On January 5, 2007, at 0728 Mountain standard time (1428 Coordinated Universal Time, UTC), N425MA, a Swearingen Metroliner (SW4), operating as Key Lime Air (LYM) flight 426 and N915FR, an Airbus 319 (A319), operating as Frontier (FFT) flight 297, were involved in a runway incursion at Denver International Airport (DEN), Denver, Colorado. The Airport Movement Area Safety System (AMASS) activated after the FFT297 flightcrew initiated a go-around after seeing the SW4 on the runway. The aircraft missed colliding by approximately 50 feet. Use your browsers 'back' function to return to synopsisReturn to Query Page
SEQUENCE OF EVENTS
At 0725:00 MST, the DEN ground controller instructed LYM4216 to taxi to runway 34 via taxiways M and AA. According the SW4 pilot, blowing snow reduced his visibility and taxiway SC was covered with snow that prevented him from seeing the centerline lighting. As he attempted to find the centerline lighting, he saw blue taxi lights, followed them, and turned onto runway 35L.
According to the recorded airport surface detection equipment (ASDE), LYM4216 entered runway 35L at taxiway M2 at 0727:06.
At 0728:10, the ground controller asked the LYM4216 pilot his location. The pilot advised he was "abeam Signature". According to the pilot, once the controller asked for his location, he noticed that he was on a runway.
The First Officer of FFT297 reported, "the Captain and I were flying back to Denver from St. Louis. The weather had deteriorated in Denver so we prepared to fly a full instrument approach. We were cleared to land on runway 35L by tower. I believe we broke out of the clouds around 600 feet. The visibility conditions were ½ mile with blowing snow. We both looked down the runway and confirmed verbally to each other that the runway was clear. We didn't see the aircraft until we were about 100 to 50 feet or so above the deck. When [the airplane] did come into sight, it was at least 2,000 feet down the runway. The winds combined with the prop wash from the aircraft along with blowing snow is what caused it to be obscured and out of sight. I immediately commenced a go around as soon as the aircraft was in sight. The tower was given warning from its collision avoidance system not more than a couple of seconds later."
At 0728:17, AMASS provided the tower controllers a verbal and visual alert for an occupied runway and 4 seconds later, the tower controller instructed FFT297 to "go around".
At 0730 MST, the DEEN weather observation was 600 overcast, 1100 broken, visibility 1/2 statute mile, light snow, mist, wind 030 at 12, runway visual range on runway 35L was 5,500 feet.
Air Traffic Control Interviews
On January 9, 2007, Federal Aviation Administration ATO-S staff interviewed the DEN local controller (LC2), ground controller (GC2), and Front Line Manager (FLM) who were working at the time of the incident.
LOCAL CONTROL 2 INTERVIEW: According to the FAA's report, LC2 stated that the weather was IFR with low ceilings and poor visibility, and that most of runway 35L was not visible. He was responsible for operations on runway 35L, and the facility was conducting triple simultaneous approaches to runways 35L, 35R, and 34R. Aircraft are transferred to his frequency at least 25 miles out. He stated that he was engaged in monitoring the final approach course on the DBRITE, obtaining and distributing Pilot Reports (PIREPs), and monitoring the runway lights because of previous lighting failures. He stated that earlier in the week the runway 35L/R condenser discharge sequence flashing lights (RABBIT) had failed and that the "speed lights" had failed. He was also watching the Runway Visual Range (RVR) because the values were reportable. He stated that there were numerous activities taking place on the parallel monitor position with speed adjustments and spacing, and he was monitoring UAL529 as the aircraft rolled out and exited the runway. He obtained a braking action report and continued to monitor the aircraft as it exited the runway, because he felt that even though the braking action reports for the runway were favorable, the taxiways may be slippery. The next aircraft was FFT297 and he acknowledged the aircraft and cleared the flight to land approximately 20 miles out, issued traffic, wind information, the RVR and the braking action report. He heard the following aircraft, GL760, being issued a speed restriction and then an instruction to reduce to final approach speed because of the spacing behind FFT297. He stated that it "wasn't working," and the parallel monitor canceled the approach clearance of GLA760. He was watching that situation to see how it was going to work out, and the ground controller pointed to the AMASS and alerted me that something was going on and at the same time the AMASS alerted and he sent the Frontier around. He stated that he saw a runway 34R departure coming off to the west and he coordinated an eastbound heading and gave the aircraft to departure.
