There are a number of ways to provide some assurance that airplanes won’t run into each other while airborne, and nearly all of them have been put to use at one time or another in the history of aviation. These days, the prevention of midair collisions—particularly those involving airliners and other high-flying jets—has become positive and proactive because of the radar control services provided by the ATC in the United States.
This sort of positive ground controlling of the flow of air traffic is the procedure being used in most locations around the world—and certainly at those places where there is a high volume of local and/or international air traffic.
A secondary backup system to the positive-separation standards of air traffic control has been the use of modern Traffic Alert and Collision Avoidance Systems (TCAS) and related electronic devices inside the cockpits of airliners, and these days, in many General Aviation airplanes. These electronic warning systems provide the pilot with a secondary line of defense from potential conflicts.
Traffic separation hasn’t always been this good. In earlier years, collision avoidance was mostly handled by following basic rules. Procedures such as odd/even altitude assignments, or simply having very little traffic being spread across very large areas had been the operating methods of the day. It took a bizarre airline accident that occurred in literally the middle of nowhere to get the subject of collision avoidance started down the road toward more positive methodology.
On June 30, 1956, two four-engine propeller airliners sat not far apart on the ramp at Los Angeles International Airport (KLAX). A red-and-white Lockheed L-1049 Super Constellation, N6902C, named Star of the Seine, was being prepared to operate as TWA Flight 2.
The 8:30 a.m. departure for Kansas City, Mo. and then on to New York was running slightly behind schedule as the 64 passengers boarded. By the time the crew of six had gotten everyone settled, started the engines and taxied out, they were half an hour behind; takeoff time from LAX’s Runway 25L was 9:01 a.m.
Across the ramp, the blue-and-white United Airlines DC-7, Mainliner Vancouver, N6324C, had already loaded its crew of five and 53 passengers for its 8:45 a.m. scheduled departure to Chicago as Flight 718. Taxiing out just behind the TWA Constellation, the DC-7 lifted off Runway 25L at 9:04 a.m.
The captains and crews of both airliners were experienced both with their companies and with the routes to be flown that day. The weather was somewhat cloudier than average for the middle portion of the flight, since there would be multiple cloud layers to about 15,000 feet extending eastward from Las Vegas for a considerable distance. In addition, there were widely scattered buildups in the Grand Canyon area with tops reported to be at or near 25,000.
Both flights were on instrument flight plans, with TWA Flight 2 assigned 19,000 feet, while United Flight 718 had gotten clearance to 21,000. The routes for the two flights were divergent at first, with TWA departing northeast toward Daggett, Calif. before turning slightly right to Kansas City, while United departed LAX headed directly east toward Palm Springs, Calif. before it was to turn northeast toward Chicago.
After crossing Daggett, TWA Flight 2 called their company radio operator on the airline’s frequency to give the position report that would then be relayed to Los Angeles Control. In addition, the crew put in a request for a cruising altitude change to 21,000 feet.
In the mid-1950s, just as it had been since the earliest days of airline flying in the United States, each airline communicated strictly with its own company to give en route position reports. Company radio operators on the ground would then pass on the information to ATC via landline, get a response and then pass that reply back to the aircraft.
Air traffic control was, effectually, a clearinghouse for route/altitude requests designed to prevent planned conflicts—the actual communications between ATC and the airliners that they handled were accomplished individually by each airline. That meant the TWA crew couldn’t hear United’s position reports, and vice versa.
Los Angeles Control refused TWA’s request for 21,000 feet “because of traffic.” After this message was relayed to the flight, the flight crew then requested that TWA’s radio operator ask for a cruising clearance for “1,000 feet on top.”
That request was granted, with the controller adding “Advise TWA 2 that traffic is United 718… estimating Needles at 0957.” The new clearance and ATC’s message were passed along to the flight.
What TWA was asking for was a legal and often used procedure. In lieu of a particular cruising altitude, a flight on an instrument flight plan was saying that they were in VFR conditions, and that they would pick whatever altitude they thought was suitable for them as long as they remained 1,000 feet beyond any clouds. Effectively, TWA was now being provided with no separation services from other known IFR traffic since they had assumed the task of traffic separation themselves.
Complicating matters further were the airspace arrangements. Controlled en route airspace was aligned in corridors with vast regions of uncontrolled airspace between them—particularly across sparsely populated sections with limited traffic.
Those areas of uncontrolled airspace were available to IFR traffic also, the only difference being that ATC would be providing no separation in those chunks of uncontrolled air—even from other known IFR flights. Collision avoidance was basically handled by a matter of odds. What could be the odds of two IFR flights attempting to occupy the same small spot in this vast uncontrolled airspace at precisely the same moment?
After passing their previous fixes on routes that were still many miles apart but with flight paths that would eventually converge, both the TWA Constellation and the United DC-7 provided their company radios with an identical arrival time for the next fix: 10:31 a.m. Had that identical time alerted anyone in ATC to a possible collision? No, for three valid reasons.
The TWA flight was not at any particular assigned altitude—the “1,000 on top” meant that it could climb or descend at whim as long as it remained clear of the clouds and utilized the see-and-avoid concept for any potential traffic.
The area that both flights were now navigating in and would be flying through for some time was uncontrolled, so neither of them, regardless of their IFR flight plans, was getting any separation services from ATC. The rules said that whether they were in the clouds or not, the crews were strictly on their own. This was an often used procedure back then, although individual airlines sometimes had different policies about how they would operate while flying through uncontrolled airspace.
