Probable Cause #39: Almost Home
An inherent problem that comes with familiarity is that it tempts us to take risks we would otherwise never consider. Think of the road you take home from work every day. Because you're familiar with its every bounce and undulation, you probably take a turn or an exit at a higher speed than if you were traveling down it for the very first time.
This same problem also creeps into our flying. Whether we are consciously aware of it or not, we are more apt to push the envelope at airports with which we are familiar. Unfortunately, this can and has led to accidents.
The Beech B55 Baron, with a pilot and two passengers, departed Worthington, Minn., at approximately 3:40 p.m. (all times EST) on a January afternoon originally bound for Milwaukee, Wisc. But when a weather check revealed that Milwaukee was fogged in, it was decided to continue on to Elkhart Municipal Airport (KEKM) in Elkhart, Ind., where the aircraft was based. Prior to leaving Worthington, the pilot had received a standard weather briefing from the Princeton, Minn., Automated Flight Service Station where he also filed an instrument flight plan.
At 6:18 p.m., the pilot made contact with South Bend Approach Control, which has jurisdiction over approaches into KEKM. The controller advised the pilot that the weather at Elkhart was reported as 1/2-mile visibility, light rain and mist with an indefinite ceiling at 200 feet overcast. The winds were from 120 degrees at 7 kts, the temperature was 9 C and the altimeter setting was 29.97. Dew point information at KEKM was not available but the temperature/dew-point spread in nearby South Bend around that time was reported as zero.
Although there are several approaches to the airport, only Runway 27 has an ILS. When asked by the controller which approach the pilot would like, he chose this one, knowing that he would be landing with a slight tailwind. Considering that Runway 27 is 6,500 feet long, this seemed like a reasonable risk given the weather conditions.
The controller began issuing vectors to the aircraft to align it with the final approach course. The flight path took the aircraft approximately eight miles south of the airport on an easterly heading before the controller turned it to the north and instructed the pilot to descend and maintain 2,500 feet. The pilot acknowledged the instruction and a subsequent radar plot verified the descent.
At 6:34 p.m. the controller told the Baron pilot that he was six miles from SOUSA, the outer marker, and to turn left to a heading of 300 degrees. The pilot was to maintain 2,500 feet until established on the localizer and was cleared for the ILS 27 approach. The pilot read back the instructions. Radar recorded the aircraft's altitude at that point at 3,300 feet and showed the aircraft turning inbound on the localizer.
At 6:37 p.m., the Approach controller instructed the pilot to contact the Elkhart Tower. At that point, radar showed the aircraft at 2,500 feet, with a ground speed of approximately 120 kts. Upon contacting the Tower, the pilot was told to report over the marker and was given a wind and altimeter check. The pilot responded, "I got everything (unintelligible)." That was the last transmission from the aircraft.
The Baron struck some trees and crashed 190 feet from the middle marker. The pilot and one passenger were killed while the other passenger received serious injuries.
The surviving passenger, who was sitting in the right rear seat (his unfortunate partner was sitting to his left) told investigators that as the aircraft was approaching Elkhart from the east he noticed "very bad fog conditions." He said that he and the other passenger were nervous because of the weather conditions. Probably not helping the situation was that they could not communicate with the pilot, as he was the only one wearing a headset.
In the interview with investigators, the passenger said that he heard the airplane's landing gear being extended and saw the three landing-gear indicator lights illuminate. He estimated that the gear came down approximately two to five minutes before the crash. He and the other passenger were trying to determine their position over the ground and he reported that they looked out the left side of the aircraft and saw lights on the ground, which he believed to be the glow of the parking lot lights from a nearby Wal-Mart that was under construction.
The passenger said he saw the pilot looking about 90 degrees out the left side of the window and then turn his head back inside. Soon after that the crash occurred.
Personnel employed at a fixed base operator at the Elkhart Airport reported that they heard the transmissions between the aircraft and the control tower. They went out to a hangar and opened the doors in preparation for the aircraft's arrival. They told investigators that when the aircraft hadn't arrived 10 minutes later, they assumed it had diverted to another airport.
A witness who was driving by the airport's east boundary in a northerly direction told investigators that, "the weather was extremely foggy in low lying areas." He told his fiancÚ that he thought the airport was closed because he estimated the visibility was only "about 300 feet to 1,000 feet."
Another witness, a Boeing 737 captain, reported that he was at a store parking lot about 1.1 miles east of the approach end of Runway 27 when he heard the aircraft on final approach. "I thought it was surprising that an aircraft would be making an approach in such low visibility," which he estimated to be between 1/2 and 1/4 miles. "I probably listened for 15 seconds or so before getting into my car and then heard no more. I never saw the lights of the aircraft either because of the low ceiling (and) poor visibility. The engine noise sounded steady. It also seems it was a normal RPM and loudness level for an aircraft on short final approach."
According to control tower reports, a Piper Aztec had landed at the airport a half hour earlier. The FBO's customer service representative told investigators that when the Aztec landed she could still see the airport beacon. "The fog got bad real fast," she said. By the time the Baron was on the approach, she could no longer see the airport beacon. One of the line service personnel said the fog was the same as when the Aztec landed, but that "it was just darker."
The reported weather at 6:45 p.m., about five minutes after the accident, was very similar to what was given to the Baron pilot earlier. The winds were from 120 degrees at 7 kts, the visibility was 1/2 mile in light rain and fog, the ceiling was indefinite at 200 feet, the temperature was 9 C and the altimeter setting was 29.77.
At South Bend, approximately 14 miles west of Elkhart, the weather at 6:34 p.m. was reported as winds from 080 degrees at 5 kts, a visibility of 1/4 mile with light rain and fog and a 400-foot overcast ceiling. The temperature and dew point were both 8 C, the altimeter setting was 29.73, and the tower visibility was 1/4 mile.
