NTSB Factual Report

HISTORY OF FLIGHT

On January 19, 2004, about 0106 eastern standard time, a Piper PA-28-181, N298PA, registered to Merrill Lynch Business Financial Services Inc., and operated by Pan Am International Flight Academy Inc., impacted with trees, a power line and the ground while on an instrument landing system (ILS) approach to the Saint Lucie County Airport, Fort Pierce, Florida. Instrument meteorological conditions prevailed at the time and an instrument flight rules flight plan was filed for the 14 CFR Part 91 instructional flight. The airplane was destroyed. The certified flight instructor and the private pilot-rated student were fatally injured. The private-pilot rated passenger received serious injuries. The flight departed from West Palm Beach, Florida, at 2345 on January 18, 2004.

The instrument training flight departed Fort Pierce about 2100 on January 18 and flew to Palm Beach International Airport to practice instrument approaches. After landing at Palm Beach, the airplane was refueled with 28 gallons of 100 LL aviation fuel and departed for the return flight to Fort Pierce via Vero Beach. The flight executed a practice approach, but did not land at Vero Beach Municipal Airport. At 0053:46, the controller cleared the flight for the ILS approach to runway 09 at Fort Pierce by stating, "two niner eight papa alpha yes sir seven west of lucie heading one two zero maintain one thousand eight hundred until established cleared i l s runway nine approach into fort pierce airport." The pilot responded, "heading one zero heading one two zero maintain a thousand two hundred and then cleared for the i l s niner." The controller did not challenge the pilot's incorrect read back of the altitude to be maintained. The controller then cleared the flight to execute a missed approach at the completion of the ILS approach. The pilot read back this clearance. At 0054:20, the controller told the pilot that radar service was terminated, and the pilot was cleared to change radio frequency. At 0054:40, the pilot responded, "miami center two nine eight papa alpha we will try to do a uh make this a full stop if not we'll come back to you." The controller acknowledged and reminded the pilot that he needed to close his flight plan upon landing. The last transmission from the pilot recorded by air traffic control was at 0054:57.

According to the passenger, who was seated in the right rear seat, "due to the fact that the visibility was so poor [the flight instructor] took control of the plane to land it." The passenger stated that "there was no visibility." The last thing the passenger recalled prior to impact was that "the VSI [vertical speed indicator] was dropping more than 500 ft per min." The passenger lost consciousness during the crash sequence. When she regained consciousness, the passenger climbed out of the airplane and was going to get help when the driver of a car saw the accident scene and stopped.

Radar data for the flight was obtained from the FAA radar sites at Melbourne, Florida and Tamiami, Florida. From 0055:21 to 0056:32, the flight proceeded along the ILS inbound course heading (094 degrees) from a point about two nautical miles west of the outer marker to the vicinity of the outer marker and descended from a mode C altitude of 2,000 feet to a mode C altitude of 1,300 feet. (Note: To correct the Mode C altitudes (which are based on a fixed altimeter setting of 29.92 inches) for the local altimeter setting (29.84 inches), reduce mode C altitudes by 100 feet.) From 0056:32 to 0057:21, the flight continued along the ILS course and maintained a mode C altitude of 1,300 feet. The flight then began to descend, and at 0058:48, when the last radar return was recorded, had reached a mode C altitude of 400 feet and was about 1/4 nautical mile west of the accident site.

The accident occurred during dark night conditions approximately 1/2 nautical mile from the threshold of runway 09. The airplane was approximately 300 feet left of the extended runway centerline when it initially struck trees and then power lines about 50 feet above the ground. The main wreckage was found at 27:29.855 N latitude and 080:23.404 W longitude at an elevation of 54 feet msl as measured using a hand held GPS unit.

PERSONNEL INFORMATION

The flight instructor, who was seated in the right front seat, held a commercial pilot certificate with airplane single and multi-engine land and instrument ratings. Additionally, he held a flight instructor certificate with airplane single and multi-engine land and instrument ratings. He held a first class medical dated November 21, 2001, with no limitations. The flight instructor received the flight training for all of his ratings at Pan Am International Flight Academy from January 2002 to February 2003. He began working as a flight instructor for Pan Am in June 2003. Review of the instructor's flight logbook revealed that the last entry had been made on January 13, 2004. As of that date, the instructor had accumulated 528 hours total flight time, 245 hours acting as an instructor, and 78 hours total instrument flight experience of which 32 hours were in actual instrument meteorological conditions.