The final reason why those identical estimates to the next position report did not alert the controllers that a midair collision was at least a possibility was that the estimate given by each flight wasn’t to a specific position—it was the time that each airliner would be crossing a “line of position.”
The area that the flights were overflying after they departed controlled airspace was predominantly uninhabited: it was, in fact, the Grand Canyon area. There were no established ground radio stations to create fixes from, so flights simply reported when they passed the line between two usable stations—Winslow and Bryce Canyon, about 175 miles apart. Theoretically, a flight could be crossing that line anywhere along its entire length, so an identical estimate of arrival time meant nothing.
Except this time. Defying all odds and throwing the concept of “collision avoidance by probability” out the window, the United DC-7 ran across the tail of the TWA Constellation at an altitude of 21,000 feet, at what was calculated to be a 25-degree converging angle. The DC-7’s left wingtip smacked into the center tail fin of the Super Constellation, followed an instant later by the 7’s number-one engine slicing off the Connie’s tail just aft of the cabin entrance door.
The Constellation, now without its tail, pitched straight down and dove directly into the bottom of the Grand Canyon, very near to the Colorado River. The DC-7, now without a good portion of its left wing, spiraled down and impacted the sheer canyon wall a little more than a mile from where the TWA wreckage had impacted.
Later that day, a small airplane from a sightseeing service flew down the canyon and discovered the first signs of the wreckage, from which there would be no survivors.
The odds were strongly in their favor that day not to collide, but events just didn’t work out. The flight crews didn’t see the other airplane for reasons that will never be known. Suffice it to say that “low odds” and “see and be seen” were just not enough of a one-two defensive line to prevent the very first commercial air disaster that claimed more than 100 lives.
Only four years later, the same two airlines collided again—this time under much different circumstances but with the same basic results. On Dec. 16, 1960, a United Airlines Douglas DC-8, N8013U, collided with a TWA Lockheed 1049 Super Constellation, N6907C—a sister ship of the Grand Canyon airplane. Both flights were inbound to New York City, on instrument flight plans, and under radar contact and positive control.
TWA Flight 266 had departed Dayton, Ohio with 39 passengers and a crew of five and was inbound to La Guardia Airport (KLGA). United Flight 826 had left Chicago O’Hare (KORD) with 76 passengers and a crew of seven, inbound to Idlewild (now JFK). It was an overcast winter day in New York with some fog and light snow that had caused air traffic to back up.
The TWA flight was on radar vectors from La Guardia Approach Control and had been turned to a heading of 130 degrees. They were maintaining 5,000 feet. The United jet had been issued a clearance to a holding pattern well south of the TWA routing, and was in radar contact with New York Center. The controller was about to hand off the United flight to Idlewild Approach Control.
During United’s approach to New York, things began to happen quickly. The United crew was faced with a piece of failed onboard radio gear—one of their two VOR navigation receivers was inoperative. ATC had also changed the DC-8’s clearance so that the route they would fly toward the holding area—an intersection called PRESTON—was shortened. In addition, the DC-8 had its airspeed up so that they were approaching the holding spot rapidly— considering the wind, at a groundspeed of 330 knots.
To locate the PRESTON Intersection required the crew to take two VOR bearings from two different radio stations. With one of their VOR receivers inoperative, that meant they had to tune the single remaining radio from one station to the other and then back again. That would be a time-consuming and potentially confusing operation.
There was some evidence that instead the captain may have attempted to use the number one ADF, which was working normally, to get a different cross bearing with which to locate PRESTON. Regardless of the cause of his error, when the United flight reported to New York Center that they were now approaching PRESTON, they had, in fact, already passed it by 10 miles.
The New York Center radar controller who was handling the flight failed to notice that the jet had already sped well past its clearance limit. He terminated radar service with Flight 826 as he told the crew to change to the next frequency.
The La Guardia Approach Controller, on a different frequency, noticed an unidentified blip on his radar scope. He transmitted to the TWA Constellation that there appeared to be “jet traffic off your right, three o’clock, one mile, northeast bound.”
The DC-8 overtook the Constellation from behind, in the clouds, at an altitude of 5,000 feet. Impact initially occurred between the jet’s right outboard engine and the top of the fuselage of the TWA flight. The Connie broke into three main sections and fell nearly straight down onto the Staten Island terrain below.
United Flight 826 continued northeast for nearly 10 miles, coming down in a crowded area of Brooklyn and bursting into flames. Although a few people were pulled out alive from both planes’ wreckage, all of them died shortly thereafter to bring the accident’s toll to 134 killed, including six on the ground in Brooklyn.
This accident, on the heels of the Grand Canyon tragedy, brought about more aviation changes. The “250 knots below 10,000 feet” rule was the immediate reaction, followed by changes in ATC procedures and, eventually, the requirement for an operating transponder in bigger airplanes.
Collision avoidance has come a long way. Large airplanes and a measurable number of smaller ones now fly around with two effective lines of defense: a transponder coupled with ATC radar, plus onboard electronic collision avoidance gear. That combination has proven to be an effective solution at keeping airplanes apart.
Editor-at-large Thomas Block has flown nearly 30,000 hours since his first hour of dual in 1959. In addition to his 36-year career as a U.S. Airways pilot, he has been an aviation magazine writer since 1969, and a best-selling novelist. Over the past 30 years he has owned more than a dozen personal airplanes of varying types. Send questions or comments to editor@www.piperflyer.com.
