Fri, Nov 6, 2009 — David Evans
Based on an unprecedented helicopter ambulance accident record in 2008, the National Transportation Safety Board (NTSB) will host a 4-day public hearing, commencing 3 February at 9 a.m., exploring virtually all aspects of helicopter EMS (emergency medical services) operations. It is obvious that NTSB special studies in 2006 and earlier, in 1988, and the resulting recommendations, have not had the desired effect – which is an improved level of safety.
Rather, 2008 witnessed 12 EMS helicopter accidents – an average of one a month – that resulted in 28 deaths and 8 injuries. It is the most dismal EMS helicopter accident rate on record:
- 5 February, South Padre Island, TX, Eurocopter AS350, 3 fatalities.
- 2 May, Hiram, GA, AS350, 3 uninjured.
- 10 May, LaCrosse, WI, EC135, 3 fatalities.
- 29 May, Grand Rapids, MI, Sikorsky S-76A, 2 seriously injured.
- 30 May, Pottsville, PA, EC135, 3 minor injuries.
- 8 June, Huntsville, TX, Bell 407, 4 fatalities.
- 26 June, Flagstaff, AZ, two Bell 407s, 7 fatalities.
- 27 June, Ash Fork, AZ, AS 350, 3 seriously injured.
- 21 July, Salinas, CA, McDonnell Douglas MD900, 3 aboard, no injuries.
- 1 September, Greensberg, IN, Bell 206, 3 fatalities.
- 27 September, Forestville, MD, AS365N1, 4 fatalities.
- 15 October, Aurora, IL, Bell 222, 4 fatalities.
The NTSB said of its public hearing:
“The goal of the upcoming hearing is for the Safety Board to learn more about helicopter EMS operations so that it can better evaluate the factors that lead to accidents. The Board will invite expert witnesses to provide sworn testimony. The majority of these witnesses will participate as part of small panels addressing particular safety issues. Additionally, several organizations will be granted ‘party status’ to the hearing so that they may question the witnesses directly. The witnesses and parties will represent a range of EMS-related communities, including pilots, medical personnel, managers, and Federal Aviation Administration (FAA) officials who provide oversight.”
It is not clear if air traffic controllers, or their union, the National Air Traffic Controllers Association (NATCA), will be present. One hopes they will be, as exactly what controllers are qualified and not qualified to do may be contentious. For example, in the Forestville, MD, crash, the controllers in the Andrews AFB tower were not qualified to conduct an airport surveillance radar (ASR) approach. Doug Church, a spokesman for NATCA, said, “The FAA does not have the staffing at Andrews to be able to train controllers there on this ASR approach in question.”
It will also be useful to have present members of the Professional Airways System Specialists (PASS), the union of FAA technicians that maintains equipment. The FAA is placing great hopes in ADS-B (automatic dependent surveillance – broadcast) for traffic separation and, in the case of EMS helicopters, for flight following by dispatchers. However, it is obvious from the documents amassed in the Forestville, MD, crash that ADS-B has problems of reliability when tracking low-altitude targets like EMS helicopters.
As part of its activities pursuant to the February hearings, the NTSB recently opened nine accident dockets, making available to the public literally thousands of pages of material on nine accidents. A few general observations may be useful:
- First, of the nine events, four occurred at night.
- A couple accidents appear to be “descent into terrain” in night visual flight rules (VFR), absent any noted defect.
- A review of the mishaps indicates that they are not unique to EMS flights, but are part and parcel of any helicopter operation. However, EMS operations involve night flying, in marginal weather, with landings at unprepared sites – with all three factors often in the same flight.
- A couple events on the NTSB’s list – certainly the AS350 helicopter – are going to be determined to be mechanical problems inherent in those models, regardless of how they are employed. One source maintains that AS350 flight controls/hydraulics kill more people each year than any other rotary wing mechanical cause, because the FAA will not conduct an airworthiness review of this model because the U.S. is not the state of certification (which is the providence of France).
