385 N666DS Citation S550 Crash in San Diego – Flying below Minimums
17:57 Share

This morning in the pre-dawn hours, a Citation jet crashed into a San Diego neighborhood while attempting to fly an RNAV approach. But it was flying below the GPS glide path, and it leveled off hundreds of feet below minimums. If you've ever flown an approach late at night, or wondered when to choose an ILS over an RNAV approach, this is a crash you need to understand. Hello again and welcome to Aviation News Talk, where we talk general aviation.

My name is Max Trescott. I've been flying for 50 years. I'm the author of several books, and I'm the 2008 National Flight Instructor of the Year. And my mission is to help you become the safest possible pilot.

Last week in episode 383, we talked about why autopilots sometimes fail to capture the glide path during RNAV GPS approaches. And two days ago, we played the first episode of my new podcast, NTSB News Talk. So if you didn't hear either of those episodes, you may want to check them out at aviationnewstalk.com slash 383 and 384.

And if you're new to this show, welcome. So glad that you found us. And now so you stick around, if you would take a moment right now and whatever app that you're using, touch either the subscribe key or in the Spotify or Apple podcast app, the follow key so that next week's episode is downloaded for free. And if you've been listening to us for a while and you feel like, yeah, you're getting some value out of this podcast, if you would take a moment, stop what you're doing right now and sign up to become a member to support the show.

It's easy to do that. Just go out on the web to aviationnewstalk.com slash support and choose one of four different ways to support the show financially. And when you do, I'll read your name on the show. Now, I had planned to talk about something totally different this week, but when major accidents occur, we talk about them first.

At about 3.46 a.m. on Thursday morning, November 666 Delta Sierra, a Citation S-550 crashed in San Diego while on an instrument approach to Montgomery Field, KMYF. After the crash, multiple homes and vehicles caught fire and an evacuation center was established at a nearby elementary school. First, my condolences to the family and friends of the pilot and any passengers that may have been on board.

as well as to the people who may have lost their homes. As I've said in the past, we talk about accidents so that we can learn from them and hopefully prevent future accidents. So far, there's no information on how many people were on board, though there is one confirmed fatality. The Citation S-550 can be flown by a single pilot, so it's possible that the pilot was the only person on board.

From the ADS-B data, it's fairly clear the instrument approach into Montgomery Field was flown incorrectly. We'll get to that in a moment, but first let me talk about the flight itself.

The flight departed Teterboro, New Jersey at 11.15 p.m. ET and stopped for fuel in Wichita, Kansas. The approach to land in Kansas was a little unusual. The pilot didn't fly an instrument approach, perhaps to save time. It overflew the airport before making a large 360-degree turn to land at the airport. The pilot can be heard attempting to turn on the lights five different times between 1.42.10 and 1.43.46 a.m. CT,

This suggests the pilot didn't see the airport until he was flying right next to it. Now, there is nothing wrong with how the pilot flew that approach to landing, though best practice at night would be to fly an instrument approach. The plane was on the ground for about 47 minutes to refuel. It then departed for Montgomery Field in San Diego.

The Montgomery Field AWOS was out of service at the time, so we don't know what the weather conditions were at the field, though they likely weren't good. I've listened to the ATC audio for SoCal Approach, and while we can't hear the pilot side of the conversation, we can hear the controller reading the weather for different airports. And this comes from LiveATC.net. Station 666, this is SoCal Approach. The Montgomery ASOS is out of service. Which approach would you like?

Neste is the IAF for both the ILS and the RNAV GPS28Rite approach. Later we hear the controller clear the pilot for the RNAV approach, so that's apparently the one the pilot requested. The pilot apparently asked about weather at other airports, and here's what the controller said. Two minutes later, the controller provided this weather.

Number six, Delta Sierra, I've got the Miramar automated weather for 0955 Zulu with wind calm, visibility one-half, and indefinite ceiling 200. Six, Delta Sierra, I drink, defend, and maintain 5,500. A minute and a half later, the pilot apparently inquired about the weather at a different airport. Here's what the controller said. Let me take a look.

The controller apparently thought the pilot may have been a little too high, so we ask about whether he could get down in time. Okay, you're going to make the scent okay from there? You want me to turn you out to the south?

6-0 to Sierra Roger, 5 miles from Nespe, cross outer above 3800, cleared RNAV runway 2, right approach. The pilot apparently said that he could get down as the controller didn't issue a vector and the flight was cleared by SoCal approach for the RNAV GPS 28R approach. The barometric pressure at San Diego area airports was close to 29.92, so the ADS-B altitudes displayed are correct within 10 or 20 feet.

