United B752 near Dublin on Oct 20th 2013, severe turbulence and pilot control inputs cause injuries
Last Update: May 10, 2016 / 16:28:47 GMT/Zulu time
Incident Facts
Date of incident
Oct 20, 2013
Classification
Accident
Airline
United
Flight number
UA-23
Departure
Newark, United States
Destination
Dublin, Ireland
Aircraft Registration
N41140
Aircraft Type
Boeing 757-200
ICAO Type Designator
B752
The return flight UA-22 was cancelled.
On May 10th 2016 the Irish AAIU released their final report reporting 13 passengers and 4 crew received minor injuries in the encounter. The AAIU concluded the probable causes of the occurrence were:
Probable Cause(s)
- Temporary blockage of the right main pitot probe due to ice crystal icing, leading to an unreliable airspeed indication.
- Non-standard Flight Crew response to a low airspeed indication.
Contributory Cause(s)
Startle effect due to a sudden unexpected indication of low airspeed, following an encounter with turbulence.
The AAIU reported that following an uneventful flight so far the aircraft was descending through FL250 when the aircraft began to encounter turbulent atmospheric conditions, St. Elmo's fire was present. Descending through FL235 the intensity of both turbulence and St. Elmo's fire abruptedly increased and there were sounds of precipitation. After the turbulence eased the first officer (43, ATPL, 12,010 hours total, 5,384 hours on type), pilot flying, noticed that the airspeed had dropped to 90 KIAS and believed the aircraft was about to stall, therefore pushed the control column forward and applied maximum thrust without disconnecting autopilot or autothrust. Following these actions the airspeed returned into the normal range, but as soon as he began to raise the nose and reduce thrust the speed dropped again into a stall, the first officer therefore commenced a second pitch down.
Following the second pitch the flight crew assessed that the captain's (59, ATPL, 11,606 hours total, 8,769 hours on type), pilot monitoring, airspeed readings were normal and the first officer's indications were unreliable through a cross check with the stand by instruments matching the readings of the captain's instruments but disagreeing with the indications at the first officer's side. The commander took control of the aircraft, stabilized the aircraft. Some time later, when the first officer's airspeed indication returned to normal, the captain handed control back to the first officer.
At about that time the crew received an EICAS indication that the center hydraulic system had lost pressure. The crew worked the related checklist. Subsequently the crew received information from flight attendants that two flight attendants and a number of passengers had received minor impact injuries in the turbulence encounter. The captain advised ATC that they had suffered a turbulence encounter, a number of passengers had been injured and medical assistance was needed upon arrival.
The AAIU reported that immediately after landing 8 passengers and 2 flight attendants reported being injured. Later 5 more passengers reported minor injuries, a written statement received from flight attendants subsequently reported 4 unrestrained flight attendants received minor injuries.
The aircraft sustained damage as listed:
- The door frame of the centre hydraulic system servicing bay and adjacent structure (Figures No. 1 and No. 2).
- The centre hydraulic system servicing bay door and hold-open rod.
- The electrical connector on the filter module pressure transmitter.
- The reservoir drain valve, quantity transmitter and associated electrical connector.
- The hydraulic service panel light.
- The wiring loom P-Clip at the rear of the hydraulic service bay.
- Some wing to body fairings had popped open.
- A ceiling panel in the rear passenger cabin had suffered impact damaged.
- There were coffee and tea stains on various ceiling panels throughout the cabin.
The AAIU summarized the captain's testimony: "In his initial interview and first written statement, the Commander made no mention that he was aware that the Co-Pilot was of the belief that the aircraft was about to stall, or that this was the reason that the Co-Pilot had pitched the aircraft down and applied engine thrust. The Commander’s first written statement provided very little detail about the event. In a later statement provided some time after the event, the Commander stated that during the remaining descent, he learned from the Co-Pilot that immediately after encountering the turbulence, the Co-Pilot’s instruments showed a very low airspeed and a potential stall, which prompted the Co-Pilot to carry out the pitch down manoeuvre. The Commander made no mention in any of his reports, of any communications between himself and the Co-Pilot regarding the manoeuvres performed by the Co-Pilot during the event, or of activation of the Stick Shaker."