- There was no second pilot in any of the EMS helicopters involved. Given the pilot workload, the absence of a second pilot may be stripping away a major safety defense. Clearly, EMS pilot workload is a major problem, as evidenced by the role assumed by medical personnel on many flights (e.g., operating radios, maintaining a lookout, etc.). The fact that the NTSB has not cited single-pilot workload in any of its EMS helicopter crash investigations is, frankly, a mystery. Medical evacuation helicopter flights in the military are undertaken by two pilots, no questions asked.
The other problem, that should be addressed, is the status of NTSB recommendations regarding EMS helicopter operations. The NTSB issued four recommendations in February 2006, the FAA and EMS industry response to which appears to be weak, to say the least. The recommendations, the response, and the NTSB characterization of the response reflects dilly-dallying to the extreme after two NTSB special studies and umpteen accident investigations:
NTSB recommendation: Require all EMS operators to comply with Part 135 operations specifications, which are more stringent than Part 91.
FAA response: a number of helipads at hospitals have installed equipment and developed procedures that enable EMS flights to use instrument approaches, however, these helipads do not have the weather-reporting equipment or personnel required for a Part 135 flight’s approach and it would impose a burden to acquire this capability. Therefore, the FAA plans to require that most of the Part 135 flight rules, including weather minimums and pilot flight and duty time limitations, be applicable only when medical crew are on board, but plans to exempt the EMS flights from the requirement for weather reporting at this destination.
NTSB status: Open – Acceptable Alternate Response.
NTSB recommendation: Require EMS operators to develop and implement a flight risk evaluation program.
FAA response: risk assessment to be added to Operations Specifications (OpSpecs).
NTSB status: Open – Acceptable Response.
NTSB recommendation: Require EMS operators to use formalized dispatch and flight-following procedures.
FAA response: the FAA is examining the role that a communication specialist could play in performing dispatch and flight following activities.
NTSB status: Open – Acceptable Response.
NTSB recommendation: require EMS operators to install terrain awareness and warning systems (TAWS) on their helicopters.
FAA response: standards do not yet exist for installing TAWS on helicopters [although a commercially available TAWS is available for helicopters]. Standards were to have been developed in 2008, but a notice of proposed rulemaking (NPRM) has not yet been issued [thus, TAWS equipage on EMS helicopters is years away].
NTSB status: Open – Acceptable Response.
Given the great paucity of action, the NTSB characterization of the FAA responses as “acceptable” prompts one to ask: what is it in any of these temporizing, legalistic and promissory responses that is acceptable?
Hard evidence that these recommendations are toothless can be found in just one of the nine dockets recently opened by the NTSB – the one dealing with the fatal crash in Forestville, MD.
Going through each of the four NTSB recommendations:
A-06-012: At the time of the accident, the aircraft was being operated under Part 91, not the more stringent Part 135 recommended by the NTSB. The Maryland State Police (MSP) were apparently exploiting a loophole. The police functions were regarded as “public use” and hence could be operated as Part 91, while ambulance operations were supposed to be operated as Part 135. The second part of the regulation was simply ignored. The State Police are now evaluating the adoption of Part 135.
Irrespective of Part 91 or Part 135, the State Police helicopter operation was subject to FAA oversight. The NTSB Operational Factors/Human Performance Specialist’s Factual Report revealed:
“A search of FAA surveillance records relating to MSP revealed no surveillance was performed during the year prior to the accident, in the following areas: en route inspections, base inspections, station inspections, simulator inspections/checks, manual reviews, training programs, ramp inspections, records inspections.”
A-06-013: The NTSB wanted a flight risk evaluation completed before missions. According to the NTSB docket on this MSP crash, pilots performed “mental assessments.” A risk assessment matrix was supposed to be used for these assessments, but “pilots did not routinely use it.”
