Landmark Accidents: Radar Service Terminated

A beautiful flying day in Florida claims four lives

October 1, 2005

One of the great things about learning to fly in Florida is the wonderful weather. One of the more challenging aspects is that everyone else likes flying there too. Florida is home to many large flight academies and several universities that certificate thousands of new pilots annually. Students come from all over the world to enjoy the weather and the relatively inexpensive cost of learning to fly in the United States. Many of the airports and the airspace around them are, by any definition, high density. The accident recounted here could happen at any nontowered airport and the lessons apply to all of us, regardless of experience or location.

One December day, a Piper Cadet with a CFI and a private pilot on board departed Daytona Beach International Airport in Florida at 9:21 a.m. Eastern Standard Time VFR to the southeast. The flight was handled by Daytona Beach's south arrival sector controller until 9:34, when radar service was terminated.

As the Cadet was taking off, a Piper Seminole with a very seasoned CFI and a commercial-, multiengine-rated pilot contacted clearance delivery and requested, "A local IFR shoot VOR 23 approach DeLand [DeLand Municipal-Sidney H. Taylor Field, a nearby nontowered airport], to be followed by NDB approach Runway 30 DeLand and then break off and come back for an ILS later." The controller cleared the flight to the DeLand airport via radar vectors. The Seminole taxied to the runway and was cleared for takeoff at 9:38 a.m.


After takeoff, the Seminole contacted the controller in Daytona Beach's north sector at 9:42 and requested, "Is it possible we can have the full approach, sir? Uh, to fly over the VOR?"

The controller responded, "Unable VOR DeLand approach."

The Seminole replied, "A mile north of Ormond [the VOR from which the DeLand VOR 23 approach begins] is fine, sir."

The controller responded, "Not for you, sir. Unable VOR 23 approaches. Fly tower-assigned heading now."

Not to be deterred, the Seminole requested a nondirectional radio beacon (NDB) instrument approach to Runway 30 at DeLand.

The controller advised, "OK. DeLand's saturated right now. Unable any approaches at DeLand. Let me re-phrase that — unable any approaches at DeLand."

The Seminole pilot then responded that he would like to execute a VOR approach to Runway 16 at Daytona Beach, and an NDB approach to a runway at the New Smyrna Beach Municipal Airport. The flight was vectored and cleared for the VOR 16 approach at Daytona Beach, and the pilot was told to expect to break it off four miles north of the airport. At 9:59 a.m. the controller asked the Seminole if he wanted an instrument landing system (ILS) approach after the VOR approach, and if he had another request. The controller then asked, "After this VOR 16 would you like an approach at DeLand now?" The Seminole replied that they'd like the full VOR 23 approach.

Into the pattern

There was some discussion between the aircraft and air traffic control (ATC) regarding the full approach, and at 10:01 the controller canceled the approach clearance and provided a heading and altitude to fly, which was acknowledged. At 10:03, there was a controller shift change and the relief controller was briefed that the Seminole was expecting the full VOR 23 approach. At 10:06 the Seminole was directed to proceed to the Ormond Beach Vortac and to join the 212-degree radial.

The Seminole was subsequently given a frequency change, checked in at 10:12, and shortly after that was cleared for the approach. At 10:15 the controller advised caution because of constant parachute jumping. At 10:17 the controller questioned if the flight would be terminating, to which the pilot responded, "Negative, sir, we're going to break off and do some airwork and then come back with an ILS, full stop. We'll advise you." The controller advised the flight to report canceling IFR. At 10:18:02, the Seminole canceled IFR and the controller responded, "Radar service terminated. Squawk 1200, frequency change is approved." This transmission was acknowledged and there was no further contact with ATC.

Later, witnesses reported seeing, at about the same time, the Cadet in the traffic pattern to Runway 5 at DeLand for two patterns and reported hearing radio calls from it in the pattern on the common traffic advisory frequency (CTAF)/unicom. Two pilots reported hearing a radio call from an unidentified airplane report "VOR 23" on the DeLand CTAF; one pilot did not recall hearing the identification of the airplane or hearing anyone announce the distance or intentions.

Several pilots reported hearing the Cadet report on the CTAF that the flight was departing Runway 5. One witness reported that the Cadet used almost the full length of the runway, became airborne, then banked to the left and disappeared behind trees. Another witness, who was located near the accident site, reported seeing an airplane flying north and another airplane flying west. At 10:20 a.m., the Seminole and the Cadet collided near the departure end of Runway 5. Both aircraft were destroyed and all aboard were fatally injured.

The CFI aboard the Seminole held an airline transport pilot certificate and the title of director of corporate flight training for the flight school where he was employed. He had more than 7,300 hours' total flight time with nearly 5,000 hours of dual instruction given.

The student in the Seminole was a commercial pilot with airplane single- and multiengine land and instrument airplane ratings. He had logged more than 5,500 hours, of which nearly 3,000 hours were as second in command. The student was receiving instruction for an airline transport pilot certificate.

