Mid-Air Collision on Approach to Centennial Airport

14 May 21 19 Comments

On the 11th of May 2021, a Cirrus SR-22 and a Swearingen Metroliner collided while on approach to land at Centennial Airport.

Centennial Airport is a busy general aviation airport in Denver, Colorado. Centennial has three asphalt runways: 17L/35R, 17R/35L and 10/28. At the time of the incident, the parallel runways 17L and 17R were in use.

Aerial view of Centennial Airport, looking south towards runways 17 left and right. Credit: Centennial Airport Photo

The Swearingen Metroliner is a twin turboprop, popular as a business aircraft. The Metroliner that day, a 43-year-old SA226-TC registration N280KL, was an air charter cargo flight operating as Key Lime Air flight 970. The pilot was the only occupant for the repositioning flight from Salida, Colorado to Centennial Airport.

The Cirrus SR22 is a high tech single engine aircraft which is known for its full aircraft parachute system. The private light aircraft registeration N416DJ, was owned by a local flight school and rental firm. The private pilot had rented the aircraft for a local area flight. The SR22 had two on board, the pilot and one passenger, and were returning to Centennial Airport after an hour’s flight in the Fort Collins area in the foothills of the Rocky Mountains.

The weather was clear and conditions were visual. The Metroliner was inbound from the north and cleared for a straight-in approach to runway 17L, the longest of the runways at Centennial with 10,000 feet (3,048 metres). He was descending through 6,400 feet about three nautical miles north of the runway threshold.

At the same time, the Cirrus SR-22 was coming in from the northwest and speaking on a different ATC frequency. He was cleared for a visual approach to runway 17R, a shorter runway parallel to the first. He was advised to watch for traffic landing on the parallel runway and warned not to veer too far to the east.

The Cirrus SR-22 initiated in a gentle turn towards the south but as he descended through 6,400 feet, he missed the turn onto final, crossing the centrelines of both runway 17R and 17L.

The Tower controller reacted immediately, asking “Cirrus 6 Delta Juliet…Did you overshoot the final? Cirrus 6 Delta Juliet, do you require assistance?”

You can see the two trajectories on this recreation of the ADS-B data from Flightradar24 with the green line of the Cirrus SR22 coming in from the left side while the Metroliner was flying down from the top:

ADS-B data of the two aircraft from Flightradar24

The Cirrus SR-22 appears to have been slightly above the Metroliner and it seems likely that the pilot, having failed to turn onto final, was looking out for the runway. Instead, he flew into the side of the Metroliner.

The pilot of the Metroliner never saw the Cirrus SR-22. Unsure of what had just happened but presumably experiencing a hard yaw to the left, the pilot immediately declared an emergency.

“We had, um looks like the right engine failed so I’m going to continue my landing.”

The pilot of the Cirrus SR-22 obviously also did not see the Metroliner until it was too late, which is a bit more difficult to justify, especially as he’d called that he had the traffic in sight.

A third aircraft, a Cessna 172 whose pilot was flying his first solo, saw the Cirrus deploy its parachute and reported this to ATC, who asked for a more specific location and then made one last call to the aircraft.

“Cirrus 6 Delta Juliet, if you hear this transmission, we have emergency vehicles in your direction.”

The Metroliner continued its approach and landed safely on Runway 17R. After landing, the pilot contacted ATC again.

“Tower, that was a definite mid-air on short final.”

Yes, it sure was.

The Metroliner after landing

I’m not sure it had really sunk in though, as when he was asked if he needed assistance, he replied cool as a cucumber.

“I’m going to taxi off here and I think I’ll just park over at Signature. I’m good, though.”

A local resident heard the collision. “I was in the kitchen and I heard a loud firecracker bang. I ran out…. I thought, ‘Is it somebody jumping out of a plane?’ And then I realised the parachute was attached to a plane.”

Another resident ran to the scene expecting the worst and was stunned when saw two men, the pilot and his passenger, in front of the plane, unharmed.

“They were just standing there like they were at a cocktail party.”

