Embraer ERJ170 incident near Atlanta, Georgia, November 7, 2019

The NTSB's narrative final · quoted from the NTSB record

The incident flight occurred during the third flight leg of the day. During the first flight leg, the engine indicating and crew alerting system (EICAS) provided the “PITCH TRIM SW 1 [switch 1] FAIL” advisory message. After an uneventful landing, maintenance personnel met the airplane to address the issue. The maintenance personnel reported that they partially removed the switch before deciding to defer the maintenance per the minimum equipment list. The maintenance personnel then reinstalled the switch but did not perform a functional test because the switch was a deferred maintenance item. They also placed a placard to indicate it was inoperative; thus, the captain’s trim switch, although not deactivated, was not supposed to be used during the subsequent flight legs. After takeoff on the incident flight, the captain instructed the first officer to declare an emergency because of runaway pitch trim. The captain stated that he conducted the single memory item on the pitch trim runaway emergency checklist in Republic Airways’ quick reference handbook. The memory item required him to push and hold the autopilot/trim disconnect button on his yoke to stop the runaway condition. The captain then instructed the first officer to push and hold the autopilot/trim disconnect button on the first officer’s yoke. The flight crewmembers reported that they did not notice any change in the runaway condition and continued to have difficulty controlling the airplane’s pitch. The captain stated that the flight crew banked the airplane to maintain control. The captain and the first officer stated that they needed to use both hands at the same time to counter the airplane’s nose-up pitch motion. As a result, neither was able to physically reference the quick reference handbook procedures to troubleshoot the problem. The captain selected the cutout button for the pitch trim system on his yoke to interrupt the nose-up trim condition and instructed the first officer to use the trim switch on the first officer’s yoke to pitch the airplane down to a normal attitude. The first officer, now in control of the pitch trim using his pitch trim switch, was able to trim and regain control of the airplane. About three minutes later the captain took back control of the airplane. For about 2.5 minutes afterwards the FDR recorded multiple pitch trim up commands from the captain’s pitch trim switch, and the airplane again went into a mis-trimmed condition. The first officer took back control of the airplane and remained in control until the end of the flight. During this time the FDR showed no pitch trim commands originating from the captain’s trim switch although there were multiple up and down pitch trim commands originating from the first officer’s trim switch. The airplane was controllable and in a trimmed condition and able to return to the departure airport and land uneventfully. Postincident examination of the captain’s pitch trim control switch identified imprint marks on the back of the switch, indicating that, at some point, the switch was installed in an inverted position. The most likely opportunity for this to occur would have been during maintenance after the first flight leg when the pitch trim switch was partially removed then reinstalled since the decision was made to defer maintenance. Since maintenance was deferred, the switch was not functionally tested. If a functional test had been completed, maintenance personnel would likely have identified the reversed switch position. On April 13, 2020, Embraer and the FAA revised their master minimum equipment lists to remove the yoke pitch trim switches from the list of deferrable items, effectively requiring pitch trim switches to be functionally tested and operational before flight. Embraer issued Service Bulletins (SBs) 170-27-0051, 190-27-0039, and 190LIN-27-0019 in February 2015 after reports about inverted pitch trim switches. The SBs recommended the installation of a support in the control yoke within the next 7,500 flight hours or 36 months after the SB’s issuance (whichever occurred first). However, the SB 170-27-0051 had not been accomplished on the incident airplane. If the SB had been accomplished, the recommended support in the control yoke would have prevented the faulty installation. In January 2020, the National Transportation Safety Board issued Safety Recommendations A-20-5 and -9 to the National Civil Aviation Agency of Brazil (ANAC) and the Federal Aviation Administration (FAA), respectively, to mandate incorporation of the SB. On May 26, 2020, ANAC published Brazilian Airworthiness Directive 2020-05-0051, which mandated compliance with the SB; as a result, Safety Recommendation A-20-5 was classified “Closed—Acceptable Action” on September 2, 2020. Regarding Safety Recommendation A209 (classified “Closed—Acceptable Action” on October 21, 2022), the FAA stated that it issued Airworthiness Directive 2020-25-08, which became effective on January 21, 2021, and mandated operator compliance with the ANAC’s airworthiness directive requirements. The installation of the pitch trim switch in reverse and the captain’s use of the deferred trim switch resulted in the pitch excursion; when the captain attempted to trim the airplane nose down, the airplane responded with nose-up inputs. Thus, to effectively control the reversed pitch trim system, the captain would have had to recognize that his trim-down inputs were causing trim-up commands. However, the captain would not have been aware that the switch had been reinstalled incorrectly and was commanding trim opposite of his inputs. The captain was aware that the pitch trim switch was not to be used because it had been deferred per the minimum equipment list. However, highly practiced behavior, such as making inputs to keep an airplane in trim, can result in typical motor actions being made automatically despite higherlevel knowledge indicating that such action might not be appropriate. During the captain’s interview he stated that it was second nature to use the trim switch on the yoke. Systems with deferred maintenance are typically deactivated to prevent any errant command input from a faulty control. However, there was no requirement, nor was there a published procedure, to disconnect or disarm a faulty pitch trim switch. Placards for inoperative equipment are typically collocated with the related control to remind pilots that the equipment should not be used. However, the pitch trim switch position on the yoke did not allow space for a placard on or near the switch. The captain’s use of an inoperative and a placarded flight control was inappropriate and caused the airplane to pitch up when the captain was trying to trim down, resulting in flight crew difficulties in controlling the pitch of the airplane. In addition, the captain thought that the airplane had a runaway trim condition, but, once he transferred control of the airplane to the first officer, the mistrim condition stopped, and the crew was able to regain control of the airplane. Thus, the captain’s trim inputs (and not a runaway trim) caused the pitch trim anomaly. In addition, FDR data show that the captain’s pitch trim only inputted nose up trim and there were no nose down trim commands. Maintenance records showed a history of “PITCH TRIM SW 1 FAIL” EICAS messages on the airplane between August 4, 2019, and the incident date, resulting in five replacements of the captain’s pitch trim control switch and two replacements of the horizontal stabilizer actuator control electronics. Also, during the first flight leg on the day of the incident, the FDR recorded multiple “TRIM FAIL” occurrences related to the captain’s pitch trim control switch. The “TRIM FAIL” occurrences and the “PITCH TRIM SW 1 FAIL” EICAS messages were caused by a short to ground of a single circuit to the captain’s pitch trim switch. The short was created by contact with the incorrectly tucked pigtail from the safety wire retaining the forward mechanical stop bolt for the captain’s control column. Although both wires enabling the captain’s pitch trim switch trim-up command were damaged and chafed and were capable of allowing continuity to the untucked safety wire pigtail (which would result in a short to ground condition), it is unlikely that the pigtail would have made continuity to both wires at the same time. Both wires would need to short to ground simultaneously to allow the stabilizer trim to activate. If both wires shorted to ground, the horizontal stabilizer control surface movement would automatically stop after three seconds, and if the condition persisted the aural alert “TRIM” would annunciate. This alert was not reported by the crew. Finally, the maintenance history for the captain’s pitch trim switch had many reports of a “Pitch Trim Switch 1 Fail” EICAS message, but no reports of a trim runaway. FDR showed that the captain's pitch trim switch only provided trim up commands when there was force being applied to the captain's control column, suggest that the captain was in control of the airplane during those times.

The complete narrative as the NTSB published it. The NTSB's docket holds the report as a PDF and any photographs, statements and other documents from the investigation.