NTSB Factual Report

HISTORY OF FLIGHT

On August 3, 1997, approximately 0705 Pacific daylight time, the gondola of a Avian Magnum IX hot air balloon, N20146, separated from the balloon envelope after coming in contact with the top of a coniferous tree near the edge of an open field near Donald, Oregon. One of the passengers received fatal injuries, and the commercial pilot and the other six passengers received serious injuries. The balloon, which was owned and operated by American Hot Airlines, sustained substantial damage. The 14 CFR Part 91 commercial flight, which departed Aurora, Oregon about 65 minutes earlier, was being maneuvered to a landing in visual meteorological conditions at the time of the accident.

On the morning of the flight, the pilot and his balloon arrived at the launch site adjacent to Isberg Mobile Home Park, near Aurora, Oregon, around 0530. With help from passengers and crew, the pilot had the balloon inflated by about 0555, and lifted off with seven passengers aboard approximately 0600. According to the pilot, his plan was to fly for 45 minutes to an one hour, land in an open field, exchange depleted fuel tanks for full ones, and then give a ride of about the same duration to a second set of passengers. After liftoff, the balloon drifted in a southerly direction, spending much of its time over or very near Interstate Highway Five (see Area Map). According to a number of passengers, as the balloon moved along the highway, the pilot varied his altitude between 200 and 1,400 feet above the ground (AGL). During the flight, the pilot explained to the passengers that he was making these altitude changes in order to find winds that would carry the balloon away from the highway and the power-lines that ran alongside the highway right of way. After approximately 50 minutes of flight, the balloon had moved about 1.3 miles south of the launch site, was just west of Highway Five, and just south of Donald Road NE. About that time, the pilot switched the balloon's burners from the two liquid tanks (nitrogen pressurized), which had been used for inflation, takeoff and maneuvering up until that time, to the two liquid/vapor tanks. According to witnesses, because of the near-calm winds, when the balloon arrived over the south edge of an open field at this location, it almost completely stopped moving horizontally. One volunteer ground crew member said "...it basically hovered over the south edge of the field..." The witnesses said that the balloon, which at that point was over a number of coniferous trees located on the southern edge of the field, eventually started to move very slowly to the north. About the same time as it began to move to the north, the balloon started to descend. Then according to the six surviving passengers, the gondola made contact with a tree. Four of the passengers said that the gondola hit, impacted, struck or went into the tree, while the two others described it as crushing the tip of the tree or brushing the tree. Soon thereafter, the gondola, with all its passengers still onboard, separated from the balloon envelope and fell down through the tree and impacted the ground. Three of the passengers recalled that after coming in contact with the tree, the gondola tipped or tilted prior to beginning its fall through the tree. The other three passengers did not make any statement about the balloon tipping or tilting, but simply remembered it falling after hitting the tree. None of the passengers recalled any tilting, tipping, or unusual/sudden movement of the gondola prior to it contacting the tree.

There were 10 witnesses on the ground who either provided written statements or were interviewed by local law enforcement officials. Six of the witness were individuals waiting to be passengers on the next flight, three were part of the ground crew for American Hot Airlines, and one was the owner of the land upon which the pilot was attempting to land. All six of the waiting passengers indicated that the balloon descended or lowered into the tree, or simply "hit" the tree. Three of these waiting passengers said that after the balloon contacted the tree, the gondola tilted before the balloon separated and the gondola fell. None of these six witness reported seeing the gondola tilt, make any unusual movements, or begin to separate from the balloon prior to the impact with the tree. The land owner said that he was watching his wife take pictures of the balloon when he heard a loud crashing noise and immediately turned toward the balloon. He said he then saw the "basket" falling down the side of the tree and thought it had "...tipped the top of the tree." He did not mention hearing any other noises prior to the breaking of the branches. Of the three ground crew members, two were volunteers and one was contracted to American Hot Airlines as a crew chief. One of the volunteers said that he was not looking at the balloon when it made contact with the tree, but heard branches breaking and looked up and saw the basket was in the tree "...coming straight down." He also said that for a short period of time he thought the balloon was going to pull the basket out of the tree, but that the balloon "ripped loose and flew up and away." The other volunteer said that the balloon had been hovering at the south end of the field, and it appeared that when the pilot started down to land he had "...no real speed forward." It appeared to this witness that as the balloon neared the tree "...the balloon appeared to hit a wind sheer." He further said that the basket went through the top foot of the tree and then seemed to be pushed down the side of the tree. He said that from where he was, he could hear the branches breaking and the balloon envelope tear as it came down.