GROUND CONTROL 2: According to the FAA's report, GC2 was asked to describe the event. He stated that the weather was "serious IFR," one-half mile visibility, 3000-5000 runway visual range, and he was working arrivals from runways 35R and 35L, in addition to arrivals eastbound on taxiway Z and runway 34R. He could not see taxiway M6 and sometimes could not see M7. He stated his main concern were taxiing aircraft on taxiways M, Z, L and ED, all crossing. He stated that it requires a lot of attention there. A number of Key Lime flights were late, because the "mother ships" were late, and that he could not see them except on the AMASS. He stated that he had one Key Lime flight come out and about 25-30 seconds later Key4216 called for taxi and both were given instructions to taxi via M and AA; north on M and then the ramp and then on AA and then the west side ground controller would assign the runway for use. He stated that "something wasn't sitting right and he looked on AMASS and asked the pilot of 4216 his location. The pilot advised he was abeam Signature. He observed one target north on taxiway M and he assumed it was him, and that "the guy in front must have turned." He then saw a target on the runway and asked again about his location. The pilot then advised "by Signature". He stated that he thought that the target that was on the runway was a typical AMASS tag that had dropped on the arrival. He stated that something "still was not right," and then he pointed to the AMASS display and the AMASS alert went off and he screamed to the local controller to send the Frontier around.
When he was asked when he first saw the Key Lime flight, he stated between taxiways M6 and M7. He stated that he saw the aircraft on the ASDE near taxiway M4, approximately 10-18 seconds prior to the alert. He stated that when aircraft are cleared from the cargo area to taxiway M, he normally monitors the turn on the AMSSS. He did not remember monitoring this flight because his attention was diverted to the north. He stated that he monitors the turns because he deals with pilots from Key Limes every day and most don't understand the signage. He stated that taxiway L is currently a small mountain range of snow dumped by the city.
When asked about his scanning technique, he stated that his scan normally starts where there are the most possible conflicts, and then to the right, then to the DBRITE to write down call signs, and then to the ASDE. When asked if he knew how long Key Lime 4216 had occupied the runway, he stated that he did not. When asked his physical location in the tower cab with respect to the ASDE display during the incident, he stated that he was in between GC2 and LC2 but was closer to LC2.
When asked where his attention was focused prior to the AMASS alert, he stated that he was doing the scan and had resolved the ramp problems, and then looked for the Key Lime flights. He assumed that the first Key Lime flight was 4216 because the pilot advised that he was abeam Signature. He stated that the pilot advised that he was abeam Signature but he was really south of Signature on the runway.
When asked his perception of AMASS, he stated that it is a wonderful tool but the only problem is that we do have a documented occurrence of tags dropping, and issues with extra targets, and we'll see things that are not there. They don't last, and there situations like now where we get a lot of ground clutter related to the snow. It is hit or miss whether we got a lot of clutter. We have had problems down there at the cargo area seeing airplanes. Departure data tags would have made a world of difference. When asked if the system alarmed in a timely manner, he stated that he thinks that the way that it is set up on runway 17 it is better. He stated that they don't see targets arriving runway the runway 35 parallels because of the way the display is set up, and he has not seen the map re-centered in the cab. He stated that he does not see the target until it hits the catch zone or the polygon, and that "all you get is the correlated data tag and no target."
