On 19 March 2008, a Dornier 328-100 being operated by Cirrus AL on a scheduled domestic passenger flight from Berlin Tempelhof to Mannheim made a long landing on the dry surface of runway 27 at destination in benign daylight weather conditions following a non precision approach before leaving the paved surface and colliding with an earth bank about 50 metres beyond the end of it. The impact severely damaged the aircraft but there was no fire and only five of the 27 occupants sustained injuries, all minor.
An Investigation was carried out by the Bundesstelle für Flugunfalluntersuchung (Germany) (BFU) and Flight Data Recorder (FDR) and Cockpit Voice Recorder (CVR) data was successfully downloaded. It was found that the inexperienced First Officer had been designated PF and that, although the aircraft commander was relatively experienced,” his leadership skills as a PIC were not always without fault”.
During the flight from Berlin, it was noted that there was considerable evidence of the aircraft commander ‘coaching’ the First Officer and that whilst he was occupied on a Company call on COM 2, a significant level bust had been made by the PF which had taken the aircraft below the prevailing Minimum Sector Altitude.
The LLZ/DME approach to runway 27 had proceeded broadly in accordance with the promulgated procedure with the aircraft configured for landing soon after passing 1000 ft agl. As the MDA of 760 feet Altimeter Pressure Settings was passed, the AP was disconnected and the aircraft controlled manually thereafter. Pitch control and power selection remained somewhat erratic but the threshold was subsequently crossed at 40 feet agl with only marginal excess speed over the calculated Reference Speed (Vref). However, by this time, a 30% torque had been set and the aircraft began to float in ground effect at about 10 feet agl. As the TDZ of the 1066 metre long runway was overflown, the First Officer handed control to the aircraft commander, who attempted to get the aircraft on the runway by manipulating the control column but without selecting flight idle power.
FDR data showed that initial ground contact occurred about 530 m beyond the threshold with about 20% Torque set. After a couple of small bounces, a final touchdown at 93 KCAS occurred. The aircraft commander was then unable to retard the power levers to get into the beta or reverse pitch range and called for the emergency/parking brake to be set. When there was no response to this, he reached over the central pedestal and selected it himself. Whilst the aircraft commander’s right hand was removed from the power levers to set the emergency brake, the First Officer had attempted to retard the power levers himself and was successful, with reverse achieved and ground spoilers fully deployed leading to a marked increase in the rate of deceleration.
The aircraft passed the end of the runway with a groundspeed of about 50 knots and then left the paved surface beyond the end of the runway at about 30 knots. The left main landing gear collided with an obstruction and buckled and shortly afterwards, the left wing and left engine collided with an earth wall aligned at an angle to the runway which caused the aircraft to turn right and come to a stop with the nose directly in front of the earth wall.
There was no PA to the cabin to brace for impact and flight crew actions in relation to aircraft shutdown and evacuation were such that it was considered that they had been "overwhelmed by the situation”. It was noted that 8 seconds elapsed before the aircraft commander gave the order to evacuate the aircraft and that he did this without a review of external risks or, in particular, first shutting down the right hand engine.
Since ‘Human Factors’ were considered central to the circumstances of the accident, the Investigation reviewed the available evidence and concluded that despite the fact that the flight was a regular scheduled flight with no unusual circumstances such as difficult weather conditions or aircraft technical defects, the flight crew had “reached their performance limit during the approach and at the end of the flight went beyond it”. It was noted that although the training documentation for both pilots “showed individual signs of below-average performance” they nevertheless could be considered to have possessed “sufficient skills”.
It was concluded that:
“A reason for (their) high stress level was primarily the non-compliance with the…operator’s SOPs. The pilots could, in several instances, not put their full attention on the respective flight phase because they had deviated from the SOPs (e.g. radio communication with the company below FL100) or implemented them too late (e.g. the aircraft was configured after the FAF instead of before the FAF).”
Significantly, the Investigation “thought it highly unlikely that these deviations were just an isolated case or limited to this particular flight or crew”.
The Investigation also noted that operations at Mannheim Airport had been facilitated by a series of approved exceptions to normal safety standards and procedures triggered by the location of the runway between two public roads. These exceptions were found to include:
- the provision of the maximum possible runway length by limiting the Runway End Safety Area at the end of runway 27
- accepting that the creation of an earth wall at the end of the reduced length runway 27 RESA would protect third parties from aircraft overrunning the runway but ignoring the need to protect the occupants of overrunning aircraft.
