the hilltop in red and issued an alert, “Terrain, Terrain. Pull up, Pull up.”

“The pilots leveled off,” missing the hill by about 98 meters, Mr. De Courville says. The jet then landed safely. Although the pilots had spotted the hill, “there were few visual cues, and the enhanced GPWS was very helpful” in confirming the danger ahead, enabling the pilots to save the plane.

[A37575]

Although the pilots had spotted the hill, “there were few visual cues, and the enhanced GPWS was very helpful” in confirming the danger ahead, enabling the pilots to save the plane.

Was there a appropriately placed DME at or near the airport that could have given the crew an approximation as to the accuracy or otherwise of the ILS three degree slope? It is called double or cross checking and DME is a very handy navaid for that?