Cirrus SR-22 crash site

Amazingly, no one was injured. The local sheriff apparently said that both pilots should buy lottery tickets immediately.

The pilot of the Cessna 172 who was on his first solo has posted a YouTube video of his conversation with ATC with a map and his photographs taken of the two aircraft once he was on the ground.

I particularly like the first solo t-shirt he was awarded for that flight that he most certainly will never forget.

If you want to hear the full ATC transcript, I recommend VASAviation’s video which includes a written transcript as subtitles:

The NTSB have confirmed that they are investigating and that the preliminary report is expected in the next 14 days.

“We are working to understand how and why these planes collided,” said John Brannen, a Senior Air Safety Investigator from the NTSB’s Central Region office and the Investigator-in-Charge for the accident investigation. “It is so fortunate that no one was injured in this collision.”

They predicted 12 to 14 months for the final report but I have to admit, looking at the data available now, it doesn’t seem like it will take that long to come up with an analysis.


  • An unbelievable story with an even more unbelievably lucky outcome.

    It seems likely, at this moment when investigations are still ongoing, that the Cirrus pilot carries the blame for this collision.
    But, having said that, his subsequent reaction seems to have been professional and no doubt saved his, and his passenger’s, life.

    The Metro pilot deserves applause for his handling of the situation. HJe was more than likely not fully aware of the extent of the damage to his plane, nor probably to what exactly had happened. By a miracle the aircraft did not break up, not were the controls affected.

    But full marks to the student pilot. Even though he was not involved directly himself, it must have been a jolt to the system on his first solo.
    If I were an airline executive and – in post-covid – in need of young pilots for a cadet programme, I most certainly would consider hiring him.

    Good to read a story with a positive end !

  • It might have been better that the Metro pilot was by himself… a copilot looking out the door and going “holy **** half the aircraft is missing!” would not have helped.

    I hope that Cirrus pilot is rich. I see lots of lawsuits on the horizon for him.

  • I wonder whether the Cirrus pilot was less cautious because he thought the parachute would get him out of trouble. — or whether he was thinking at all; the view of the airport shown above suggests he shouldn’t have had any trouble seeing the right place to turn to line up, if he’d been paying attention.

  • What is not mentioned is the Cirrus Jockey was doing 150+ knots on base. In order to not miss final he would have had to initiate the turn fairly early. Sounds to me like he may have been in a hurry, maybe head in the cockpit fooling with this TV screens.

  • I suppose it will come out in the ATC transcript when the NTSB publishes their report, but I’m curious about the terminology “visual approach” being used for the Cirrus. A “visual approach” is an IFR approach clearance, as opposed to a VFR pattern entry instruction such as “Enter right downwind.” I rather doubt the Cirrus was actually on a visual approach because if they were they would have had to report the Metroliner in sight a long time before the events shown in the videos.

    • Hmm. This was my screw up. Both were described as “approved for approach” so I think I jumped the gun instead of interrogating what that might mean. Now that I’m paying attention, the Cirrus was told to follow another aircraft on base and then cleared to land, there was no “approved for approach” and the flight was clearly VFR.

  • I’m wondering at must have been a loss of separation even if things had gone as planned. I’m aware that bigger jets sometimes land in parallel at busy airports, e.g. Los Angeles, but I’m surprised ATC is scheduling that for approaches involving General Aviation aircraft. It’s probably convenient if you have traffic at very different speeds, but that doesn’t seem to have been the case here. I’m also surprised that they didn’t require the aircraft to have each other in sight mutually.
    Did any of these planes have TCAS equipped?

    • It looks like the Metroliner is just below the weight limit at which TCAS equipment becomes mandatory.