The AAIU summarized the testimony of the first officer: "During initial interview, the Co-Pilot, who was the PF at the time, said that when the turbulence was encountered, the aircraft experienced a “series of very rapid major jolts”. He said that following this he noticed that his airspeed was showing low, “well into the red warning” and “probably somewhere around 90 knots” so he “initiated a stall recovery procedure”. He stated that following this, within just a few moments, the airspeed went back up into the normal range, so he “pulled the power back” and started to level off but then the aircraft went “back into a stall, the indication of a stall”, leading him to repeat the pitch down manoeuvre. The Co-Pilot stated that at this point, it was realised that the Commander’s instruments and the standby instruments showed that the airspeed was “well above the stall” even though the airspeed as indicated on his instruments was “below [the stall] and the stall warning was still on”. Regarding this “stall warning”, a subsequent report received from the Co-Pilot, stated that it was later realised that the warning heard was an overspeed warning and not a stall warning as initially thought. Control was then passed to the Commander."
The AAIU summarized the lead flight attendant's testimony: "The In-Flight Service Manager (ISM) reported that the FAs were in the process of stowing galley equipment and checking the aircraft cabin when the event occurred. The ISM and another FA managed to secure themselves into their crew-seats adjacent to the main passenger door at the front of the aircraft (door 1 left), but four other FAs “hit the ceiling”. The ISM stated that “the drop occurred twice”."
Another flight attendant told the AAIU: "Another FA confirmed that the fasten seatbelt sign was illuminated when moderate turbulence was encountered. The FA said that he made a Public Address announcement, urging passengers to return to their seats and fasten their seatbelts. He stated that seconds later “the turbulence went from moderate to severe” and that there were “two abrupt drops that threw everything around that wasn’t latched down”."
Weather Services had issued a significant weather forecast charts indicating scattered to broken thunderstorms and light turbulence.
The actual weather conditions at the time of the occurrence were: "Low pressure with a centre of 988 hPa18 at the mouth of the Shannon estuary maintained an unstable south easterly flow across the country. A well-defined showery trough was moving through the region at the time of the incident with embedded cumulonimbus cloud present. Archived radar echoes indicate some heavier embedded showers were also present”. It was also stated that: “The radar and satellite evidence clearly shows that there was a high possibility of significant (moderate to severe) turbulence associated with cumulonimbus cloud that was present in the area at the time the incident occurred.”"
The AAIU stated, that the cockpit voice recorder had not been preserved (although the FDR had been preserved) and quoted the first officer's statement:
A lady from Dublin (I assume she was an agent) came to the cockpit and asked us to pull the CVR and FDR circuit breakers. As there was other personnel in the cockpit speaking to the captain who was on the phone with dispatch at the time, I proceeded to locate and pull the breakers. At this time there were at least four persons in the cramped cockpit and with difficulty I tried to locate the breakers to be pulled.
It was late and just getting light and we were tired and very busy attempting to manage all of the duties placed upon us at that moment with coordinating with the company, the medical personnel, the maintenance staff, the flight attendants and the Irish Safety Authorities. I finally located the FDR breakers but apparently I had completely missed the CVR breakers. I advised the Agent and the Captain that I had pulled the breakers thinking that I had pulled all of the breakers asked of me. I didn’t even realize this mistake until the next day when the agent present on our outgoing flight mentioned that the CVR tapes were not usable due to the breakers not being pulled. At that time I searched the circuit breaker panel and found the CVR breakers and realized that I missed them the night before. This was not done intentionally.