The student pilot, who was seated in the left front seat, held a private pilot certificate with an airplane single engine land rating, issued on December 12, 2003. He held a first class medical certificate dated September 4, 2003, with no limitations. He received flight training towards his private pilot certificate at Pan Am International Flight Academy from September 27, 2003, to December 12, 2003. He enrolled in a 14 CFR Part 61 instrument flight training course at Pan Am International Flight Academy on December 15, 2003. Review of the student's Pan Am instrument course records indicated that the accident flight would probably have been Mission 25 in the course's flight training syllabus. The stated objectives of this lesson were, "review of previously practiced procedures in the airplane. Subject to the equipment available in the airplane, the student will review the GPS, ILS, and VOR approaches. The student will also be introduced to partial panel approaches." Review of the student's flight logbook revealed that the last entry had been made on January 17, 2004. As of that date, the student had accumulated 132 hours total flight time of which 10 hours were in simulated instrument conditions.

AIRCRAFT INFORMATION

Examination of the airplane's maintenance records revealed that the 2002 model Piper Archer III had been operated by Pan Am International Flight Academy since it was new. The airplane was placed on Pan Am's progressive inspection program on December 26, 2002, at which time it had accumulated a total airframe time of 558.7 hours. It received its most recent inspection, an Event Number 3 inspection, on January 14, 2004, at a total airframe time of 1,780.6 hours. As of this inspection, the engine, a Lycoming O-360-A4M, S/N L-38496-36A, had accumulated 1,780.6 hours total time. When the accident occurred, the airplane had been flown about 25 hours since this inspection. Review of the checklist for the Event Number 3 inspection revealed that it was similar in scope and detail to a 100 hour inspection of the engine combined with a detailed inspection of the wings and landing gear. The checklist did not list any discrepancies found with the airplane during the inspection. The most recent transponder test required by FAR 91.413 was performed on February 5, 2002, and the most recent altimeter and static system test required by FAR Part 91.411 was performed on March 16, 2002.

Review of the discrepancy log kept in the airplane revealed no evidence of any uncorrected maintenance discrepancies. A discrepancy dated December 29, 2003, indicated the vacuum pump was inoperative. The corrective action block for this discrepancy stated that the vacuum pump was removed and replaced. Further investigation revealed that the vacuum pump had not been replaced. According to a written statement, provided by the mechanic who worked the discrepancy, he started the airplane's engine and performed a ground check of the normal and auxiliary vacuum systems and found both systems operated normally with no anomalies. He released the airplane back to service, signed off the discrepancy in a computerized record keeping system (CASSi), but forgot to sign off the discrepancy log in the airplane. On December 30, 2003, a student pilot noted the unsigned off discrepancy and questioned a dispatcher about it. The dispatcher checked in CASSi, found that the discrepancy had been corrected, and made the assumption that the vacuum pump had been replaced. The dispatcher made the entry in the corrective action block and entered the mechanic's name in the corrected by block of the discrepancy log. There were no further reports of discrepancies with the vacuum system.

METEOROLOGICAL INFORMATION

The National Weather Service surface analysis chart for 0100 showed a trough of low pressure extending from the southwest Atlantic Ocean to the vicinity of Miami, Florida. Then the trough curved southward through the Florida Keys. A weak pressure gradient was indicated over Florida. The plotted stations indicated patchy fog and haze with generally light and variable winds throughout the state.

At 0053, the weather conditions at Saint Lucie County Airport reported by the unaugmented Automated Surface Observing System (ASOS) were wind 270 degrees at 7 knots, visibility 5 statute miles, present weather - mist, sky condition - overcast 100 feet, temperature 17 degrees C, dew point 17 degrees C, altimeter setting 29.84 inches.

At 0053, the weather conditions at Vero Beach Municipal Airport, located approximately 10 nautical miles north of the accident site, reported by the augmented ASOS were wind 250 degrees at 7 knots, visibility 4 statute miles, present weather - mist, sky condition - overcast 100 feet, temperature 17 degrees C, dew point 17 degrees C, altimeter setting 29.84 inches.

The closest Doppler weather radar was at Melbourne, Florida, about 39 nautical miles northeast of the accident site. At 0109, the radar indicated no reflectivity in the vicinity of the accident site. For further weather information, see the Meteorology Factual Report in the public docket for this accident.

AIDS TO NAVIGATION

Review of the approach plate for the ILS runway 09 approach indicated that the glide slope intercept altitude at the outer marker was 1,800 feet msl and the decision height for the approach was 223 feet msl, or 200 feet agl. After the accident, on January 21, 2004, the ILS runway 09 was flight checked by the FAA. The localizer, glide slope, distance measuring equipment, locator outer marker, and middle marker all inspected satisfactory.