- the installation of a Visual Approach Slope Indicator Systems set to a steeper angle than the Instrument Landing System (ILS) GS as a means to ensure a sufficient altitude when crossing a public road just prior to the runway 27 threshold which required flight path ‘correction’ at a low level and was contrary to the responsible regulator’s own policy of encouraging the conduct of stabilised approaches.
It was also noted that the runway markings were not in compliance with the applicable regulations in that there was neither the required aiming point marking nor the recommended TDZ markings.
The aircraft in its final position (Reproduced from the Official Report)
The difficulty in retarding the power levers, whilst not central to the cause of the accident, was found to be a longstanding problem which had not been properly handled by any of the parties involved.
The Investigation concluded that the Causes of the overrun were:
- the failure to initiate a rejected landing when it was clear that a safe landing was not assured
- the fact that after touchdown the power levers could not be pulled into ground idle or reverse by the PIC (PF)
- the fact that the power levers were not retarded to flight idle during the flare (because) neither pilot was aware of their position
- the fact that the crew deviated from the SOPs and as a result reached their performance limit and at the end went beyond it.
The following Contributory Factors were also identified:
- the Operator’s non-precision approaches and the landings at Mannheim City were, with a significant frequency, conducted other than in accordance with the requirements of the OM.
- the aircraft manufacturers recommended procedures were not incorporated into the Operators OM Part B or OM Part D.
- the practical training of the flight crew by the Operator in respect of the prevention of an erroneous operation of the power levers using instructions published by the aircraft manufacturer had been inadequate.
- the crew conducted a non-precision approach which did not comply with the Operators OM requirements or the AIPs procedure.
- the design of the power levers was not sufficiently fault-tolerant.
- the existing risks caused by problems during the use of the power levers were not recognised correctly and remedied by the responsible authorities and type certificate holders in spite of several occurrences and safety recommendations.
- the TDZ at Mannheim City Airfield was not marked.
Safety Action indicated by the accident and already taken by the time the Investigation was completed was noted to have included:
- The development by the aircraft type certificate holder of Service Bulletin SB 328-76-486 ‘Engine Controls – Modification of Power Lever Assembly and introduction of Crew Aural Alerting Device’ to minimise the risk of an crew operating error which has subsequently been adopted by EASA as part of AD 2009-0196.
- The provision of painted aiming point markings on the runway by the Mannheim City Airport Operator.
The following eight Safety Recommendations were issued as a result of the Investigation:
- that the Operator should, in order to qualify and quantify more accurately and permanently, remedy the existing deficits in the adherence with the SOPs within the Do 328 fleet, extend the existing flight data analysis program for aircraft with a MCTOM of more than 27 tonnes in accordance with ICAO Annex 5 and EU OPS 1 requirements to the Do 328 fleet (and particularly use) this program (to collect and review) data which document adherence to the criteria for stabilised approaches and (touch down within) the touch-down zone. [26/2012]
- that the Mannheim City Airfield operator should mark the runway in accordance with the ICAO recommendations. [27/2012]
- that the Mannheim City Airfield operator should extend the runway end safety areas according to existing regulations and recommendations or carry out suitable
constructional measures like the installation of energy absorbing surfaces or shorten the runway distance available. [28/2012]
- that the Luftfahrt-Bundesamt (LBA) should define a written process for the handling of safety recommendations which ensures that they are processed, assessed and answered in a documented fashion. [29/2012]
- that the European Aviation Safety Agency (EASA) should define a written process for the handling of safety recommendations which ensures that they are processed, assessed and answered in a documented fashion. [30/2012]
- that the Operator should modify its fuel policy such that the special requirements of Mannheim City Airfield are sufficiently taken in to account. Special attention should
be paid to the adherence of the maximum allowable landing mass. [31/2012]
- that the Luftfahrt-Bundesamt should ensure that in all air operators’ operations handbooks, distinct criteria for stabilised approaches for all kinds of approaches, including in Visual Meteorological Conditions (VMC), are described. [32/2012]
- that the Operator should ensure that in its operations handbooks distinct criteria for stabilised approaches for all kinds of approaches, including in VMC, are described. [33/2012]
The Final Report of the Investigation: BFU 1X001-08 was completed on 8 August 2012 and published in German the following month. It was subsequently made available in English translation.