SEA05MA085

Prior to the accident flight, the air tanker airplane had flown 9 times on the day of the accident, for a total flight time of 5 hours and 46 minutes. The purpose of the flights was to provide recurrent training for pilots scheduled to conduct fire-fighting operations for the United States Department of Agriculture Forest Service (USFS) during the upcoming fire season. Pilots aboard the airplane during the earlier flights reported no mechanical problems with the airplane. As was the mission for other flights that day, the accident flight, the tenth flight, was a training flight to conduct practice drops of water over an area of rugged mountainous terrain located north of the airport. Aboard the flight were the captain, the copilot, and the company's Chief Pilot, who was providing flight instruction. The Chief Pilot had been on board all the flights that day. Radar data indicated that the flight departed the airport and proceeded about 10.5 nautical miles to the north where it began maneuvering in a manner consistent with the conduct of practice water drops. The data showed the airplane entering and then following a right-hand racetrack pattern oriented northeast-southwest, during which it appeared that practice drops were being made on the southwest leg. During the last minute of the flight, the airplane was on the northeast leg of the racetrack pattern, flying up a valley that was oriented northeast-southwest with uphill being to the northeast. The airplane's flight path was initially near the middle of the valley. During the last 36 seconds of the flight, the airplane's flight path began to deviate towards the rising terrain on the eastern side of the valley. The last two radar hits show the airplane at an altitude of less than 100 feet above ground level. The airplane had completed approximately 1.75 circuits of the racetrack when the data ended with the airplane at 2,900 feet msl, heading northeast. The initial impact point was located about 2,150 feet northwest of the last radar data point at an elevation of about 2,450 feet msl, indicating the airplane entered a left descending turn and completed about 90 degrees of turn between the last radar hit and the impact. No distress calls were received from the airplane. According to local authorities, witnesses observed a "fire ball" at the time of the accident. Examination of the wreckage site revealed that the airplane impacted on about a 304 degree magnetic heading in a 40 degree left bank with the left wingtip striking the ground first. The airplane was severely fragmented and a severe post crash fire burned most of the structure and surrounding vegetation. Remnants of the entire airplane were accounted for at the wreckage site. There was no evidence of pre-existing structural failures or impact with foreign objects in any of the wreckage. All four of the engines and propellers were found at the wreckage site. The engines did not have any indications of an uncontainment, case rupture, or pre-impact in-flight fire. All four engines had damage to the compressor and/or turbine rotors that was consistent with engine operation. The cockpit engine instrumentation gages indicate that all four engines were running at about 2,200 shaft horsepower at impact. Review of the airplane's maintenance records did not reveal any chronic issues with the airplane nor any system/component anomalies that would have contributed to the accident. The airplane was equipped with neither a cockpit voice recorder (CVR) nor a flight data recorder (FDR) and Federal Aviation Regulations did not require the airplane to be so equipped. The reason for the in-flight collision with terrain could not be determined.

History of flight

On April 20, 2005, about 1850 Pacific daylight time, a Lockheed P-3B air tanker airplane, N926AU, registered to and operated by Aero Union Corporation, Chico, California, impacted terrain while maneuvering about 12 nautical miles north of the Chico Municipal Airport (KCIC). The three crewmembers on board, the airline transport rated captain, the commercial rated copilot and the operator's chief pilot, who was providing flight instruction, were fatally injured, and the airplane was destroyed by impact forces and fire. The training flight was being conducted under 14 CFR Part 137. Day visual meteorological conditions prevailed, and a company flight plan was filed. The flight departed KCIC at 1838.

According to Aero Union Corporation records, prior to the accident flight, the airplane had flown 9 times on April 20, for a total flight time of 5 hours and 46 minutes. The purpose of the flights was to provide recurrent training for pilots scheduled to conduct fire-fighting operations for the United States Department of Agriculture Forest Service (USFS) during the upcoming fire season.

The accident captain flew the first two training flights of the day. In an interview conducted by Safety Board investigators, the copilot on those flights stated that although the accident captain had a lot of time in P-3s, he did not have much time in the low level environment. On flights 1 and 2, they dropped 2 loads of water. The accident captain briefed the copilot well regarding the drop area and escape routes. Normal drop speed was between 130 and 140 knots, and the mission was flown at an altitude about 200 feet above ground level (AGL). There were no open minimum equipment list (MEL) items and no "abnormals" were encountered during the flight.