The only on-ground witness who reported seeing any anomaly prior to the gondola contacting the tree was the crew chief who was contracted to American Hot Airlines. According to his hand-written statement, dated the day of the accident, he saw daylight between the basket and the trees, and then the "...right side of the basket dropped." He also added that the "Left side of basket broke away immediately after." Although his written statement indicates an anomaly prior to contacting the tree, the official report of Deputy Russell J. Isham, Marion County Sheriff's Office, states the following in reference to the interview of the Crew Chief (Jim):

Jim said that Schwiegart said he was preparing to throw out a drop cable to help secure the landing. Jim said he looked up and saw the balloon coming straight down toward the tree. Jim said he turned just prior to the balloon hitting the tree and told people to , "Call 9-1-1." Jim said that the balloon itself did not "fold" until after the basket hit the tree, it came straight down through the branches. Jim said he was not aware of any problems with the balloon prior to the crash.

On Saturday, September 27, 1997, approximately eight weeks after the accident, the pilot of the balloon was interviewed while still being treated in the hospital. At that time, along with a number of other questions, he was asked what the rate of fall was prior to contacting the tree. He responded that it was initially slow, similar to negative buoyancy. He was also asked how far the balloon dropped prior to contacting the tree, and he said that the balloon had started its fall from about 20 feet above the tree, and that he had activated the burners before tree contact, first just one, and then both. During the interview he did not mention anything about the balloon malfunctioning or beginning to separate from the gondola prior to contacting the tree. At a later date, the pilot was contacted again to review the accuracy of the answers transcribed from the notes taken during the initial interview. At that time, he asked that the phrase "... and prior to tree contact, the basket tipped to one side" be added to the interview transcript.

WRECKAGE AND IMPACT DAMAGE

The gondola of the balloon separated from the envelope at the point where the gondola cable attach rings were held in the loops formed at the bottom end of the vertical load tapes. These loops were formed during manufacture by folding the vertical load straps back against themselves and securing them by sewing a combination lock-stitch/zig-zag-stitch seam. Except for the seam at vertical load tape 13/14, the seam along the lower 13 inches of every vertical load tape had separated. Although the lower seven and one-half inches of the seam on vertical load strap 13/14 had separated, the last two inches of the seam held, and the load strap itself failed approximately nine and one-half inches from the bottom of the loop. Of the 23 vertical load tape seams that separated, all seams from 14/15 through 22/23 contained at least some threads that had melted and formed droplet heads at their tips.

The horizontal load tape and the hem/piping tape around the mouth of the balloon both failed at different points along the bottom of gore number three. The load tape had separated,either partially or completely, from the mouth of the balloon along the bottom of all gores (except gore number 15). Melted seam threads were found on the horizontal load tape along the bottom of all gores from number 12 through number 23 (except for gores 14 and 15). A number of load tape seam threads in gores number 17 through 22 had well defined droplet heads on their tips.

The Nomex lower panel and the nylon envelope fabric of gore three had been torn loose from the two adjoining gores (gore 2 and gore 4) from the mouth of the balloon up to the horizontal load tape located between panels 13 and 14. Gore load tapes 2/3 and 3/4 remained attached to the fabric of gore three. It was noted that gore tapes 2/3 and 3/4 are where the two ends of the parachute deflation line pulley bridle are attached. While inspecting these attachments, it was found that the 5/8 inch diameter parachute deflation line had been tied to the burner support structure, and a two foot section of that structure remained attached to the line.

The nylon envelope fabric of gore 15 (above the lower Nomex panel) was also separated from its adjoining gores, and load tapes 14/15 and 15/16 were still attached to gore 15. Vertical gore seam 15/16 was torn from the mouth of the envelope to the horizontal load tape between panels five and six, and vertical seam 14/15 was torn from just above the lower Nomex panel upward to horizontal tape 5/6. It was noted that gore tapes 14/15 and 15/16 were where the two ends of the bridle for the 1/4 inch diameter parachute deflation line extension are attached.

