NTSB Factual Report

HISTORY OF FLIGHT

On March 23, 2007, approximately 2105 mountain daylight time, a Dassault Aviation Mystere Falcon 900C business jet, N129KJ, sustained substantial damage during a runway overrun while landing at Garfield County Regional Airport (RIL), Rifle, Colorado. The certificated airline transport pilot, co-pilot, and flight attendant were not injured. The airplane was operated by XOJET, Inc., McClellan, California, and registered to Ohana Aircraft Ltd. LLC., Redwood City, California. Night visual meteorological conditions prevailed, and an instrument flight rules flight plan was filed for the Title 14 Code of Federal Regulations (CFR) Part 91 positioning flight. The flight departed Scottsdale, Arizona, and was destined for RIL.

According to statements submitted by the pilot and co-pilot, prior to departure, the airplane was fueled with approximately 17,500 pounds (lbs) of fuel, which allowed a landing weight at RIL of greater than 2,000 lbs. under maximum gross landing weight. The crew stated that they had operated in and out of RIL many times in the past several days, and the pilot "was very comfortable with the weather [and] performance numbers planned." The pilot remembered observing a weather report that reported light rain at RIL; however, the report was not current after the flight departed Scottsdale. The pilot "did not use wet runway performance numbers...[and] did note the 135 landing distance requirements as an aid for safety margin; approximately 6,600 feet."

The crew reported the en route portion of the flight was uneventful. As the airplane approached RIL, the crew checked the automated surface observing system (ASOS) report which reported the sky overcast at 4,000 feet and no rain. During the downwind leg to runway 26, the crew could see the runway. Prior to the final approach fix for the instrument landing system (ILS) approach to runway 26 and approximately 8 miles from the airport, the crew had the runway in sight. The crew communicated their intentions on the RIL common traffic advisory frequency (CTAF). The pilot flew the airplane on the glideslope approximately Vref (Final Approach Speed) 128 knots plus 10-15 knots, then below 1,000 feet, Vref plus 10 knots. The pilot stated that the airplane touched down approximately 1,500 feet from the approach end of runway 26. The co-pilot reported the airbrake was in the #2 position and they had "good anti-skid indication" after touchdown. The pilot "began braking, deployed the thrust reverser, [and] continued to increase brake pressure as it didn't seem like a normal deceleration rate. [Co-pilot] said, 'Get on it' and [the pilot] replied, 'I am' as I had max brakes applied and full reverse." With approximately 2,500 feet of runway remaining, the airplane continued to not decelerate normally, and the pilot knew they did not have enough runway to execute a go-around. With approximately 1,000 feet of runway remaining, the pilot pulled the parking brake to the second detent, and the aircraft slid off the end of the runway into the dirt and muddy terrain. The crew shutdown the engines and exited the airplane via the forward cabin door.

A witness reported that during the accident, he was on the ramp across from the airport gate 2 and was waiting to the airplane after landing and taxi. He stated that the ramp was wet from rain that had ended about 30 minutes prior to the accident. The witness reported, "The approach speed for the aircraft looked to be fast. The [aircraft's] main wheels touched down between [taxiway intersection] alpha 4 and alpha 5, right about straight north of the airport gate 2. When the aircraft reached alpha 4, the nose wheel touched down. I heard the thrust reverser deploy about the 3,000 foot marker...The runway and taxiway alpha were wet from the rain." Alpha 4 taxiway intersection was located approximately 2,000 feet from runway 26 threshold and Alpha 5 was located approximately 2,700 feet from runway 26 threshold.

PERSONNEL INFORMATION

The pilot, age 38, held an airline transport pilot certificate with type ratings for Boeing 757, Boeing 767, Cessna CE-500, Cessna CE-750, and Dassault DA-50 airplanes. She also held a commercial pilot certificate with a rating for airplane single-engine land. Her most recent first-class medical certificate was issued on November 28, 2006, with no limitations or restrictions.

On March 2, 2007, the pilot completed a Federal Aviation Regulations (FAR) Part 135 Airman Competency/Proficiency Check. The check was conducted in a Cessna CE-750, Level D simulator, and satisfactorily completed. On August 4, 2006, the pilot completed a FAR Part 135 Airman Competency/Proficiency Check. The check was conducted in the Dassault DA-900 EX, Level D simulator, and satisfactorily completed.

According to the Pilot/Operator Aircraft Accident Report (NTSB Form 6120.1), the pilot had accumulated 8,496 total flight hours and 350 in the accident airplane make and model, of which 290 were recorded as pilot-in-command (PIC). In the 90 days preceding the accident, the pilot had accumulated 195 total flight hours, of which 193 were in the accident airplane make and model, and 75 total hours in the 30 days preceding the accident.

