Go-Around Gone Bad

Unfortunately, the Safety Board, as is too often the case, took a simplistic approach in its probable cause statement and never really did provide the kind of analysis that's helpful to pilots seeking to learn from the unfortunate experiences of others. The Board determined that the Cessna 340A crashed because of the pilot's failure to maintain airplane control due to spatial disorientation in low-visibility conditions while maneuvering during a missed approach. It said that the contributing factor was the pilot's ineffective use of the onboard GPS equipment.

The NTSB has been wringing its collective hands lately because general aviation accident rates have yet to be reduced to the extremely low levels achieved by the airlines. For example, the NTSB's stats for 2013 show that the Part 121 scheduled air carriers had 20 total accidents, only one of which involved fatalities, killing a total of two people. General aviation had 1,222 accidents, 221 of which involved fatalities, in which 387 people in the aircraft and on the ground were killed. That works out to 0.117 accidents per 100,000 flight hours for the air carriers, compared with 5.85 accidents per 100,000 flight hours for general aviation.

The Safety Board did some analysis of its accident files, and found that between 2008 and 2014, about 47% of fixed-wing general aviation accidents in which someone was killed involved the pilot losing control of the airplane. What came next shouldn't be a big surprise. The Safety Board made "prevent loss of control in flight in general aviation" a key feature of its "Most Wanted List of Transportation Safety Improvements" for 2016.

The Safety Board tells us that "pilots can reduce these accidents through education, technologies, flight currency, self-assessment, and vigilant situational awareness in the cockpit." It recommends additional training in both stall recognition and how to apply recovery techniques before a stall can become fully developed. It suggests installing, and getting trained in using, an angle of attack indicator. It suggests that the biennial flight review by itself is inadequate, and all pilots should take regular recurrent training. It says that pilots need to manage distractions so they don't interfere with situational awareness.

What the Safety Board doesn't seem to emphasize is that the stage often is set for a loss of control accident long before the aircraft actually is allowed to go out of control. A stall leading to an unrecoverable spin, for example, may be the culmination of a series of adverse events or faulty decisions, which finally came together with an obscenely high angle of attack. Getting the angle of attack wrong seldom happens in isolation, although the NTSB probably is right in thinking an angle of attack indicator could be a lifesaver if the pilot is lucky enough to look at it at the right time. Parenthetically, I fly with one and look at it only occasionally even though it's right there on the glareshield.

Had this been an airline accident, we surely would have been treated to a lengthy "analysis" section in the report exploring several scenarios and identifying the points at which the flightcrew could have conceivably altered the outcome. In this report, no such luck. It's almost as if the difference in the depth of accident reporting parallels the difference in accident rates between the airlines and general aviation. For now, if we want to help reduce the general aviation accident rate by not repeating the Cessna 340A crash, we'll have to figure out on our own how various elements stacked up to set the stage for the loss of control.

The Cessna 340A is a pressurized six-seat, twin-engine airplane that was first certified in late 1975. Just under 950 of the airplanes were built. Two Continental TSIO-520-NB engines power the 340A, each putting out 300 horsepower. The airplane has a maximum speed of 244 knots, stalls at 82 knots in landing configuration and has a service ceiling of 29,800 feet.

The airplane had been en route from Fort Lauderdale Executive Airport in Florida to Hampton Roads Executive on the day of the accident. The pilot, age 61, held a commercial certificate with multi-engine, single-engine and instrument airplane ratings. There were three passengers onboard. All four were killed when the plane crashed shortly after the pilot had missed the GPS approach to runway 10 at KPVG. Runway 10 is 5,330 feet long by 100 feet wide. KPVG is an uncontrolled field.

It's not as if the pilot was a novice. On his most recent application for an FAA second-class medical on June 3, 2013, he reported 4,256 total flight hours. On an application for insurance filed a month before the accident, he reported 5,541 total hours, with 3,076 in multi-engine aircraft, including 600 hours in the Cessna 340A. The Safety Board couldn't establish his instrument flight time, but did interview an instructor who had checked him out in the 340A a month before the accident. The instructor said he flew the airplane well, but they didn't perform instrument procedures. According to the instructor, the pilot didn't have experience before that flight in operating the airplane's Garmin GTN 750 GPS system. The Safety Board didn't gather information indicating how experienced the pilot really was with GPS systems. In trying to learn from this accident, we have to ask, "Why would the pilot accept a GPS approach if he had doubts he could fly it?" Airports near to KPVG, Chesapeake Regional (KCPK) and Norfolk International (KORF), both offered ILS, localizer and VOR/DME, in addition to their own RNAV (GPS) approaches.

KPVG is about 8 miles northwest of the accident site, at a field elevation of 28 feet MSL. The weather observed at KPVG at about the time of the approach was IFR, with an overcast ceiling at 500 feet, visibility 3 miles in moderate rain showers, and the wind 360 degrees at 9 knots gusting to 15 knots. The area forecast, AIRMETs, terminal forecasts and hourly observations would have alerted the pilot to expect to be in the clouds with possible light to moderate turbulence and possible thunderstorms. The NTSB found no evidence that he had received a weather briefing from Flight Service or through DUATS. It's possible the pilot received preflight weather information from some other source or was relying on in-flight updates displayed on the GTN 750 unit. The Safety Board didn't mention that as a possibility in its report. In our analysis, we remind ourselves that composite radar images sent to panel displays are time-delayed and never a substitute for real-time airborne weather radar.