The aircraft flew the beginning of the approach correctly, crossing Nesty at 3,800 feet as instructed, and then descending and leveling off briefly at 2,500 feet before crossing Penny, the final approach fix. And here's a problem the pilot would have encountered. You may recall that last week in episode 383, I explained that an RNAV glide path diamond turns magenta while crossing the fix immediately before the FAF.

Another way of saying that is that it turns magenta when the final approach fix becomes the active waypoint. And if you cross the fix before the FAF at the designated altitude, you will be on the glide path. However, the designated altitude for Nesty is 3,600 feet, and the controller had the pilot cross Nesty at 3,800 feet. Thus, the jet was probably above the glide path, and the autopilot would not have coupled.

So the pilot would have had to use the autopilot's vertical speed mode or hand fly the airplane to get down below the glide path. So he would have had to descend to 2,500 feet, the glide path intercept altitude, before reaching the FAAF Penny. And it appears that he did just that, that he got down to 2,500 feet and leveled off at Penny. At that point, if the autopilot were in the approach mode, it should have coupled to the glide path.

but apparently it didn't as ultimately the pilot flew a track that went below the RNAV glide path for this approach. The aircraft crossed Neste at 200 knots and about a minute later it was still at that speed as it leveled off at 2,500 feet as it was approaching the FAF. During that short level off at 2,500 feet, it was able to slow at 175 knots as it crossed the FAF and started its final ascent on the glide path.

During the descent along the glide path, the aircraft slowed from 175 knots to 124 knots just before it crashed.

I have only a little citation time, but to me that seems like an atypical way to fly an instrument approach. Generally, I like to slow to approach speed prior to the FAF, and then at the glide slope or glide path intercept, descend at that speed and remain at that speed until I reach minimums. So this pilot appears to have flown the approach fast and at a constantly changing speed.

One problem with flying at constantly changing speed is that the descent rate required to maintain the glide slope or glide path is also constantly changing. At higher speeds, you need a higher descent rate, and at lower speeds, you need a lower descent rate. Also, if the speed is continually changing, then the amount of drag is also changing. And that means that the power you need to maintain the glide path is also going to continually change.

Ideally, you'd like to fly the glide slope or glide path at a constant speed so that your descent rate and your power setting remains constant. And if it's not constant, then it becomes much harder to fly the approach as your speed, descent rate, and power required will change throughout the approach. And that increases pilot workload unnecessarily.

But there was a bigger problem than speed on this approach. I've plotted the aircraft's altitude versus a normal 3-degree glide path starting at Penny, and the aircraft appeared to be flying this approach as if it were a non-precision approach. In other words, instead of tracking the glide path,

there were two distinct level-offs during the approach. It's as if the pilot didn't have a WOS-capable GPS receiver, but was still trying to fly to LPV minimums. However, in doing so, he went as much as 511 feet below the glide path, and even at the last data point we have, he was still 376 feet below the glide path. I've used the plot as the image for this episode, so you may be able to see it in your podcast player.

For about the first mile and a half after Penny, the aircraft tracked the glide path fairly closely. Then the aircraft went below the glide path, and then it leveled off at between 1,300 and 1,400 feet, and it stayed level for about six-tenths of a mile. Then, as the aircraft reached Palos, it resumed a steep descent below the glide path.

Now, it's significant that the aircraft leveled off prior to Palos and started descending again after Palos, because that's exactly how a pilot would fly this approach if he were flying it as a non-precision approach to LNAV minimums. The approach plate shows that pilots must remain above 1,380 feet until reaching Palos, and there's an asterisk next to that altitude.

Elsewhere on the chart, another asterisk says LNAV only. That means that this altitude is to be ignored if you're flying the approach with a WASC-capable receiver and you're flying to LPV minimums. In other words, you'd fly this approach as a series of step-downs rather than tracking the glide path.

So the fact that the pilot leveled off at between 1,300 and 1,400 feet and later started descending again after Palos strongly suggests he was flying this as a non-precision approach.

If he were doing that, he could only descend to 940 feet, which would be 517 feet AGL, because of a NOTAM that raised the minimums, but given that a nearby airport was reporting 200 feet overcast, that wouldn't be good enough. At Palos, the aircraft was 130 feet below the glide path, but then it began a rapid descent, taking it at one point to 511 feet below the glide path.

The aircraft then leveled off for a second time, this time at 500 feet MSL. It stayed essentially at that altitude with one brief excursion back up to 600 feet,

for the next half of a mile when data for the aircraft was lost. At 500 feet, it was 173 feet below the LPV minimums for this approach, and it was 440 feet below the LNAV minimums. And the aircraft was flying a level as if it were at an MDA, which is what you would do when flying a non-precision approach.

