Radar Defects Cited in Probe of Air Disaster

The radar system used to monitor air traffic east of Los Angeles International Airport exhibited defects shortly before and after the Aug. 31 collision between an Aeromexico jet and a small private plane, according to documents introduced Wednesday at a National Transportation Safety Board hearing.

From the night of Aug. 25 to the morning of Aug. 28, one of the radar system’s two channels was inoperable for nearly 37 hours.

Three days after the crash, the radar system’s other channel failed to properly track a small plane that was test-flown along the same route that a Piper Cherokee had taken before it collided with Aeromexico Flight 498.

The crash, which occurred more than 6,000 feet above Cerritos, resulted in 82 deaths.

The question of radar reliability became a critical issue in the NTSB inquiry Tuesday when the Federal Aviation Administration controller who directed Flight 498 said his radar screen never showed the presence of the Piper, which had entered the L.A. airport’s Terminal Control Area without authorization.

In highly technical testimony Wednesday, three FAA staff members who work with the radar that guided Flight 498 acknowledged that the system has defects. However, each refused to say that the radar system is unsound. Instead, they described it in terms ranging from “acceptable” to “entirely safe.”

The staff members echoed a point that Flight 498’s controller, Walter R. White, made on several occasions Tuesday: radar is not foolproof, and despite various safeguards it is technically possible that at any given moment an airplane simply will not appear on a radar screen.

Installed in 1960s

The radar system in use when the crash occurred is known as the ASR 4, and was installed in the early 1960s. Its antenna receives and transmits the images of airplanes in a large area east of the airport. These signals, processed on one of two channels, are sent to a computer in the FAA’s Los Angeles Terminal Radar Approach Control (TRACON), which is located on the ground and is responsible for guiding jets toward the airport.

TRACON uses a second radar system, ASR 7, installed in the early 1970s, to guide planes from the other side of the airport. When one system is out of use, the other can quickly substitute.

Before the crash, the FAA had begun planning to install a more advanced system to replace the radar equipment now in use. Installation is scheduled next year.

TRACON’s air traffic manager, Richard A. Cox, Wednesday described the older ASR 4 radar as “an old piece of equipment (that has) had more than its share of failures over the past year.”

From January until July, the system was plagued by “a significant increase” in incidents in which data about planes was not properly transmitted to controllers’ screens, Cox said.

Specifically, the radar was not recording the presence of some small airplanes that use altitude-encoding transponders, devices which allow a controller to see the plane’s altitude as well as its presence.

Even after this problem was corrected, Cox said, the system was still occasionally failing to spot airplanes.

“There are limitations to radar,” he said. “It is not an unusual occurrence that there is a plane in the system not displayed on radar.”

Problems Are Cited

The trouble grew exceptionally bad in late August on the ASR 4 system’s Channel A. During a period of 55 hours, 29 minutes, Channel A was unusable for 36 hours, 44 minutes, according to FAA records.

The problems in the radar included visual static, waving signals and “uncancellation,” in which radar begins displaying non-moving targets.

Cox said that safety was assured by simply using the system’s Channel B. However, at one point--for 33 minutes around 5 a.m. Aug. 28--Channel B was also out of service. That forced controllers to temporarily abandon the ASR 4 system and switch to the ASR 7 system.

NTSB air control specialist Allen Lebo attempted several times to elicit testimony from Cox and from TRACON radar technician Allen Davis that the image of the Piper seemed certain to have appeared on controller White’s screen.

Computer Overloaded

But neither witness agreed with that assertion.

“There are times,” said Davis, “when the (computer that processes information from the ASR 4 radar) gets overloaded (and) may not process a target like that.”

Davis characterized the radar system as “acceptable.”

Added Cox, “I view the system as being entirely safe.” And, despite the system’s past and current troubles, Cox said safety is “absolutely not” compromised.

Cox’s statement was greeted with obvious skepticism by NTSB Chairman James E. Burnett Jr., the hearing’s presiding officer, who said, “I don’t think that squares” with Cox’s earlier acknowledgements that the system sometimes fails to show planes.

Test Flight Planned

In the days after the crash, NTSB investigators said they believed the image of the Piper had been on White’s screen and that--despite his denials--White had simply failed to pay enough attention.

At the time, the investigators said they had used another Piper to retrace the path of the small plane that collided with the jetliner and found that the TRACON radar screen picked up the signal.

However, according to a report on one FAA test introduced Wednesday, the signal that appeared on the radar screen was too faint to be considered useful during the moments when the plane approached the point of impact.

That test, conducted Sept. 3, involved a Piper Cherokee that was flown along the same route followed by William K. Kramer, the pilot whose plane collided with the jetliner. Kramer had taken off from Torrance with plans to fly to Big Bear and--for reasons that still remain unclear--flew a course that took him over Cerritos.

Scale of Sharpness

The test was run on the ASR 4 system’s Channel B. Technicians watching a radar screen at TRACON kept track of how sharply the test plane registered. They used the FAA’s inspection scale of 0 to 3, in which 0 is “unusable,” 1 is “barely visible” and also regarded as unusable, and 2 and 3 are progressively more usable.

As the plane climbed northwest, it registered well, receiving a score of 2 or 3 at 58 of 69 monitoring points, and a 1 at the other 11 points.

But as it climbed the last stretch--from 5,800 feet to the where the collision occurred at about 6,200 feet--the image of the test plane faded, becoming barely visible on the radar screen. It registered only a 1 in intensity on each of the last six monitoring points.

Lawrence Sump, a regional vice president of the Professional Airways Systems Specialists, the union which represents the technicians who service FAA radar, said in an interview he found the test “very persuasive” evidence that the radar system is not sufficiently reliable.

‘The Radar Works’

However, Frank del Gandio, deputy manager of the FAA’s accident investigation division, said in an interview that it was impossible to duplicate flight conditions exactly and that despite the isolated low scores, “the tests prove the radar works.”