WPR13LA384

A partial loss of engine power during initial climb for reasons that could not be determined because postaccident examination revealed no evidence of preimpact mechanical malfunction or failure that would have precluded normal operation.

The pilot was receiving a checkout in the airplane from a flight instructor. This was their second flight together in the airplane, and the pilot's second flight in that airplane make and model. According to the pilot and the instructor, the preflight inspection detected no anomalies, and the airplane had sufficient fuel for the flight. Engine start, taxi, and engine run-up were all normal. According to the pilot, during the climbout, when the airplane was about 500 feet above the ground, the engine "sputtered," recovered briefly, and then "lost thrust." The flight instructor reported that the engine was running "roughly." The pilots briefly attempted to restore power. Their efforts, which included changing the position of the fuel selector handle, were unsuccessful; they then attempted to land in a clearing on a golf course adjacent to the airport. The airplane struck trees and came to rest inverted on the golf course. Postaccident examination of the airframe and engine confirmed that the engine was developing little or no power at impact but did not reveal any preimpact conditions or failures that would have precluded continued engine operation. Before the takeoff, the engine was idled on the ground for at least 8 minutes. Although the carburetor heat was off (the recovered checklist did not reference carburetor heat) and the ambient conditions were conducive to "serious icing" at glide power, both pilots reported that the taxi, engine run-up, and takeoff roll were normal. Thus, it is unlikely that carburetor ice was present at the time the engine began to lose power. The airplane's original fuel selector handle and placard design were ambiguous with regard to the selection indication, particularly to a person unfamiliar with the airplane; the selector could inadvertently and unknowingly be set to the "OFF" position for flight. The manufacturer issued two service letters (in 1971 and 1972), which recommended modification of the fuel selector, and subsequently issued a mandatory service bulletin (in 1986) that called for the replacement of the handle and placard to reduce the possibility of pilots inadvertently selecting the "OFF" position for flight. Although the service bulletin was applicable to the accident airplane, the handle and placard had not been replaced. Because the pilot manipulated the fuel selector following the power loss and a first responder manipulated it after the accident, neither the takeoff nor the preimpact setting of the fuel selector valve could be determined. The investigation was unable to determine if the fuel selector valve was inadvertently set to the "OFF" position at some point before the takeoff, allowing some limited high-power engine operation using residual fuel. The investigation also considered the possibility that the fuel selector valve was improperly set so that the selector valve ports were partially, instead of fully, aligned, allowing sufficient fuel for low-power operation and limited-time high-power operation. However, because power was not restored after the pilot repositioned the fuel selector after the first indication of an engine problem, it is unlikely that a mis-set fuel selector valve caused the power loss.

