A Fundamental Failure Of The FAA

The NTSB found serious deficiencies in the training of controllers

When the NTSB finished its investigation of the September 7, 2015, Beech A36 Bonanza crash near Kernersville, North Carolina, in which the pilot and both passengers were killed, it was just six days after the agency had released an alarming Safety Recommendation Report. The subject of the report was "Emergency Training for Air Traffic Controllers." It demonstrated that you can't always count on FAA air traffic controllers to have the training and skill to help bail you out of trouble if you ever need to ask. The NTSB highlighted five general aviation accidents in which the controllers didn't provide effective emergency assistance to pilots and sometimes made bad situations worse. The Kernersville accident turned out to be yet another accident with inadequate controller training as a factor.

Sadly, the Safety Board had urged the FAA to upgrade emergency training for controllers as far back as Safety Recommendations issued on September 24, 2001. In the opinion of the NTSB, 15 years later, as of the new Safety Recommendation Report released on August 25, 2016, the FAA still hadn't gotten it right.

The first of the five accidents cited in the new report involved a Cessna 152, which crashed into shallow water at New Smyrna Beach, Florida, on January 13, 2015. The pilot was killed. She strayed into instrument conditions and asked for help. Controllers vectored her toward the New Smyrna Beach Airport, which was IMC, instead of toward airports to the west, which were VFR.

Next was the April 11, 2014, crash of a Piper PA-32RT-300T near Hugheston, West Virginia, in which both occupants were killed. The pilot was on an IFR flight plan and was attempting to navigate around weather. Despite 15 minutes of airspeed and altitude fluctuations while the pilot was sounding confused and distressed in radio transmissions, controllers failed to recognize and treat the situation as an emergency.

The Safety Board pointed to an accident on January 4, 2013, in which a Beech H35 lost all power while on approach to Flagler County Airport in Palm Coast, Florida. The pilot had earlier advised air traffic control of vibrations from the propeller and problems with oil pressure. Although it was close to the airport, the controller assumed the airplane still had some power and vectored it away from the airport. All three onboard were killed in the crash.

The fourth accident cited by the NTSB was on December 16, 2012. The pilot of a Piper PA-28-160 was on an IFR approach in instrument conditions when he started having trouble staying on course and advised the controller he was "no gyro." The controller didn't understand the implications and urged the pilot to try another instrument approach rather than vectoring him to VFR conditions. The pilot was killed when the plane crashed at Parkton, North Carolina.

The fifth accident involved a Beech V35B, which made a forced landing at Effingham, South Carolina, on August 11, 2012. The airplane entered an area of severe weather and the pilot lost control. The controller failed to warn the pilot about the weather and didn't know what the pilot meant when he radioed that the airplane's "AHRS" (attitude and heading reference system) had failed. The pilot got the airplane under control and executed an emergency landing to a field. Although the airplane was damaged, both occupants escaped injury.

If your blood pressure hasn't yet jumped up a few points, consider that what we're talking about here isn't cost overruns on new computer systems or bureaucratic foot dragging on reforming medical certification. The NTSB has spent years trying to get the FAA to do something about what it believes is a genuine safety deficiency, which has resulted in multiple failures of controllers to perform effectively. If you look beyond the government prose in the Safety Recommendation Report, which sometimes is used to soften or even obfuscate something that's hard to swallow, you can conclude that the FAA's failure in training controllers is at least partially responsible for the deaths of pilots and their passengers.

"Here's where it looked as though the pilot was becoming confused. He radioed, ’Established if I could have vectors to final, please.' Of course, if you're established on the localizer, you don't need vectors to final; you're on it."

As if to underscore the importance of its recommendations, on September 1, 2016, the NTSB issued a Safety Alert directly to pilots. It urged pilots not to hesitate declaring an emergency, but cautioned, "If pilots are not specific and only provide ATC with minimal information about the situation, ATC might not fully understand the seriousness of the situation and, as a result, may be unable to provide timely, relevant assistance or recognize an emergency." In other words, don't assume that a controller will be able to detect when you're in trouble from your voice or actions. Speak up in words of one syllable, and be specific about the kind of help you need from the controller.

So, where does the Bonanza accident near Kernersville fit in? There's a coincidence in the timing: The Probable Cause was adopted a day before the Safety Alert was issued and six days after the Safety Recommendation Report was published. And, the scenario and outcome reinforce the doubts the NTSB has as to the adequacy of the FAA's training offered controllers in handling emergencies. The Probable Cause states that the Kernersville accident was due to "...spatial disorientation, which resulted in an aerodynamic stall/spin. Contributing to the accident was deficient air traffic control training on recognition and handling of emergencies, which led to incorrect controller actions that likely aggravated the pilot's spatial disorientation."

