On August 17, 2001, at 1920 central daylight time, a Cameron Balloon Shuttle 90, N8094S, piloted by a commercial pilot, was destroyed during a descent for landing when the balloon's envelope caught on fire. The balloon was participating in a balloon festival in Centralia, Illinois. The pilot reported looking up prior to a burn when he saw the entire mouth of the balloon on fire. Visual meteorological conditions prevailed at the time of the accident. The 14 CFR Part 91 business flight was not operating on a flight plan. The pilot and passenger received minor injuries. The local flight originated from a city park in Centralia, Illinois at 1830. The pilot reported aborting a landing attempt due to high weeds. Fifteen minutes later he approached a second field for landing. At least one minute later, he looked up prior to performing a burn and noticed that the mouth and pressure cone of the balloon was on fire and sustaining a flame. The wind was from the northwest at one knot. He stated that he had both burner pilot lights on for the flight and may have shut one of them off when he saw the balloon on fire. The pilot said that he didn't know exactly what happened nor does he know how it caught on fire. The pilot, age 27, was employed as a contact pilot for the operator. The pilot held a commercial pilot certificate with a lighter-than-air free balloon rating. He reported a total flight time of 720 hours all of which were in lighter than air aircraft. Fifteen hours of his total flight time was in the accident balloon. The balloon, serial number 4822, was type certificated under Federal Aviation Regulations (FARs) Part 31. The balloon received a standard airworthiness certificate on July 25, 2000, and was registered to Ballunar Liftoff Festival, Inc., on August 18, 2000. The balloon received an annual inspection on July 19, 2001, at a total time of 10 hours and accumulated a total time of about 15 hours. The balloon envelope is shaped like a space shuttle with a vertical height of 48 meters and a wing span is 32.74 meters. The balloon had an actual volume of 137,000 cubic feet and an active volume of 95,000 cubic feet. According to the Cameron Balloons, the mouth was constructed of Nomex and had a diameter of 10 feet 2 3/4 inches. The vertical distance from the burner frame to the end of the Nomex was 22 feet. Typical mouth diameters for other balloons are: N-90, 12 feet; Champaign-90, 11.8 feet; Lightbulb-100, 9.14 feet. The typical distances for the burner frame to the end of the Nomex on these balloons are: N-90, 13.8 feet; Champaign-90, 12.9 feet: Lightbulb-100, 12 feet. The balloon was equipped with a MK IV Ultraburner, serial number B5438-B5439, double burner. The burners were attached to the burner frame. The crossover valve was in the off position, both whisper valves were in the on positions, one pilot valve was in the on position, and the other pilot valve was in the off position. Testing of nine samples of balloon fabric and fabric from the accident balloon was performed by the Federal Aviation Administration at the William J. Hughes Technical Center. Testing was based upon Appendix F of Federal Aviation Regulations Part 25 in which a specimen is supported vertically and exposed to a flame from a burner with a nominal diameter of 3/8 inch inside diameter tube adjusted to give a flame of 1 1/2 inches in height. The minimum flame temperature must be 1,550 degrees Fahrenheit. The lower edge of the specimen must be 3/4 inch above the top edge of the burner. The flame must be applied to the center line of the lower edge of the specimen. Descriptions of the nine samples tested and their results are as follows: Sample 1 was taken from the top of the fuselage approximately between the rudder and banner area. The sample was also located 18 inches from the an area completely consumed by fire. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were recorded as 18 seconds, 14 seconds, and 15 seconds. In all of the three test runs, the specimens were consumed. Sample 2 was taken from the fuselage near the leading edge of the right wing. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were recorded as 15 seconds, 13 seconds, and 13 seconds. In all of the three test runs, the specimens were consumed. Sample 3 was taken from an inner former of the left wing. Two test runs were performed due to a limited amount of material and an ignition time of 12 seconds was recorded for both. The corresponding flame out time was recorded as 5 seconds and 15 seconds. In both test runs, the specimens were consumed. Sample 4 was taken from the bottom of the fuselage. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were recorded as 10 seconds, 10 seconds, and 15 seconds. In all of the three test runs, the specimens were consumed. Sample 5 was taken from the upper surface of right wing. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were 28 seconds, 25 seconds, and 23 seconds. In all of the three test runs, the specimens were consumed. Sample 6 was polyurethane coated rip stop nylon not from the balloon. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were 0 seconds, 49 seconds, and 12 seconds. The burn length during the first run was 2 1/4 inches. During the second run, the specimen was consumed and was noted to have multiple drippings which were self extinguishing when they hit the floor. Sample 7 was single side silicone coated rip stop nylon not from the balloon. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were 32 seconds, 43 seconds, and 43 seconds. In all three test runs, the specimens were consumed. Sample 8: Ripstop nylon silicone coated on both sides. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were 20 seconds, 8 seconds, and 12 seconds. In all three test runs, the specimens were consumed. Sample 9: Ripstop nylon with urethane coating. Three test runs were performed and an ignition time of 12 seconds was recorded for each run. The corresponding flame out times were all 0 seconds. The burn lengths were 3 inches, 2 1/2 inches, and 1 3/4 inches. FAR Part 31 prescribes the airworthiness standards for manned free balloons to receive U.S. type certification. FAR Part 31 addresses strength requirements and design construction but does not prescribe materials flammability requirements.