On flight 1, they climbed out of Chico to an altitude of 3,500 feet. The Chief Pilot assumed the role of Attack Captain and selected a simulated spot fire area that needed the first portion of the drop across the head of the fire. The Chief Pilot noted that the accident captain never called for the Before Drop checklist. When conducting a drop, basically, they are trying to simulate a landing pattern with the landing gear retracted. They were at 120 knots approaching a "saddle" (between two peaks) before the push over to make the drop across the simulated head of the fire. The second part of the drop was to be made down the left or right flank; therefore, the accident captain attempted to do the drop out of a right hand turn. They had to go around because they were not lined up on the target. The copilot stated that the go-around decision was made late. The accident captain then decided to do the drop from a left hand pattern. The third drop was also made from a left hand pattern. When they departed the drop area and headed back to Chico, the accident captain faked a heart attack as a cue for the copilot to take over and land the airplane.

On flight 2, the Chief Pilot selected a drop area that was along a road that curved. The accident captain selected a right hand pattern for the drop but had to make a go-around. He then selected a left hand pattern and made the drop while in a turn. The Chief Pilot did not like that the accident captain made the drop while in a turn and briefed him accordingly, stating the dangers of possibly hitting a snag or other hazards. The copilot stated that he did not remember the accident captain making a go-around call but he did remember him saying "check flaps approach" as the go-around was initiated. The copilot waited for a minimum of 140 knots before retracting flaps beyond Takeoff/Approach setting.

The copilot stated that they try not to exceed 30 degrees of bank angle. The accident captain got close to that at times but did not exceed it. Final approach entails a constant descent to a level altitude. On these runs, the lowest altitude was perhaps 200 feet above terrain, but certainly not less than 150 feet. In case of a go-around, there is no lag in power; power is available right away.

The copilot stated that the accident captain was not as comfortable in the low level environment as he was in the normal flight regime, but the accident captain was confident. The copilot stated that he believed that the accident captain was task saturated in the low level environment, but that was not uncommon for other captains, as well. There was not much talk on final approach and the copilot stated that he did not think there should be. There were no simulated emergencies during the drop. They had already done a lot of simulated emergences at other times during training.

In an interview conducted by Safety Board investigators, the captain of flights 3 and 4 on the day of the accident stated that they took on a load of water and did three drops in a triangular area with a simulated boxed-in area of fire. They normally train in the area where the accident occurred, using a variety of different targets in that general area. The captain does all the flying during the actual fire-fighting activity. There were no mechanical problems with the airplane; it was a good flying airplane. He wrote up a couple of light bulbs after the flight but there were no other problems noted.

Flight 7 was a type rating check ride. The accident captain occupied the copilot seat on the flight and the Chief Pilot was conducting the check from the center seat. According to company records, the flight lasted for 1 hour and 56 minutes.

In an interview conducted by Safety Board investigators, the captain of flights 5, 6, 8 and 9 on the day of the accident stated that he flew the accident airplane for a total of 2 hours and 2 minutes on these four flights. The captain did not report any anomalies with the airplane.

The accident airplane had been fueled to 22,500 pounds just prior to flight 9. After flying for 28 minutes, flight 9 returned to KCIC at 1815. After returning to KCIC, no additional fuel was added but the airplane's retardant tank was loaded with 2,550 gallons of water.

The accident flight, the tenth flight of the day, taxied out for takeoff at 1834 and departed at 1838. As was the mission for other flights that day, the accident flight was a training flight to conduct practice drops of water over an area of rugged mountainous terrain located north of the airport. Aboard the flight were the captain, the copilot, and the Chief Pilot. The Chief Pilot had been on board all 10 flights that day.

According to radar plots of data provided by the U. S. Navy, the flight departed the airport and proceeded north while climbing to and maintaining between 3,000 and 3,500 feet msl. The airplane turned northeast and approximately 10.5 nautical miles from the airport, descended to between 2,700 and 2,900 feet msl and executed a left-hand 270-degree turn. The airplane then executed a second level left-hand turn of about 270 degrees coming out of the turn on a southwest heading. While on the southwest heading, the airplane descended to 2,300 feet msl and then began climbing. The airplane continued to climb to an altitude of 3,200 feet msl as it executed a right-hand race track pattern oriented northeast-southwest. In the race track pattern, as the airplane returned to a southwest heading, it again descended to an altitude of 2,300 feet msl and then began climbing. As the airplane continued to follow the race track pattern and turned from the southwest to the northeast it climbed to 2,700 feet msl. The last minute of radar data showed the airplane continuing to head northeast with its altitude varying as follows:

Time Mode C Altitude

18:49:55 2,700 feet

18:50:07 2,500 feet

18:50:19 2,500 feet

18:50:31 2,700 feet

18:50:43 2,600 feet

18:50:55 2,900 feet

The airplane had completed approximately 1.75 circuits of the racetrack when the data ended at 1850:55 with the airplane at 2,900 feet m