A-06-014: The NTSB wanted EMS operators to use formalized dispatch and flight-following procedures. The state police duty officer, in discussing the mission on the radio with the pilot, said “it is up to you” whether or not to fly the mission. The record of radio communications does not reflect “formalized dispatch” or anything remotely akin to joint decision-making about whether or not to fly the mission.
As far as MSP duty officers (dispatchers) using flight following information to track the progress of the helicopter, the post-accident interview conducted by the NTSB with Sgt. Robert Adams, the MSP Operations Supervisor, revealed a lack of confidence in ADS-B, the principal means of tracking the helicopter. According to the NTSB notes of the interview:
“[Sgt. Adams] stated that the GDL-90 equipment installed on the helicopters for ADS-B tracking ‘does not function well at low levels.’ There had been many, many examples of when ADS-B position stopped at the end of a runway and it turned out that the helicopter actually landed at the other end of the runway or somewhere else. This led to a type of ‘conditioning’ among personnel and a ‘lack of confidence in the low level position.’ …
“Sgt Adams said that a loss of ADS-B signal with a helicopter required the duty officer to make contact with the helicopter immediately. (The helicopter icon on the screen turned red and an audible alarm was heard when the signal was lost.) However, in practice, because of the problems with the low level position, this was only done when the signal was lost in cruise flight…”
Recall that this helicopter crashed while descending to land at Andrews. Thus, loss of ADS-B signal would have been ignored by the duty officer.
A-06-015: Recall the NTSB wanted a requirement for EMS operators to install terrain awareness and warning systems (TAWS) on their helicopters and that the FAA responded that it is developing standards for a helicopter-installed TAWS.
Three other State Police helicopters, based in the hilly portion of the state, were equipped with TAWS but the accident helicopter was not.
Of interest here is how three of the MSP helicopters came to be equipped when the FAA has not formally adopted a standard yet. It is known that at least one avionics manufacturer produces a TAWS for helicopters; perhaps they were installed on the MSP helicopters by the Supplemental Type Certificate process. If so, the criteria for award of a Supplemental Type Certificate could form the basis of a performance template for retrofitting all helicopters. It is
clear that where operators perceive the need, they have gained approval for installing TAWS.
What follows is an account of the crash gleaned from numerous documents in the docket.
An air traffic controller was removed from his regular duties after giving the pilot of a state medevac helicopter a weather report based on hours-old information shortly before the helicopter crashed near Andrews AFB in Maryland, according to records made public by the NTSB.
The pilot decided to land at Andrews AFB after the controller reported clouds starting there at 1,800 feet. In fact, conditions were far worse. Fifteen minutes later, at 12:06 a.m., a different controller – based at Andrews – noted that thick, dense clouds started at 500 feet.
By then, the helicopter had already gone down. Four people died in the crash, including the pilot and a patient who was being flown from an automobile accident in Waldorf, MD. A second patient, also being transported on the helicopter, survived the cash. Even so, it was the worst medevac crash in Maryland state history.
While no conclusions were offered about the cause, the records indicate that a large-scale search was not initiated until nearly an hour after the helicopter disappeared from radar and that the second controller was unable to provide searchers with the helicopter’s last-known coordinates to guide their efforts.
Guidelines call for the state police-run helicopter service not to initiate flights at night when the cloud ceiling is less than 800 feet above ground level and/or the visibility is less than three miles.
Kendall Young, who relayed the misleading, old weather report, has been assigned to administrative duties, according to the records. Young told investigators that the weather report is supposed to automatically update, and the only way to tell if it is not updating is to check the time stamp. Young has been an FAA controller for almost 30 years.
On September 27, the night of the accident, at 11 p.m., a duty officer at the helicopter dispatch office contacted pilot Stephen Bunker was a request to transport two car accident victims to Prince George’s Hospital Center in Cheverly.
It was stormy, and the two discussed flying conditions. The officer left the decision to Bunker, according to a transcript of the exchange. This would not be the situation in airline practice, where the decision to make a particular flight is arrived at jointly, by the dispatcher and the pilot in command.
