airtraffic

Author Topic: EVA 28 Too Low at KSFO 23 July  (Read 27029 times)

Offline PHXCONXrunner

  • Newbie
  • *
  • Posts: 37
EVA 28 Too Low at KSFO 23 July
« on: July 25, 2013, 03:09:52 AM »
Heard about this from a crewmember on a flight that had just landed on 28R, apparently Eva Air flight 28 from TPE set off the low altitude alert and wound up going missed during a visual 28L approach... the same approach and runway as Asiana 214.

Pulled from the KSFO#2 feed, July 24 0330Z-0400Z, action occurs around 22 minutes into the clip.  Short clip with relevant transmissions attached.


Disaster averted?



Offline svoynick

  • Full Member
  • ***
  • Posts: 120
Re: EVA 28 Too Low at KSFO 23 July
« Reply #1 on: July 27, 2013, 03:47:34 AM »
Disaster averted?
I suppose in one way of thinking about it, every successful landing is a disaster averted.

I would love to know what happened there, though.  She didn't sound completely certain when she first keyed the mic.  I wonder if the situation in the cockpit was still developing at that time...

Offline flyflyfly

  • Full Member
  • ***
  • Posts: 143
Re: EVA 28 Too Low at KSFO 23 July
« Reply #2 on: July 27, 2013, 07:07:40 AM »
Aircraft was a 777-300ER. According to flightaware, at 1500ft they were decending with 2200ft/min. At 800ft they were still decending with 1920ft/min. At about 600ft they initiated the go-around. Seems they were still not stabilized and decending much too fast at low altitude.

http://flightaware.com/live/flight/EVA28/history/20130723/1530Z/RCTP/KSFO

Edit:
http://avherald.com/h?article=465e38db&opt=0
They were down to 600ft at a distance of 3.8nm from the runway...  :roll:
« Last Edit: July 27, 2013, 08:20:12 AM by flyflyfly »

Offline svoynick

  • Full Member
  • ***
  • Posts: 120
Re: EVA 28 Too Low at KSFO 23 July
« Reply #3 on: August 01, 2013, 03:38:57 AM »
Aircraft was a 777-300ER. According to flightaware, at 1500ft they were decending with 2200ft/min. At 800ft they were still decending with 1920ft/min. At about 600ft they initiated the go-around. Seems they were still not stabilized and decending much too fast at low altitude.

This inspired me to do a bit of a graphical display with the track log data off FlightAware.  This is a crunched-down screen-grab.  There's a larger version of the same graph here:  http://www.selmar.com/atc/


Offline svoynick

  • Full Member
  • ***
  • Posts: 120
Re: EVA 28 Too Low at KSFO 23 July
« Reply #4 on: August 01, 2013, 04:05:26 AM »
I've just doctored the graphic one last time, to include a red dotted-line extension of the flight path from the 4.8 nm point, imagining if the approach had continued beyond that point at the same descent angle.  Extrapolating the descent rate from the previous data points, they would have reached the water about 1 nm further on (at the 3.8 nm point), which would have been about 20 seconds after they passed the 4.8 nm point.  

I was going to let the graphic stand and shut up and let folks comment on it without trying to start anything myself, but now that I noticed this, I have to ask (rhetorically) did anyone in the crew have eyes outside the cockpit?  Was anyone situationally aware of what was going on?  

I wonder if the low altitude warning from ATC was their first notice that they were off glideslope.  Note that the descent angle was increasing over the 3 nm (or roughly 1 minute) prior to the 4.8 nm data point.  The radio call from the aircraft in response to ATC's warning came 20-ish seconds after the warning was issued, and didn't sound real confident and certain - as if they were still processing the warning.

I originally thought, OK, they got a little low and went around, no huge deal.  Now that I see this data graphically, it chills me a little bit, and I have to echo PHXCONNXrunner's question in his original post:  disaster averted?
« Last Edit: August 01, 2013, 04:17:42 AM by svoynick »

Offline sykocus

  • Sr. Member
  • ****
  • Posts: 348
Re: EVA 28 Too Low at KSFO 23 July
« Reply #5 on: August 01, 2013, 05:31:53 AM »
I wonder if the low altitude warning from ATC was their first notice that they were off glideslope.  Note that the descent angle was increasing over the 3 nm (or roughly 1 minute) prior to the 4.8 nm data point.  The radio call from the aircraft in response to ATC's warning came 20-ish seconds after the warning was issued, and didn't sound real confident and certain - as if they were still processing the warning.

The equipment at SFO definitely has newer then anywhere I've worked but where I do work the radar functions this way: It has database the terrain mapped out and it projects the position and altitude of an aircraft based on their descent rate and ground speed in reference to the terrain. Based on it's calculations it decides to activate the MSAW (minimum safe altitude warning). There is a box programed around the approach path that is used in the calculations, but predominately it's the projection of the aircraft's position base on speed and descent rate. I got an an MSAW on an aircraft descending out of FL190 over water. It was an F18 descending at close to 10k fpm and 450kts ground speed. The radar determined in x number of seconds if that trend continued that it would be in unsafe proximity to terrain.

On your chart you can see an increase in their descent rate between 7 and 6 miles (compared to the trend  from 10+ to 7 miles) this corresponded with a sharp increase in groundspeed.  I'd say this is what triggered the low altitude alert. The relationship to the glideslope was probably secondary in that it only meant they weren't very high and didn't have a lot of time to make up for a loss of altitude especially considering the ILS is still OTS at SFO as far as I know.
« Last Edit: August 01, 2013, 05:37:56 AM by sykocus »