The flight data recorder revealed that the aircraft was in a steady descent at 300 KIAS and about 1500 fpm rate of descent when turbulence began to set in and then increased followed by first control nose down inputs, that caused the pitch to reach 9.1 degrees nose down, engines were accelerated to 100% N1. As the airspeed increased the engines were pulled back to idle, the pitch was increased to 8.2 degrees nose down, the airspeed continued to increase. A second nose down control input caused the nose to drop to 16.2 degrees nose down at 340 KIAS. During the nose down control inputs vertical accelerated of -0.16G, -0.36G and -0.22G were encountered which changed rapidly to +1.62G over a period of 2.4 seconds. The airspeed exceeded Vmo (350 KIAS) and reached 380 KIAS until the attitude of the aircraft was recovered, a maximum rate of descent of about 12000 fpm had been reached during the descent.
The AAIU reported: "The FAA has defined ‘Startle’ as “uncontrollable, automatic muscle reflex, raised heart rate, blood pressure, etc., elicited by exposure to a sudden, intense event that violates a pilot’s expectations”. It also defines ‘Surprise’ as “An unexpected event that violates a pilot’s expectations and can affect the mental processes used to respond to the event”. They go on to state, “Because upsets that occur in normal flight operations are unplanned and inadvertent, pilots may be startled or surprised, adversely impacting recognition or recovery”."
The AAIU analysed:
Shortly before the event, the aircraft was in a stable descent towards EIDW at 300 kts and descending at approximately 1,500 fpm. The Seat Belt sign was reported as being on. The aircraft then entered IMC and encountered turbulent atmospheric conditions, which the Co-Pilot described as “severe”. There was a marked increase in the intensity of St. Elmo’s Fire. The Co-Pilot, who was PF, then noticed that his ASI indicated a low airspeed. Fearing that the aircraft was in danger of stalling, he initiated what he termed “a high altitude stall recovery manoeuvre”. During this action, the Co-pilot stated that his indicated airspeed started to increase, but then dropped again. As a result the Co-pilot repeated the actions.
During these manoeuvres the actual airspeed of the aircraft increased, eventually reaching 380 kts, 30 kts above the Vmo. Shortly after the second pitch down manoeuvre, the overspeed audio warning activated. This was initially considered and reported by the Co-Pilot to be a “stall warning” and likely compounded his belief that he was responding to a stalling condition.
Following the second pitch down manoeuvre, there was a realisation that the Commander’s airspeed agreed with the Standby ASI, and control of the aircraft was transferred to the Commander.
During these manoeuvres, the aircraft experienced negative ‘g’ loads, followed rapidly by positive ‘g’ loads. During the negative ‘g’ manoeuvres some unrestrained persons, and items not secured within the aircraft, impacted parts of the aircraft interior such as the ceiling. The sudden return to positive ‘g’ conditions resulted in persons and objects falling back down. It is likely that the injuries sustained by the passengers and cabin crew occurred during these rapid changes in ‘g’ load.
FDR data shows that the hydraulic quantity decreased sometime after 19,685 seconds. At this time the aircraft was undergoing the second pitch down manoeuvre, after which the aircraft underwent a rapid change in normal load factor from -0.22 g to + 1.62 g, as the airspeed increased through the Vmo of 350 kts. The computed airspeed remained above 350 kts for the next 50 seconds, peaking at 380 kts. It was likely that the centre hydraulic access panel became detached during this period, causing the damage that resulted in loss of centre hydraulic system fluid and pressure.
The normal acceleration forces recorded on the FDR during the event did not exceed the certified limit load factor envelope of the B757 aircraft.
The AAIU analysed: "The Flight Crew stated that they also saw St. Elmo’s Fire after the aircraft had entered IMC, and the Co-Pilot reported that the St. Elmo’s Fire intensity increased just prior to the event. St. Elmo’s Fire is an indicator that the atmosphere in which an aircraft is flying is highly electrostatically charged and is often associated with nearby convective cloud such as cumulonimbus or thunderstorm activity. It may also be an indication of the presence of ice crystals. This, and the other meteorological evidence, is consistent with the likely presence of ice crystals in the area through which the subject aircraft tracked."
