On 4 March 2016, the crew of an ATR ATR-72-500 (G-COBO) being operated by Aurigny Air Services on a scheduled passenger flight from Manchester to Guernsey found as they conducted their take-off and climb that an abnormal amount of nose down pitch input was continuously required to maintain the appropriate pitch attitude and a diversion to East Midlands was made without further problems, with the control problem dissipating prior to commencing the final approach there.
After a three day delay in notification because neither the aircraft Captain nor the airline involved considered that the occurrence was notifiable, the UK AAIB was advised and assessed that it had been one ‘which could have caused difficulties controlling the aircraft’ and therefore constituted a 'Serious Incident'. The Operator had already downloaded both the FDR and the CVR after the diversion and the data obtained were passed to the AAIB.
It was noted that the 58 year old-Captain had joined the Operator two months earlier already type rated and current on the ATR42/72 and had 8,276 hours total flying experience including 928 on type. He had accumulated "most of his flying experience while based in Guernsey where, due to a mild climate, de-icing/anti-icing is seldom necessary and then usually only after overnight frost" and so "his experience of operating in freezing precipitation was not extensive". The 30 year-old First Officer was in his first airline job and had 920 hours total flying experience including 620 hours on the ATR72 all with the same operator. He had "no previous experience of flight in snow or of de-icing/anti-icing between flights" and the diversion during the investigated flight was his first during commercial operations.
It was established that the aircraft had landed at Manchester after a flight from Guernsey at 0810 and the turnround prior to the investigated departure had been more than an hour, including almost half an hour waiting to get onto their allocated parking stand because the aircraft on it was being de-iced. As they began the taxi in after landing "the crew's perception was that light, wet snow was falling and melting on taxiways, although some was lying in the adjacent grass areas". The Captain was recorded as having remarked at that time that "it doesn't appear to be sticking […] so I think we can get away without de-icing”, adding that he would “have a good look”. Later, whilst waiting for the stand to become free, the Captain "said he did not see any snow settling on the aircraft and suggested snow visible on other aircraft had probably accumulated overnight". In fact, the snow had only begun falling shortly before they had landed as the temperature fell to 0°C.
Throughout their time on the ground, it was snowing and both the Air Temperature and Dew Point remained at 0°C. The First Officer stated that with no previous experience of ground operations in snow, "he had been trained that an aircraft was clear of ice if none was visible on the Ice Evidence Probe, the leading edge de-icing boots or the propeller spinner and he saw no ice in these locations while taxing-in”. The two cabin crew were Guernsey residents who "seldom saw snow […] and they took photographs while the aircraft was parked" which "showed snow lying on the fuselage" and on the coat of the senior member of the cabin crew who "later recalled brushing snow from her coat when passengers began boarding". The Captain had carried out the pre-flight external inspection and although aware from the 0850 ATIS that the air temperature was 0°C, he had advised the handling agent that "no de-icing procedures were required". The aircraft pushed back from the stand at 0910 with the Captain as PF and subsequently began its take-off 9 minutes later.
When rotation was attempted "the Captain found he needed to apply less aft pressure on the control column than he anticipated and, once airborne, had to push forward on the column to achieve the (required) climb attitude". He used the electric pitch trim switch several times to trim nose-down and the aural warning which is activated when this action continues for more than one second was triggered. The AP was engaged one minute after take-off but disconnected after about two minutes. When it was immediately re-engaged, it disconnected after one minute and the First Officer "noticed that the pitch trim indicator showed full nose-down deflection". In response to the 'Pitch Mistrim' message was displayed on the AFCS, reference was made to the corresponding QRH drill but this required only that the AP should be disengaged. The First Officer "suggested the problem might have been caused by contamination" and the Captain responded by saying that "the reason he had been happy to depart was that the snow was not sticking to the aircraft and, because it was wet, he expected it to have blown off".
