Launch site: Columbia Scientific Balloon Facility, Palestine, Texas, US
Balloon launched by: National Scientific Balloon Facility (NSBF)
Balloon manufacturer/size/composition: Zero Pressure Balloon SF3-424.37-080-NSCHR-04
Balloon serial number: W29.47-3-08
Flight identification number: 1482P
The balloon was launched on June 5 1989, using the dynamic method, with assistance of launch vehicle (Tiny Tim), from the balloon base at Palestine, Texas.
The flight developed without troubles, but in the last part of it, the balloon had gone over a thunderstorm west of Fort Worth, Texas. By unknown reasons an uncommanded release of the payload occured along with the separation of the parachute from the gondola, wich started a freefall back to earth from 120.000 feet. Normally when the gondola is attached to the parachute it takes about 50 minutes to reach ground but this time it took only two minutes. The payload hit the ground at an estimated speed of 700 miles per hour, on a farm located four or five miles outside Graham, Texas.
The NSBF's Site and Operations Manager were monitoring closely the telemetry from the balloon when the abnormal termination occured so they almost inmmediately realized the problem they had. Rapidly got a vehicle and drove to the zone where a NASA recovery team was already waiting for the balloon arrival. While driving toward Graham, they received a report that people had seen the parachute impact on the ground near Weatherford, Texas, about 80 miles to the southeast. Some people thought it was a hot air balloon falling and reported that they had seen people waving their arms. The large parachute only had about 100 pounds of weight on the bottom so it drifted a very long way and came down very slowly.
Meanwhile, the balloon impacted on the north shore of the Possum Kingdom Lake. It was reported by a lonely fisherman in a small boat on the lake who was very shaken because he had seen the balloon coming down through the storm clouds and thought the wrath of God was coming to get him. According to his statement, the 3000 lbs of balloon material (wich had rolled up as it fell and create a chevron shape) made a lot of noise and sounded like a locomotive. At impact it covered an area of about 900 square feet.
When the NSBF's managers arrived to Graham, they meet the recovery team and started searching for the payload. Afetr renting a small Cessna airplane they put aboard one men of the team who was familiar with the terrain as an observer. During this time a TV station from the Fort Worth/Dallas area had received a report about the balloon incident and had sent a video crew to find the NSBF people and if possible to take some video of the payload or anything that looked interesting. That was when a sort of "mouse and cat" game started between the two groups. NSBF moved the people of the recovery team to another town nearby, being forced to coordinate the search effort and to avoid the television reporters at the same time.
Finally, after two days of search, the Cessna spotted the payload in the farm and directed the rest of the team to it. It was located between the trees, of a fruit orchard.
The NSBF people contacted the land owner asking for authorization to allow the recovery team to get into the orchard and retrieve the gondola. After access was granted, they reached the crash point. Althought the heavy impact, the cylindrical-shaped payload was found essentially in one piece because during the fall it was trailing steel suspension cables, wich kept it vertical rather than tumble so when it hit the ground, it impacted on the bottom which was slightly curved. The impact caused about a six to eight inch deep crater in the soil, but no debris thanks to the pressure vessel wich acted like a shock absorber. The nitrogen contained inside it, was pressurized and blew out the optical window of the instrument allowing the gas to escape and absorbing some of the impact forces. The sides of the pressure vessel were made from aluminum and folded up a bit, but they did not fracture or come apart. All of the scientific systems inside the vessel were of course crushed by the impact about 6 inches on the crater inside floor.
The recovery team collected the remains of the crash and loaded it on a trailer, covered the load with canvas, and returned back to the launch base at Palestine, Texas. Before leaving, the NSBF paid for the damages to the property and obtained a signed release of liability from the land owner. By the time the TV crew figured out what was going on the NSBF personel had vanished from the scene. What would be turned into a nightmare of bad press for the agency, turned into a a short story in the evening news for a day or two. Then, nothing more was heard or seen of the incident.
However, that was not the end of the history as far as NASA or NSBF was concerned. Initially the technicians were very puzzled about why the payload had been released from the parachute.
The electronic systems recovered were examinated and it was obvious looking at the circuit boards that large electrical currents had flowed through them. The transistors and integrated circuits were opened and analysed in the NASA laboratories where experts determined that lightning had not directly struck the suspension system, but induced currents had caused the damage.
Apparently a bolt of lightning had traveled up from the clouds close enough to generate an electromagnetic pulse in the flight electronics which caused several events to happen at the same time. First, the parachute was released from the balloon. Second, were turned on the motors which pulled the safety pins from the parachute/payload release system (used to separate them once on the ground to avoid the gondola from being dragged across the soil) and third, the explosive cutter used to release the parachute from the payload was fired.
It took 3 seconds for the motors to pull the safety pins meanwhile takes 7 seconds for the parachute to open and take the load of the payload. The design at that time was that if the parachute was open and there was a shear load on the pins due to the payload weight, the clutch on the motor will not pull the safety pins and release the payload. So in this situation the payload was in free fall for seven seconds and there was no load on the safety pins to keep them from retracting. When the parachute opened up, the payload was released and it free fell to the ground from 120,000 feet.
Initially in the system design, the time for the motor to retract the safety pins was 10 seconds, more than enough time to enable the parachute to open and take the weight of the payload. With this time window the clutch would not enabled the motor to retract the pins further, thus securing the parachute to the payload for a safe land. As turned out during the investigation, over time and for some unknown reason, the safety pin retraction time was reduced to three seconds.
One of the side effects of this incident was to give some credence to the theory that lightening can indeed travel upward from the clouds to space, a phenomena formally discovered a few years later and actually widely known as "SPRITES". But, without doubt, the most important derivation of the investigation was the change of the NASA policy on balloon flights over storms. Prior to this event, stratospheric balloons often traveled over clouds without incident with the only concern of the height loss due to the albedo or temperature of the earth below the balloon became much colder than before causing the colding of the gas. After that, the policy became to not launch balloons near thunderstorms unless could terminate the balloon flight safely before got to the thunder storm area. Also the agency became more conservative about launching a balloon if the high level wind forecast and the weather forecast resulted in a chance that it aproached to a thunderstorm area.
As a footnote, beside the little publicity that the incident had, after the COLUMBIA reentry disaster in 2003, an article appeared on several newspapers theorizing about the possibility that the Shuttle was struck by a SPRITE before desintegration. In the final part of the article the author made mention of this balloon incident but made a wrong location of the impact site stating that the payload freefall on "an angry Dallas resident front yard" wich of course was not true, as the nearest dwelling was about 1/2 mile away from the crash point and Graham was located far from that city.