On November 26, 1997, at 0722 hours local island time, a Piper PA-32-300, N1256H, experienced an engine failure while performing a maintenance test flight at the Marianas International Airport, Saipan, Marianas Protectorate. The commercial pilot subsequently ditched the aircraft approximately 1/2 mile off the southern tip of Saipan Island in the Saipan Channel. The aircraft, owned and operated by Pacific Island Aviation, Inc., sank and was not recovered. The pilot sustained minor injuries and was rescued by a private fishing vessel. Visual meteorological conditions prevailed for the local test flight and no flight plan was on file. Use your browsers 'back' function to return to synopsisReturn to Query Page
A Federal Aviation Administration (FAA) airworthiness inspector from the Honolulu Flight Standards District Office traveled to Saipan, interviewed the pilot and witnesses, and reviewed the aircraft maintenance and company flight operations records. In addition, the Safety Board obtained copies of certain maintenance and flight department records for review.
Another company pilot flew the accident aircraft in a revenue passenger flight on November 24, 1997. In an interview with the FAA inspector on December 5, 1997, this pilot stated that after leveling off and setting the engine power to cruise, he felt vibrations. He said that he did not know if the vibrations were coming from the engine or the airframe. The pilot reported that he heard a pitch change in the engine, reduced the power, and the vibrations continued. The pilot aborted the flight and returned to base, where he verbally reported the problem to the maintenance department. He stated that he did not write up the incident in the discrepancy log.
The maintenance department conducted an engine ground run and found the engine to be rough. The pilot who reported the problem stated in his interview that the aircraft was shaking during the maintenance ground run at full power.
The Safety Board conducted a review of the Operators' Inspection Discrepancy Records (IDR) and noted that there were two IDR's for this aircraft dated November 24, 1997. These two documents are appended to this file. The FAA inspector obtained the first document when he first reviewed the records. The company provided the second one to the Safety Board on May 25, 1998. Examination of the documents disclosed that they were essentially the same, except for item number 3. On the first document found by the FAA inspector, item 3 discrepancy states: "#5 cyl fuel injector nozzle uneven spray need replacement." The corrective action section states: "none in stock." On the second document provided to the Safety Board, discrepancy item 3 reads: "As per item 1." The corrective action section states: "Removed and re-cleaned fuel nozzles, reinstalled, leak tested. . .no leak, IAW PA-32-300 svc man 7A-22 to 7A-23. . .flushed servo unit screen, reinstalled, safetied IAW PA-32-300 svc man 7A-19. . .ground run up found normal, a/c needs fcf." The first document contains a corrective action entry number 4 not found in the second document, which states: "Cleaned fuel injection nozzles, performed compression check, flushed servo unit screen, ground run up ok. . .fcf."
Correspondence was sent to the Vice President of Technical Services for an explanation. A written statement was provided by the Vice President of Maintenance stating that he was not satisfied with the initial record and had asked the mechanic who had completed the work to describe the discrepancy and corrective action in more detail.
A review of the historical aircraft maintenance logs and IDR's further revealed that on August 8, 1997, a pilot reported that the magnetos dropped 300 rpm when the aircraft was set at full rich. The corrective action was to adjust the mixture and injectors and clean the clogged injectors. On the same day, it was also written up that the mixture control lever was stiff. Corrective action was to readjust and lubricate the mixture control lever.
On August 9, 1997, a pilot reported that the engine was running lean at full rich. Corrective action was to readjust the mixture control lever. On August 16, 1997, it was reported that the difference between the left and right magneto check was more than 100 rpm. Corrective action was to retime the magnetos to the engine.
On December 4, 1997, the FAA inspector interviewed the company mechanic who worked on the aircraft on November 24, 1997. This interview was recorded on tape and the tape is appended to this file. The mechanic said that there was a discrepancy report on the aircraft reporting vibration, low manifold pressure, and low fuel flow pressure. He stated that he checked the fuel nozzles and cleaned them for approximately 1/2 hour in acetone. He then blew air through the nozzles and reinstalled them. He then tested the flow divider and no discrepancies were noted. The mechanic reported that he completed a compression check and checked the manifold for leans. No discrepancies were noted. The mechanic stated that a ground run-up was conducted and the engine ran rough with vibration.
The mechanic further stated that the fuel nozzles were cleaned again and reinstalled. Another ground run was conducted with the same results, a rough running engine with vibration. The aircraft was grounded on November 25th so the night crew could continue working on it. The engine was ground run that evening, and no anomalies were noted. The mechanic stated that a suggestion was made for a functional check flight for further troubleshooting.
In his interview, the mechanic stated that the engine was ground run approximately four times. The last engine ground run was normal. After flushing the servo filter, he did note some water and dirt, but not a lot. The mechanic did note that the number 5 nozzle had a different spray pattern than the other nozzles, but that it was not a concern to him. Corrective action in that case was to clean the number 5 nozzle and reinstall it on the aircraft.
In an interview with the FAA inspector on December 5, 1997, a company dispatcher reported that she had asked the pilot involved with the accident to do a functional check flight to help out the company, since the aircraft was scheduled for a revenue flight later that morning. She stated that the pilot asked her what kind of test flight to do and she stated that she did not know and to check with maintenance. After the accident the dispatcher reported that someone from maintenance had called asking why the pilot had not come down to see him before taking the aircraft.
In the pilot's written statement, he reported that while conducting his preflight he noted that maintenance personnel had indicated that they had performed an operation's check on the aircraft and that it was marked as okay. The company dispatcher instructed the pilot to perform a test flight, as the aircraft was to return to service as a 14 CFR Part 135 operation that same day.
The pilot reported that he completed the ground check checklist that included checking the magnetos and the propeller. He did not note any malfunctions. After the ground check he received weather information and contacted Saipan Air Traffic Control Tower (ATCT) with his intention of conducting a test flight. The pilot reported that there were no discrepancies with engine performance while in the pattern. The ATCT informed him that he was number two for landing. The pilot executed a 270-degree turn to set up on the base leg. Once established on the base leg, the pilot reported that the aircraft engine "banged three times and then parts flew through the engine cowling." The pilot stated that the cowling flew off, and over the cabin, and struck the empennage. He reported that it was difficult to maintain control over the aircraft. The pilot said that he set up for an emergency landing, but had to ditch the aircraft into the ocean.
Review of the maintenance records disclosed that the Lycoming IO-540-K1G5 engine, serial number L-16288-48A, completed a major overhaul on July 25, 1996. The engine was installed in the airframe on November 14, 1996, and had accumulated about 593 hours since overhaul.