As the aircraft continued to climb in Instrument Meteorological Conditions (IMC) and icing conditions, two further attempts to engage the AP were made but it disconnected after three minutes the first time and less than two minutes the second time. The Captain asked the First Officer to "check the load sheet" and said that they would divert "if he was still having control difficulties once they levelled-off". Once level at FL170, the First Officer reported that the load sheet "looked very similar to the one from the previous flight" (which had been similarly loaded) and soon after this, the Captain decided to divert to East Midlands because of the continued need to maintain abnormal forward pressure on the control column. The diversion request was made to ATC without any declaration of urgency or emergency and the reason given was "a pitch trim problem". The aircraft flew clear of cloud soon after the descent began.
As the descent continued, the aircraft became easier to fly and the Captain was recorded as saying "I reckon it was ice, I’ve got the pitch trim back”. A normal approach to and landing on runway 27 at East Midlands, where the air temperature and dew point had been given on the ATIS as temperature 3ºC and 0ºC respectively, was made. After shutdown, the horizontal tailplane was inspected and when no ice was found, the Captain "decided the problem might have been mechanical and placed the aircraft unserviceable". After an engineer had inspected the aircraft later that day and found no faults, the Operator had downloaded the FDR and the CVR and returned the aircraft to service.
The Investigation was able to eliminate both mis-loading of the aircraft or a lack of airworthiness as factors in the pitch control difficulty experienced. Initial considerations of a possible cause for the control difficulty also included the possible effects of re-hydrated thickened de/anti ice fluid residues. It was noted that the Operator stated that because it seldom needed to use thickened fluids, their maintenance requirement had, for several years, been that a check for residues was only required "after an airframe had received 10 applications of thickened fluid". In fact, it was found that the single time the aircraft had been treated with thickened fluid during the 2015/16 winter season had been four days before the flight and although no specific check for residue had been triggered by this, it was noted that the engineer who inspected the aircraft after the flight (which was the subject of the Investigation) saw no evidence of residue on the tailplane. It was noted in passing that the Operator's residue inspection practice was contrary to the explicit requirement in Part 'A' of their Operations Manual which was found to state that such an inspection was required "within three days of (any) thickened fluid application".
The focus of the Investigation therefore turned to the possibility of frozen deposits affecting the operating and trim of the elevator. Based on the evidence available, it was confirmed that "for much of the flight the elevator had a tendency to deflect upwards and that the aircraft would have adopted an undesired nose-up attitude but for the maximum nose-down pitch trim which had been applied and the pilot’s control input". It was noted that the mechanically controlled elevators were each linked mechanically to trim tabs.
An analysis by the aircraft manufacturer based on the available FDR data concluded that the "abnormal nose-up pitching tendency was consistent with the aerodynamic effects of upper surface icing on the horizontal tailplane". The manufacturer’s aerodynamic explanation for this was that "with ice contamination to the upper surface of the horizontal tailplane, which may also extend over the upper surface of the elevator (and possibly over the elevator tab as well) […] ice on the upper surface thickens the airflow boundary layer". This reduces the downward pressure on the upper surface and since there is no comparable effect on the lower surface, any deflection of the elevator in this condition will mean that less downward aerodynamic force will result, so that additional input to provide the downward force which would normally result with a clean elevator is required. Dependent on the extent of the additional downward force required, the elevator trim may reach its stop before the desired result has been achieved, so that only additional force applied manually through forward movement of the control column could maintain the desired pitch attitude. The diagram below illustrates this.
The result of a reduction in downward force on the elevator is an increase in the pitch up for a given column position and a reduction in the amount of corrective elevator trim tab movement available. [Reproduced from the Official Report slightly reduced]
During the Investigation, it also became clear that the elevator trim tab had run to a maximum nose down position of +1.76º whilst under normal control instead of being limited to +1.5º. Tests indicated that this had been the consequence of a previously unidentified failed microswitch within the left trim tab actuator with the unappreciated-at-the-time consequence of this being that the pitch trim system had been assisting the Captain "to a greater extent than would be expected under normal control" by responding as if the standby trim switch on the central console had been used, which it was not.
Two previous AR-72 events attributed to a similar cause were identified, one of which was a 2012 fatal accident to a Russian aircraft in which the crew lost control after failing to de/anti ice prior to departure.
