Posted on the wall of my organic chemistry lab at the University of Wisconsin was a sign that read, Expectations cloud observations. Its a postulate thats as true in aviation as it is in science.
In a chem lab, if you expect the solution to have a bluish tint, your eyes may fool you and spot blue when theres really none to be found. In doing an aircraft preflight, a fuel tester may look like blue avgas rather than water simply because you have an eyeful of sky.
The unforgiving nature of aviation makes pilots, air traffic controllers and even regulators less than tolerant of human foibles that would go unnoticed almost anywhere else. Perhaps because the stakes are high pilots demand perfection from themselves and those around them.
Unfortunately, perfection is a standard to which no one can perpetually rise.
Take, for example, a tragic incident in Sarasota, Fla., in March 2000. The airport was fairly busy on a nice spring day. There were business jets and trainers and transients lined up to come and go on runway 14, a 7,000-foot runway thats 150 feet wide.
With about 200,000 operations a year, Sarasota Bradenton International Airport is a Class C airport whose airspace intrudes into the Class B of nearby Tampa International Airport. Like many Class C airports, it combines flight training, light jets and scheduled airlines.
There are two large FBOs there, Dolphin Aviation and Jones Aviation. The two FBOs are on opposite sides of runway 14/32, and each uses a different taxiway to shuttle airplanes to the runway. Complicating the picture is an intersecting 5,000-foot runway.
As the morning progressed, there were four controllers and the air traffic manager on duty, working ground and local control. In the rapid-fire world of air traffic control, breaks and briefings are frequent. Fully qualified controllers rotate among the different positions. There are ATIS recordings to be made and operational concerns pilots seldom see.
It was in this environment that one of the controllers fell down on the job. Literally.
When going down the stairs from the tower cab, a controller fell, cutting his arm badly. After getting first aid, the injured controller left the facility to get further medical attention. The tower manager cleaned up the blood and restocked the first aid kit.
In the tower, the regular flow of breaks and briefings hiccupped, but the controllers took it in stride. Outside, pilots came and went, oblivious to the scene in the tower cab.
The first link in an unfortunate chain appeared shortly after the controllers spill down the steps. Whether the distraction contributed or was merely a coincidence is moot, because a single link is seldom enough for an accident chain to be considered complete.
The taxiways used by the two large FBOs feed into the approach end of runway 14, but not at the same place. Coming from Dolphin, pilots use taxiway alpha, which goes right to the runway threshold. Coming from Jones, pilots use foxtrot, which meets 14 about 1,200 feet down the runway.
When 14 is active and a pilot calls for taxi clearance, the ground controller creates a flight strip that includes a reference as to whether the pilot will get to the runway via alpha or foxtrot. If alpha, the flight strip will include the runway, in this case 14. If coming from Jones, light plane pilots are generally able to start their takeoffs from the intersection of foxtrot and 14 rather than crossing 14 and continuing down alpha to the end of the runway. In this case, the strip contains a nonstandardized notation such as F/14, 14/F or 14xF.
On this morning, the pilot of a Cessna 172 called for taxi instructions and reported he was coming from Jones. The ground controller issued the taxi instructions and the supervisor, who was about to relieve the ground controller, prepared the flight strip. The ground controller then left, leaving the supervisor handling ground control duties.
Listen to the radio much at airports and youll frequently hear the kinds of mistakes that happened next. Pilots identifying their airplanes incorrectly. Pilots stepping on controller transmissions. Similar call signs being confused. This morning was no different.
In the midst of this, a Cessna 172, N79960, called for taxi clearance. He chopped off an IFR clearance, but then waited his turn. He reported he was at Jones and ready for taxi. The ground controller issued him a squawk code, then cleared him: taxi to runway 14. He marked the strip 14, but the pilot was planning a departure from foxtrot.
Aboard the Cessna were an 81-year-old, 13,000-hour flight instructor in the right seat and a 75-year-old, 2,000-hour Canadian private pilot in the left seat. They taxied to the runup area.
Across the airport, an 80-year-old student pilot was heading out for another lesson in a Cessna 152, N89827. He was accompanied by a freshly minted 26-year-old flight instructor. She had about 340 total hours and had given about 15 hours of dual instruction. Her instructor ticket was less than two months old.
They taxied to the end of runway 14 via taxiway alpha, where they joined a short line of airplanes waiting to depart. An Anheuser-Busch Falcon jet came in. A Mooney landed because the door had popped open. That pilot wanted to taxi back and take off again.
