On September 29, 2001, approximately 1126 mountain daylight time, a Piper PA-30 airplane, N954S, impacted a cattle pasture approximately 3/4 mile south of the approach end of runway 34 at McCall Airport, McCall, Idaho, while on approach to runway 34 at McCall. The airplane was substantially damaged by impact forces and a post-crash fire, and the private pilot, who owned the airplane and was its sole occupant, was fatally injured. Visual meteorological conditions were reported by the airport's automated surface observation system (ASOS) at 1050, and no flight plan was filed for the 14 CFR 91 personal flight from Friedman Memorial Airport, Hailey, Idaho, which departed Friedman Memorial at 1017.

Another pilot, who was operating a Mooney airplane in the McCall traffic pattern at the time, reported to an FAA investigator that he heard the accident pilot call left base for runway 34 on the airport's common traffic advisory frequency (CTAF). This pilot reported that the accident pilot's base call sounded normal, with no sign of distress in the accident pilot's voice. The Mooney pilot, who reported he was on downwind when he heard the accident pilot's base call, reported that he did not see the accident aircraft until it had crashed.

Three individuals provided witness statements to the Valley County, Idaho, Sheriff's Department. The first witness reported seeing a small plane "that looked like it was having trouble", and (the witness thought) "flying too low." This witness reported that the airplane was flying south and "made a sharp turn to the East, but was dipping its wings as it appeared to be making a landing." The second witness reported that the aircraft "came from a south east heading on approach to the McCall airport directly overhead." This witness stated that about one mile to the north, the airplane drifted to the right (east), then corrected to the left (west), and at that point "the left wing went down tail up." The airplane then went out of this witness's line of sight behind a ridge. The third witness reported that he saw the plane "coming in very low, to land then it started swaying from side to side and then turned towards the Northwest." The third witness reported that the aircraft then "went into a spen [sic] sideways", then went out of his line of sight. The third witness reported that he then heard the crash.

The accident occurred during daylight hours at approximately 44 degrees 52.2 minutes North latitude and 116 degrees 6.3 minutes West longitude.


The pilot held a private pilot certificate with airplane single-engine land and airplane multiengine land ratings, with a date of issue of June 23, 2000. A burned copy of a pilot logbook found in the aircraft wreckage was sufficiently readable to extract a total pilot time of 372.7 hours as of the last entry, and contained an endorsement for high-performance and complex airplanes (the date of this endorsement was burned away.)


According to the FAA aircraft registry, the accident airplane was manufactured in 1968, and was registered to the pilot on April 17, 2000. Copies of the aircraft logbooks furnished by the pilot's wife indicated that the aircraft received its last annual inspection signoff on July 17, 2001, at 2,100.9 hours airframe total time (the July 17, 2001 annual inspection signoff indicated 2.048.6 hours total time and 169.3 hours aircraft recording hour meter time; however, the previous aircraft log entry, dated September 9, 2000, gave the aircraft total time as 2,027.2 hours and the recording hour meter time as 95.6 hours.) The annual inspection was performed by Shore & Shore Aviation of Caldwell, Idaho. Copies of the engine logbooks furnished by the pilot's wife indicated that at the time of the last annual inspection, both of the aircraft's turbocharged Lycoming IO-320-C1A engines had 2,100.9 hours (the July 17, 2001 annual inspection signoff in the engine logs indicated 2,048.6 hours total time and 169.3 hours aircraft recording hour meter time; however, the previous aircraft log entry, dated February 16, 2001, for work accomplished on the aircraft in Tucson, Arizona, gave the aircraft total time as 2,046.1 hours and the recording hour meter time as 114.5 hours.) The July 17, 2001 annual inspection entry was the most recent entry furnished with the aircraft logbook copies, as well as with both engine logbook copies, submitted by the pilot's wife. Textron Lycoming Service Instruction No. 1009AQ gives the manufacturer's recommended time between overhauls (TBO) for IO-320-C series engines as 1,800 hours.