The co-pilot, age 42, held an airline transport certificate with type ratings for Beech BE-1900, Cessna CE-500, Cessna CE-750, and Dassault DA-50 airplanes. He also held a commercial pilot certificate with a rating for airplane single-engine land. His most recent first-class medical certificate was issued on December 11, 2006, with a limitation for corrective lenses.

On November 21, 2006, the co-pilot completed a FAR Part 135 Airman Competency/Proficiency Check. The check was conducted in the Cessna CE-750, Level D simulator, and satisfactorily completed. On May 20, 2006, the pilot completed a FAR Part 135 Airman Competency/Proficiency Check. The check was conducted in the Dassault DA-900 EX, Level D simulator, and satisfactorily completed.

According to the NTSB Form 6120.1, the co-pilot had accumulated 7,878 total flight hours and 897 in the accident airplane make and model, of which 763 were recorded as pilot-in-command (PIC). In the 90 days preceding the accident, the pilot had accumulated 37 total flight hours, of which all were in the accident airplane make and model, and 28 total hours in the 30 days preceding the accident.

AIRCRAFT INFORMATION

The 2000-model Dassault Aviation Mystere Falcon 900C, serial number 184, was issued an airworthiness certificate on October 3, 2000. The airplane was configured to carry 14 passengers, one cabin crewmember, and two pilots. The airplane was equipped with three Honeywell TFE 731-58R turbofan engines rated at 4,750 pounds (lbs) of thrust. The airplane was maintained under a continuous airworthiness program, and the most recent inspection was completed on March 8, 2007. At the time of the accident, the airframe had accumulated 3,130 flight hours.

METEOROLOGICAL INFORMATION

At 2053, the RIL ASOS reported the wind calm, 10 statute miles visibility, sky overcast at 3,900 feet, decreasing rain, temperature 8 degrees Celsius, dew point 6 degrees Celsius, and an altimeter setting of 29.98 inches of Mercury.

AERODROME INFORMATION

The Garfield County Regional Airport, RIL, was a public, uncontrolled airport located approximately 3 miles east of Rifle, Colorado, at 39 degrees, 31.578 minutes north latitude, and 107 degrees, 43.616 minutes west longitude, at a surveyed elevation of 5,544 feet. The airport featured one ungrooved asphalt runway, Runway 8/26, which was 7,000 feet by 100 feet. The runway had a 1.25 percent downslope gradient to the west. According to the airport/facilities directory, the airport remarks section notes, "Runway 08-26 slick when wet, airport manager recommends landing uphill on runway 08 when able."

FLIGHT RECORDERS

The Cockpit Voice Recorder (CVR) installed on the accident airplane was a Honeywell 6022 SSCVR 120, serial number 3037, digital CVR. The CVR had sustained no heat or structural damage, and the audio information was extracted from the recorder without difficulty. The 2-hour recording contained the entire flight from Scottsdale to Rifle. The 30-minute recording contained the last 30 minutes of the flight. Both recordings were of excellent quality. A CVR group was not convened and a summary report was prepared using the 30-minute recording.

The Flight Data Recorder (FDR) was a Honeywell Advanced Recorder FDR (AR-FDR), model 980-4710, serial number 00624, which recorded airplane flight information in a digital format using solid-state flash memory as the recording medium. The recorder was in good condition and 26.9 hours of data were extracted normally from the recorder. As per CFR Part 121.344a, the accident airplane was required to be equipped with a FDR that recorded 57 parameters, including lateral control surface, or elevator control position. The NTSB recorders specialist determined that the elevator control position was non-functional. Review of the previous flight control ground check revealed the elevator position data were not working correctly.

WRECKAGE AND IMPACT INFORMATION

Dassault Falcon Jet Corporation, and XOJet, Inc. personnel responded to the accident. Examination of the accident site revealed the airplane came to rest upright and a measurement from the aircraft's tail to the end of runway 26 was 268 feet. Skids marks were noted 3 feet from the edge of the end the runway, no other marks were noted. The terrain between the runway and aircraft resting point was reported as wet clay.

Examination of the airplane by Dassault personnel revealed that the right main landing gear trunnion sheared from it's attach points, rotated up and punctured wing skin. The main landing gear strut impacted the aft spar and sheared rib 3, which compromised the wing fuel tank. The forward main landing gear trunnion exited the forward trunnion bore, and the aft spar was fractured. All 4 main landing gear tires indicated "evidence of skidding: one flat spot on each tire..."

During May 8th to May 10th, the brake system was examined by representatives from XOJet, Inc., Dassault Falcon Jet Corporation, and Messier-Bugatti. According to all the tests and inspections performed on-site, the braking system "shows no failure."

TESTS AND RESEARCH

NTSB Performance Study

Available data pertaining to the accident included the FDR. The study used the FDR data to examine the approach and landing roll performance of the flight. Calculations were performed to determine the ground track from approximately 5,800 feet mean sea level (msl), or 400 feet above ground level (agl) to where the overrun occurred. The landing point and the location of critical FDR events were determined and landing distance calculations were performed.