The AAIU analysed: "Analysis of the atmospheric conditions through which the aircraft flew indicated the presence of ice crystals. This is supported by flight crew statements of increasing St. Elmo’s Fire. Similar to other events outlined in this Report (Section 1.12 and Appendix C), the airspeed discrepancy on the subject aircraft was most likely due to a temporary blockage of a pitot probe. The Aircraft Manufacturer stated that: “It is suspected that a pitot probe that supplied pressure information to the First officer’s [Co-Pilot’s+ ADC had a temporary blockage which caused the inconsistent wind data and most likely an airspeed disagree indication between Captain’s [Commander’s] and First Officer’s instruments”. Following the event, the Operator performed a check on the Pitot-Static, Air Data, and Probe Heat Systems. No faults were found. The Investigation is therefore of the opinion that the most likely cause of the temporary blockage of the Co-Pilot’s Pitot probe was ice crystal icing. ... The B757 aircraft type was certified in 1982, under the icing certification criteria requirements in existence at that time. Those criteria did not account for ice crystal icing. This event shows that the B757 aircraft pitot probes may be susceptible to ice crystal icing under certain conditions."
The AAIU analysed: "As the turbulence decreased, the aircraft’s rate of descent, airspeed, pitch and roll, had not altered significantly. Thus, the change in airspeed indicated on the Co-Pilot’s ASI should have been recognised as an unreliable airspeed by the Flight Crew. If this had been recognised, the first action required by the Airspeed Unreliable checklist in the Operator’s QRH, is to “Check the Pitch and Thrust”. The checklist goes on to tell flight crew, if pitch and thrust are not normal for the phase of flight, to disconnect the autopilot and autothrottle. These actions were not carried out by the Flight Crew during the initial stage of the event. Furthermore, when it was recognised that the Co-Pilot’s indicated airspeed was unreliable, these actions were not carried out."
With respect to the IAS disagree warning the AAIU analysed: "The IAS DISAGREE and its associated MASTER CAUTION alert does not operate unless there is a discrepancy of more than 5 kts between the Co-Pilot’s and the Commander’s airspeed for more than 5 seconds. Analysis of the FDR data suggests that the discrepancy between the two airspeed sources started to develop approximately 5 seconds before the Co-Pilot pushed the control column forward for the first time. Therefore, when the control column was moved forward for the first time, it is likely that the IAS DISAGREE and the MASTER CAUTION alert were not ‘on’. It is possible that when the IAS DISAGREE alert was eventually displayed, it was not noticed/recognised due to the Co-Pilot’s belief that the aircraft was stalling, the possible effects of startle factor, and his focus was on trying to recover the aircraft. At no stage during their interviews, or subsequent statements, did the Pilots mention seeing or commenting on, the presence of an IAS DISAGREE alert (or an associated MASTER CAUTION). The Investigation found that in two previous events involving IAS DISAGREE on a similar aircraft (B767) investigated by the TSB Canada (see Section 1.12 and Appendix C, item 7), the flight crew did not notice the EICAS alert. The Investigation is therefore of the opinion that the EICAS IAS DISAGREE alert may not provide a sufficient level of warning to flight crews in the event of an Unreliable Airspeed occurrence."
The AAIU analysed with respect to human factors: "The Co-Pilot believed that an aircraft upset had occurred, that this had resulted in a loss of airspeed and that the aircraft was about to stall. This perception may have been compounded by the sensation of the turbulence, the sight of a very low indicated airspeed, and the advancing of the power levers by the Autothrottle. In contrast, the Commander referred to the aircraft experiencing “two drops” that he initially thought were caused by windshear. This perception may have been compounded by the negative ‘g’ forces he experienced and the increase in airspeed he saw. As a result of the two different perceptions of the same event, it took approximately 20 seconds before appropriate corrective action commenced, and approximately 30 seconds before aircraft pitch attitude was returned to zero degrees."
The AAIU analysed: "During the second pitch down manoeuvre the overspeed warning horn sounded, which the Co-Pilot initially believed was a stall warning as this corresponded with the low airspeed that he was seeing on his instruments. This likely further reinforced his belief that the aircraft was stalling, and as a result delayed the recovery of the excessive nose-down attitude of the aircraft, which resulted in the Vmo exceedance."