In a letter to the NTSB investigator-in-charge dated February 20, 2002, the attorney for the pilot's wife stated:

...Following the annual check up [in July 2001], [the pilot] took the
plane in the latter part of July on a short flight from Caldwell to
McCall. He felt that the plane's left fuel/engine was still not working
correctly and brought it back to Shore & Shore for them to evaluate
further. The plane remained at Shore & Shore for continued
maintenance until September 28, 2001....

The NTSB obtained a copy of Shore & Shore Aviation invoice No. 005498 dated September 28, 2001 (the day before the accident), that indicated that Shore & Shore Aviation had performed the following maintenance on the accident aircraft:


The invoice total included a charge for 7.75 hours of labor. No entries in the aircraft or engine logbooks documenting this work were noted, and the invoice itself did not contain the name of the person performing the work, nor the signature, certificate number, or kind of certificate held by the person approving the work. Discrepancies in the engine serial numbers were also noted on this invoice, with one engine serial number being recorded as "L3654-55" and the other being recorded as "L3654-55A" (the actual left engine serial number was L-3654-55A, and the actual right engine serial number was L-3653-55A.) When the NTSB investigator-in-charge asked the owner of Shore & Shore Aviation to describe how the fuel pressure adjustment was accomplished, the owner stated that the invoice entry must have been made in error, and that his firm did not adjust the fuel pressure on the aircraft. The attorney for the pilot's wife submitted a copy of a check (number 166), dated September 28, 2001, made out to "Shore & Shore" and signed by the pilot and his wife, in the amount of the invoice total for Shore & Shore Invoice No. 005498. The copy of this check indicated that Shore & Shore Aviation had endorsed the check.

The accident aircraft's fuel system consisted of six fuel tanks (three in each wing). Each wing contained a main tank (30 gallons capacity/27 gallons usable), auxiliary tank (15 gallons capacity) and tip tank (15 gallons capacity). The total fuel capacity was 120 gallons, of which 114 gallons were usable. Fuel tank feed is selected by means of two fuel selectors, one for each side. Each selector can be positioned to four different positions: OFF, AUX, MAIN, or CROSS FEED. Feed from the auxiliary or tip tank on each side is accomplished by placing the fuel selector on that side to AUX and selecting an additional, two-position switch on that side's fuel selector to AUX or TIP, as desired. The Piper Turbo Twin Comanche B Owner's Handbook (Piper Part No. 761 452), March 1971 revision, specifies that fuel from the auxiliary and tip tanks is to be used during level flight only, and that fuel should be used from the main fuel cells for takeoff, landing, climb and descent. The handbook gives the aircraft's range as 820 statute miles (713 nautical miles) based on the following parameters: 10,000 feet density altitude; 75% power with best economy mixture (producing a total fuel consumption of 17.2 gallons per hour at 2,400 RPM); standard temperatures; "full fuel at take-off with 84 [gallons] usable"; and 45 minutes reserve fuel at cruise. It was not determined where the aircraft last refueled.

The Piper Turbo Twin Comanche B Owner's Handbook gives the following performance figures for the aircraft, based on flight at maximum gross weight and standard sea level (or stated altitude) conditions: single-engine minimum control speed (Vmc) 90 MPH; stalling speed (gear and flaps down, power off) 69 MPH; stalling speed (gear and flaps up, power off) 76 MPH; single-engine rate of climb 165 feet per minute; best single-engine rate of climb speed (Vyse) 105 MPH; single engine absolute ceiling 12,600 feet; and single engine service ceiling 8,800 feet.


The 1050 automated METAR observation at McCall reported conditions there as winds from 210 degrees true at 4 knots, clear skies with 10 statute miles visibility, temperature 9 degrees C, dewpoint 6 degrees C, and altimeter setting 30.26 inches Hg. At 1150, conditions at McCall were reported as calm winds, clear skies with 10 statute miles visibility, temperature 12 degrees C, dewpoint 4 degrees C, and altimeter setting 30.26 inches Hg. Based on the temperature and altimeter setting reported at 1150, and the McCall airport elevation of 5,020 feet, density altitude at the airport at 1150 was computed to be 5,425 feet.


