The Silent Killer
The NTSB’s latest safety recommendation targets the dangers of carbon monoxide leaks caused by defective exhaust systems
Against the background of an aging fleet of general-aviation, piston-powered airplanes, the NTSB suggested that it's time for the Federal Aviation Administration (FAA) to take a closer look at engine mufflers and do more to eliminate potential hazards posed by mufflers that have deteriorated. So on June 23, the NTSB issued a safety recommendation to the FAA dealing with the subject. NTSB safety recommendations often are the first step along the path leading to FAA regulatory edicts, such as airworthiness directives.
In its recommendation, the NTSB expressed that although it's routine for engine-exhaust systems to be periodically examined using visual inspection or pressure testing when called for, defects sometimes go unnoticed. Cracks and pinholes can allow carbon monoxide (CO) and other exhaust gases to leak into the cabin, causing one part of CO per 20,000 parts of air to seep into a person's system.
The effects are dangerous. They include shortness of breath, headache, fatigue, nausea, disorientation, unconsciousness and, ultimately, death. In addition, hot exhaust gas escaping from a bad muffler or other part of the exhaust system into the engine compartment can damage hoses, wires and other components.
The recommendation revealed that the NTSB examined its own accident database and found that over the last 30 years, there were 125 accidents or incidents involving muffler failure on single-engine, piston airplanes that resulted in 42 fatalities and 27 serious injuries. When the NTSB looked for accidents involving CO poisoning, it found 58 accidents or incidents. These resulted in 84 fatalities and five serious injuries. In addition, the NTSB looked at the FAA's database of service difficulty reports filed by mechanics. Between 1974 and 2001, the FAA received 232 reports of cracked or leaking mufflers. Many of the reports noted that the cracks or signs of deterioration couldn't be detected through a surface visual inspection.
By submitting its safety recommendation, the NTSB wants the FAA to evaluate the methods that are currently used for inspecting mufflers and exhaust systems, and determine whether or not they're sufficient. Furthermore, the NTSB thinks that the FAA should require replacement of exhaust systems in general-aviation aircraft after a certain time in service and the installation of carbon-monoxide detectors on all single-engine reciprocating-powered airplanes with forward-mounted engines and enclosed cockpits. In addition, the NTSB wants the FAA to establish standards for detectors.
There are several types of CO detectors that are currently available. But whichever type of CO detector you purchase, there's no denying the fact that bringing one with you in the cockpit is cheap insurance, especially when flying in winter with an enclosed cabin and the heat blasting away.
One of the accidents cited by the NTSB in its safety recommendation involved a Piper PA-28-236. On January 17, 1997, at 1:35 p.m. EST, the single-engine airplane was destroyed when it collided with trees near Alton, N.H. The pilot and one passenger were fatally injured. Visual meteorological conditions prevailed. The personal flight was conducted under Part 91. The flight originated from Farmingdale (FRG), N.Y., at approximately 11:15 a.m. The intended destination was Saranac Lake, N.Y.
According to the air-traffic-control records, at 11:17, the pilot radioed New York ATC, advised the controller that he was 4 ½ miles north of FRG at 3,000 feet and requested flight following. The flight then proceeded uneventfully for the next 25 minutes.
At 11:39, the controller radioed the pilot and instructed him to contact the Boston Air Route Traffic Control Center (ARTCC), but there was no response. At 11:41, the private-pilot-rated passenger transmitted that they were in trouble and that the pilot wasn't responding. The controller instructed her to do a shallow left turn southbound and told her that he would talk her down to an airport. He also advised her to descend if she could and to stay clear of the clouds.
The passenger advised the controller that she could see the ground while making the turn and that the pilot was throwing up. The controller informed the passenger that she was over the Danbury Airport in Danbury, Conn. During the next 16 minutes, the controller and the pilot of another airplane provided directions and instructions to the passenger.
At 11:56, the passenger radioed that she was getting tired and that she was nauseated. The pilot of the other aircraft asked the passenger to lower the nose of the airplane and to start descending, and the passenger responded that she was trying to "activate" the pilot. The pilot of the other aircraft told the passenger where the control wheel was and directed her on how to trim the airplane.
At 12:04, the controller advised the pilot of the other aircraft that the Piper had turned northbound and that it was slowly climbing. The controller said that the passenger was responding with transponder idents.
Three minutes later, the controller advised the pilot of the other airplane that the Piper was at 8,200 feet and that there were no further idents. At 12:09, he noted that the airplane was holding its heading and altitude of 8,800 feet. Then, the controller radioed the passenger and told her to open a vent on the airplane and get some fresh air, as there might be CO in the cockpit. He also announced that maybe the autopilot was on, making it difficult to descend. The transmissions weren't acknowledged.
At 1:13, the pilot of the other aircraft reported that he had the airplane in sight. Shortly thereafter, he noticed that there was no one who was sitting up, the entire cabin was full of smoke and there was smoke coming out of the engine.
Ten minutes later, the pilot reported that the airplane was descending rapidly. At 1:26, he reported that the Piper had crashed into trees.
The accident pilot held a private certificate and was instrument-rated. His logbook was incomplete, but investigators determined that at the time of his last medical examination, about 18 months before the accident, he had a total of 700 flight hours. The passenger held a private certificate. Her current flight times weren't determined.
All major components of the airplane were accounted for at the accident scene. The rear, fresh-air vents were in the closed position. One front vent was closed while the other was partially opened. The overhead vent knob and the overhead vent were in the open position, and the panel-mounted fan switch was in the high position. The heater control box remained attached to the fuselage and was in the open position. The two forward-floor heat vents were also in the open position.
The engine was removed from the wooded area and examined in a hangar at the airport in Wolfeboro, N.H., under the supervision of the NTSB. When the muffler shroud was removed, investigators saw a large hole in the muffler that measured approximately 4 ½ inches long and one inch wide.
The NTSB's metallurgist found that the inlet end of the muffler case on the left side was crushed and contained a large crack that extended approximately 2â3 of the muffler circumference. Most portions of the crack were darkly discolored, as if they had been exposed to exhaust gases for a period of time. Other areas were shiny, as if they were recently created.
The toxicological testing report from the FAA Toxicology Accident Research Laboratory in Oklahoma City showed that the pilot's lung fluid had a CO saturation of 43%, while the passenger's blood was at 69%. To give you an idea of how much CO saturation is too much, generally, a person breathing air with a 0.5% concentration of CO for 30 minutes will obtain a blood-saturation level of 45%.
The maintenance manual published by the airplane's manufacturer stated, in part, that "the entire exhaust system, including the heat-exchange shroud, muffler, muffler baffles, stacks and all exhaust connections must be rigidly inspected at each annual or 100-hour inspection. The possibility of exhaust system failure increases with use. The system also must be checked carefully before winter operation when the cabin heater will be in use." The manufacturer explained that if any component can't be inspected visually, a pressure check or a ground check using a carbon-monoxide detector should be conducted. This involves running the engine at full-static RPM with the cabin heat valves open and taking air samples at several locations within the cabin. The manufacturer says that if the detected CO concentration exceeds 0.005%, the muffler must be replaced. The NTSB determined that the probable cause of the accident was an exhaust gas leak because of inadequate maintenance, which resulted in carbon-monoxide poisoning, and the incapacitation of the pilot.
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, write to: NTSB Reporter, Subscription Dept., P.O. Box 831, White Plains, NY 10602-0831.
Subscribe to Our Newsletter
Get the latest Plane & Pilot Magazine stories delivered directly to your inbox