Radar data showed that the airplane crossed the final approach fix at 1,200 feet MSL, which was 400 feet below the published minimum altitude for the crossing of 1,600 feet MSL. The pilot got down to the published minimum descent altitude of 420 feet MSL for GPS with only lateral guidance, which was 392 feet AGL. At the missed approach point, radar showed that the pilot began a missed approach. The pilot flew southeast into an area where weather radar showed there had been no precipitation. The missed approach procedure called for flight in the southwest direction, with a climbing right turn to 2,500 feet MSL. According to radar data, the flight had climbed to 1,100 feet MSL, then descended to 700 feet, then made an abrupt right turn and began a climb to 1,600 feet. From 1,600 feet, it descended to 1,000 feet and turned about 45 degrees left, which was the wrong direction for the missed approach procedure. Air traffic control called the pilot four times, but there was no response. The airplane subsequently reached 2,700 feet MSL in a climb, which was gradual compared with other maneuvers. At about 12:07:13, it turned right and began descending to 1,600 feet. It then abruptly climbed to 2,800 feet where it leveled off with a ground speed of 107 knots.

At 12:08:35, the pilot radioed ATC, "...we're on a miss and unfortunately in some really bad weather, uhm, is there an airport that's closer that's got better than 500-foot visibility?" Did the pilot mean to say "ceiling" rather than "visibility"? If so, were this and flying the missed approach to the southeast instead of southwest signs that he was already confused? We could have used some NTSB analysis of those points.

Radar data showed that the airplane's altitude, rate of turn and ground speed remained erratic. The controller asked the pilot to ident, and advised that Norfolk International was showing 1¼ miles visibility in light rain and mist with 500 feet broken, and that the airport was about "twelve miles to the northeast of your current position." A few seconds later, the controller radioed, "...you're radar contact six miles to the southeast of Hampton Roads at one thousand one hundred [feet], did you want to try to come in to Norfolk International?" At 12:09:12, the pilot responded, "Standby, we're fighting some bad weather, and it's causing us to lose altitude tremendously." By then, the airplane had completed a turn from 030 degrees to 231 degrees. At 12:09:38, the controller radioed, "...radar contact lost, say altitude." The controller made several calls to the 340A, but there was no reply. At 12:10:28, the controller asked an airline flightcrew to listen for an ELT, but they heard nothing.

An NTSB meteorologist reviewed the forecasts, pilot reports, weather observations and weather radar images applicable to the flight. Although there were light to moderate echoes over and around the accident site at the time of the crash indicating light to moderate rain, no lightning strikes were recorded. There was no convective activity nor heavy rain. There might have been some wind shear and light to moderate turbulence below 2,000 feet, but nothing as dramatic as the pilot's radio call at 12:09:12 suggests, according to the NTSB meteorologist. We could have used some analysis of whether the erratic altitude and flight path changes were due to weather phenomena, the pilot becoming disoriented, other factors or all of the above.

The NTSB's report glosses over a medical issue, which might explain a lot. I find this a bit odd since the Safety Board has been pushing for heightened medical fitness in all modes of transportation and hasn't gotten behind efforts to reform the FAA third-class medical requirement. The NTSB's Chief Medical Officer authored a report on the pilot, but it wasn't mentioned in the NTSB's published narrative on the accident. The medical report reveals that, in 1998, the pilot reported to the FAA that he had diabetes, but never reported using any medication to control the disease. In June 2011, when the pilot went to an FAA examiner for a second-class medical, his urine was loaded with sugar and the examiner deferred issuing a certificate. At that time, the urine strip reading was +3, the equivalent of a blood glucose level of 1,000 mg/dL, about 10 times normal. Follow-up tests found his blood glucose was 331 mg/dL (normal is 65-99) and his Hemoglobin A1C level was 13.1%. Normal for that test is below 5.6%. A week later, the pilot went to a different FAA examiner who issued his certificate, reporting that his urine test was normal. Yes, it's possible to control and even reverse diabetes. But, how likely is it to happen in a week? Could that testing by the first FAA AME have been an aberration? In this accident, we'll never know how the pilot was doing medically because the FAA never followed up on the information it had. If the pilot was flying the accident flight with poorly controlled diabetes, resulting in blood sugar levels behaving erratically in tense situations and high workloads, the pilot might have been in physical distress, which affected his ability to function and maintain aircraft control. We could have used some NTSB analysis on that score, coupled with a full look at his personal medical history. There likely are other factors in addition to those I've highlighted here, which might have been examined. For example, the NTSB had nothing to say about the pilot's activities the night before the accident, and whether he was sleeping well or had been fatigued. In a fact sheet on "loss of control" accidents, the Safety Board says that "all stakeholders should recognize the importance of their roles in the reduction of loss of control accidents." The NTSB could enhance its contribution to the cause by giving us more facts and more of its analytical expertise in GA accident reports like this one on the Cessna 340A.

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.