Nightmare Fire In An Air Ambulance

On April 9, the NTSB adopted the probable cause and issued its final report on the inflight fire and crash of a Piper PA-31T Cheyenne II at McKinleyville, California, in which all four on board were killed. Why did I begin by pointing out the date that the agency completed its work on this accident? So that I can point out that the Safety Board was doing its job of proactively promoting aviation safety long before the report was issued. The FAA, too, started to move once it became clear that serious safety issues were being uncovered, so serious that it would be a mistake to wait for the investigation to run a typical two-year course.

The wreckage debris path was 2,400 feet long, consistent with an inflight breakup. The forward fuselage showed damage consistent with an inflight fire. There was fire damage in the floor area between the two front seats near the main bus tie circuit breaker panel. There was some tearing of the aluminum structure, consistent with the metal being weakened by thermal exposure. Medical examination showed that the pilot had been exposed to fire from the right. The right leg of his flight suit was damaged from heat, part of his right boot was melted, and his right hand and right lower leg had been burned.

As they worked on the McKinleyville accident, NTSB investigators decided to examine two other PA-31T airplanes, and FAA investigators examined four more, and they found some of the same things they were uncovering in the accident investigation: damaged electrical wiring.

The accident occurred on July 29, 2016. By December 16, 2016, investigators had confirmed enough for the FAA to issue a Special Airworthiness Information Bulletin (SAIB) alerting owners and operators of PA-31T's that the condition of all electrical wiring in the floor-mounted circuit breaker panel area needed to have a special inspection at the next scheduled maintenance.

Within days, the NTSB became convinced that the potential hazard was so serious that the FAA's SAIB wasn't enough and what was needed was an Emergency Airworthiness Directive. The NTSB noted that there had been an April 12, 2015, accident involving another PA-31T in which the pilot had reported smoke in the cockpit to ATC shortly before it crashed, killing all four on board, but the wreckage didn't yield enough evidence to determine what went wrong.

So, on January 5, 2017, the Safety Board issued a Safety Recommendation pegged to the McKinleyville accident, calling on the FAA to issue an Emergency AD. On February 7, 2017, the FAA issued AD 2017-02-06. But, being the FAA, it allowed operators to keep flying their PA-31T's without having to do anything until the AD became effective on February 22, and even after that to keep flying for another 30 days until they ran up against the deadline for AD to be accomplished.

The probable cause in the Safety Board's McKinleyville report reflected what investigators had figured out comparatively early on with the help of on-scene examination of the wreckage as well as laboratory analysis of recovered bits and pieces. The accident, said the NTSB, was due to an inflight fire in the floor area near the main bus tie circuit breaker panel that resulted from chafing between an electrical wire and a hydraulic line and/or airplane structure.

The evidence showed not only damaged and melted wiring and damage to the circuit breaker panel but also damage and melting of the four nearby landing gear hydraulic lines. All the lines showed signs of heat exposure and melting and were missing sections of material. Missing sections of lines provide a way for hydraulic fluid to flow out and fuel a fire.

The airplane was being operated as an air ambulance flight by Cal-Ore Life Flight under Part 135. At the time of the accident, the company operated seven PA-31T aircraft and employed a dozen pilots. It had three bases in northern California. According to the company's website, its CEO is an ATP-rated pilot, is qualified to fly helicopters and is a certified EMT. He also is a flight instructor, Part 135 check airman and FAA Safety Team Representative. The company's management includes a chief pilot and a director of quality assurance for aircraft maintenance. It also has a director of maintenance and an air medical clinical manager.

The accident airplane was manufactured in 1981. It was powered by two Pratt & Whitney PT6A-135A turboprop engines producing 620 horsepower, each turning controllable pitch propellers. The airplane had 7,309 total hours, and the engines each had 3,712 hours since new. The props had been overhauled 1,662 flight hours before the accident. The airplane was being maintained in accordance with a progressive inspection plan and had its last inspection 15 days before the accident.