An on-site investigation of the aircraft wreckage was conducted by investigators from the NTSB, FAA, New Piper Aircraft Corporation, and engine manufacturer Textron Lycoming from September 30 to October 1, 2001. The aircraft had crashed into a level to rolling cattle pasture about 0.7 nautical mile south of the McCall runway 34 approach end, slightly west of the runway 34 extended centerline. The wreckage area originated at a ground scar (hereafter referred to as the "initial ground scar") containing wing tip tank fragments, and from the initial ground scar continued generally northwest for about 50 feet to the main wreckage, which had come to rest upright on a magnetic heading of 070 degrees. A second ground scar, with an embedded fragment of the right propeller tip, was located about 7 feet north of the initial ground scar. The aircraft's separated right wing tip tank was located just to the west of the second ground scar. A curved slash in the ground was located about 7 feet north and slightly west of the second ground scar. Sections from the aircraft's nose and windshield were located between the curved ground slash and the main wreckage. An approximately 3-foot-long, straight ground scar, with a round indentation at one end that was about the same diameter as one of the aircraft's propeller spinners, was located about 30 feet north and 10 feet west of the initial ground scar.

The main wreckage of the aircraft comprised generally the complete aircraft less the separated sections of the nose and the separated right wing tip tank. The cabin section of the aircraft had been largely consumed by fire, with only steel components in this area such as seat frames and control cables surviving. The wings and tailcone, although surviving to a greater extent, also exhibited significant fire and/or heat damage. Beyond the central fire area, the major aircraft components within the main wreckage generally maintained their normal positions within the airframe, and the main wreckage overall was readily identifiable as the burned remnant of an airplane. Most instruments and controls were burned beyond readability, with the following exceptions. The aircraft's dual-needle tachometer was found captured at a left engine reading of 800 RPM and a right engine reading of 2,650 RPM (NOTE: according to the aircraft owner's handbook, the maximum engine RPM is 2,700.) Additionally, a pair of gauges in the engine instrument cluster, in a location corresponding to that of the oil pressure gauges on an illustration of the instrument panel in the aircraft owner's handbook, was found with the left needle at bottom of scale and the right needle in midrange (the dial faces of the gauges were burned/melted away). The aircraft's rudder trim actuator was found approximately halfway between the neutral and full nose-right positions. The landing gear and flap actuator jackscrews were found in positions corresponding to gear down and flaps fully down. A small quantity of fuel was found in the separated right tip tank (approximately 1/2 pint of 100LL aviation gasoline in the downhill portion of the tank, although the tank was breached open) and left auxiliary tank (approximately 1/2 inch depth across the bottom surface of the tank). The other four fuel tanks were breached and/or burned.

The aircraft's right propeller exhibited significant aft curling and torsional twisting, and the tips of both its blades were broken off. Fragments of the separated right propeller blade tips were found at three different locations in the vicinity of the crash site; one fragment was found 194 feet northwest of the main wreckage (this propeller fragment was the piece of wreckage found furthest from the main wreckage), one was found 72 feet south of the main wreckage, and one was found embedded in the second ground scar in the wreckage path. The aircraft's left propeller was nearly intact, with only relatively minor chordwise scratching, minimal leading edge damage, and polishing on one propeller blade. The left propeller was not feathered; the blade pitch angle at the propeller tip was measured at approximately 20 degrees from the plane of rotation.

No evidence of inflight structural failure, flight control malfunction, or inflight fire was noted during the on-site examination. Following the on-site examination, both engines and propellers, and the aircraft's fuel selectors, were removed from the wreckage and taken to the facilities of McCall Air Taxi at the McCall airport for further examination (refer to the TESTS AND RESEARCH section below).


Refer to WRECKAGE AND IMPACT DATA section above.


