NTSB Factual Report

HISTORY OF FLIGHT

On May 9, 1999, at 0746 hours Pacific daylight time, an Aerostar International, Inc., Rally RX-8, N70290, operated by Balloons by the Sea, Marina, California, drifted into power lines during the initial takeoff climb from an open field in Morgan Hill, California. Visual meteorological conditions prevailed, and no flight plan was filed for the 14 CFR Part 91 aerial sightseeing passenger flight. The balloon was destroyed. The commercial pilot was seriously injured, and the two revenue passengers were fatally injured. The planned local area flight commenced from the field about 0745.

The pilot reported to the National Transportation Safety Board investigator that he drove onto the open field and chose the location where the balloon was set up for the flight. During a previous occasion, he had used the same field to launch the balloon, and he was aware of the field's proximity to the power lines.

The pilot's sole ground crewmember reported to the Safety Board investigator that he was not a pilot. The crewmember stated that the pilot set up the balloon for takeoff. The crewmember identified the location where the balloon had been set up. This location was subsequently determined to be approximately 298 feet west of the east side of the rectangular shaped field. Power lines are located along the field's east side, where the entranceway to the field is located.

The pilot reported to the Safety Board investigator that he was unable to complete the pretakeoff check of the balloon before taking off because the wind speed increased rapidly. However, he believed that all of the balloon's systems were functional, and no mechanical malfunctions were experienced during the flight. As the balloon ascended it drifted into the power lines.

The ground crewmember stated that a fire erupted within seconds following the wire impact. Flames were visible emanating from the basket in horizontal and vertical directions. While the basket was still airborne, the partially burned pilot exited the basket and fell to the ground.

Responding fire department personnel reported that the pilot stated he had jumped out of the basket. The passengers remained in the basket. The basket, which was engulfed in fire, came to rest in a broccoli field about 121 feet downwind (east) of the impacted power lines. Both passengers subsequently died from thermal injuries.

OTHER DAMAGE

Pacific Gas and Electric Company personnel reported that they examined the impacted 21,000-volt distribution electric lines, which are oriented in a north-to-south direction at the eastern edge of the field. Evidence was observed that, at 43 feet above ground level (agl), the north A phase power line had been pushed eastward until contacting the middle B phase line. This action created a fault. The associated circuit relayed and locked out, thus interrupting power to 3,558 customers. Power was restored by 1355.

PILOT INFORMATION

The pilot was issued a student pilot certificate on March 5, 1992 with a limitation that it was for balloons only. He soloed on September 2, 1992.

Upon completion of the balloon school's course of instruction, the pilot was issued a commercial pilot certificate on September 8, 1992. The certificate bore the following ratings and limitations: "lighter-than-air, free balloon, limited to hot-air balloon with airborne heater." When the certificate was issued, the pilot's solo, dual, and total balloon flight time was 5.5, 14.1, and 19.6 hours, respectively.

The pilot first flew an Aerostar RX-8 model balloon when the registered owner of the accident balloon checked him out in it, on February 15, 1999. According to the pilot's logbook, by the accident date his total flying experience in the balloon was amassed during 10 flights between February 15 and April 25, 1999. By the accident date, his total flight experience in the balloon was recorded as 11.9 hours. The pilot's total balloon flying experience was approximately 122.7 hours.

BALLOON INFORMATION

FAA Certification and Ownership.

Aerostar manufactured the balloon in 1994. The Federal Aviation Administration (FAA) issued the balloon a standard airworthiness certificate indicating it was constructed in conformity with the requirements of Type Certificate (TC) No. A15CE.

The new balloon, with its Aerostar, HP III dual burners, having a rated power of 15,000 BTU per hour per burner, and a Model RWS-454 wicker basket (gondola) was sold to the current registered owner in November 1994.

The balloon envelope's serial number is RX8-3214. In February 1997, basket serial number RWS-442 and two fuel tanks, bearing serial numbers V853 and V843, were placed in the balloon.

The balloon was designed with a parachute style panel covering the deflation port opening in the top of the envelope. Pulling on the deflation line allows for the rapid discharge of hot air.

Maintenance & Inspection History.

The operator reported to the Safety Board investigator that he contracted out for the balloon's maintenance. On May 19, 1998, at a total time of 75:50 hours, the balloon received its last annual inspection. No subsequent evidence of any maintenance was observed in the logbook.

The last flight listed in the logbook is dated April 25, 1999. On that date, the balloon envelope's recorded total time was 89:40 hours. The basket's total time was approximately 289 hours.

Logbook Entries.

The following statement appears in the FAA Approved Flight Manual (AFM): "All time during which the balloon is inflated and buoyant/upright . . . must be recorded in the envelope log book."