The six-seat Beech A36 was manufactured in 1981. It was powered by a Continental IO-520-BB engine, producing 285 horsepower. It had a three-blade constant-speed propeller. The pilot had purchased the Bonanza about two weeks before the accident. It was equipped with a glass panel setup, including GPS, autopilot and flight director.

The glass panel was new to the pilot. A lot of his 800 hours of flight experience, especially when IFR, had been in a Piper PA-28-180 Cherokee that didn't have modern avionics. A flight instructor with whom the pilot had regularly flown told investigators he gave the pilot a checkout in the Bonanza, which included six hours of flight time and one-and-a-half hours of ground instruction. He said he demonstrated how to use the electronics for instrument approaches, but this was done VFR and the pilot wasn't under the hood. The instructor reported he had flown IFR with the pilot in the Cherokee, and thought he was a very good and conscientious pilot.

The airplane was registered to a corporation in Paxinos, Pennsylvania. Three days before the accident, the pilot flew to Florida with his wife, father-in-law and a friend onboard. The friend wouldn't be making the return flight to Pennsylvania. The pilot had asked his flight instructor to come along, but the instructor couldn't go, and warned the pilot not to fly in actual IFR conditions. The trip to Florida was uneventful.

On the morning of the accident, the pilot phoned flight service and filed an IFR flight plan from Sarasota-Bradenton International Airport (KSRQ) to Piedmont Triad International Airport (KGSO), Greensboro, North Carolina. The flight was supposed to take 3 hours and 30 minutes. The pilot asked for a hazardous weather briefing. The briefer reported there was an AIRMET for IFR conditions over central and northern Florida, Georgia, South Carolina and North Carolina. The pilot asked about thunderstorms nearing Sarasota, and tried to end the briefing so they could get started and beat the storms. The briefer explained he had more adverse weather to report, and the pilot shouldn't cut him off. The pilot agreed to let the briefer continue, but soon ended the call.

The flight departed Sarasota at about 8:19 a.m. Most of the flight was flown at 9,000 feet MSL, an altitude where some pilots might benefit from supplemental oxygen. Things went smoothly until the flight was nearing the destination, at about 11:33 a.m. The pilot was handed off to Greensboro Approach Control and reported to the controller handling the south radar position that he was descending through 5,700 feet. The controller advised that ATIS Delta was current, the altimeter setting was 30.19, and the pilot should expect an ILS to Runway 5 Right. The controller then told the pilot to report receiving information Delta. The pilot responded that he had Delta, and asked if he'd be landing on Runway 5 Left. The controller repeated that the runway was 5 Right. A few seconds later, the pilot asked if he was cleared down to 5,000 feet. The controller radioed that if the previous controller had cleared him to 5,000 feet, that was still his assignment.

At 11:35:19, the pilot asked for clarification if he should expect Runway 5R or 5L. The controller replied, "The right side, R-I-G-H-T." The pilot responded, "Five Right." The controller then told the pilot to fly a heading 360 for sequencing, and the pilot confirmed. The controller working the south radar position then handed off the flight to the west radar controller.

At 11:45:24, the west radar controller cleared the flight down to 3,000 feet and advised of traffic, type and altitude unknown. The pilot radioed back, "Roger, 36HT is about to go IMC." About 20 seconds later, the controller told the pilot that traffic was no longer a factor, and he should turn right to 010 degrees. The pilot read back the heading, then reported level at 3,000 feet and asked if he should stay there. The controller radioed, "Affirmative."

The controller then gave the pilot another course adjustment, then cleared him for the ILS approach to Runway 5R. At 11:52:48, the pilot radioed the controller, "How do you like this route of flight?" The controller responded that he was a bit right of course and to turn to 360. The pilot asked whether he should turn left or right, and the controller radioed, "...left to 360." A minute later, the controller asked if the pilot was established on the localizer.

Here's where it looked as though the pilot was becoming confused. He radioed, "Established if I could have vectors to final, please." Of course, if you're established on the localizer, you don't need vectors to final; you're on it. The controller again asked, "Are you established on the localizer?" The pilot said he believed he was, but the controller responded, "Now you look like you just actually went through the localizer." The pilot again asked for a vector to final. The controller responded by canceling the approach clearance, and telling the pilot to maintain 3,000 feet and turn left heading 320 degrees for sequencing.