The AAIU analysed: "Therefore, it is possible that the Co-Pilot experienced a startle response due to the sudden onset of the erroneous airspeed indications on his flight instruments, coupled with the turbulent conditions being encountered and the fear that an aircraft upset had taken place. This startle response likely elicited the immediate, reflex actions performed by the Co-Pilot. Likewise, the Commander may have been startled by the ‘g’ forces he was experiencing as the Co-Pilot manoeuvred the aircraft in response to what he (the Co-Pilot) thought was a stalled condition. A startle response by the Commander may have resulted in him not perceiving the IAS DISAGREE or its associated MASTER CAUTION and delayed his analysis and crosscheck of the instruments and the action required to return the aircraft to normal flight conditions."
The AAIU analysed:
Notwithstanding this, the Co-Pilot was of the belief that the aircraft was stalling.
The Stall Recovery actions are not outlined in the Operator’s QRH, but stall recovery actions are outlined in their Supplementary Procedures Manual, where it states that initial recovery is accomplished by “disconnecting the autopilot and autothrottles”. The Aircraft Manufacturer’s stall recovery procedures (see Appendix B) require the control column to be held firmly, and to disconnect the autopilot and autothrottle.
These actions were not performed, as shown by FDR data, which recorded that the autopilot and autothrottle remained engaged throughout the event.
Furthermore, if the Co-Pilot had carried out the first actions of the Aircraft Manufacturer’s Stall Recovery Checklist, which are to hold the control column firmly, and disconnect the autopilot and autothrottle, it is likely he would have sensed the control loads on the control column, especially as he applied forward pressure. The control loads and pitch rate become more pronounced as the IAS of the aircraft increases. This would have given some tactile feedback to the Co-Pilot in the initial stages of the manoeuvre and assisted in his analysis of what was occurring.
The AAIU analysed further with respect to human factors:
Interviews conducted with the Flight Crew and further statements provided by them after the occurrence, suggest that the level of inter-crew communication was such that neither pilot was able to fully assess what had happened to the aircraft, or what action the other pilot was taking. There was no indication that an initial cross-check took place in the cockpit with regard to the low airspeed condition that was indicated on the Co-Pilot’s instruments, or the Co-Pilot’s belief that the aircraft was in danger of stalling prior to the initial pitch down manoeuvre. In his initial interview and first written report, the Commander made no mention of any communication from the Co-Pilot during the event.
Consequently, there appears to have been no shared picture or shared understanding by the Flight Crew members of what had initiated the event, or what actions were being performed to address it. Following the initial pitch down manoeuvre by the Co-Pilot, in response to a perceived stall condition, it took approximately 30 seconds before the aircraft pitch was restored to approximately zero degrees.
With respect to injuries the AAIU analysed: "As part of the analysis of the ‘g’ forces experienced during the event, the Investigation sought to determine if they were due to control column input alone. The Aircraft Manufacturer attempted to separate the effects of control column input and the weather conditions, through the use of an engineering simulation. It was concluded that “both the column input and wind profile contributed to the upset”."
With respect to the loss of center hydraulic pressure the AAIU analysed: "Regarding the loss of quantity/pressure in the centre hydraulic system, it is likely that during the event, the hydraulic service door became dislodged and opened, causing the hold open rod to disconnect and strike the dump valve key. The degree of damage to the dump valve key was such that it led to a loss of hydraulic fluid from the centre hydraulic system, which was signalled in the cockpit. As the aircraft continued its descent into EIDW, the hydraulic service door was subjected to further airflow damage resulting in the door bending and breaking in half, but remaining attached to the aircraft. The B757 aircraft has three hydraulic systems. In this case, the loss of the centre system had no bearing on the continued safe operation of the aircraft."