Links Two and Three
Working 10 aircraft, the local controller was trying to get the departures out without seriously slowing or diverting the incoming traffic. As an arrival landed, he would taxi airplanes into position and hold, then release them when the arrival cleared the runway.
He knew of two airplanes ready to depart. Then a third one called.
Ah, tower, this is 960. Were number two, ready for takeoff.
Nine six zero, roger. Number three for departure. Hold short.
He cleared a Cessna into position and hold at the foxtrot intersection and at the end of the runway. He let the intersection airplane depart, then cleared the airplane at the end to depart.
The Cessna pilot had reported being second in line. He was at the intersection. The controller glanced at the flight strip. It reported Cessna 960 at 14 – the end of the runway.
There were two airplanes in the air for the local controller to worry about – one in the pattern and another inbound on a practice ILS approach. He cleared the 152 to take off, then looked at the runway end and saw another Cessna 172 behind it waiting to go. Thinking it was 960, he cleared 960 to taxi into position and hold eight seconds after clearing the 152 to take off.
The 152 gathered speed as only a 152 can. Then, as it was about to rotate, Cessna 960 pulled onto the runway ahead of it.
Available evidence is inconclusive but appears to show the 152 was going between 50 and 60 knots when the 172 pulled out. The Sheriffs Department later measured 63 feet of skid marks. After skidding 63 feet, the pilots aboard the 152 apparently realized they wouldnt stop and tried to either take off or hop over the encroaching airplane.
The 152 went up, apparently stalled, and crashed onto the 172. The prop sliced into the right wing fuel tank of the 172 and, as the 152 nosed over on top of the 172, both airplanes burst into flames, killing all aboard.
The Blame Game
It was the first operational error reported at the Sarasota tower in more than 2 years. The FAA immediately conducted reviews of the facilitys operations and operations support. It found only minor, generally administrative shortcomings.
Although the tendency may be to blame the ground controller for marking the strip wrong or the local controller for clearing the wrong airplane into position, there are two other distinct links that could have broken the accident chain. A certain amount of blame falls squarely on the shoulders of the pilots in each aircraft.
First, the pilots in the 172. They displayed a remarkable lack of situational awareness by taxiing into position and holding without waiting for the 152 to go by. They were listening to the radio. They knew they were at the front of the line on foxtrot, yet when the controller cleared an airplane to take off they didnt make the connection that it would be rolling right at them. Neither did they look to their right to make sure the runway was clear before taking position.
As a minor point, perhaps if they had reported ready to go in sequence rather than number two the controller might not have been so quick to assume the airplane was at the end of the runway.
They abdicated command authority to the controller, perhaps subconsciously figuring that if the instructions were optional the guys in the tower would be called air traffic advisors or air traffic suggestors. Such complacency often appears in cockpits containing two (or more) experienced pilots. Perhaps each one is thinking the other is covering that base.
For that reason, its critical that pilots flying with other pilots have clear understanding of what each will do and not do on the upcoming flight. Designating one pilot as the captain helps prevent these kinds of errors.
But the pilots in the 152 were not without fault. Perhaps because the instructor had only 15 hours instructor time as well as 15 hours in the 152, she may have been slightly unsure of her responsibilities and the capabilities of the airplane.
The student pilot had more than 150 hours, although the NTSB report does not say over what period of time the hours were accumulated.
At least one of the pilots in the 152 – probably both – should have been alert for a potential incursion, especially if they routinely operated at an airport where such intersection takeoffs were common.
In addition, the 152 could have made it out of trouble without colliding with anything. The airplanes wing flaps were retracted and the airplane was nearing or at rotation speed. Estimates from radar data and witness accounts put the speed at more than 50 knots, which is the normal rotation speed.
But the bottom of the green arc is 40 knots, so the airplane would have flown without stalling if the pilots had flown away rather than skidding 63 feet and then trying to fly away.
The local controller had an expectation of what he would find, and he sought out the data to confirm it. That is, he saw a high-wing airplane where he thought Cessna 960 was. The pilots in both airplanes expected the runway to be clear, but did not seek to verify that.
The result is clear, but the lessons are broad. Trust the controller, but verify the instruction. Look out the window. Try to separate your observations from your expectations. Hey, if its good enough for a scientist it ought to be good enough for you.
-by Ken Ibold