The CFI in the Cadet was employed as a full-time flight instructor by a local university and had logged 827 hours' total time, with nearly 400 hours of dual given. The student was a private pilot with an instrument rating. She had logged 168 hours.

The weather was not a factor other than being typical for a good VFR flight. The 9:53 a.m. weather observation taken at Daytona Beach, 14 nautical miles from the accident site, indicated calm wind, visibility of 10 miles, and a broken ceiling at 4,800 feet.

Frequency congestion

The DeLand CTAF was the very common 122.8 MHz frequency. According to a statement by a pilot and airplane owner, an air traffic controller who was flying on the day of the accident stated, "As a tower controller, I am experienced in determining aircraft positions through radio communication. After 17 years in the business, I have never witnessed frequency congestion as bad as it was on December 3. Both 122.7 and 122.8 were so unmanageable and absolutely chaotic that it is very understandable how no one on either frequency could understand what was going on." He stated that he heard radio calls from pilots at seven different airports on the DeLand CTAF.

According to the university aviation safety program manager, the school began operating 10 airplanes at the DeLand airport a few months before the accident and recognized the frequency congestion problem. There had been "a dialogue" with the city engineer of the DeLand City Commission regarding existing frequency congestion from other airports on the same frequency. The university urged the city, which was the CTAF license holder, to request a frequency change.

Accident data

According to the FAA, after the accident, examination of the radios from the Seminole revealed they were inoperative. One radio was set to frequency 122.8 MHz, and the second set was tuned to 125.57. However, if the radio was inoperative no communication with ATC would have been possible, so there is clearly some ambiguity here.

A recorded radar plot prepared by the NTSB indicates that the transponder of the Seminole was observed from 9:39:57 a.m. until the secondary radar target disappeared at 10:18:08, when radar service was terminated and the transponder code was noted to change to 1200. The airplane was located at 210 degrees and 1.23 nautical miles from the DONGS Intersection at that time. The Seminole descended to approximately 600 feet mean sea level (msl), leveled off momentarily, descended to approximately 500 feet msl, then climbed to approximately 600 feet msl, where the last radar target was noted approximately 0.6 nautical mile from the approach end of Runway 23. The Cadet proceeded to the DeLand airport and several traffic patterns associated with the airplane were noted. Just prior to the accident, the Cadet departed Runway 5 and turned to the northwest.

The NTSB finding of probable cause was, "The inadequate visual lookout by the pilot in command (PIC)/certified flight instructor (CFI) of both aircraft. Contributing factors in the accident were: 1) the frequency congestion of the CTAF; 2) the poor in-flight planning decision by the PIC/CFI of the Seminole for his continuing a practice instrument approach to within approximately 0.6 nautical mile from the approach end of the runway with opposing airplanes departing on the upwind leg; and 3) the absence of guidance in the Aeronautical Information Manual [AIM] and advisory circulars as to how or when to terminate a practice instrument approach to an airport that does not have an operating control tower."

The lessons

At nontowered airports, collision prevention should be foremost on the mind. We need to be looking, listening, and talking appropriately. Courtesy and common sense work really well in these circumstances — in fact, I'm hard-pressed to think of a circumstance when they wouldn't. But it bears repeating, because in the heat of routine operations you'll see a lot of "Hey, hey, hey, get outta my way." A favorite CFI response to "I don't see anyone" is "It's the ones you don't see who'll get you."

In this accident, the Cadet crew obviously didn't expect to meet an aircraft at the departure end. Why not? Especially in calm wind conditions, it's entirely possible that someone has a legitimate reason for using a different runway. The Seminole crew was engrossed in their approach or missed approach, against traffic, and apparently had forgotten the rules of engagement. Why? Practice instrument approaches usually take place in visual conditions. The rules, even when operating under an IFR flight plan, are VFR. Just as VFR traffic must defer to IFR under instrument meteorological conditions, the opposite is true. There are some pilots who think that IFR somehow confers superior status. In my opinion, in VFR conditions it confers additional responsibility. Here is where common sense and courtesy come into play. If a local airport is saturated with VFR traffic, does it make sense to inject yourself into that, especially when flying contrary to the flow? How much better would it be for all concerned to schedule IFR approach practice to busy nontowered facilities when they are not so busy?

The CTAFs at many nontowered airports are a mess. What worked 20 years ago may not be workable today if traffic has grown appreciably. If you're flying regularly at one of these locations, start working with the authorities to make the frequencies usable. Flight schools, instructors, and the insurance community definitely have a vested interest. There may be whining from the CTAF operator, fixed-base operator, or municipality, about needing new ground radios since the old ones cannot be converted or about how the local pilots will be confused because Podunk Municipal has always been on 122.8. These concerns pale in comparison to the alternatives. If the frequency is overcrowded, do something. Be sure to put your concerns in writing so that a paper trail is evident. It encourages people to be more responsible.