WRECKAGE AND IMPACT INFORMATION

The initial point of contact with trees was marked by a section of the left flap that was lodged in a tree. About 75 feet from the initial point of contact, the left wing was found, and the main wreckage was found about 225 feet from the left wing. The measured magnetic heading of the wreckage path was about 170 degrees. The fuselage was lying on its left side oriented on a magnetic heading of 270 degrees. The left side of the fuselage was crushed inwards, and the forward portion of the floorboard was crushed upwards near the rudder bar. The cabin door was open, and the latches were latched. The flap handle was at the 25 degree notch. The empennage remained attached to the fuselage. The outboard 22 inches of the left stabilator separated and was found near the initial tree impact. The separated section of stabilator displayed a 10-inch circular deformation in the leading edge. The rudder and vertical fin were bent to the right about 45 degrees near the top rudder attachment point.

The left wing was separated at the wing root. The leading edge near the wing root displayed a 6-inch circular deformation. Outboard of the fuel tank, the leading edge was crushed aft. The aileron and the inboard section of the flap remained attached to the wing. The left wing fuel tank was compromised. The right wing was bent over the right side of the fuselage. The outboard section of the wing was heavily damaged and partially separated. Fuel, blue in color and consistent with 100 LL aviation fuel, remained in the right wing fuel tank. The aileron and flap remained attached to the wing.

Both aileron cables were attached to the aileron control chain. The right aileron cables were cut by recovery personnel at the wing root. The left aileron cables were separated near the wing root and displayed evidence of tensile overload. Stabilator and rudder continuity were established from the control surfaces to the cockpit controls.

The fuel selector valve was selected to the left tank position. The fuel selector valve was field tested by applying low pressure air through its ports. The valve appeared to be functional, and all ports were free of obstruction. The electric fuel pump was field tested by applying external battery power. The pump was heard operating, and fuel was observed pumping out of the outlet port. The fuel pump screen was free from obstruction.

The engine remained attached to the firewall. The fixed pitch propeller remained attached to the engine crankshaft mounting flange. The outboard 3 inches of both propeller blade tips were separated. One blade displayed severe leading edge damage and chordwise twisting. The other blade sustained chordwise scratching. The engine was intact with all accessories remaining attached. All four cylinders were borescoped, and the piston domes and valve heads were observed intact. The crankshaft was rotated by hand, and valve train and rear gear continuity were observed. Suction and compression were obtained at all cylinder positions. The magnetos were removed, and both produced spark at all leads when manually rotated. The carburetor fuel inlet screen, the oil suction screen and the oil filter element were observed to be free of debris. Fuel that was blue in color and consistent with 100 LL aviation fuel was found in the fuel pump, fuel lines, and the carburetor bowl. Fuel was expelled out of the fuel pump during manual rotation of the crankshaft.

The engine vacuum pump and auxiliary vacuum pump were removed from the airplane and installed onto another airplane for testing. The test airplane's engine was started, and the suction gage indicated 4.9 inches for both pumps, which is within normal parameters.

MEDICAL AND PATHOLOGICAL INFORMATION

Autopsies of the flight instructor and student were conducted by the District 19 Medical Examiner Department, Fort Pierce, Florida. Toxicological tests were performed by the FAA's Toxicology and Accident Research Laboratory. Toxicological test results for the flight instructor were negative for carbon monoxide, cyanide and ethanol and detected 28.28 ug/ml acetaminophen and 61.41 salicylate in urine. Toxicological test results for the student were negative for carbon monoxide, cyanide, ethanol and drugs.

TESTS AND RESEARCH

On January 20, 2004, the following equipment was removed from the airplane and tested at a certified avionics repair station in Fort Pierce, Florida: Garmin GTX-327 transponder, Garmin GNS-430 GPS/NAV/COM, Garmin GI-106A VOR indicator, King KI-208 VOR indicator, artificial horizon, and altimeter. The transponder met the specifications of FAR 43 Appendix F. When the GNS-430 was turned on, the message "user magnetic variation in use" was displayed. All modes of the GNS-430 checked satisfactory. The KI-208 indicator unit functionally tested satisfactory. The altimeter was partially tested in accordance with FAR 43 Appendix E up to 8,000 feet. Prior to the testing, it was noted that the altimeter was indicating -50 feet, and the Kollsman window was reading 29.82 inches. The measured scale errors of the altimeter at altitudes of 0, 500 and 1,000 feet were -35, -25, and -20 feet, respectively. The allowable scale error listed in Appendix E for each of these altitudes is +/- 20 feet. The artificial horizon had sustained impact damage; when tested, it failed to fully spool up

NTSB Probable Cause

The flight instructor's failure to comply with the instrument approach procedure in that he descended prematurely below decision height resulting in an in-flight collision with trees, power lines and the ground. Contributing factors were the dark night light condition and the low ceiling.