It was the middle of a short December day when a call came to a helicopter emergency medical service (HEMS) firm in Soldotna, Alaska, requesting transport for a patient to Anchorage from a clinic in Cordova. Soldotna is south and west of Anchorage; Cordova is 127 nm east. The helicopter would first have to position to Cordova before loading the patient. By the time the flight would be ready to leave there, the sun, which would have set before 4 o'clock, would be gone, and the long northern twilight almost ended. The moon, a waning crescent, would be down, and the unpeopled wilderness between the few towns along the route would be utterly dark.
Normally, the hospital in Cordova would have requested an airplane, which could go IFR, for the transfer; but all medical transport airplanes were already booked. This HEMS operator flew VFR only. Part 135, under which the flight would operate, requires that helicopters operating in Class G airspace under VFR have visual contact with the surface or, at night, with lights on the surface. Night vision goggles were used to comply with this requirement.
The 42-year-old pilot on duty, a recent hire with 2,700 hours but little experience of the Alaska winter -- and who had not flown to Cordova before -- had the right to decline the flight. He expressed some concern to a mechanic about the darkness and the unfamiliar route, and he told the flight nurse to be sure to bring along an extra pair of night vision goggles to help with terrain avoidance. He did not talk with an FAA briefer, but he did check FAA weather cameras (akweathercams.faa.gov/sitelist.php) along the route. The weather was generally VFR, and, after all, anyone might feel a few butterflies before a night flight in remote and unfamiliar surroundings. The pilot accepted the flight, and the Eurocopter BK117 left Soldotna at 1:40 p.m. for the 157 nm trip to Cordova. There were three aboard: the pilot, a male flight nurse and a paramedic. After embarking the patient at Cordova, the flight left for Anchorage at 4:40.
The first half of the way from Cordova to Anchorage is over the waters of Prince William Sound. Beyond the town of Whittier, one crosses a narrow neck called Portage Pass, whose elevation rises to about 700 feet above sea level. Once on the west side of the pass, one can again descend, if need be, almost to sea level, and follow the shoreline to Anchorage.
While the flight was en route in VFR conditions from Soldotna to Cordova, the weather around Whittier was deteriorating rapidly. Another helicopter, passing through the same area about an hour before the HEMS flight left Cordova, encountered "waves of snow squalls and near zero visibility." The pilot had to land several times to wait for weather to pass. He did not file a pilot report, however; and in any case it was not clear that the HEMS pilot rechecked the weather when he was at Cordova.
While en route, the HEMS pilot checked in with the hospital dispatch center at 10-minute intervals via satellite phone. His last report came from 25 miles east of Whittier; he was estimating Anchorage in 27 minutes. At that time, the weather observation in Whittier was 300 feet, sky obscured, with one mile visibility in light snow and mist.
The helicopter was not heard from again.
Hampered for several days by bad weather, searchers eventually located the body of the nurse, as well as a few pieces of the helicopter and the helmets of the nurse and the pilot, floating in an inlet called Passage Canal, three miles east of Whittier. Each helmet had a pair of night vision goggles attached to it by a lanyard; both sets were switched on. The autopsy of the nurse determined that he had died of drowning, not of trauma. Nevertheless, the National Transportation Safety Board concluded that the helicopter most likely entered an area of reduced visibility in snow squalls, and that the pilot, despite having both a ground proximity warning system and a radar altimeter, probably "flew the helicopter under control into the ocean." The helicopter and its three other occupants were never found.
The NTSB's probable cause for the accident was compendious: "The pilot's decision to continue VFR flight into night instrument meteorological conditions. Contributing to the accident were the operator's failure to adhere to an FAA-approved and mandated safety risk management program, the FAA's failure to provide sufficient oversight of the operator to ensure they were in compliance with the risk management program, the pilot's lack of experience in night winter operations in Alaska, and the operator's lack of an EMS dispatch and flight following system."
The "safety risk management program" was a procedure designed to remove, as much as possible, the subjective element from go/no-go decisions. Such programs, in various forms, are in use by many HEMS operators, including this one. But investigators found that the program, in this case, "was not well understood, and not monitored."
In principle, it required pilots to complete a "risk assessment form" that assigned numerical scores to various aspects of the proposed operation -- crew, weather, equipment, environment, etc -- and then, on the basis of the total score, provided a cumulative risk factor -- low, caution, moderate, high. A low-risk flight required only the pilot's approval; caution required the pilot to take steps to mitigate the risk; moderate risk required consultation with the company; and so on. Naturally, this type of bureaucratic effort to reduce highly complex, subjective judgments to computer-friendly zeros and ones tends to devolve into an empty and irritating routine, and it seemed to have done so at this operator. The NTSB interviewed a number of pilots, and found no "continuity" among them. Some filled out a risk assessment form daily, some for each flight. The fate of the forms, once completed, was unclear; the main office was unable to produce any completed forms, including the one for the accident flight.
The chief NTSB investigator assigned to the accident filled out one of the risk assessment forms, based on the information that was likely available to the pilot. The risk score came out "moderate," meaning that company management should have been consulted about the flight. It had not been. Another company pilot, more experienced than the accident pilot, told investigators that, given the lack of ground reference lights and of weather information, he would not have accepted a transport over Prince William Sound at night. Of course, he had the benefit of knowing how this particular flight had turned out.
Apart from the evident ineffectiveness of the risk assessment procedure, the NTSB noted a broader lack of coordination among the participants. As a matter of policy, intended to avoid pressuring EMS pilots into rash heroics, the hospital did not provide them with any information about the urgency of the transport. In this case, the patient was not in critical condition and could have waited for an airplane to become available. Fixed-wing and helicopter services were provided to the hospital by different companies, and they did not communicate with one another. By the time the helicopter had flown from Soldotna to Cordova, one of the previously booked fixed-wing airplanes had in fact become available and could have taken over the transport. But the dispatch center was operated by the hospital, not the HEMS company, and did not concern itself with weather, only with keeping track of the whereabouts of patients. Thus, even if the pilot had checked the weather while he was at Cordova and had learned of the snow flurries at Whittier, he would have had no straightforward way of acting on the information.
In January the NTSB published its findings in this investigation, along with reports, many of them still incomplete, from eight others involving HEMS services. It was an effort to call attention of the conspicuously high accident rate among HEMS operators. A search of NTSB reports since the start of 2000 turns up 33 accidents with 96 fatalities; 35 fatalities occurred in the 13 months prior to the NTSB's presentation. Darkness and weather play a role in many of these accidents.
The FAA has announced that it will propose new rules requiring that medical helicopters be provided with helicopter-specific terrain avoidance systems and radar altimeters, as well as collision avoidance systems; that pilots receive enhanced training, particularly in recovery from inadvertent entry into IMC, and more frequent proficiency checks; and that services with 10 or more aircraft establish control centers capable of providing useful dispatch information and timely en route guidance to their pilots. All these requirements sound constructive, although their cost may put some operators out of business.
The stage for the Alaska accident, however, was set by the pilot's decision-making. His inexperience, combined with the limited information available to him and his failure to actively seek more, apparently led to his eventually finding himself flying at low level over icy water in snow, mist and total darkness. It is possible that he could see the glow of Whittier's lights ahead of him, and that he intended to land there; it is also possible that he had already made a decision to turn back, or was in the process of doing so. Why the helicopter ended up in the water -- and how this could happen without traumatic injuries to the flight nurse -- will remain a mystery.
This article is based on the National Transportation Safety Board's report of the accident and is intended to bring the issues raised to the attention of our readers. It is not intended to judge or to reach any definitive conclusions about the ability or capacity of any person, living or dead, or any aircraft or accessory.