    • Local ATC was probably dealing with what it was handed. On long flights at IFR-only altitudes planes can be asked to slow down to reduce traffic jams at their destinations — IFF en-route centers figure out that this is necessary; if a couple of smaller planes on local flights ask for clearance when they’re the same number of minutes out from a twin-runway airport, ATC has no reason not to have them land simultaneously — that’s why the airport has two runways. (Wikipedia describes Centennial as “one of the busiest general aviation airports in the United States with an average of 874 operations per day”. Yes, they could use one runway for takeoffs and one for landings — if all the lightplane pilots would conform to a schedule….) At least there were two runways for planes to land on; at Worcester (KORH), in the 1970’s, when it had commercial service, I was once given a straight-in on 29 at almost the same time as a full-size passenger plane was cleared on 11. We didn’t come at all close but I was being VERY watchful for traffic (air and ground) on approach. I suspect there are times that ATC at Hanscom AFB (KBED, where I learned), with about as many movements as Boston, would have really liked a parallel runway.

    • In Class D airspace there is no minimum ATC-provided separation between IFR and VFR aircraft, only traffic advisories and safety alerts. If the Cirrus was VFR (which I believe to be true) there was no loss of separation up until the actual point of collision, and perhaps not even then depending on how technical you want to be.

  • My comment has only minor relevance to this amazing story and Sylvia’s great article and photos of it. Here in Ontario Canada, in the area I’m in anyway, the student pilot wears a white shirt to his/her first solo and if the solo is successfully done, upon landing and taxiing in, the instructor or senior instructor cuts the collar off the shirt and writes on the shirt the date, name of student, type of aircraft, and “First Solo”. A lot of these shirts ended up hung on the hangar walls, but I took mine home and have it to this day.

  • I’m impressed by the professional ATC approach shown in the video. It seems that the Cirrus chute was deployed pretty much immediately after the collision and it handled the quite high speed well. However, the pilot doesn’t seem to have done any sort of handling check or considered a forced landing. Is this rapid deployment the recommended protocol for chute equipped aircraft?

    • Several years ago, Cirrus’ safety team looked at their accident statistics and concluded that pilots were not pulling the chute when doing so would have saved their life, so modern Cirrus training emphasizes the pull-first approach, which seems to have flattened the fatal statistics, despite an increase in aircraft sales.

  • There are many little side issues emerging here. Interesting comments about “visual”. Don’t forget that in VMC all pilots are obliged to keep a look-out. I have personal experience about this in France, where during the course of my career I was framed no less than three times by ATC.

    Relevant to this is:
    I was on an IFR approach into Nice in a Citation. It was CAVOK. The initial approach then was via Cannes. We were level at the initial approach altitude of 2000 feet, established on the LLZ for, if I remember correctly, runway 05. Left or right? Not relevant in this case. What was relevant was that we were still doing 200 kts IAS, as said IFR, on a straight-in.
    Suddenly a speck appeared in our windscreen. Before we had time to identify what it was, let alone to react, we passed a red-and-white Robin at the same altitude, opposite to our track. We could clearly see the pilot.

    At a closing speed of more than 300 kts there was virtually no time in which we could take avoiding action. The little speck became an aircraft in a matter of seconds. The Robin must have missed us by 20 or at the most 30 metres.
    We declared a near-miss.

    When I went to file the report after landing, speaking to the ATC controller on the phone, the conversation went as follows (I will attempt to write it as verbatim as I can remember):

    ATC, (strong French accent): “Ah, you are the pilot of the Citation who wanted to file a near miss. Are you certaine that you want to file a report? Because maybe we will file a report against you. It was VMC, yes? So he had a perfect right to be there. Maybe you did not keep a proper look-out yourself, hein?”

    Needless to say, we did not pursue the case any further.

  • PS: When we were on the LLZ and maintaining 2000 feet, reducing speed to 200 kts, ATC never mentioned the presence of VFR light traffic on a dead-opposite track, at the exact same altitude. And Nice (LFMN) is a relatively major airport.

  • Did the Cirrus strike the Metroliner level, or was it banking? Either way if he had struck it lower down, he could very well have ripped the tail off, and we would not have the happy ending that we did.

