Blind Guidance

False glideslope indication seals fate of Learjet on approach in low IFR conditions


The world of charter pilots is different from that inhabited by those who fly for fun, people who use their planes for business travel, or even airline pilots.

Many are contract pilots who get paid when they make a trip, putting pressure on them to complete trips under almost any conditions. They may fly a variety of equipment in varying condition, despite Part 135 rules that may say otherwise.

Charter customers are a demanding lot. They expect airline-style performance with private plane-style convenience. They want to be on time, and they vote with their checkbooks.

With that kind of environment, the safety record of Part 135 passenger flights borders on remarkable. But like any other field of human endeavor, even safe operations can sometimes break down.

A Houston company was hired to fly five passengers to Fargo, N.D., and back in a Lear 25B. The captain who was assigned the trip was a 52-year-old man who had accumulated more than 2,500 hours of Learjet time and more than 8,700 hours total time. He was approved to be PIC of a Learjet 24 or 25 for three Houston Part 135 operations, with one check airman calling his skills well above average.

The first officer for this trip was a 39-year-old woman with more than 2,200 hours total time and 350 hours in Learjets. She was a part-time flight instructor and was also approved to fly Learjets for three area Part 135 operators. Check airman who had flown with the first officer in Learjets made comments that ranged from [she] flew like she was ready to get her Learjet type rating to [she] was at minimum proficiency and capability.

The day before the flight, the airplane was fueled and the captain taxied it from the ramp into a hangar. There, a stretcher was removed and two seats were installed in its place. The captain and first officer discussed the next days flight, although its uncertain just how detailed they got.

On the morning of the flight, the captain called Flight Service at about 5:30 a.m. He filed an instrument flight plan for an initial leg from Houstons Hobby Airport to Intercontinental Airport across town, where he would pick up the passengers, and then to Fargo. He also filed a similar return route.

The proposed times of departure – 7:45 a.m. from Hobby and 8 a.m. from Intercontinental, were big questions. A cold front extended from southern Texas to central Louisiana, with Houston just south of the front. As the cold front moved in, wide areas of fog and mist were reported, as well as scattered rain and thunderstorms.

At the time of the briefing, Intercontinental was reporting -mile visibility with a ceiling of 100 in fog. Houston Hobby was nearly 0/0.

Im not that brave. Ill wait till it gets better, the pilot told the briefer.

As he concluded the briefing, the pilot added, Im just going to take my time and mosey on out to the airport after it gets a little better. Ill get an update with you, then well go.

Thirty five minutes later, the pilot called again to get weather information. The situation had not improved. Another half hour passed, and the pilot made another call to Flight Service. There was still no improvement in the reported conditions, although the pilot said he thought conditions at Hobby were better than were being reported.

Weather or Not, Here I Come
Fifteen minutes later, at shortly after 7 a.m., the captain called again. Intercontinental was still reporting an overcast ceiling at 100 feet and visibility of one-quarter mile.

During his conversations with Flight Service, the captain had mentioned the possibility of getting the passengers to drive from Intercontinental to Hobby, a tedious trip during the morning rush hour.

By 7 a.m., the first officer joined the captain at the airport. They moved the airplane out of the hangar and began preparing for their trip. The cockpit voice recorder began recording at 7:39 and the Learjet took off exactly on time.

After contacting departure, they were immediately vectored to make an ILS approach to runway 26. The current ATIS called the ceiling 200 broken, visibility five and a runway visual range of more than 6,000 feet.

The flight was handed off to another controller and, after a few turns, the jet was cleared for the approach at 7:51 and told to contact the tower.

The airplanes cockpit voice recorder showed the crew did not verbally brief the approach, but the flight began descending when cleared to land by the tower. It was about 1 miles inside the final approach fix at 1,600 feet and on the centerline of the localizer when the captain said, Ive got a compass flag.

The airplane immediately departed the centerline and took up a heading about 25 degrees left of course. Still descending, it got to 700 feet and was less than a mile from the approach end of runway 27, a nearly parallel runway that was also handling traffic.

Penalty Flag
At the instruction of the captain, the first officer called a missed approach and the tower controller gave them altitude and heading instructions. For the next few minutes, the two pilots worked to figure out why the captains directional gyro was flagged. At first the captain decided to go back to Hobby, saying, We cant do a trip like this.