In terms of the crew decision making involved in the investigated event, it was noted that according to aircraft ground handling providers at Manchester, "all other commercial aircraft which departed on the morning of 4 March 2016 sought de-icing/anti-icing before start-up" and that "de-icing providers had difficulty coping with the demand for their services" with flights either delayed or cancelled as a result. This was not unexpected given that with both a temperature less than 3ºC and visible moisture present, it was apparent both ‘atmospheric icing conditions’ (FCOM) and ‘freezing conditions’ (OM) existed. These were noted to be unequivocally conditions in which, whatever opinion the flight crew had held, the guidance was "to de-ice/anti-ice an aircraft completely and check it afterwards". It was considered that had the First Officer previously experienced winter operations, "he might have questioned the […] decision not to de-ice/anti-ice" before departure.
It was noted that pilot decision making can be affected by various cognitive biases which can interfere with a balanced consideration of all the available evidence. If an initial mental model of a situation is incorrect, pilots' can be prone to "confirmation" or "plan continuation" bias which may then prevent them from "accepting clear evidence that contradicts their initial assessment". It was additionally noted that pilots are also susceptible to "optimism bias" in which only positive outcomes to problems are envisaged and that this "can be the subconscious result of overcoming previous difficulties" which may support a belief that they are relatively less prone to risk than others. It was considered that the Captain's "decision to depart without being certain the aircraft was free of ice" tended to suggest that the pilots in this case, especially the Captain, had been affected by "optimism bias" having "only foreseen a positive outcome" that any snow on the airframe would "blow off".
It was surmised by the Investigation that since the pilots had been trained for conditions like those encountered at Manchester, "a contributory factor in deciding not to de-ice may have been that their training was less effective than it might have been". It was found that their conversion courses - which had occurred at the same Approved Training Organisation (ATO) - may not have adequately covered airframe icing risk. It was also noted that despite a reliance placed by the Operator on pilot self-study of the ATR 'Cold Weather Operations' booklet prominently referenced in the annual winter operations "Flying Staff Memo", which both pilots had signed as having read, neither had done so. No recurrent classroom-based training was provided on icing risks and "the knowledge amassed" from the assumed self-study was "not tested other than through participation in recurrent simulator checks".
Finally, it was found that the manufacturer’s ATO included a scenario for "a badly de-iced tailplane" in its conversion course but that the ATO at which the two pilots involved had done their type conversion was not one of those using simulators which incorporated this profile - which in any case only represented one potential consequence of inadequate ground de-icing.
The formally-stated Conclusion of the Investigation was as follows:
"Ice contamination affected the tailplane and caused pitch control difficulty after the aircraft rotated, on departure. The evidence indicated that this would have been avoided if the aircraft had been de-iced/anti-iced and then inspected carefully before flight.
The crew considered, before parking, that de-icing was probably going to be unnecessary. It may then have become difficult for them to change their assessment because of ‘confirmation bias’, even though they were in freezing conditions and snow was falling. A contributory factor may have been the crew’s lack of experience operating aircraft in such conditions.
The commander optimistically thought that lying snow would blow off the aircraft before rotation; an assessment that was flawed and a possible reflection on the training the pilots had received for such winter conditions. The operator has recognised that recurrent winter training for pilots may have been over-reliant on self-study and has taken remedial action."
Safety Action taken as a result of the event and known to the Investigation included the following:
- The Approved Training Organisation which was responsible for the pilots' type conversion training is enhancing this by incorporating in it the aircraft manufacturer’s simulator profile for a "badly de-iced tailplane".
- The Aircraft Operator has enhanced its winter awareness training for pilots by introducing a computer-based training module which all pilots must now complete prior to each winter season and are then tested on their knowledge of its content.
- The Aircraft Operator has extended its own type conversion input by adding a day of ground training with a Training Captain before Line Training leading to release to unsupervised operations commences to ensure that time is spent discussing technical issues relating to line operations such as winter operations and de/anti icing in detail.
The Final Report of the Investigation was published on 9 February 2017. No Safety Recommendations were made.