  • Rudy, Nice ATC, wow! Lovely safety culture. Not nice. ;-)

    Re: Denver accident, looking at the VASAviation video and aircraft track, I’m going ask “could this happen to me, or you?” I’m gonna say yes. Why do I say that? First, the Cirrus entered a relatively wide pattern (or Cessna was very close in), and Cirrus was not really following the Cessna, but offset to left. We’ve probably all done that. Speed control is going to be a huge issue which started boxing him into a corner, and I’ve been fast in a slick aircraft at one point. It looks like the Cirrus turned to base abeam the Cessna as soon as the Cessna turned final. Now the box is closing. He’s on base, attempting to descend and slow down, trying not to catch up to the slower Cessna, all the while identifying the smaller runway centerline, keeping the Cessna in sight, scanning for the Metroliner, and configuring for landing. I think that turn from downwind to base just about quadrupled his workload. Now he’s trying to figure out when to turn final, but he’s hesitating because he needs spacing from the Cessna. Now in a normal single runway pattern, overshooting final is not good, but it’s recoverable. In a parallel runway environment overshooting final could have been deadly. What do we learn from this, before the final report? I’m going to ask myself on every downwind to base turn: “am I properly configured and do I need more time”, because that’s typically the best time to get extra spacing.

  • Well Hans, the other two were also rather dramatic but bear no relevance to this story.

    Waiting again for another “smashing” story from Sylvia. And hopefully, with a similarly good outcome.

    If the Cirrus had struck the Metro just a foot lower, by the looks of it, it would have broken up in mid-air. Amazing, and a credit to Swearingen for building the aircraft that strong.
    I have some time on a Metro II myself. “Flying drainpipe” seems to describe it. It had the reputation of having been a “hot ship”, but it was not too difficult to handle. We flew it with two pilots with passengers, but single crew on cargo operations.

    When being towed the front door with air stairs had to be closed, otherwise, so I was told, the airframe could deform. So I am doubly surprised that it held together and remained controllable with that massive amount of damage.

    Another anecdote before Sylvia closes this conversation for the next episode:

    I was based at Shannon in the early ‘eighties. My primary job was captain on an Aerospatiale Corvette. It was owned by Guinness Peat Aviation, an aircraft leasing company headed by Dr. Tony Ryan, founder of Ryanair. But it was operated by an air taxi company, Shannon Executive Aviation. The Corvette was also available for charter flights, and SEA had other aircraft such as a Metroliner II, a Shorts Skyvan and a Piper Aztec.
    One particular afternoon we returned from a charter flight from Inverness in Scotland. Normally we would sign off-duty, but there was a little problem: The Metroliner was doing a scheduled passenger flight from Shannon to Dublin. There, a ground crew removed the chairs and it was used for a parcel cargo flight to Luton, return to Dublin and do the early morning passenger flight back to Shannon.
    The Skyvan was also sitting at the ramp at Dublin, waiting to do a late afternoon mail flight to Liverpool.
    So far so good. My copilot and myself parked the Corvette but there was a panic: The Metroliner was at point of departure, passengers ready to board but… there was no crew. They had not reported for duty, the phone was not answered and the mobile phone had not been invented yet. So we were pressed back in action: We took the Metroliner and flew it to Dublin. A few of the passengers that had been on the Corvette were also on board. They must have wondered if we would also drive their taxi in Dublin and open the hotel door for them.
    A hitch: My copilot could go home in Dublin, I was told to wait a few hours, then do the night cargo flight to Luton.
    I had to refuse. I would violate the maximum allowable duty times, by a margin of several hours.
    Instead, I took the Skyvan. That one departed earlier and for some reason needed a pilot too.
    Meanwhile, the captain of the Metroliner had been located and was told to fly himself to Dublin in the Aztec to take over the (single-crew) cargo operation with the Metroliner. After this flight the pilot would get a bed for some rest and then take the early flight back to Shannon. My copilot was asleep at home and would act as F/O on that leg.
    So I took the Skyvan to Liverpool and back, a relatively easy hop. When I returned, the Aztec was sitting at the ramp, so I took that to get back to Shannon.
    In quick succession I flew four different aircraft – on the same day, within a 24 hour time period:
    A Corvette jet, a Metroliner, a Skyvan and an Aztec. Not many pilots can claim that !

Post a comment:

Your email address will not be published. Required fields are marked *


This site uses Akismet to reduce spam. Learn how your comment data is processed.