The controller directed the pilot to turn left to 230 degrees for vectors to the ILS Runway 5R approach. About 45 seconds later, he asked the pilot to verify that he was flying a heading of 230 degrees. The pilot replied, "Negative, and 36HT is close to..." The controller then took a phone call from another controller, which was interrupted when the pilot radioed again asking for vectors. At 11:57:57, the controller radioed, "Bonanza 36HT, you need to be level at 3,000, altitude shows you at 2,500, Greensboro altimeter 30.19." The pilot responded that he was climbing to 3,000 feet. The controller asked him to verify his heading, but there was no response. At 11:59:17, the controller asked the pilot to verify his heading. The pilot replied that it was 166 degrees and said, "We need a descent, we are almost disoriented...."

The controller asked the pilot if he could accept a no-gyro turn to final. The pilot said he could. At 11:59:43, the controller told him to start a left turn, then to maintain 2,500 feet, then, at 12:00:16, to start a right turn. At 12:00:47, the controller radioed a low-altitude alert to the flight and asked the pilot for his altitude. There was no response, but the airplane was observed on radar completing a 360-degree turn. The investigation didn't establish whether the pilot was hand-flying or using the airplane's autopilot, but a properly operated autopilot and navigation system should have had no problem tracking a localizer.

At 12:01:24, the controller told the pilot to "...climb and maintain 4,000, I can see if I get you back up into the, uh, above the clouds." The controller followed with, "N36HT Greensboro approach, you up? Last altitude showed 2,100." The pilot answered with the tail number, in what sounded to investigators like an agitated voice. The controller answered, "N36HT just climb and maintain 4,000, I'll block altitude for you." The controller added, "Last tops were reported at 3,500, if I can get you up to 4,000, maybe you can square it back off." The airplane's altitude, however, remained at 2,100 feet. At 12:02:54, the pilot radioed, "Is there a nearby field for 36 ho-" The controller responded that the nearest airport was Greensboro, "...off your left wing and off your current heading 7 miles. Right now showing 1,000 overcast for the ceiling." At 12:03:21, the controller asked for the flight's altitude. At 12:03:34, screams are heard on the radio. The airplane had crashed next to a rock quarry, about 6 miles southeast of the airport.

When investigators asked the west radar controller about no-gyro procedures, he said he wasn't familiar with standard rate turns, but knew to watch the aircraft's track and call a "stop turn" when the aircraft appeared to be close to the desired heading. He said he wouldn't consider a VFR pilot in instrument conditions to be an emergency situation unless the pilot thought so. He said he had never worked a situation similar to the accident situation, but had worked emergencies in the past. He said he thought this situation was a priority, but not an emergency. He couldn't recall the last time he received training on emergencies or determining when an emergency exists. He said he didn't know how to use emergency obstruction video maps, which show controllers the locations of mountain peaks, man-made obstructions and the location of hospitals.

After this accident, the KGSO ATC Facility's Safety Committee, composed of union and management personnel, developed some training on their own. All personnel received a briefing on no-gyro vectors, presentations on handling aircraft emergencies and equipment failures, a presentation specifically focused on VFR aircraft emergencies, and a review of disseminating pilot reports and weather information.

Nationally, the NTSB wants the FAA to use standardized, scenario-based, recurrent training to be sure controllers recognize emergencies, determine what help pilots need and take actions that help pilots safely resolve a situation. It says this should include lessons learned from events including power losses and fuel emergencies; control difficulties; loss of navigation capability; loss of flight instruments or other critical aircraft systems; weather encounters; and pilot medical issues, such as hypoxia.

Was it naive to believe the FAA was already doing all of this?

Peter Katz is editor and publisher of NTSB Reporter, an independent monthly update on aircraft accident investigations and other news concerning the National Transportation Safety Board. To subscribe, visit www.ntsbreporter.us or write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.

Want to read more proficiency analysis from Plane & Pilot? Visit our After The Accident archive.

Peter Katz is editor and publisher of NTSB Reporter, an independent monthly update on aircraft accident investigations and other news concerning the National Transportation Safety Board. To subscribe, visit www.ntsbreporter.us or write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.
Comments(0)

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


Subscribe to Our Newsletter

Get the latest Plane & Pilot Magazine stories delivered directly to your inbox

Subscribe to our newsletter