The Nomex panel on the bottom of gore 3 had torn from the envelope at the time of the accident and remained caught near the top of the tree until recovered at a later date. The seams securing the vertical load tapes along both sides of the panel (tapes 2\3 and 3\4) had failed, and the tapes had separated from the panel. The seam along the top of the panel, connecting the Nomex to the rip-stop nylon fabric above, remained intact, but the nylon fabric itself tore along the upper thread-line of the seam. At the bottom of the panel, the Nomex cloth tore along the seam where the horizontal load tape and the mouth piping had pulled away. There were four locations on this panel where the cloth had been punctured and torn. The first was located about 14 inches from the top of the panel, where there were three small holes, each less than 1/4 inch in length. These holes were along a two and one-half inch long crease in the cloth that ran parallel to the horizontal seams. A white/tan residue was found along the edge of the crease and around the holes.

The second location was approximately twenty-four and one-half inches from the top of the panel, and consisted of five holes varying from 1/4 to 1/2 inch in length. These holes where also located along a crease, approximately five inches in length, and the residue was also present along this crease and around these holes.

The third location was about twenty-seven and one-half inches from the top of the panel, and consisted of one hole about 1/4 inch in length. This hole was located along a seven inch long crease that also ran approximately parallel to the horizontal load tape. What appeared to be the same type of residue was also found along this crease, and there was one 1/8 inch-long chunk of residue material located about 1/4 inch from the hole.

The fourth location, which was centered about two to three inches above the bottom of the panel, was made up of two separate tears. The first (tear 4a) was located along a crease that was about three and one-half inches long. Both the crease and the tear, which was about 3/4 inch long, ran approximately parallel to the horizontal tapes. This area also contained the white/tan residue. The second tear at this location (tear 4b) had a horizontal component, about five inches long, which was along a crease that ran on a line about 20 degrees from being parallel to the horizontal tapes. This tear also contained three separate vertical components, one of which was six and one-half inches long, and two of which were about one and one-half inch long. These vertical tears were approximately parallel to the vertical load tape seams. The cloth surface around this large tear also contained areas where the white/tan residue was present.

The balloon's gondola and burner support assembly sustained damage as it fell through the tree and impacted the ground. There were several areas of damage to both the vertical and horizontal wicker weave of the basket. In addition, damage was sustained by the floor, the skids (underneath the basket), both the bottom and top rails, the burner supports, and the vertical supports. The fuel tanks did not rupture, and their respective valves were turned off immediately after the accident. As part of the investigation, all four fuel tanks were weighed in order to determine the remaining volume of propane. The two nitrogen-pressurized liquid tanks were found to hold approximately 9.5 gallons and 10.9 gallons of propane respectively. The liquid/vapor tanks were determined to hold about .70 and .46 gallons. According to these volumes, the balloon still held 21.56 gallons of the approximately 60 gallons it held at the beginning of the flight.

TESTS AND RESEARCH

As part of the post-accident inspection of the condition of the balloon envelope structure, porosity and strength tests were conducted on the nylon fabric of the envelope cap and envelope gores. The strength tests were conducted using an approved FAA Repair Station-produced testing rig, and were completed in accordance with Avian Balloon Company's Instructions for Continued Airworthiness. The porosity testing was conducted using an FAA-approved Aerostar Porosity Tester, and was completed using the Aerostar Instructions for Continued Airworthiness. Both test procedures were carried out by FAA Repair Station HT6R599N (Go Lightly, Inc.). All required fabric strength tests met or exceeded FAA airworthiness standards, and all fabric porosity tests indicated "0" porosity on the porosity tester gauge.

In addition to the aforementioned tests, a section of a vertical load tape, as well as portions of the balloon's horizontal load tape, mouth piping tape, and seam thread were sent to the University of Massachusetts Dartmouth for testing, examination, and comparison to load tape, piping, and thread samples provided by the balloon manufacturer. These items were examined by Alton R. Wilson, Associate Professor of Textile Sciences. The following is a description of the items provided for testing:

a. Two sections of the horizontal load tape. One section was 36 inches long and the other was 42 inches long. Each section contained one side of the tape failure.

b. A 16 inch section of the envelope mouth piping. One end of this piping contained a failure that occurred near the same location as the failure on the horizontal load tape.

c. A portion of the vertical load tape running between gores number 13 and 14. This is the only vertical load tape in which the load tape itself failed.

d. A portion of the vertical load tape which ran between gores 20 and 21. This tape exhibited extensive meltin

NTSB Probable Cause

The pilot's failure to maintain clearance from a tree that was located on the edge of the field in which he was attempting to land.