The accident flight crossed the runway 26 threshold at 150 knots airspeed and touched down 2,300 feet from the threshold at 141 knots airspeed, with a minus 1 foot per second vertical speed. Immediately at touchdown, the spoilers were deployed. Approximately 4 seconds later and 3,260 feet from the threshold (3,740 feet of landing distance remaining), the thrust reverser was fully deployed with engine 2 fan speed (N1) reaching about 100 percent 9 seconds later. The airplane continued to roll down the runway with the thrust reverser deployed and exited the end of the runway at about 65 knots ground speed coming to a stop in the safety area approximately 268 feet from the end of runway 26.

The Airplane Flight Manual (AFM) was reviewed to determine the landing distance required for the accident flight. Using the landing weight of 40,000 lbs (39,857 actual), and a field pressure altitude of 5,500 feet, a landing distance of 3,700 feet was calculated. According to the Federal Aviation Administration (FAA) CFR, the landing field length is the actual landing distance times a safety margin. Appendix A contains CFR section 91.1037 of Part 91, Subpart K, which defines the landing field length and safety margin. Part of the regulation includes a provision for operators who conduct flights in accordance with an approved Destination Airport Analysis Program (DAAP), which this operator was using, allows flights to land and come to a full stop within 80 percent of the runway, rather than the conventional 60 percent.

Using the 80 percent criteria, the calculated dry runway landing field length for a similarly configured airplane to the accident flight was approximately 4,625 feet, and since the reported conditions at the time of the accident indicated precipitation was falling, an additional 15 percent margin was required, which increased the landing field length to a total of 5,319 feet. The AFM does not provide any data on runway slope effect on the landing distance, therefore, the distances mentioned above would increase for landing on runway 26 which has a negative (down) 1.25 percent slope.

Data from the previous flight was examined and compared to the accident flight (see NTSB Performance Study figures). The longitudinal acceleration data indicated the two flights deceleration was fairly consistent for the majority of the ground roll until the accident flight started its excursion. The thrust reverser for the previous flight indicated deployment (transition) begins instantaneously at weight on wheels (WoW) indication and is fully deployed in a total time period of about 1.5 seconds. The accident flight deployment did not occur until 2.5 seconds after WoW indication and full deployment was completed in a total time period of 4 seconds after WoW. The thrust reverser in the Falcon 900 series airplanes is not an auto deploy system. In order for the thrust revesers to deploy, pilot action on the throttle quadrant is required once the nose wheel touches down.

Brake Assembly Testing

On November 12, 2007, at the facilities of Aero Precision Repair and Overhaul (A-PRO), the brake assemblies were examined. Prior to the examination, the brake assemblies were disassembled and the heat pack was inspected during the on-scene portion of the investigation. The brakes were then reassembled and shipped to A-PRO for examination. According to the engineering report provided by A-PRO, the #1 position brake was initially found to have mud and stones embedded in the torque tube and around several pistons. The initial operational test could not produce an acceptable running clearance between the heat pack and pressure plate due to debris. Once cleaned, the running clearance was recorded slight below the minimum requirement. A-PRO reported, "Additional cleaning and removal of the pistons will be required to get an acceptable running clearance.

No additional discrepancies were noted with the other 3 brake assemblies during testing.

ADDITIONAL INFORMATION

According to the RIL airport manager, since 2001, 12 business jet aircraft have experienced a runway overrun on runway 26. All but one of the overruns included "wet" runway conditions. In the one "dry" runway conditions overrun, the airplane experienced a hydraulic failure.

The airport manager reported to the NTSB that runway 8/26 was grooved during the summer of 2007. The runway grooves extended 40 feet to the left and right of centerline, with 10 feet on each edge remaining ungrooved. The manager stated that the "runway now looks damp instead of wet...there is no pooling of water, and the runway has good drainage." In addition, the manager reported the airport was expected to receive a grant for engineering to begin on a runway improvement plan with a start date of 2008, and a completion of 2012.

On August 13, 2007, the airport manager conducted runway friction tests using a Dynatest Consulting Inc. Runway Friction Tester. Three tests were completed on runway 26, and three were completed on runway 8, at a speed of 40 miles per hour (mph). Three sections of each runway were tested; Section 1: 10 feet offset of runway centerline, Section 2: 20 feet offset of runway centerline, and Section 3: Ungrooved section. Each runway section was then tested in three sections of 2,000 feet, and measurements were taken at 250 foot intervals. Those measurements were th

NTSB Probable Cause

The pilot's improper decision to land with excessive airspeed during the approach and failure to obtain the proper touchdown point, which resulted in a runway overrun. Contributing factors were the pilot's failure execute a missed approach, to use available reverse thrust in a timely manner, and the wet, ungrooved, downsloped runway.