Business & GA, Commercial

Safety in Avionics: EMI: A Serial Killer?

By David Evans | November 1, 2000
Send Feedback

The airspace over the Atlantic Ocean immediately east of New York’s John F. Kennedy International Airport is sort of a "Bermuda triangle" of electromagnetic interference (EMI). This is the clear inference of Harvard University researcher Elaine Scarry. In three articles, she suggests that EMI from external sources may have played a common role in the 1996 crash of TWA Flight 800, the 1998 crash of Swissair Flight 111, and the 1999 crash of EgyptAir Flight 990.

With jet airliners coming to depend ever more heavily on electronic systems for vital functions, Scarry believes their vulnerability to EMI poses a potentially lethal threat to air safety.

In her April 9, 1998, essay in the New York Review of Books, Scarry raised the possibility of EMI in the fall of TWA Flight 800. In articles appearing in this year’s Sept. 21 and Oct. 5 editions, she raised the potential role of EMI in the Swissair and the EgyptAir tragedies. All three airplanes departed from John F. Kennedy International Airport, flying through a region heavily covered by various ground-based radars and adjacent to areas where the U.S. military has been known to conduct operations, including electronic warfare exercises.

Fingerprints of a Serial Killer?

Scarry points out that TWA Flight 800 and Swissair Flight 111 shared at least eight features:

  1. They took off from the same airport.

  2. They took off on the same day of the week, Wednesday, at virtually the same minute, 8:19 p.m.

  3. They both traveled along the Bette flight route.

  4. For both airplanes, the first signs of trouble occurred about 12 to 14 minutes into the flight (when radio contact was lost with the TWA jet, and for the next six to 13 minutes, radio contact was lost with the Swissair jet).

  5. Both airplanes appear to have suffered an electrical catastrophe.

  6. The cause of the catastrophe remains elusive and mysterious, even after years of rigorous inquiry by U.S. and Canadian accident investigators.

  7. Both airplanes flew during a week when military exercises were scheduled in areas adjacent to the Bette route, designated W-106 and W-105.

  8. They both flew when certain specific transmitters, notably U.S. Navy P-3 patrol aircraft and submarines, were in the region (Scarry also mentions that an Aegis cruiser with its powerful Spy-1 radar was about 180 miles [290 km] away). Some P-3s are configured as electronic warfare aircraft. Messages to the submarines from the 2-million watt Navy transmitter at Corbett, Maine, Scarry postulated, might have interfered with the functioning of the passenger airplanes’ electronics.

As she pointed out in her September article, explicitly comparing TWA 800 with Swissair 111, "The literature on electromagnetic interference is full of stories about unwanted electrical upsets that recur in the same space at the same time." Hence, her implication that a common electronic event on two nights at the same place in the sky may have affected both airplanes.

Similarities with EgyptAir Flight 990 are more tenuous. Same departure airport, to be sure, but a different day of the week, a different departure time, and a different route. Not along Bette, but right into areas W-105 and W-106. Although no military exercises were on tap for that evening, Scarry points out that powerful U.S. Air Force radars at Riverhead, on Long Island, and Truro, on Cape Cod, were tracking the Egyptian B767.

Here, too, Scarry hints at the role EMI may have played. Certain details, she asserts, are consistent with EMI: disconnect of the plane’s autopilot, the steep dive, the split elevators, and the engines shutting down.

"The uniformity of the region in which the accidents occurred suggests that the region itself–the environment external to the (air)plane–should be included among the causes to be investigated," Scarry posits. Three cases, a total of 676 dead–and EMI may be the serial killer.

The Link Inside

Peter Ladkin, Ph.D., challenges Scarry’s assertion that EMI could be the link in the TWA 800 and the Swissair 111 disasters. Ladkin is a professor of computer networks and distributed systems at the University of Bielefeld in Germany. There is a common feature of both accidents worth looking at, Ladkin believes: "Defective wiring showing traces of arcing."

With respect to TWA 800, Ladkin points out that the National Transportation Safety Board (NTSB) retained the services of the National Aeronautics and Space Administration (NASA) to assess interference-induced ignition of flammable vapors inside the center wing tank, where the explosion occurred, that downed the airplane. The NASA inquiry found that the maximum energy produced by all known external sources (including secret ones) at the location of the TWA 800 explosion was less than 0.1 millijoule (mJ).

During the long course of the TWA 800 investigation, experts concluded that the minimum energy needed to ignite flammable vapors was 0.2 mJ. The NASA study noted further that the actual energy delivered to the center wing tank (CWT) was several orders of magnitude less than 0.1mJ. In fact, based on the technical details of the NASA study, the best estimate of the energy available to the wiring in the CWT was between 20,000 and 600,000 times less than the minimum energy required for fuel vapor ignition.