A copy of the audio recording between the SAC air traffic control tower and the airplane was obtained and reviewed. Approximately 6 minutes 20 seconds elapsed between the time the pilots requested taxi clearance and the time they requested takeoff clearance. Due to landing traffic, another 2 minutes elapsed before the airplane was issued a takeoff clearance. One minute 54 seconds after the takeoff clearance was issued, the tower controller queried the flight as to whether it needed assistance, due to his observation of its turnback and descent. No response was recorded from the airplane. Review of the recording revealed that after the pilots acknowledged the takeoff clearance, there were no other transmissions from the airplane. According to FAA information, runway 2/20 was the longest runway at SAC, and measured 5,500 feet by 150 feet. Airport elevation was listed as 24 feet above mean sea level. Carburetor Heat and Emergency Procedures Guidance Review of the "Normal Procedures Checklist" and the "Emergency Procedures Checklist" that were recovered from the airplane revealed that the checklists were not published by Piper, and that they were designated for a PA-28R-180 airplane, which is equipped with a fuel-injected engine. Since the PA-28R-180 was not equipped with carburetor heat, the checklists' "Normal" and "Engine Failure" procedures did not contain any references to, or procedures for, application of carburetor heat. Review of the "Ground Check" portion of Section III (Operating Instructions) of the Piper-produced Owner's Handbook (OH) for the PA-28-180 contained the statement "Carburetor heat should also be checked prior to take-off to be sure that the control is operating properly and to clear any ice which may have formed during taxiing." The OH did not contain any amplifying or clarifying information regarding carburetor icing during taxi. Review of the Piper OH and the FAA-approved AFM revealed that neither document contained any guidance or checklists regarding emergency procedures, including loss of engine power. According to representatives from Piper and the FAA, the OH and AFM were in compliance with the applicable certification regulations. Cockpit Fuel Selector The cockpit fuel selector had three settings; a detented position for each (left and right) tank, plus a 90-degree arc for the OFF position. The handle could be rotated through 360 degrees in either direction. The original handle and placard design was ambiguous with regard to proper use/setting by pilots; a pilot could inadvertently select the OFF position while interpreting the selected setting as being for the left or right tank. In 1971 and 1972, Piper issued two separate Service Letters (SL 588 and SL590, respectively), which recommended modifications to the cockpit fuel selector to make it less susceptible to inadvertent incorrect operation. The basic contents of those two SLs were integrated into the 1986 Piper Service Bulletin (SB) 840 entitled "Fuel Selector Valve Cover Replacement. SB840 called for the replacement of the existing handle and placard with redesigned versions, and Piper designated SB840 as "mandatory." According to SB840, the purpose was "To announce the availability of a new fuel selector valve cover assembly and mating handle, which reduces the possibility of pilot mismanagement of the fuel system through inadvertent selection to the "OFF" position, resulting in power interruption or stoppage." SB840 compliance time was specified as "To coincide with next regularly scheduled maintenance event, but not to exceed the next 25 hours time in service." Although SB840 was applicable to the accident airplane, it was never accomplished on that airplane. In November 2013 (subsequent to the accident), SB840 was reissued as SB840A to revise certain Piper part numbers. On August 20, 2013, the FAA issued a Notice of Proposed Rulemaking which proposed an Airworthiness Directive based on SB840. In June 2014, the FAA withdrew that proposed AD, due to an FAA re-evaluation that determined that "an unsafe condition does not exist that would warrant AD action." Fuel Starvation Tests Because the installed fuel selector was subject to inadvertent improper operation by pilots, and was the subject of a Piper service bulletin, the investigation attempted to locate data regarding how long an engine would operate once the fuel was shut off. Such testing was not required for FAA certification. Piper did not provide any data on this topic, and searches for relevant or similar data yielded very sparse results, none of which were for a PA-28 with an O-360 engine. With the assistance of the FAA and another flight school, a dedicated test using a PA-28 equipped with an IO-360 engine was able to be accomplished. However, because there was no data from an O-360 equipped PA-28, and the previous and new results displayed significant data scatter, the results were deemed to be inapplicable for this investigation. No tests were conducted for a partially mis-set selector valve, where the valve was positioned so that the valve ports were partially, but not completely, aligned. In such a mis-set position, the port misalignment would limit the maximum fuel flow, and the maximum fuel flow would be governed by the degree of misalignment. A misaligned port could permit sufficient fuel flow to allow the engine to run continuously at a low power setting, but not permit sufficient fuel flow to sustain continuous high-power operation, such as for takeoff or climbout. In such a case, the fuel in the carburetor bowl would provide a small reserve to allow high power operation for a limited time, until that fuel was depleted, when the misaligned port then becomes the limiting factor for fuel flow. The SAC 1853 automated weather observation included winds from 200 degrees at 7 knots, visibility 10 miles, clear skies, temperature 27 degrees C, dew point 11 degrees C, and an altimeter setting of 29.74 inches of mercury. Temperature and dew point values for the approximate time of the accident indicated that the relative humidity was approximately 35 percent. When the intersection of the two temperature values was located on a chart that depicted carburetor ice envelopes, the point was in the region of the chart denoted as "Serious Icing at Glide Power," near the boundary of the region denoted as "Icing - Glide and Cruise Power." The wreckage was located on a golf course that was situated adjacent to the southern boundary of SAC. The fuselage came to rest inverted. Both wings had fracture-separated during the impact sequence, but the fuselage and cabin remained essentially intact. The engine remained attached to its mount, which remained partially attached to the fuselage. There was no fire, but firefighters did spray the wreckage with water due to the fact that the airplane was leaking fuel. The site was photographically documented by first responders, and the wreckage was then recovered to a secure facility for detailed examination. Representatives from the NTSB, the FAA, Lycoming Engines, and Piper Aircraft conducted a detailed examination of the wreckage in December 2013. All major components of the airframe and engine were accounted for. Flight control continuity was established for all primary and secondary flight controls. The flaps were determined to be up/retracted at impact, and the stabilator (pitch) trim was in the 'neutral' setting. The engine primer control was found in the 'in and locked' position. The throttle, mixture, and carburetor heat controls were all found in their full forward positions, and were securely attached at their respective control arms in the engine compartment. Damage to the carburetor heat airbox and actuation cable was consistent with the carburetor heat in the OFF (heat not applied) setting at impact. The ignition switch was found in the OFF position, and the key was absent. The master and fuel pump toggle switches we