Early in the fog-blanketed morning of June 18, 1994, a Learjet 25D smashed into tall trees a mile south of the threshold of Runway 1R at Dulles International Airport near Washington, D.C. The crew was making its second attempt at the ILS approach for that runway when the accident occurred, killing all 12 occupants.
Learjet XA-BBA was operated at the time by Transportes Aereos Ejecutivos, S.A. (TAESA), a Mexican corporation. There was no post-crash fire, but the airplane was nevertheless destroyed by the impact.
The charter flight had originated in Mexico City, with a planned refueling and Customs stop at New Orleans' Lakefront Airport. The first leg of the journey proceeded uneventfully and, except for an unexpected holding pattern assigned by Washington Center, the cruise portion of the second leg proceeded routinely, as well. At 5:53 a.m., the aircraft was cleared out of the holding pattern and the crew contacted Dulles Approach Control.
Over the course of the next several minutes, the pilots were advised by Approach Control to expect an ILS approach to Runway 1R and were issued vectors for sequencing. ATIS Delta, recorded at 12:50 a.m., reported, "... weather: indefinite ceiling six hundred, sky obscured; visibility one half, fog; temperature seven-one, dew point seven-one; wind one-four-zero at four ...."
A short while later, the approach controller advised two inbound air carrier aircraft that the Runway 1R RVR had deteriorated to 1,200 feet. Soon afterwards, he announced that it had dropped to only 600.
The published minimums for the 1R Category I ILS approach are one-half mile, or 2,400 feet RVR. Minimums may be decreased to 1,800 feet RVR if centerline lights and touchdown lights are operative, which they were on the morning of the accident. Where existing visibility is reported to a pilot in both RVR and miles, however, RVR becomes the controlling value. The reason for this is that a visibility report from a control tower observer, often located far from a particular runway, can bear little resemblance to the visibility existing on that runway. A current report of 600 feet RVR in the touchdown zone meant that the weather for the approach was below Cat I approach minimums anyway, based on the visibility as reported on the ATIS.
Runway 1R can also support Cat II and III operations for aircraft and crew that are qualified to utilize them. Both allow significantly lower RVR minimums to be used. During the 15-minute period prior to the accident, two air carrier aircraft made successful Cat III autolandings. A third, whose crew was qualified for Cat II only, executed a missed approach because of the low visibility. In between the two autolandings, Lear XA-BBA, whose crew was not Cat II or III qualified, made its first unsuccessful approach attempt.
Lear XA-BBA had neither a cockpit voice recorder nor a flight data recorder. Radar data and crash scene analysis, however, helped to rule out several possible causes for the accident. Final approach airspeed was sufficiently high to preclude a stall. Coarsely chopped vegetation, similar to sawdust, packed in both engine compressor sections, was a sign that the engines were operating at the time of the accident. Leaking fuel observed at the crash site and records of fuel boarded on previous legs suggested that fuel starvation was not a factor.
Investigators soon focused on the crew itself. The 27-year-old captain had a total of 1,706 hours' flying time, of which 1,314 were in the Lear. Of these, only 87 were as captain. He had just recently completed upgrade training at FlightSafety International's Tucson facility, where TAESA crews often trained.
The pilot's FlightSafety training records reflected well below average performance as a captain during four simulator periods. His instrument scan was deficient, and he made poor use of the flight director. He had difficulty prioritizing workload and delegating tasks to the first officer, especially during emergencies. Instrument approaches generally did not meet ATP standards. During one simulator period, he failed to stop the aircraft on the runway on each of several rejected takeoff attempts. The instructor's written notes indicated that his student needed more crew resource management training in order to become competent as a PIC. His final comments, written after the fourth simulator session, stated that the candidate was "below FSI standards for PIC."
The instructor offered the pilot two additional simulator periods at no charge, and without any formal notification to be made to TAESA. The captain declined this offer, believing that his company needed him back to fly the line.
FlightSafety later sent TAESA a written evaluation of the pilot. The report described his performance but was worded in a more diplomatic manner than the instructor's original notes, which were intended for internal FlightSafety use only. The report sent to TAESA did, however, contain FlightSafety's conclusion that "...we do not think he is ready to upgrade to pilot in command."
By contrast, the accident first officer received high marks during his FlightSafety training. He was described as well qualified for first officer duties. When the accident occurred, he had a total of 852 hours, of which 426 were in the Lear.
Investigators suspect that language differences led TAESA officials to misinterpret the meaning of the FlightSafety report, written in English. Instead of delaying his upgrade to captain, TAESA completed his line training in Mexico and advanced him to the left seat.
The NTSB believes that the captain was at the controls during both approach attempts. The first can best be described as highly erratic. Radar plots show the aircraft diverging from the localizer in a series of large left and right course changes. Vertical control was even more unstable, with the Lear flying well above and below the electronic glidepath. At one point its descent rate exceeded 2,000 fpm. (A stabilized descent rate of about 700 fpm would be normal for the Lear under the conditions.) The captain did not execute the missed approach until well north of the published missed approach point, and then only when queried by the controller as to his intentions.
The second approach attempt started off better than the first. Localizer bracketing was not as capricious initially. But shortly after intercepting the glideslope, the aircraft descended rapidly. It leveled off briefly well below glideslope, barely 100 feet above the ground. At one point it flew within 70 feet of power lines. It then climbed abruptly, until finally diving at 3,000 fpm into the ground.
The NTSB determined that the probable causes of this accident were "the poor decision making, poor airmanship, and relative inexperience of the captain in initiating and continuing an unstabilized instrument approach that led to a descent below the authorized altitude without visual contact with the runway environment." The lack of a ground proximity warning system was cited as a contributing factor.
Did the crew have options other than attempting a below-minimums approach? Certainly. They could have held nearby in the hope that the weather would improve enough to land. They might have flown to their alternate. They could even have requested a Cat I ILS approach to Runway 19L — which, with a reported touchdown zone RVR of 3,000 feet, presented a considerably better, and legal, chance of success. They did none of these — and, instead, succumbed to fatal misjudgment.
Related Articles