The airplane crashed in a creek bed 0.4 miles from the approach end of Runway 27. The southeast corner of the middle marker was located approximately 190 feet from the wreckage on a bearing of 300 degrees.
Investigators had the aircraft's ILS equipment inspected and found no discrepancies that could not be considered impact damage. The glideslope receiver flag operated normally, and the glideslope deviation indicator was normal except the glideslope center was 20 millivolts high, which would have placed the aircraft a few feet higher than normal on the glide slope.
The pilot's altimeter was tested and it read 760 feet when the test equipment was set for 777 feet. The altimeter tested 20 feet low at test altitudes of 500, 1000, 1,500, and 2,000 feet.
The aircraft's engines were inspected and no problems were found with them, either. It was estimated that there were 76 gallons of fuel on board at the time of the accident.
The 50-year-old pilot held an Airline Transport Pilot certificate with single-engine land, multiengine land and sea ratings. His logbook showed that he had approximately 7,400 hours of flight time. The pilot first flew the accident aircraft on Aug. 16, 1997, and had approximately 135 flight hours in it. In the 90 days preceding the accident he flew 102 hours with about 20 hours in the accident aircraft. The pilot owned an aircraft similar to the Baron that was also based at Elkhart.
The NTSB blamed the accident on all too-common and obvious reasons: The pilot's failure to maintain a proper glide path and terrain clearance. But that alone does not answer why the accident happened. For that, we have to consider other potential contributing factors.
We can safely surmise that the pilot was familiar with the airport, the approach to Runway 27 and the surrounding area. Once one becomes intimately familiar with an approach, the concept of having the "runway environment" in sight takes on a whole new meaning. It's only natural to look for familiar landmarks in order to monitor an approach's progress, especially when the ceilings and visibility are bad. This, in turn, has tempted many pilots to bust minimums, justifying themselves by the fact that "I know where I am."
Could it be that the pilot was caught in this trap?
The Wal-Mart that the passengers -- and, apparently, the pilot -- saw is located just 0.8 miles from the approach end of Runway 27. While still under construction, the building was far enough along to have a roof with skylights and electrical power. While the surviving passenger only mentioned seeing the lights of the parking lot, on that night the store's interior lights also illuminated the skylights. Remember that the passenger was sitting on the right hand side of the aircraft and therefore did not have a full view out the left side of the airplane.
Could it be that the pilot saw the lights from the Wal-Mart, whether they were the parking lot lights or skylights, thought he knew exactly where he was and decided to descend below the glide slope in the hopes of seeing the runway environment? Or could it be that the Wal-Mart lights distracted the pilot from his instruments? Perhaps while he was looking out at them, the airplane descended below the glide slope and the pilot never recognized the fact. Another possibility is that the pilot mistook the Wal-Mart lights for part of the approach lighting system and he thought he could continue to descend to 100 feet above the touchdown zone elevation, as per the FARs.
All this does not adequately explain one disturbing aspect of this accident, however, which is that the airplane was below the glide slope towards the final stages of the approach.
Recorded radar data indicates that at 6:37 p.m., as the aircraft passed 0.2 miles east of the outer marker, it was at 2,200 feet. This is acceptable, since an airplane on the glide slope would theoretically cross SOUSA at 2,178 feet. There were an additional eight radar hits on the Baron between SOUSA and the point at which radar contact was lost due to altitude and terrain. The radar hits were recorded at 10-second intervals.
The last six radar hits indicate that the aircraft's glide path was approximately 100 feet below the glide slope. When the last radar hit was recorded at 6:39 p.m., the Baron was about 2 nm from the approach end of Runway 27 at an altitude of 1,300 feet, which put it approximately 130 feet below the glide path.
Again, did the pilot get distracted trying to look for familiar landmarks or did he deliberately go below the glide slope? We'll never know for sure.
Regardless whether the pilot's actions were deliberate or not, there are steps we can take to prevent similar situations.
The most obvious one is that if you are ever tempted to bust minimums to get on the ground when weather conditions are not good, don't do it. If you have a known set of personal limitations and you respect those limitations, such an accident will never happen.
By busting minimums you are compromising the safety features that are built into approaches. Standard ILS approaches are designed to put you in position to make a safe landing in conditions of low visibility and/or cloud cover. They are not designed to be all-weather solutions unless the airplane, crew and approach are certified to much higher levels. This is typically the purview of commercial and military operations, and not the average general-aviation pilot.
If it is not possible to make a safe landing, the missed approach procedure is calculated from the missed approach point at the decision height. If you are below decision height at the missed approach point, you may compromise your ability to fly the procedure safely due to terrain or other issues.
One must also consider the implications of flying an approach to minimums with a tailwind. Higher groundspeeds will require faster descent rates. If you've never flown an approach with a tailwind, or are uncomfortable flying one, experimenting on a foggy night is not the right time.
The accident report did not describe the general weather conditions for the region, but it appears that on the evening of the accident, at least the airports near the Lake Michigan shoreline were at or near minimums due to fog. Still, the aircraft had plenty of fuel on board and should have been able to continue on to a reasonable alternate.
Sometimes the fear of failure to get passengers where they want to go results in pilots doing things they may not do otherwise, such as busting minimums at the home airport. Even pilots flying by themselves have done the same thing in the past, simply because they wanted to get home.
When you travel by air, you must realize that there are times when getting to the original destination is not an option. That means you will be late getting where you need to go. Sure, you can try ducking under the glide slope or descending below the MDA on a non-precision approach, but that might make things much worse than just being late for a meeting.