While the AIM might not have specific guidance on when to begin a practice missed approach, it might make sense to start the miss before getting too deeply into the VFR pattern if it's jammed. Arbitrarily, that might be two miles or so from the runway, and perhaps we ought to turn away from where the VFR traffic is going. If it's a nonprecision approach, an easy, low-conflict idea is to fly it 1,000 or 2,000 feet above the published altitudes. Copy the approach chart and modify the altitudes as appropriate — this cuts the collision potential way down while allowing the full procedure to be practiced. There are many different ways to adapt and avoid. The main thing to remember is that Newton was right!


The loss of life in this accident is tragic and can be characterized in no other way. However, in our society fault is measured in dollars. The family of the Cadet flight instructor sued the FAA for negligence in authorizing the Seminole to practice an IFR approach into DeLand's Runway 23 without warning that the airport was saturated with training aircraft landing in the opposite direction on the same runway, Runway 5, and "failing to timely release the Seminole pilot from...ATC radio frequency to enable him [the Seminole pilot] to warn pilots in the traffic pattern of his location and intentions...."

Reviewing the briefs filed by both the plaintiff and the defendant, there are some interesting interpretations of the operating rules and procedures under which we operate. At press time the case was awaiting judgment — up to $3 million is on the line. Here are some highlights that, because of space limitations, are summarized but contain the essence of the argument.

Plaintiff: Tracon controllers could determine from radar that Runway 5 was in use at DeLand.

FAA: When the Seminole cancelled IFR, no aircraft were displayed on radar, in the pattern or about to depart from any runway at DeLand. Note that ATC radar does not have ground coverage at DeLand.

Plaintiff: Controller never advised the Seminole of traffic using Runway 5 and thus created a potential collision hazard. In the AIM, "advisories are issued to alert pilots to other known or observed air traffic which may be in such proximity to the position or intended route of flight...."

FAA: At nontowered airports, pilots are free to choose any runway they wish and ATC does not monitor nontowered CTAFs. "It is not correct or feasible for ATC to determine a runway in use at a nontowered airport by observing possible aircraft targets on a radar scope.... ATC would be incorrect to assume any particular runway was the only one that might be in use at DeLand and transmit that information to any aircraft." Twenty-nine seconds before the collision, radar detected a primary target with no altitude or transponder code which was later determined to be the Cadet. Four seconds before the collision, there was a single radar hit on the Cadet showing a Mode C readout of 500 feet. ATC was not in contact with the Cadet after it arrived for VFR pattern practice at DeLand. ATC terminated the Seminole's IFR flight plan and radar coverage and was not in contact with either aircraft in the two minutes preceding the accident.

Plaintiff: ATC delayed the Seminole's change from Daytona Beach Tracon frequency to CTAF such that the Seminole was unable to warn DeLand traffic of their location and intentions. ATC manuals require controllers to release IFR traffic from the ATC frequency "at a sufficient distance from the airport to permit the pilot to change to the appropriate frequency for traffic and airport information."

FAA: Authority to terminate the IFR flight plan rested solely with the Seminole pilot. If there was concern about switching to the CTAF he could have terminated his IFR flight plan sooner. ATC advised the Seminole pilot at 10:17 about cancellation procedures, and at 10:18, when the Seminole was 3.8 miles from the accident site, the pilot cancelled IFR and was immediately cleared from the ATC frequency. "The Seminole and Cadet collided at 10:20:17...approximately two minutes and four seconds after the Seminole last communicated with ATC."

Note: According to witnesses, there was no radio call from the Seminole identifying that a practice IFR approach was in progress or giving their present location.

Plaintiff: The ATC manual states that controllers are to "ensure that neither VFR nor IFR practice approaches disrupt the flow of other...traffic" and to "refuse practice approaches as conditions require." ATC must do more than idly watch as a pilot creates an unnecessary collision risk. Despite a denial earlier, the relief controller's decision to authorize approaches in a traffic pattern that was saturated was "clearly negligent."

FAA: "When the Seminole pilot announced that he intended to break off the approach and would not be landing, no aircraft were displayed on the controller's scope in any DeLand traffic pattern.... ATC has no duty to advise ...of unknown traffic...."

Note: Approximately 17 minutes had elapsed from the denial of a practice approach by the first controller to the offer of a DeLand approach by the second controller. Traffic conditions may have changed significantly in that time.

Bruce Landsberg is the executive director of the AOPA Air Safety Foundation.

AOPA Online Safety Center: Operations at Nontowered Airports

Less than 5 percent of the more than 13,000 airports in this country are served by air traffic controlled towers — chances are that if you fly, you'll be flying to or around a nontowered airport.

FAA-recommended procedures (FAR 91.113, FAR 91.126, and FAR 91.127) establish traffic-flow rules at nontowered airports. The Aeronautical Information Manual (AIM) and FAA Advisory Circular 90-66A expand on these procedures. However, regulations and procedures cannot cover every situation. To help you interpret and understand these regulations and procedures, the AOPA Air Safety Foundation offers a comprehensive Safety Advisor downloadable from the AOPA Online Safety Center, Operations at Nontowered Airports. Information on other safety-related issues is available 24/7 in the AOPA Online Safety Center. — Julie Summers Walker