Then he reconsidered and said that since the weather in North Dakota was severe clear, theyd try again at Intercontinental.

Although the cockpit voice recorder indicates they made some progress in dealing with the problem, its unclear exactly what that progress was. Several times the captain said I got it figured out now.

At 8:06 the flight was again cleared for the approach and told to contact the tower. Again the pilots did not discuss the approach or how they planned to fly it.

As they approached the outer marker, the first officer made a comment that apparently the glideslope [isnt] working. For the next 30 seconds they tried to get established on the ILS, but were having trouble with both heading and altitude. Finally, at 8:09:07 the captain said, All right, can you fly it?

Yeah, I think so, the first officer responded.

At that point, the airplane was 3.8 miles from the runway, slightly below glideslope and on the localizer centerline. A few seconds later, the captain told the first officer she was descending too much and she replied, Am I? …We are way above glideslope.

The captain then agreed and told her to ease it on down.At this point, however, the airplane was between 200 and 300 feet below the glideslope.

The last radar position was about 2.2 miles from the runway and about 400 feet below the glideslope at an altitude of about 400 feet msl – 300 feet agl. At 8:09:54 the cockpit voice recorder picked up the sound of the Learjet hitting trees, and two seconds later it stopped recording.

When the wreckage was located more than two hours later, the recovery team discovered that a post-crash fire consumed most of the plane, including the crew and the cockpit instrumentation. Some of the navigational instruments were remotely mounted in the nose of the airplane and survived the impact and the fire.

All components were tested for failures that could have explained why the captains HSI flagged, but no anomalies were found. Some of the instruments could not be tested due to impact damage, but tests of the circuit boards revealed no answer on why the captain got the compass flag.

The Smoking Gun
Investigators next turned to the glideslope. The first officers glideslope module was intact, and when tested displayed a full fly down deflection regardless of the channel selected and warning flags were erratic. A check of the schematics revealed that a component failed downstream of the glideslopes flag circuitry. Collins, the manufacturer, produced records that showed that the same component had been replaced 12 times between January 1990 and March 1998 in other installations of the VIR-30A. Six of those cited a glideslope deviation problem.

Two months before the accident, avionics technicians had looked into reports that the #1 VOR would not channel two frequencies and that the glideslope pointers would stick.

One maintenance entry read, …found glideslope pointer in both ADI and HSI to be sticking, co-pilots position. No action taken at this time. Another inspection card referring to the same squawk said, Note: advised customer, repairs deferred until such time available to send units out to mfr. In fact, the 24-year-old plane had flown nearly 9,000 hours, but only about 30 hours in the 4 months prior to the accident.

Another Dangerous Discovery
The accident investigation revealed another unlikely shortcoming – in air traffic control. Intercontinentals Tracon is equipped with Minimum Safe Altitude Warning software designed to monitor aircraft altitude and show alerts if the airplane gets too close to the ground or obstacles. The system has several capabilities, including monitoring the approach path and sending a visual and aural warning if the flight path is below or predicted to be below certain minimum altitudes.

The alert altitude is found by subtracting the airport elevation plus a 100-foot buffer from the lowest MDA associated with non-precision approaches serving the runway. Runway 26 is served by a GPS approach with an MDA of 600 feet. The alarm should have sounded when the Learjet was predicted to go below 402 feet agl on two consecutive radar passes.

The software was incorrectly configured, however, in that the MDA from the precision ILS approach was used instead of the non-precision approach. It was set to sound alarms at 100 feet.

Although the error appears glaring, testing the Learjets radar returns against the new Minimum Safe Altitude Warning configuration shows it wouldnt have helped the crew much. Although it would have detected a deviation 21 seconds before impact, the deviation was not repeated on the next scan, so no alert would have been issued to controllers. Had it been properly configured, an alert would have sounded no earlier than 8:09:50, just four seconds before the plane hit the first tree.

Like most accident chains, the one involving the Learjet could have been broken in several places. Prompt repair of the co-pilots glideslope months earlier would have been a good place to start. Other spots include limiting the pressure the captain felt to begin the flight on schedule and having the crew call altitudes on the descent instead of fixating on the glideslope indication.

-by Ken Ibold


Please enter your comment!
Please enter your name here