An autopsy on the pilot was conducted by a forensic pathologist under the authority of the Valley County, Idaho, Coroner on October 2, 2001. The autopsy report concluded that the cause of the pilot's death was "BLUNT FORCE TRAUMA DUE TO AN AIRPLANE CRASH" [emphasis in original], with "No medical diseases [noted] on gross exam."

Toxicology tests on the pilot were performed by the FAA Civil Aerospace Medical Institute (CAMI), Oklahoma City, Oklahoma. The CAMI toxicology tests screened for carbon monoxide, cyanide, ethanol and various legal and illegal drugs, and none of these substances were detected.


A field disassembly/functional examination of the aircraft's engines at the facilities of McCall Air Taxi in McCall, conducted on October 2, 2001, under the supervision of the NTSB IIC, did not reveal any evidence of pre-impact mechanical problems with either engine. A compressed air flow test of the aircraft's two fuel selector valves disclosed that the left fuel selector was in the AUX position, and the right fuel selector was in the MAIN position. Following the engine examination, the engine-driven fuel pump, fuel injection servo, flow divider, and fuel injection nozzles were removed from the left engine for further examination.

Functional testing of the left propeller and left propeller governor at Precision Propeller Service in Boise, Idaho, conducted on October 3, 2001, under the supervision of the NTSB IIC, demonstrated that the propeller feathered smoothly and quickly when oil pressure was removed from the hub, and that the governor cut off oil pressure to the propeller at a slightly lower RPM than specified by the Hartzell service manual (1,510 RPM observed versus 1,575 to 1,600 RPM specified) but otherwise functioned nominally.

Examination of the Precision Airmotive model RSA-5AD1 fuel injection servo, the flow divider, and the fuel injection nozzles, from the left engine was conducted at the facilities of the component manufacturer, Precision Airmotive Corporation, Marysville, Washington, on November 16, 2001, under the supervision of the NTSB IIC. During the examination, the fuel injection servo, flow divider, and three of the four fuel injection nozzles were flow-tested against production specifications (one fuel injection nozzle, marked as "A" during the examination, was deformed at the inlet where the fuel line attaches, precluding flow testing of this nozzle.) No disassembly was conducted during the examination. The fuel injection servo was first flow-tested in the "as received" condition, and then tested again with idle adjustments set to original specifications. The servo's spring-loaded throttle stop, incorporating a spring tension set screw in the throttle lever that is factory-adjusted and then spiked in position (this setting is not intended to be field-adjusted), was noted to have been reset "much too soft", such that the throttle would stay on the idle stop with the spring fully compressed with no pressure on the throttle lever. The set screw on the throttle stop was visibly out from its normal position (this condition requires the factory-set spikes in the screw head slot, which secure the screw in the proper position, to be forcibly broken out). During the test in the "as received" state, the servo idle mixture was noted to be set excessively lean, and idle air flow was difficult to determine because of the soft spring setting in the throttle stop. The servo flowed within specifications after the idle settings were corrected. After flow-testing with the idle adjustments set to original specifications, the servo idle settings were readjusted to re-establish the "as received" flow parameters observed on the test bench, to enable further testing. The three fuel injection nozzles that were tested (marked "B", "C", and "D" during the examination) all passed production tests. The flow divider passed a high-pressure, mass flow test, but had two ports which were below specification on the low pressure test (with the test instrumentation adjusted to indicate a flow rate of 100% on the highest-flow port, all four ports should flow between 90-100%, but ports "B" and "C" flowed at 88% and 84%, respectively.)