With respect to lack of CVR data the AAIU analysed: "The non-availability of the CVR denied the Investigation the opportunity to fully understand the cockpit environment at the time of the occurrence and to accurately account for the communications, discussions and actions of the Flight Crew as they responded to the unfolding events. As stated in Section 1.9.2 of this report, the Investigation notes and accepts the Co-Pilot’s statement that he thought the circuit breaker for the CVR had been pulled."
Aircraft Registration Data
Incident Facts
Date of incident
Oct 20, 2013
Classification
Accident
Airline
United
Flight number
UA-23
Departure
Newark, United States
Destination
Dublin, Ireland
Aircraft Registration
N41140
Aircraft Type
Boeing 757-200
ICAO Type Designator
B752
This article is published under license from Avherald.com. © of text by Avherald.com.
Article source
You can read 2 more free articles without a subscription.
Subscribe now and continue reading without any limits!
Read unlimited articles and receive our daily update briefing. Gain better insights into what is happening in commercial aviation safety.
Send tip
Support AeroInside by sending a small tip amount.
Related articles
Fleet type match
United B752 near Gander on Oct 7th 2023, engine vibrations
A United Boeing 757-200, registration N41140 performing flight UA-317 from Newark,NJ (USA) to Dublin (Ireland), was enroute at FL370 about 210nm…
Fleet type match
United B752 at Belfast on Oct 24th 2016, unsafe nose gear indication
A United Boeing 757-200, registration N41140 performing flight UA-76 from Belfast Aldergrove,NI (UK) to Newark,NJ (USA) with 121 passengers and 9…
Fleet type match
United B763 near Providence on Apr 22nd 2024, suspected fuel leak
A United Boeing 767-300, registration N644UA performing flight UA-108 from Washington Dulles,DC (USA) to Munich (Germany), was enroute at FL330 about…
Fleet type match
United B789 near San Francisco on Apr 19th 2024, cargo smoke indication
A United Boeing 787-9, registration N15969 performing flight UA-801 from Los Angeles,CA (USA) to Hong Kong (China), was enroute at FL340 about 30nm…
Fleet type match
United B753 enroute on Apr 10th 2024, Rockies' coach in captain's seat
A United Boeing 757-300, registration N73860 performing charter flight UA-3779 from Denver,CO (USA) to Toronto,ON (Canada) with the Colorado Rockies…
Fleet type match
United B738 at Las Vegas on Apr 10th 2024, dropped panel in flight
A United Boeing 737-800, registration N24224 performing flight UA-1960 from Los Angeles,CA to Las Vegas,NV (USA), concluded a seemingly uneventful…
Fleet type match
United B772 at Rome on Apr 5th 2024, engine compressor stall
A United Boeing 777-200, registration N209UA performing flight UA-885 from Rome Fiumicino (Italy) to Washington Dulles,DC (USA), was climbing through…
Newest articles
Red Air MD82 at Miami on Jun 21st 2022, runway excursion causes gear collapse on landing
A Red Air Dominicana McDonnell Douglas MD-82, registration HI1064 performing flight L5-203 from Santo Domingo (Dominican Republic) to Miami,FL (USA)…
Fleet type match
Southwest B737 near Dallas on Apr 25th 2024, engine shut down in flight
A Southwest Airlines Boeing 737-700, registration N793SA performing flight WN-4195 from Denver,CO to San Antonio,TX (USA), was enroute at FL410 about…
Subscribe today
Are you researching aviation incidents? Get access to AeroInside Insights, unlimited read access and receive the daily newsletter.
Pick your plan and subscribePartner
A new way to document and demonstrate airworthiness compliance and aircraft value. Find out more.
ELITE Simulation Solutions is a leading global provider of Flight Simulation Training Devices, IFR training software as well as flight controls and related services. Find out more.
Your regulation partner, specialists in aviation safety and compliance; providing training, auditing, and consultancy services. Find out more.
AeroInside Blog
Popular aircraft
Airbus A320Boeing 737-800
Boeing 737-800 MAX
Popular airlines
American AirlinesUnited
Delta
Air Canada
Lufthansa
British Airways