The airplane had a 28-volt direct current electrical system. It had been retrofitted with a stretcher system for carrying patients, which had been installed in accordance with a supplemental-type certificate. It required the installation of two additional circuit breakers to the airplane's main bus tie panel, one rated at 30 amps and the other rated at 50 amps. The breaker panel, located on the floor between the front seats, already had 12 circuit breakers rated from 50 amps to 200 amps. The Safety Board found that the majority of the wiring of the stretcher system was not heat or fire damaged; the insulation on one of its wires was discolored from heat in two spots. The Safety Board did not cite the stretcher installation as playing a role in starting the fire.

The landing gear hydraulic fluid specified for use in the airplane was MIL-H-5606, the military specification for an older fluid that is mineral oil-based and highly flammable. The military and others now use hydraulic fluids (MIL-PRF-83282 and MIL-PRF-87257), which have higher flash points and will not keep burning once the ignition source is removed. They're billed as being compatible with 5606. So, why isn't there a mandate to eliminate MIL-H-5606 from aircraft? Good question.

The single-pilot of the accident airplane was 54 years old and held an airline transport pilot certificate for multi-engine land airplanes. He had a commercial certificate for single-engine land airplanes and rotorcraft-helicopter. He had a type rating for Beech 1900 airplanes and was an instructor for single-engine, multi-engine, instruments and helicopters. He had 7,300 hours with 125 in type. His second-class medical certificate was current and required him to have glasses available for near vision.

The flight departed Jack McNamara Field (KCEC) at Crescent City, California, at about 12:40 a.m. The destination was Oakland International Airport (KOAK). The flight was carrying a 35-year-old female patient being transferred from Sutter Coast Hospital in Crescent City to the California Pacific Medical Center in San Francisco. On board were a 49-year-old female flight nurse and a 30-year-old female certified flight paramedic.

The flight used the call sign Lifeguard November 661TC. According to FAA information, at 12:41:38, the pilot radioed Seattle Center, "Medevac six six one tango charlie with you, off Crescent City, looking for IFR to Oakland." The controller responded and told the pilot to squawk code 6625. The pilot confirmed and, at 12:42:50, the controller radioed, "...radar contact a mile west of the Crescent City VOR, say altitude." The pilot replied, "that checks, six thousand four hundred in the climb, six six one tango charlie."

At 12:43:11, the controller cleared the flight to Oakland as filed, and she told him to "climb and maintain one five thousand and, ah, Crescent City altimeter two niner eight eight, can you say altitude leaving?" The pilot responded, "Okay, ah, two niner eight eight, cleared to Oakland as filed, ah, climb and maintain one five thousand, and we're just passing through five thousand two hundred in the climb for six six one tango charlie."

At 12:43:38, the controller radioed, "Cheyenne one tango, I'm sorry, medevac one tango charlie roger, that checks out now, thanks."

There were no further radio calls until 12:58:13, when the pilot radioed, "And Seattle Center, Cheyenne six six one tango charlie, we're going to turn back to Crescent City, we're got, ah, we're smelling smoke in the cockpit." The controller responded, "Cheyenne six six one tango charlie, roger, cleared direct Crescent City or cleared to Crescent City via direct and descend at pilot's discretion, maintain niner thousand and, ah, I'll, ah, if you need anything else let me know."

At 12:58:42, the pilot radioed, "Okay, looks like we're going to lose some power here shortly, so I'll try and keep you posted as long as I can, six six one tango charlie." The pilot did not specify whether he was talking about electrical power or engine power, but the reference to keeping the controller posted could be interpreted as him realizing something electrical was going wrong and his radios would soon go out.

At 12:58:53, the controller radioed, "...the Crescent city altimeter two niner eight eight and, ah, if you could give me fuel remaining and people on board, um, and, uh, anything else you need."

After a pause of about seven seconds, the pilot radioed, "Yeah, we've got smoke in the cockpit, declaring an emergency, we're depressurizing and, ah, heading back to Crescent City, call the fire department, please."

At 12:59:24, the controller said, "Cheyenne one tango charlie, wilco, how many people on board?"

According to FAA radar data, from about 12:58:14 until this point, the airplane's Mode C altitude had fluctuated between 14,900 feet and 15,400 feet. At 12:59:26, the Mode C radar returns were lost, likely indicating that the electrical system was in the process of being seriously compromised.