Now, it's unclear why the pilot would have gone below minimums. It does appear the pilot wasn't using the autopilot near the end of the approach, as there is some variability in the lateral track, and it would be unusual for a plane to climb back up to 600 feet and back down to 500 feet if the autopilot were being used. Total time from departing Teterboro until the accident in San Diego was about seven and a half hours, so fatigue may have been a factor.

The accident occurred at about 3.45 a.m. Pacific time, which would have been 6.45 a.m. in Teterboro where the flight originated. Circadian lows occur between 2 a.m. and 6 a.m., and people make more mistakes then. That's why I turn down flights that are going to be late at night. It's just not worth the added risk. If you have to fly in the wee hours, then make sure you have a second pilot.

I made a Facebook post about the accident, and Paul Bertorelli commented, quote, I did enough night freight to appreciate that the hours between 2.30 and about 4 a.m. are the worst. Homo the sap is just not meant to be aroused and awakened then. Totally agree, Paul. The night accident rate is significantly higher than the daytime accident rate, and that has to be partially because people are tired, and their judgment, and their performance are poor at night.

Here's another comment on my Facebook post. Someone wrote,

This could explain the level off. Or two, the plane had JetTech STC for the Garmin 750s. Great setup with LPV, but you lose the VNAV functionality of the autopilot, and in some cases you have to trick the autopilot into flying vertical guidance by simulating an ILS-type approach. Either one could be difficult after an all-night flight.

I quit flying single pilot in the citation for just these reasons. And after I recorded this first part, someone familiar with the plane in his citation posted this information about the plane.

He wrote in part, I knew all of his planes. He had several pretty well. And I have the same Jet Tech Garmin setup in my citation sold recently that he had. And it's a known issue on LNAV approaches that as you get nearer to decision height, it starts to chase the glide slope. So most guys, at least guys I know with the same setup,

disengaged the autopilot at some point before the decision height, and began hand-flying the decision height on the LNAVs, which may explain his up-down variances from glide slope.

But Max, if you're going to report on this, I'd also look at how Sample Street, where the plane reportedly landed or pancaked in the middle of, lines up pretty perfectly and just above on a mini-mesa, runway 28R. Fatigued and below the lineslope, I'm wondering if he picked up the lights of Sample Street as runway lights and went for them.

There are approach lights with the Rabbit beyond there, but given 200-foot ceilings, as reported a couple miles away at KSEE, and perhaps still in the gunk, he didn't see the Rabbit but saw the sample streetlights. He was a very experienced and relatively young pilot with a number of planes, and he flew consistently in a lot of bad weather to one of the Finger Peninsulas in Alaska, where he had some land. So all kinds of day flying and weather were very familiar to him.

Rest in peace and praying that the rest of his family wasn't on board. Well, thanks so much for that post. So yes, this was a 40-year-old airplane and probably didn't have the most modern avionics. And in a situation like that, where the minimums were the same for the RNAV approach and the ILS, why not fly the ILS? Even in older planes that are not set up well for flying modern RNAV approaches,

The autopilots almost always do a decent job of coupling to an ILS's glide slope. And when you're tired in the middle of the night, it's easy to misread details on an RNAV approach, such as the PALOS fix only applying to LNAV minimums and not to LPV minimums. By contrast, ILS approaches are relatively simple.

The only real gotcha is remembering to switch the CDI from GPS to the NAV receiver so that you track the localizer and not the GPS signal. So while I've mentioned in the past that I prefer RNAV approaches versus an ILS, this is a perfect example where, when flying in an older airplane, it might have made more sense to fly the ILS. Remember, flying is a dynamic environment, and an answer that makes sense for one airplane in one situation matters.

might not be the best solution when flying a different plane. And just a reminder that I love hearing from you and I read many of your emails on the show. If you'd like to send me a message, just go out to aviationnewstalk.com, click on contact at the top of the page. That's absolutely the best way to send me a message. And of course, I also want to thank everyone who supports the show in one of the following ways. We love it when you join the club and sign up at aviationnewstalk.com slash support to

To support the show financially, you can also do that at aviationnewstalk.com slash PayPal. We also love it when you leave a five-star review on whatever app that you're listening to us on now. And of course, if you're in the market for a headset, please consider buying a Lightspeed headset and using one of the links in our show notes. Because if you use those links, they will donate to help support the show. So until next time, fly safely, have fun, and keep the blue side up. And remember that you can always go around. Go around and down.

You can all