Safety Case Studies

Safety Case Studies examines aviation accidents that involve maintenance as either a probable cause or factor. The information in Safety Case Studies is obtained from National Transportation Safety Board reports, regulatory au-thorities and accident investigation bureaus, and research by the Aviation Maintenance staff. The information in Safety Case Studies is presented here to help the aviation maintenance industry learn from accidents and incidents and prevent future mistakes.

Wrong Fuel Gauges Installed
Tuninter ATR-72, ditching off Palermo, Sicily, August 6, 2005. Onboard: 35 passengers, four crew. Fatalities: 15 passengers, one flight attendant. On a flight from Bari, Italy, to Djerba, Tunisia, the Tuninter ATR-72 had a double-engine flameout, and the pilots were forced to ditch the airplane in the ocean about 23 nautical miles northeast of Palermo.

Italy's Agenzia nazionale per la sicurezza del volo (ANSV--accident investigation bureau) reported that the incorrect fuel quantity indicator (FQI) was installed in the instrument panel of the crashed ATR-72. The indicator was an Intertechnique part number 749-158, which is designed for the smaller ATR-42. "The FQIs for the ATR-72 and ATR-42 have the same dimensions and installation interface," according to the ANSV. An FQI for the ATR-42, therefore, could be installed erroneously on an ATR-72 model, and vice versa. The only visible difference between the two models is that the fuel quantity figure is printed in small white letters above the digital display for each gauge. The letters on the ATR-42 FQI read L.TK: 2250 and R.TK: 2250, and on the ATR-72 FQI, L.TK: 2500 and R.TK: 2500.

The ANSV tested ATR fuel systems and found that with an ATR-42 FQI installed on an ATR-72, the resulting error is "non-conservative," meaning that the fuel quantity onboard the aircraft is less than what is indicated on the FQI. "With zero fuel on board," the ANSV reported, "the FQI ATR-42 readings for each tank is 900 kg (e.g. total fuel on board indicated by the FQI is not less than 1,800 kg)."

While the accident investigation is not yet complete, the ANSV issued this report as a safety recommendation to the European Aviation Safety Agency (EASA) requiring a fleet inspection of all ATR-42s and ATR-72s to make sure the correct FQI is installed and suggesting mandating a modification of the ATR FQI installation to prevent confusion. In early September, ATR issued two all-operators messages to inform them of the correct procedures in the ATR-72 maintenance manual. If followed, these procedures don't allow installation of the incorrect FQI.

See the ANSV recommendation resulting from this accident at: www.ansv.it/cgi-bin/eng/TS-LBB%20RS%20ENG.pdf

Ignored AD Causes Fatal Crash
Hughes 269A accident Louisburg, North Carolina, May 14, 2004. Onboard, one pilot, one passenger. Fatalities: one passenger. The flight was conducted as part of a drug-eradication program, according to the pilot. The National Transportation Safety Board (NTSB) noted that the pilot was not able to show he was FAA certificated or that he had obtained helicopter training, and that this was a contributing factor in the accident.

After takeoff, the pilot told the NTSB, "he felt two vibrations and then heard a loud bang from the back of the helicopter. He stated the helicopter yawed to the right, and he entered an autorotation and maneuvered toward a small clearing in the wooded area. The main rotor blades struck a pine tree and impacted the ground on the helicopter's left side."

The NTSB found that, though the helicopter had undergone an annual inspection two months prior, there was no record of compliance with an old airworthiness directive--AD80-05-05 on the tailboom saddle fitting. The NTSB investigators found that the helicopter's saddle attachment was broken and the "tailboom support fitting was fractured as a result of fatigue initiated at the base of a large corrosion pit on the forward wall's tip surface." This surface, which was in direct contact with the tailboom, was only partially covered by paint and contained extensive pitting damage. The NTSB determined that the accident's probable cause was "fatigue failure of the tailboom saddle fitting, which resulted in loss of aircraft control, and there was non-compliance to an airworthiness directive requiring inspection of the saddle fitting."

The NTSB report is available at: www.ntsb.gov/ntsb/brief.asp?ev_id=20040519X00615&key=1

Vacuum Pump Failure is Fatal
Piper PA32R-300, inflight breakup, near Rachel, Texas, September 9, 2004. Onboard, one commercial pilot. Fatalities: one. The Piper Lance was being flown on a night cargo run. The National Transportation Safety Board (NTSB) found that the vacuum pump had failed, causing a loss of suction to the gyroscopic instruments.

The NTSB reported that the Piper's vacuum pump had been overhauled and installed on May 7, 2004, slightly less than 300 tachometer hours prior to the accident flight. A measurement of the wear on the pump's vanes, however, showed that the vanes probably had been operating for approximately 1,380 hours, based on pump manufacturer Parker-Hannifin's vane-wear rule-of-thumb of 0.025 inches of wear per 100 hours of operation.

The NTSB report noted that Parker-Hannifin reiterated its recommendation that this particular model vacuum pump--a 211CC--be replaced after 500 hours of operation or six years from the date of manufacture.

The NTSB report on this accident is available at: www.ntsb.gov/ntsb/brief.asp?ev_id=20040928X01501&key=1

B-nuts Cause Engine Losses
National Transportation Safety Board investigator Edward Malinowski is concerned about two engine failures in turbocharged Cessna 210s caused by loose turbocharger pressure oil line B-nuts.

The first accident occurred on March 2, 1999. The loss of oil caused the engine to seize and the propeller bolts to fail, allowing the propeller to fall off. On September 27, 2004, an engine failure resulted in a forced landing and substantial damage to the airplane and serious injuries to a flight instructor. Malinowski (malinoe@ntsb.gov) has investigated both cases and found other cases in FAA records. He hopes to spread the word about this problem so that mechanics are aware of the loose B-nut problem.