FAA, NTSB Investigate United 777 Post-Takeoff Close Call


The NTSB and FAA are investigating a Dec. 18, 2022, incident in which a United Boeing 777 appeared to depart controlled flight into a steep descent after departing Maui for San Francisco, according to the industry data site The Air Current. The aircraft climbed to 2200 feet after departure and then entered the descent, coming within about 800 feet of the Pacific Ocean.

According to data distilled from FlightRadar24, the 777 descended at up to 8600 FPM before recovering, resuming the climb and continuing to San Francisco without incident. Flight loads maxed at 2.7 G’s and the entire incident occurred over a 45-second period. Weather at Maui was reported as heavy rain and a broken layer at 900 feet with a 2000-foot overcast. The incident occurred on the same day that a Phoenix to Honolulu Hawaiian Airlines flight encountered severe turbulence, injuring 25, six seriously. That flight also landed without incident.

United did not report the incident to the NTSB, but the crew filed a routine safety report after it landed in San Francisco, presumably through the FAA’s voluntary Aviation Safety Action Program and/or NASA ASRS program. The FAA reviewed the incident and said it took “appropriate action.”

Other AVwebflash Articles


  1. Wow. I think the crew should have returned and handed the plane back to United saying, “we’ll be drinking the rest of the day”. 2.7 is a heavy pull in an airliner I would think.

    • Apparently I’m missing something in this and the AA runway incursion…when the crew or the aircraft malfunction to the point that it narrowly avoids a mishap, why does that seem like a good time to proceed as if nothing is wrong? Reminds me of the BA 747 that shutdown an engine on climbout from LAX to LHR.

      This UA, the AA incursion and the FedEx/SW/ATC incident imply to me that we’re rapidly draining the community “luck” bucket with an unhealthy amount of complacency. Too bad the avoidance of “inconvenience” will not permit a system wide safety standdown, if this trajectory continues we’re about to permanently “inconvenience” at least a planeload.

  2. “Wait… that wasn’t the autopilot button!”

    I bet they were surprised when they descended out of that layer and saw ocean in front of them.

  3. PF was heads-down maybe? Will be interesting to see what comes out of this, but I hope there’s an effort to maintain a just culture

  4. I just fly small planes but 800 feet seems like not a margin at all for something that weighs 1MM pounds plus. Holy smokes

  5. Sounds like they took off into a descending airmass that was descending faster than they could climb. Scary, but I think such a mass of air will flatten out and spread as it gets down near the surface, allowing you to fly out of it and climb. It might be an issue if there were obstructions in your path, but unlikely to push a plane into the ocean.

    The media is portraying this as a “nose dive” which is annoying, since it definitely wasn’t that.

  6. The maximum flight maneuvering load limits for a B777 is only 2.5 positive…. It seems to me that airplane should have undergone a detailed inspection prior to release for next flight.

  7. This is almost identical to the Qatar Airways B787 incident departing Doha, Qatar earlier this month. In that case, it appears that spatial disorientation on the part of the Pilot Flying (PF) was the cause of a 3,000 fpm rate of descent to 800′ above the water.

  8. This 777 UA incident incurred an 8600 fpm descent. Same day, same Wx system that affected the Hawaiian Airlines Flight from Phoenix to Honolulu which caused that flight to incur 25 injuries due to severe turbulence.
    It would seem that Wx definitely played a roll in this UA 777’s takeoff and climb out.

  9. Sounds like way too much conjecture here. Why not get all the facts before passing judgement. If weather was the cause, I completely understand why the crew would not go back. 2.7g was probably exciting but not an over ‘G’ if the flaps were up.

  10. Another commenter on his YT page, who is a FO on a 777, said it is possible that the crew did not set up the autopilot correctly and put in “0” feet on the altitude selector. Thus, at 2200 feet, they switched on the AP and the aircraft did what it was told and tried to descend to ZERO feet. Ooops…

    • So when you engage the autopilot in a 777 with a lower altitude set, it descends at 8,600 feet per minute? I don’t think so, Tim.

  11. I thought commercial pilots were to report such “down drafts” or “micro bursts” for the safety of other planes… maybe too shook up to think clearly. NTSB will get all the answers!

  12. Did you hear this as well: CNN reported that a passenger on the flight stated that before the descent, “…the plane climbed at “a concerning rate” for a few seconds. It felt like you were climbing to the top of a roller coaster.” No mention of deck angle in the climb, though.

  13. Weird air out there…, once I was body surfing on a black sand beach on the big island, at dawn, and with zero wind at the shore I saw TWO waterspouts form about 1/4 to 1/2 mile away. That still freaks me out years later. Between seeing that and the severe undertow at the site, I got out of there ASAP? Can you imagine the remainder of that flight, the mood of the passengers on board? Had to be surreal.

  14. Haleakala tops 10,000 feet only ~15 miles away and generates great sailplane conditions. Would make sense you sometimes will be encountering some mongo rotor activity. I’m sure we’ll get details eventually, maybe they mishandled it in some way but for now it’s pure speculation.

  15. I’ve heard a couple of versions – both similar – from United 777 pilots. It has also been reported in the media both pilots received additional training, which points to human error. It appears this error lead to an undesirable aircraft state. The investigation will reveal the “why”. My hunch, as with most incidents involving highly trained and experienced crews flying very capable aircraft, is it will involve multiple links in the error chain, to include human factors, recency of experience (the FO was reported to be on IOE), procedural errors, non-standard takeoff configuration (word is it was a flaps 20 takeoff), environmental factors (heavy rain/turbulence), and fatigue.

  16. Totally agree with J.D.G. Like most incidents or accidents, there is a chain that keeps growing until the holes in the cheese line up. Horrendous convective weather, inexperienced FO, who was probably the PF (Captain flies out, FO flies back.) If the FO was on IOE he may have flown both legs. 2.7 G in a heavy transport airplane is not to be sneezed at. Boeing builds ’em tough, but there are limits.

  17. It can be agreed that WX played a role here. Looks like pilots did keep nose down to avoid stall and recovered. On one hand the pilots did a fantastic job in this event, on the other the NTSB et al will determine what actions could have been taken to avoid a repetition. Interesting if ‘additional training’ included sim work to enact the same scenario. Wonder if they put out a PIREP? Human factors definitely part of this.

  18. Interesting scenario. Windshear, perhaps? I know we practiced this a lot in the sim (bizjets). Increasing headwind on climbout (from outflow) with corresponding increasing airspeed. This could explain the steep climb; followed by downdraft and tailwind and airspeed falling off a cliff. On another note, any of you veteran pilots experience this? An acquaintance was cruising along in the flight levels in a G-V somewhere in the south Pacific, when they experienced a sudden and unusually rapid increase in OAT. Said the airplane experienced a rapid drop in altitude.

  19. Considering that the flight continued to SFO, would the CVR in this aircraft have preserved the recording from the time of the incident?