WPR20LA246 · Air Tractor AT802 A

The pilots for both airplanes departed on an approximate 30-minute flight to a fire traffic area (FTA) under contract with the Department of the Interior (DOI) to extend a fire retardant line near a wildland fire for the third time that day. One airplane was flying as the lead, and the other airplane was in trail. After they received permission to enter the FTA and instructions with the drop heading, the pilots of both airplanes followed a third airplane to the drop area on a path that resembled an airport traffic pattern.

The pilot in trail had maintained about 1,500 ft distance from the lead airplane for most of the flight until they were on the final leg of the drop when the trail airplane started to encroach on the lead. The trail airplane was about 500 ft behind and 100 ft below the lead airplane about 5 seconds prior to the drop. After deploying the retardant, the lead airplane began a climb while the trail airplane flew through the lead airplane’s retardant plume. The pilot of the trail airplane announced over the radio that he had retardant on his windshield, was executing a go-around, and possibly performing an emergency dump of his load, as well. The trail pilot then dumped his load and immediately entered a climb. At this point, he was 50 ft behind and 125 ft below the lead airplane. The airplane in trail then collided with the lead airplane from below and they subsequently impacted the ground. Both airplanes were destroyed by postcrash fire.

Postaccident examination of the trail airplane did not reveal any preimpact mechanical anomalies. Portions of the lead airplane’s flight control system had either been destroyed by fire or were missing; however, the airplane’s flightpath and climb were consistent with a controlled flight and do not indicate the airplane experienced any preimpact anomalies. The propeller blades for both airplanes displayed signatures consistent with high power settings at the time of impact. Two of the lead airplane’s propeller blades displayed yellow paint transfer marks consistent with the mechanical cuts observed on the trail airplane forward of the engine firewall. The orientation and position of the cuts, along with rubber marks on the cockpit step suggests the trail airplane impacted the lead airplane from below, slightly ahead, and 20° left of the lead airplane’s heading.

Both pilots were qualified to fly the airplanes by regulatory standards and their contractual obligations with DOI. However, there was no record either pilot had received overrun procedures training in the 4 years prior to the accident.

A video of the trail airplane’s go-around maneuver showed the airplane in a controlled climb; however, the direction was not consistent with the exit procedures that had been briefed prior to the retardant deployment. Had the pilot executed a left turn and climbed as prescribed by the brief, he would have likely remained clear of the lead airplane.

The amount of retardant that contacted the trail pilot’s windshield could not be determined for lack of evidence. However, the pilot’s radio call indicated he had retardant on his windscreen and that he was performing an emergency procedure to go-around and dump the retardant indicated the pilot was concerned about the loss of visibility.

The wildland fire procedures stated that the pilot in trail would have been responsible for maintaining separation from the lead airplane. The pilot in trail had at least one reported event that captured him flying too close to a lead airplane during another flight. However, the investigation was unable to confirm if there were other events and an established pattern of this behavior. The available evidence is consistent with the trailing pilot’s failure to maintain separation from the lead airplane during a fire retardant deployment, which resulted in his flying through a plume of retardant that likely obscured his view through the airplane’s windshield and led to a collision with the lead airplane.

Verbatim NTSB analysis and probable cause from the NTSB dataset