An examination of the balloon's logbook indicated, that at the time of the accident flight, the balloon had been operated for 13:50 hours since its last inspection. In particular, between February 15, and April 25, 1999, the balloon had been operated for a total of 9:20 hours during 8 flights, according to its logbook record.

No record was observed in the balloon's logbook for the pilot's two, 1-hour-long flights in the balloon on March 9 and 16.

Helmets.

The AFM contains a listing of items that must be on board during flight. In pertinent part, the equipment list states that "helmets for all occupants" must be carried as required equipment.

On May 13, 1999, the Safety Board investigator interviewed the operator and inquired if helmets were available for use by the occupants in the accident balloon. The operator stated that his company had helmets, but that they were kept "in the office" and were not used.

Fuel Tanks.

Aerostar personnel and balloon records indicate that upon delivery, the balloon was equipped with two, 15-gallon vertical stainless steel propane storage fuel tanks, serial numbers V868 and V880. Aerostar reported that its tanks are especially manufactured by the company for use in its hot air balloons.

According to the TC, the accident balloon's fuel capacity was limited to two or three or four V-10 (10-gallon capacity) or V-15 (15-gallon capacity) vertical tanks. No authorization was issued for utilizing a combination of both 10- and 15-gallon tanks in the basket.

In February 1997, stainless steel fuel tank numbers V853 and V843 were installed in the basket to replace the original tanks.

The Aerostar participant reported to the Safety Board investigator that to the best of his knowledge, the manufacturer has never experienced a rupture to one of its steel fuel tanks during a wire strike occurrence.

The pilot reported that prior to the day on which the accident occurred, he had placed the extra 10-gallon aluminum fuel tank in the basket. The pilot stated that the balloon's owner had informed him that it was permissible to use the extra tank in the balloon.

No evidence of a Major Alteration and Repair, FAA Form 337, or other logbook entry was found for the addition of a third fuel tank in the balloon's basket. No weight and balance amendment was found for its added weight.

METEOROLOGICAL INFORMATION

The closest aviation weather observation station to the accident site is at the Reid-Hillview of Santa Clara County Airport, in San Jose, California. The airport is located at an elevation of 133 feet mean sea level (msl), and is about 15 miles and 304 degrees (magnetic) from the accident site. At 0745, the airport reported that its surface wind was from 310 degrees at 5 knots (6 mph).

According to the operator, balloon pilots flying in the Morgan Hill area frequently obtain the local wind velocity from a weather sensor. The sensor consists of a privately owned and maintained anemometer, which measures the wind direction and speed. The device is located about 60 feet agl and records wind direction based upon a 16-point compass (nearest 22.5 degrees of azimuth). The data is automatically downloaded for retrieval via telephone. The anemometer serves as an adjunct to other, official, weather reporting facilities. The anemometer is known as the Morgan Hill Weather Robot.

The Robot is located about 2.4 statute miles and 145 degrees (magnetic) downwind from the accident site, at the following global positioning satellite (GPS) coordinates: 37 degrees 07.66 minutes north latitude by 121 degrees 37.83 minutes west longitude. The Robot samples the wind velocity at 15-minute intervals.

The operator reported that following the accident he contacted the Robot via telephone, and he made a written record of its wind history information. Between 0735 and 0750, the wind was from the north at 5 mph, with 9 mph gusts.

The ground crewmember and the pilot reported that upon arrival at the launch site field the surface wind was calm. The ground crewmember stated that no pibals (helium filled balloons) were utilized at the site to determine the wind velocity aloft.

The pilot reported that he was uncertain whether a pibal had been launched; however, he recalled throwing a wad of dirt and grass into the air. The wind was initially calm. Minutes later, during the envelope inflation process, the wind speed increased. The envelope started rolling side-to-side, and the pilot entered the basket. By the time the passengers entered the basket, the balloon was being dragged by the wind over the ground.

The ground crewmember indicated that because of the wind, during the takeoff and initial climb, the balloon ascended diagonally.

COMMUNICATION

According to the operator and the pilot's sole ground crewmember, the pilot did not have a communication transceiver with him in the balloon. The ground crewmember indicated that he had no method of communicating with the pilot during the planned flight.

WRECKAGE AND IMPACT INFORMATION

Accident Site.

From an examination of the accident site and balloon wreckage, and from witness statements, the balloon was found to have collided with energized power lines at the following GPS coordinates: 37 degrees 09.63 minutes north latitude by 121 degrees 38.69 minutes west longitude. The elevation of the grass-covered open field from which the takeoff commenced is about 400 feet msl.