"Translating this into concrete terms," Ladkin informs, "suggests we would need 20,000 to 600,000 P-3’s or Aegis cruisers in the region to achieve ignition, rather than those that were actually there."

Ladkin points out that external EMI would have to be considerably stronger to cause severe and catastrophic arcing in the relatively new wiring on Swissair 111.

"Maybe one of these P-3’s flying alongside SR 111, and emitting at full strength, could cause such arcing," Ladkin surmises. If so, the Swissair MD-11 most probably would have suffered a major avionics failure at this point. Energy fields of sufficient intensity to cause arcing in insulated wiring would have been likely to disturb the airplane’s electrically-driven "glass cockpit" displays also, Ladkin believes. To be sure, radio contact with Swissair was lost for as much as 13 minutes. But is EMI the source? Canadian investigators are aware of the 13-minute gap, but "there is no evidence of EMI," according to an official with the Transportation Safety Board (TSB) of Canada. He relates that there were "no anomalies" on the flight data recorder during this period of time.

"Is there any known difference between these two flights and all the tens of thousands of others?" Ladkin asks. Yes, he reiterates: Look to the defective wiring. TWA 800 manifested problems associated with improper wire repairs, with routing high and low power wires in the same bundle, and with the slow deterioration of insulation caused by tight bends, overstretching, and chemical changes with age. In the newer Swissair jet, inappropriate installation and routing of wiring have surfaced during the course of the still-incomplete TSB investigation. Investigators have found some 20 wires with evidence of arcing, and Swissair internally has proposed rerouting some wires in the cockpit area to increase separation and redundancy of vital circuits.

In comparison to a speculative case of EMI, the wiring problems common to both aircraft are documented.

Triad of Tragedy

Scarry argues that EgyptAir did not fly the Bette route, but rather directly into the middle of a military exercise area (although the airspace was cleared for civilian flights on the night of the accident). Her assertion that the "stark facts known about the fall of EgyptAir 990 are themselves consistent with electromagnetic interference" is, of course, music to the Egyptians. Immediately after Scarry’s Oct. 5 article appeared, Egyptian Transport Minister Ibrahim el Demiri advised the Cairo press that although Scarry’s scenario was hypothetical, electromagnetic waves could have confused the computer of the Boeing 767, leading to the crash.

Ah, but another scenario also could "explain" the crash. Deliberate actions, notably of relief pilot Gamil el Batouti, also explain the airplane’s plunge into the waters 60 miles south of Nantucket. The airplane’s performance was perfectly normal until the autopilot was disconnected, and it continued to perform normally for another eight seconds, until the elevators deflected, pitching the airplane into a 40� dive toward the water 30,000 feet below. A full minute before the autopilot was disconnected, Batouiti prayed, "I rely on God." Was he clairvoyant about an impeding mechanical problem?

According to Canadian psychiatrist Daniel Cappon, MD, who has written about the need for routine psychiatric screening of pilots, from the evidence available Batouti was an "acute psychotic." Cappon maintained that Batouti’s mood swings, his occasional quarrelsomeness, his hotel indiscretions (listening through the closed doors of female guests, exposing himself to teenage girls), were signs of a troubled individual. Attributing the EgyptAir crash to EMI seems to be pushing the same argument further along a more slender thread.

Avoiding an Ugly Surprise

For perhaps one overarching reason, Scarry’s concerns about EMI should not be dismissed. The industry’s comforting assumptions keep getting broken. In the TWA accident, investigators belatedly found that earlier estimates of the energy required to ignite flammable fuel vapors were optimistic (too high) by a factor of 10. In the Swissair accident, additional wire problems exposed in the course of the TWA 800 investigation were revealed in a comparatively new airplane.

EMI could be another lurking surprise. For this reason alone, perhaps it should be included in the structured process that is part of every accident investigation. Aircraft maintenance, atmospheric weather (windshear, icing, fog, etc.), and other factors routinely are examined in detail, even if their relationship to the initially suspected probable cause is problematic. Perhaps EMI should be considered as a form of "electronic weather." Three aspects pertain: "space weather," such as the 11-year sunspot cycle, at its peak now, that could affect navigation systems and radios; EMI from ground and airborne sources; and finally, interference from portable electronic devices carried by passengers in the cabin. In all three cases, the literature is growing about unwanted electrical interference and system malfunctions. Heaven forbid that EMI should someday bring down an inadequately shielded "electrified" jet loaded with passengers, but the industry has been caught off guard in other areas.

Scarry’s arguments are debatable on the particulars, but her general concern may serve as a cautionary call against complacency.

David Evans is editor of the award-winning newsletter Air Safety Week.

Comments should be directed to [email protected].

Receive the latest avionics news right to your inbox