In this case, we have a reluctance on the part of the dispatcher to assume responsibility. “That’s up to you,” the officer said, “do you think you can fly it?”
Bunker reviewed the cloud levels in College Part, MD, and at National Airport in VA, and remarked that a MedStar helicopter had just landed at Washington Hospital Center in the District of Columbia.
“If they can do it, we can do it,” Bunker said.
“Okay, it is up to you,” the officer said (again, abjuring any responsibility for the decision).
“Yeah, we ought to be able to do it,” Bunker replied.
Bunker and onboard medic Mickey Lippy, who was in the co-pilot’s seat on the outbound flight, reported a smooth flight to Waldorf, where they picked up the two teenage female patients and a volunteer emergency medical technician.
At 11:47 p.m., Lippy called the dispatcher’s office (note, he was performing this communication function while occupying the co-pilot’s seat, indicating that the workload for a single pilot – at night in bad weather – is too much). “PG Hospital is completely fogged in, and we can’t land there,” he said, according to a transcript. Lippy said they decided (note, not the pilot alone, but apparently a joint decision) to land at Andrews instead and asked that one or two ambulances meet them there.
About the same time, Bunker contacted Young, who relayed a weather report the NTSB said “was almost five hours old.”
Bunker then switched to communicating with the controller at the Andrews tower, Teal Hyman, who joined the FAA in 1985. Hyman told investigators that, throughout the night, the windows of the tower were so fogged that she could not see out.
As Bunker prepared to land, he told Hyman that the helicopter was not picking up the radio signals that would guide it down.
“It’s showing green on the panel,” Hyman replied. “But you’re the only aircraft we’ve had in a long time, so I really don’t know if it’s working or not.”
Bunker then asked if he could attempt a different kind of landing, an airport surveillance radar (ASR) approach, but Hyman said she was not qualified to do so. Hyman watched the helicopter on radar until it disappeared. The NTSB records say Hyman told investigators she “remembers giving herself a few minutes to see if the helicopter arrived.”
Just before midnight, the base’s fire chief asked Hyman about the whereabouts of the helicopter. She told him it was “somewhere out on final” descent and she couldn’t see anything because of the fog.
The fire chief checked for the helicopter in a hangar. More than 15 minutes after the helicopter disappeared, the fire chief contacted the state police to tell them it had not landed. He then checked o see it it had landed at another part of the base.
Hyman and others began calling controllers at other possible landing sites to ask whether they knew anything about the missing helicopter. Shortly before 1 a.m., authorities attempted to send out a U.S. Park Police helicopter to search, but it was turned back by clouds.
A massive search effort began. Hyman, however, told a helicopter dispatcher that she could not provide the last known coordinates.
“They don’t give us that training,” she said more than an hour after the crash. “I wouldn’t know how to do that.”
About 2 a.m., a team of state troopers on foot found the helicopter in Great Mills Regional Park and rescued the one survivor, Jordan Mills, who had been injured in the earlier car crash.
Note the systemic organizational, technical and procedural problems uncovered in this one crash investigation. The NTSB released factual reports on eight other investigations in the run-up to its February hearings on EMS operations. If those accident dockets read like this one, the problem of EMS helicopter safety is profound and urgent.
What should be obvious from this account is that FAA oversight is derilict, that EMS helicopter operations require formalized procedures akin to the airlines, and that great hopes placed in technology – ADS-B in this case – can only be justified by thorough and realistic testing in the operating environment. ADS-B in its present state will not allow reliable flight following by dispatchers, and in fact was ignored when the helicopter was in the most dangerous phase, descending to land on a dark and stormy night.
Major Andrew McAndrew, commander of the MSP aviation department, contended in an interview with NTSB investigators that “safety is their number one priority.” His declaration is belied by the docket. The question is how many other EMS operators believe they have a sound safety program, when in fact they are operating on luck?