The launch site field's total size is about 922 feet long by 368 feet wide. Because of the terrain and proximity to neighboring homes, the field's usable size is about 798 feet long by 368 feet wide, with the long axis of the field oriented approximately parallel to the balloon's easterly (109 degree, magnetic) flight path.

Based upon the width (about 41 inches) of ground scar (flattened weeds) that was the approximate size of the 35-inch wide basket, disturbed soil, and the ground crewmember's statements, the balloon was set up for takeoff about 298 feet upwind (west) of the power lines. Beginning at the setup local, ground scar evidence was observed in a downwind direction that had an appearance consistent with the size of the basket. The ground scar evidence was also consistent with the size of the basket throughout its ground travel (drag length) over an approximate 84-foot-long distance (at which point the basket became airborne). The ground scar ended about 214 feet upwind (west) of the power lines.

At the point of impact, the power lines were about 43 feet agl level. The basket came to rest about 121 feet downwind from the lines.

The wicker basket was virtually consumed by fire. The balloon's envelope was substantially fire damaged.

MEDICAL AND PATHOLOGICAL INFORMATION

A witness reported that, on May 8, the pilot had departed a restaurant about 2230. Thereafter, based upon the distance between the restaurant and the pilot's domicile, the earliest that he could have retired was 2300.

The pilot reported to the Safety Board investigator that he had consumed "beers" during his birthday party at the restaurant. Upon his arrival at his domicile, he made at least one phone call using his computer before retiring for the evening. The pilot stated he does not recall when he retired. The pilot stated that he awoke at 0430 on May 9. The FAA reported that at 0503 on May 9, the pilot commenced receiving a weather briefing via telephone.

The pilot possessed an expired third-class aviation medical certificate that was issued in 1990. No medical certificate is required for operation of the balloon. The pilot also reported that he had no physical impairments or physical limitation that would inhibit his actions or result in him being unable to fly the balloon.

TESTS AND RESEARCH

Balloon Examination.

The examination of the burnt remains of the balloon and basket revealed that the balloon's deflation panel was in the closed position. The associated rip line was found functional.

The Aerostar participant reported that the balloon appeared to be in a "standard configuration" with one notable exception. That was the presence of an extra fuel tank and a manifold hose that was connected to the tank. The extra tank had a 10-gallon capacity and an aluminum structure. It was observed broken open and ruptured on one side. The two other standard fuel tanks were of steel construction, and each had a 15-gallon capacity. They were visually examined and were found structurally intact.

The pilot light control to one of the burners (referred to herein as the number one burner) was found in the closed position. The pilot light control to the number two burner was found in the open position.

The steel fuel tank (serial number V843) was found with its fuel hose not connected to the number one burner's inlet fitting. The steel fuel tank (serial number V853) was found with a fuel hose connected to the number two burner and with a manifold to the aluminum fuel tank.

No current inspection or certification data was found for the aluminum fuel tank. This tank bore the name "Lenox" stamped in its side, and was also noted with the numbers "0435." Aerostar did not manufacture or supply this fuel tank to the operator.

The operator subsequently located a record for this Lennox tank. The record indicated that in June 1989, it was included in an annual inspection for a Raven S66A balloon, N5732H, which at that time was being operated by Balloons by the Sea.

Wind Speed Evaluation.

The Aerostar participant reported that, based upon the size of the balloon, the statement of the ground crewmember, and balloon dragging data, the estimated wind during the launch was between 7 and 10 miles per hour (mph). The participant verbally reported that, it was his experience, in a 6-mph wind the basket would not have been dragged through the field.

ADDITIONAL INFORMATION

Aerostar FAA Approved Flight Manual (AFM).

The manufacturer provides an AFM for use in operating the balloon. In the AFM's preamble, a description is provided for the three levels of information importance contained within the manual. The levels are listed as Notes, Cautions, and Warnings. Regarding Warnings, the AFM states: "A WARNING alerts the reader to information and instruction which are imperative for the safe operation of the balloon system. Failure to adhere to these warnings may result in severe damage, injury or death."

Regarding power lines, the AFM provides, in pertinent part, the following warning statement: Power lines represent a major source of

NTSB Probable Cause

The pilot's selection of an inadequate takeoff site that did not allow sufficient horizontal clearance from obstacles in the takeoff path, and, his failure to ensure that all burners were connected to fuel sources, which reduced the available climb capability of the balloon. Also causal was the pilot's failure to activate the emergency vent panel and land prior to contact with the power lines, and, his use of an unapproved aluminum fuel tank, which ruptured during the power line contact and ignited a fire. A factor in the accident was the pilot's failure before takeoff to use pibals to quantify the wind speed.