E145, Stuttgart Germany, 2009
E145, Stuttgart Germany, 2009
On 5 January 2009, a Flybe Embraer 145 made a late touchdown with slight snow falling on a runway pre-notified as affected by slush deposits and failed to stop until it had overrun into the RESA where it finally stopped on a heading 25º off the runway alignment. The Investigation concluded that although the airport operator process for determining braking action was flawed and two Safety Recommendations were made in that respect, the overrun of the 3045metres LDA was attributable to flight crew action and that operator guidance was deficient.
Description
On 5 January 2009, an Embraer 145 being operated by Flybe on a scheduled passenger flight from Birmingham to Stuttgart overran runway 07 at destination after a touchdown in normal day visibility and had only come to a stop in the subsequent Runway End Safety Area. None of the occupants were injured and all subsequently left the aircraft by use of the passenger stairs. Minor damage was caused to both the aircraft and to airfield lighting.
Investigation
An Investigation was carried out by the German BFU. DFDR and Cockpit Voice Recorder (CVR) information were available and successfully downloaded. It was noted that the aircraft commander had been PF and that their considerable aircraft type and overall experience was in marked contrast to that of the co pilot who had little flying experience which had mainly been obtained on the accident type.
It was noted that a Cat 2 Instrument Landing System (ILS) approach had been flown despite ATC clearance for a Cat 1 approach. This had required a landing flap setting of 22º rather than the 45º which was mandated for a landing on a contaminated runway and which would have reduced the LDR. FDR data showed that a stabilised approach at the appropriate speed had been flown but that touchdown had occurred nearly 700 metres from the start of the runway, slightly beyond the ideal point but not significantly so. Some directional control difficulties had been experienced soon after touchdown but more significantly, maximum possible braking had not been initially attempted and in addition, all braking had subsequently ceased for an 8 second period. As the end of the runway approached, the parking /emergency brake had been applied by the co-pilot without any such instruction from the PF and directional control, substantially maintained until this point, was then lost. It was noted that use of this system is prohibited on wet or contaminated runways by means of an instruction in the aircraft type volume of the Operations Manual.
It was noted that the runway braking action report on which the crew exclusively relied - ‘good’ as included in the ATIS first broadcast 20 minutes prior to landing - was only one part of the information available on what to expect and on how to react when braking action was initially less than normal. It was noted that “the crew did not take into account other present signs for the worsening condition of the runway. There was continuous snowfall at Stuttgart Airport and the runway was covered with a thin but visible layer of snow. This runway condition suggested a limited braking action. Since the estimation ‘braking action good’ was broadcast by ATIS at 1409 UTC, already 20 minutes prior to the landing, it could have been assumed that the conditions had deteriorated since then. Moreover, the reduced braking action detected during the first rollout phase should have been considered for all further actions.”
The absence at the aircraft operator Flybe of any “special provisions for winter operation” and in particular the absence of any “special procedures in case a reduced braking action was broadcast” were noted.
The TWR controller on duty stated that none of the five aircraft which had landed following snow removal had reported decreased braking action. However, it was concluded that by the time of the overrun, twenty minutes after the previous landing aircraft, the braking action on most of the runway was much less than that recorded during the most recent measurement made and broadcast on the Automatic Terminal Information Service (ATIS). The exception was at the very end of the runway where snow had not been removed, although in this small area braking action in dry snow was better than in slush formed after falling snow mixed with runway de-icing fluid, and this may have served to slightly reduce the extent of the overrun. The failure to update the ATIS once conditions began to change was considered inappropriate.
In respect of the airport operator’s procedures for measuring and communicating braking action on a runway surface which was affected by slush and in a situation where it could be reasonably presumed that conditions were changing after snow removal due to new snowfall onto a de-iced runway surface, the Investigation noted that:
- there had been a failure to recognise that deteriorating conditions coupled with the absence of any landing aircraft for 20 minutes prior to the accident indicated that “another measuring run should have been undertaken”.
- according to the responsible airport manager “pilot messages were used in the decision-making process in particular during ongoing flight operations. As long as operations continue without any reports on restrictions, then normally no friction measuring run is conducted so as not to interfere with traffic. As soon as there are first signs of a deterioration of braking values, a friction measuring run is undertaken".
- the equipment routinely used by the airport to measure braking action was expressly declared by its manufacturer as not to be used in slush because it “does not provide reliable values.”
With the aircraft ELW of 15400kg, well within the MLW of 18700kg, it was calculated during the Investigation that the LDR for the aircraft as loaded and configured in the conditions actually prevailing was approximately 2120 metres, comfortably within the LDA of 3045 metres.
The Cause of the overrun was determined as “the fact that the crew did not brake the aircraft consistently down to a safe speed”.
Three Safety Recommendations were issued as a result of the Investigation:
- The Flughafen Stuttgart GmbH should develop procedures permitting a timely and flexible transmission of the prospective braking action to the responsible air traffic control centre. [03/2012]
- The Flughafen Stuttgart GmbH should modify the Service Instruction D No. 43 so that the prevailing weather conditions are taken into account when decisions regarding measurement runs are made. When runways are contaminated additional corrections, which take the inaccuracy of the measuring method into account, should be made when classifying the friction coefficients. [04/2012]
- The Aircraft Operator (Flybe) should extend the instruction of the crews with respect to winter operation so that indications on a possibly reduced braking action are recognized and considered for the planning of the landing. [05/2012]
The Final Report of the Investigation BFU EX001-09 was published in August 2012