The engine-driven fuel pump from the left engine (identified by a data plate on the pump housing as a Lear Romec model RG15980) was examined at the facilities of the pump manufacturer, Crane Lear Romec, Elyria, Ohio, under the supervision of an inspector from the FAA's Cleveland, Ohio, FSDO on November 26, 2001. The examination revealed that the exterior of the pump had sheet metal fabricated "ducting" attached by means of two valve cover housing screws, which Crane Lear Romec's report stated was not a part of the pump assembly. The Crane Lear Romec report characterized this ducting as "A field modification...leaving [the pump] in an unapproved configuration and with indeterminate performance characteristics." According to the report, the RG15980 pump was designed to operate when rotated counter-clockwise while viewing the drive coupling. When tested according to the test specification applicable to the model RG15980 pump, the pump produced no flow or pressure when rotated counter-clockwise. However, when rotated clockwise, the pump produced flow and pressure, with incidental shaft seal leakage (the rate of this leakage was not quantified.) According to Crane Lear Romec's report, the pump model specified for installation on a Lycoming IO-320-C1A in the Piper PA-30 application is the RG17980A, which normally rotates clockwise. Crane Lear Romec's report stated that "When rotated clockwise, the subject pump produced flow and pressure. The performance values obtained exceeded those required of an RG15980 and were closer to the typical performance of a [sic] RG17980A. However, testing to actual RG17980A specifications was not performed."

Following examination of the individual fuel system components from the left engine at their respective manufacturers, the left engine fuel injection servo, flow divider, fuel injection nozzles, and engine-driven fuel pump were sent to the Textron Lycoming plant in Williamsport, Pennsylvania, for further testing. This testing, conducted under the supervision of the NTSB IIC on February 22, 2002, consisted of mounting the accident fuel system components (engine-driven fuel pump, fuel injection servo, flow divider, and fuel injection nozzles B, C, and D) to an exemplar "core" IO-320-C1A engine that had been returned to the plant for overhaul, and conducting an engine test cell run in an attempt to determine the overall/cumulative effects of the accident fuel system components on engine performance. The "core" engine's logbook indicated that it had 2,099.7 hours since overhaul at the time of the test. Prior to conducting the run with the accident fuel system components installed, a "baseline" test cell run was performed with "slave" fuel system components used for engine testing at the plant, to verify satisfactory engine operation for the purposes of the test. This "baseline" engine run was conducted successfully, with stable engine idle demonstrated at 648 to 683 RPM at the "close throttle" test points. The "slave" fuel system components were then removed from the engine, the accident fuel system components were installed and the test run was attempted. While installing the accident flow divider and fuel injection nozzles, it was determined by matching lines to the corresponding labeled flow divider ports that nozzle A (the deformed nozzle, not flow-tested or used for the engine run) was installed on cylinder 2, nozzle B was installed on cylinder 4, nozzle C was installed on cylinder 1, and nozzle D was installed on cylinder 3. During the first attempts to start the engine, the engine-driven fuel pump was noted to be leaking at a rate deemed excessive for safe testing. The accident fuel pump was then removed and a "slave" pump re-installed. During the test run with the remaining accident fuel system components, the engine stopped each time the throttle was closed. The lowest stable engine speed attainable prior to adjustments was 1,000 RPM. The idle stop set screw on the fuel servo throttle lever (the one found set "much too soft" during the examination at Precision Airmotive) was then turned 2 1/2 turns clockwise to adjust it to a position approximating the normal factory setting. This reduced the lowest attainable stable engine speed to 800 RPM; however, the engine continued to stop when the throttle was fully closed. The idle mixture adjustment wheel on the fuel servo was then adjusted a net of 10 clicks lean. After this adjustment, stable engine idle was attained at 673 RPM with the throttle closed.


The great circle distance from Caldwell Industrial Airport, Caldwell, Idaho (where the aircraft underwent repairs just prior to the accident flight), to Friedman Memorial Airport in Hailey (origin point of the accident flight) is approximately 102 nautical miles. The great circle distance from Friedman Memorial to McCall Airport is approximately 114 nautical miles.

The aircraft wreckage was released to Mr. Steven Maier, engineering assistant for Aeroscope, Inc., Broomfield, Colorado, on July 10, 2002. Aeroscope, Inc. is a consulting firm retained by counsel for the pilot's wife, who purchased the aircraft wreckage following the accident.

Use your browsers 'back' function to return to synopsis
Return to Query Page