While descending into Dublin during a flight from Newark, New Jersey, the crew of a United AirlinesBoeing 757-200 acted in a “nonstandard” manner in responding to an unreliable airspeed indication—an issue most likely caused by ice crystals in the right-side pitot probe.
The crew’s actions—two successive pitch-down maneuvers to counter what the first officer incorrectly thought was an aerodynamic stall—contributed to the minor injuries 13 passengers and four flight attendants suffered on Oct. 20, 2013. An excessive dive speed caused damage to external sections of the aircraft and one of three onboard hydraulic systems.
The incident raises a broader question about the safety of pitot probes (which measure airspeed) when flying through ice crystals, an issue connected with numerous recent accidents or incidents. While pitot probes for new air transport aircraft certified after January 2015 must be shown to work at conditions that would have been seen on the Newark–Dublin flight, aircraft certified before that date do not.
In its final report on the incident, published May 10, Ireland’s Air Accident Investigation Unit (AAIU) called for the FAA, as the certifying authority of the 757, to “study whether a safety deficiency exists in pitot-probe icing protection” for aircraft certified before January 2015. The recommendation is one of eight issued by the AAIU in the final report, along with two probable causes.
The first probable cause was a temporary block of the right main pitot tube due to ice-crystal icing, leading to an artificially low airspeed on the first officer’s display. The second probable cause was the crew’s “nonstandard response” to the low-airspeed reading. Contributing to the incident was the “startle effect” of that low airspeed on the first officer, just after the aircraft had experienced turbulence.
Investigators said the aircraft flew through, or close to, an area of “convective cloud activity,” with ice crystals predominantly in the “water phase,” as the aircraft descended through 25,000 ft. At the time, the 757 was 80 nm. southwest of Dublin, with the first officer at the controls. The seat belt sign was activated in the cabin.
After a bout with turbulence, the first officer told investigators the airspeed on his display decreased to about 90 kt., a speed that would likely denote an aerodynamic stall. He “immediately” pushed the control column forward and applied full power without disengaging the autopilot or autothrottle, or alerting the captain. The airspeed increased, but as soon as the first officer began raising the nose, the speed dropped again and he commenced a second pitch-down maneuver.
Data from the flight data recorder (FDR) showed the aircraft experienced a maximum negative acceleration of -0.36g during the push-overs and a maximum pull-up of 1.72g, forces that partly caused the injuries to unbelted passengers and flight attendants. The aircraft’s speed increased from 300 kt. during the initial descent to as high as 380 kt., with vertical speeds as high as 12,000 ft. per min. during the maneuvers.
Investigators said the high speed—30 kt. beyond the aircraft’s maximum operating speed—damaged the wing-to-body fairings. Excessive speed also dislodged and damaged the center hydraulic system servicing-bay door (on the underside of the aircraft) and the actual center hydraulic system, causing a loss of hydraulic fluid.
After the second pitch-down, the pilots determined that the captain’s airspeed indication was correct, and began leveling the aircraft.
The AAIU recommendations for United include reviewing weather radar operational guidance (Investigators said the convective weather that led to the turbulence should have been visible if the aircraft’s weather radar was appropriately adjusted.). The AAIU also urged United to review its “unreliable” airspeed training (Standard operating procedures call for checking the captain’s and first officer’s gauges before acting.).
United further needs to emphasize to pilots the importance of using standard callouts, especially during “non-normal” flight maneuvers, the AAIU stated. Evidence that the crew did not make standard callouts came only from postincident statements from the two pilots, rather than the cockpit voice recorder (CVR), which had been overwritten. The CVR has a 2-hr. duration.
According to a written report from the copilot, a “lady from Dublin” arrived in the cockpit after the uneventful landing and asked the crew to “pull” the CVR and FDR circuit breakers, a standard practice called for by the International Civil Aviation Organization to preserve information after incidents or accidents. But due to the commotion in the cockpit at that time, the first officer said he had missed shutting off the CVR breakers.
The confusion led to an AAIU recommendation for United to review its operations manual. Investigators said the manual could be interpreted as requiring the preservation of CVR data only for accidents, and only if requested by airline officials, excluding the pilots.