He stated that the system worked to perfection that day, and that the two controllers and the AMASS averted a huge disaster. He stated that Frontier went around because they aircraft was sent around and because he came out of the clouds after he was issued the go around and saw the aircraft. He stated that the alert parameters need to reset. When asked if he had any discomfort with the fact that the aircraft was on the runway for approximately a minute, he stated that he did and that he would have "loved to have caught it earlier."
FRONT LINE MANAGER: Runway 35 L opened approximately 10 minutes prior to the incident. He was monitoring local control 2 (LC2) and moving back and forth across the cab between the 2 sides trying to keep up with the traffic, and was walking from the east side when the AMASS alerted. He walked four to five steps and went to see why the alert happened. He overheard the ground controller tell LC2 to send the Frontier around, and he then looked at the ASDE and saw a target just south on taxiway M4 on the runway. He stated that he heard the AMASS aural alert. He stated that we had already issued a go around and Frontier showed 5500 on the BRITE. He stated that it looked like the aircraft were going to hit and Frontier came out on the other side. He said that he was unable to see the event because of the reduced visibility that he described as variable and about a mile. He said that he could barely make out taxiway M7 and couldn't see M6. He stated that he could see the arrivals rolling out and exiting if "you knew where to look." He stated that he had 3 locals open and that he was listening to LC2. He felt that it was unreasonable that the aircraft was on the runway for approximately 60 seconds prior to the event and that he expected the ground controller to see it "pretty quickly." He stated that ground control was probably focused on the taxiways near the air carrier terminal. He stated that ground usually turns the aircraft departing the cargo area north, but they need to make sure that the aircraft turns on the taxiway. When asked if he would have expected the local controller to see the aircraft, he stated yes. He would expect the local controller to look at the ASDE as part of the scan to make sure the runway was clear. He stated that he stresses the use of a good scan during bad weather conditions. He said that controllers use the ASDE to ensure that aircraft clear the runway.
DEN Acting Chief Aviation Operations Manager Interview
On January 31, 2007, Mr. Mark Kullas, Acting Chief Aviation Operations Manager, provided the NTSB information about the condition of the airfield on the day of the incident. At the time of the incident, he and Mr. Ron Morin, Denver Airport Field Maintenance, were on the airfield observing snow removal activities - the airport was in "Snow Alert". They were not made aware of the incursion until 0918, according to the clock in the truck. At that time, they checked the runway signage and Runway Guard Lights to determine whether or not they were obscured by snow; they weren't. Shortly before 1100, they returned with a camera and took pictures of the signs and runway guard lights at taxiways A, M2, and EA. There was a NOTAM in effect stating that signage was "occasionally" obscured by snow because, according to Mr. Kullas, the wind in snowstorms is typically out of the north or northeast. However, the day of the incident the signs aligned north/south did not have snow obscuring them.
When asked whether he was aware that the snow piles were blocking the radar return for the ASDE, Mr. Kullas explained that the "snow dump" areas typically used were filled from a previous storm. Several days before this incident, Mr. Morin had asked Mr. Kullas where to pile snow from the Cargo Ramp. Mr. Kullas told him that they would close taxiway L between taxiways EA and SC and pile it there. Mr. Kullas explained that this taxiway was not part of the original airport construction, but was added on some years later to give the tower ground controller an option for access to and from the Cargo Ramp during busy periods. He further explained that taxiway M is the lone taxiway down in that area so although they could not provide two way traffic (and still can't between taxiways EA and AA), it allows for an option to move an aircraft off taxiway M for a short distance and keep traffic moving to some degree. Mr. Kullas believed that taxiway was non-essential. He was not made aware that the snow pile caused a problem for the ASDE until sometime in the week after the incident. The day of the incident was the first time they had piled snow in that location. Mr. Morin arranged for a contractor to come out and work 12-hour shifts non-stop to remove the snow piles to allow the taxiway to be reopened.
Mr. Kullas instructed the duty managers that if they need to pile snow other than the normal areas if must be outside/outboard of the movement areas so as not to interfere with "line of sight" from the two ASDE antennas (one located on top of the control tower and the other on the east airfield north of taxiway ED).