At 12:59:28, a transmission from the pilot was received, saying, "Three on board." Within about 10 seconds of that transmission and for the next minute and 12 seconds, primary radar returns indicated the airplane was on a northwest heading, then entered a left turn to a southeast heading. From about 1:01:26 to 1:02:02, the airplane flew on the southeast heading, then turned heading west. At 1:01:54, the controller asked, "...would you like to try and land at Arcata?" There was no reply. At 1:02:21, the controller radioed, "...I'm not getting at any transmissions from you, would like to go to Arcata instead of, ah, Crescent City, I see you turning around."

At 1:02:50, the airplane's primary return was lost from radar. It was about 54.5 miles south of Crescent City and about 6 miles northeast of the Arcata Airport. The 1:07 a.m. weather observation at Arcata was low IFR, with wind from 180 degrees at 4 knots, visibility 1/2-mile in mist, ceiling 200 feet overcast, temperature 55 degrees F., dew point 54 degrees F., altimeter 29.85 and Runway 32 visual range 4,500 feet variable to 6,000 feet. Saying the odds were against the pilot making a safe landing in those conditions with a fire in progress would be an understatement at the very least.

The airplane was equipped with a handheld fire extinguisher and smoke goggles. Did the pilot kill whatever remained of the electrical power, don the smoke goggles and try to use the extinguisher? I can't tell from the NTSB's report; maybe there wasn't enough evidence for investigators to figure out the pilot's actions.

About 15 minutes after Lifeguard November 661TC disappeared from radar, another Cal-Ore Life Flight Cheyenne came up on frequency. The controller asked that pilot for the phone number of someone to contact at the company and also asked if the pilot would divert to the area of the Arcata Airport "and, um, just tell me if you see anything." The controller was hoping that the missing airplane had landed at Arcata, but the weather made the idea of observing anything from altitude impractical. At 2:10 a.m., an Alert Notice was issued by the FAA. U.S. Air Force Search and Rescue, the California Highway Patrol, the U.S. Coast Guard and the local sheriff's office joined in an extensive search effort, which found and identified the wreckage by 10:30 in the morning.

Although the terrain was heavily forested, the wreckage was in an area covered with brush. While the forward section of the fuselage between the front seats showed evidence of heat damage and soot was found on the fuselage's skin, evidence indicated that only the inboard sections of both wings had been exposed to fire that erupted after the crash. Some other pieces of wreckage, found away from the main cabin area, also showed signs of exposure to post-crash fire.

The Safety Board did not explain why it thinks the airplane suffered an inflight breakup other than large pieces of wreckage being widely separated from one another. Investigators did find evidence of some localized compromising of aluminum structure due to the fire, such as signs of a stringer (a part of the structural skeleton) being heated to almost the melting point before it failed. But we don't know from the report whether the pilot was maneuvering to get down quickly and exceeded safe operating parameters, whether he was being burned so badly that he was no longer able to keep the airplane under control, whether the aircraft structure had been weakened, or something else.

A witness who was camping told investigators about hearing an airplane circling overhead and then going toward the west, followed by about 15 seconds of silence, then seeing a flash followed by a second flash and a loud rumble.

The wreckage was moved to a facility in Pleasant Grove, California, for detailed examination. One thing they ruled out as a cause of the fire was the combustion heater in the airplane's nose. A large wire bundle that ran through the center tunnel in the floor of the cockpit area had a section where the insulation was melted, discolored or missing.

At the NTSB's Materials Laboratory in Washington, multiple pieces of wiring and hydraulic lines were examined. An electron microscope, electron dispersive spectroscopy and other techniques were used. The lab produced photographs showing wiring damage and melted metal in extreme close-up.

The evidence allowed the NTSB to pinpoint a wire that had chaffed on a hydraulic line to the point that its insulation was worn away and where electrical sparking occurred. The FAA, for one, advises against electrical wires and lines containing flammable fluids being allowed to come into contact and calls for a minimum of 1/2-inch clearance.

While this accident did result in corrective actions and repetitive inspections being ordered for PA-31T airplanes, it also represents something of a opportunity the NTSB might use to renew its crusade for aging aircraft awareness throughout the entire fleet. And, it should be a reminder for us to tell our mechanics to pay particular attention to the security and condition of aging electrical wiring inside of the aging airplanes we fly.

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 ntsbreporter.us or write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.