On February 4, 1995, about 1550 central standard time (cst), a Piper PA-28-140, N694SK, was destroyed when it crashed into an abandoned buildings roof following a total loss of power. Visual meteorological conditions prevailed at the time of the accident. The personal 14 CFR Part 91 flight was not operating on a flight plan. The private pilot received minor injuries and the two passengers were seriously injured. The flight departed St. Louis, Missouri, at 1540 cst.

During the preflight inspection the pilot drained the left fuel tank sump twice. He said he did this because the first sample revealed a small amount of water. He reported that no other abnormalities were noted during the preflight inspection. The pilot attempted to start the engine without success. He said after the first startup attempt he primed the engine twice and tried to start the engine. The engine did not start on the second attempt.

The pilot said his third attempt at starting the engine was successful. He said he primed the engine and then stowed the primer knob. The pilot said he placed the mixture control to the "fuel cutoff" position, and increased the throttle before the engine started. Carburetor heat was checked before takeoff, and it operated properly. The pilot said the engine runup was normal "except I stayed a little longer to get the oil temperature to rise."

The electric fuel pump was turned off during the climb. The pilot said he leveled off at 2,300 feet mean sea level (msl). Approximately 10 minutes after takeoff the pilot reported the engine exhibited a "quick sputter." This was followed by a drastic drop to 1,500-1,600 RPM. The RPM then oscillated between 1,500-1,900 RPM. The pilot said a scan of the engine instruments revealed normal fuel and oil pressure. The pilot turned on the electric fuel pump and moved the fuel selector from the left to right tank position. He said he applied full carburetor heat and left it on. The pilot said he cycled the mixture control and moved the magneto's switch "left" and "right" and noted no power change.

The pilot said "sometimes the RPM would go up [a little]... and then drop again." The engine ceased producing power a short time later. The pilot said he turned the electric fuel pump and fuel selector off and initiated a left turn for an emergency landing. The pilot said he switched off the magnetos and master switch during the descent. The airplane impacted the roof of an abandoned two-story building in a nose low attitude. The pilot stated the propeller was windmilling until impact, and the flaps were in the full up position.

One eye witness said he saw a "...pencil-like stream of smoke coming from the aircraft." A second witness said the engine "...sounded as if it were running out of gas." A third witness said the airplane was dumping fuel. He said a fluid from the airplane landed on his car windshield and evaporated.

The on-scene investigation was conducted by a Federal Aviation Administration (FAA) Principal Maintenance Inspector (PMI). The PMI reported that both of N694SK's wing fuel tanks were torn open and contained no fuel. A small amount of fuel was recovered from the left wing fuel tank lines when air was blown through the lines. Examination of the fuel system revealed the fuel strainer had separated and the bowl was filled with pieces of roof material.

The engine inspection revealed no mechanical anomalies that would prevent production of power. A cold compression check of the cylinders showed the following values for the respective cylinders: #1, 78/80 psi; #2, 74/80 psi; #3, 74/80; #4, 70/80 psi. The left magneto produced spark when turned by hand. The right magneto failed to produce a spark until it was disassembled and turned by hand. Disassembly of the magnetos revealed the contacts and detents exhibited normal wear and continuity of the ignition switches. The carburetor's two-piece venturi exhibited sooting over most of its surface.

Examination of the spark plugs revealed carbon-like deposits on the #2 cylinder top and bottom spark plugs. Lighter carbon-like deposits were observed on the other three sets of spark plugs. All spark Plugs had varying degrees of sooty discoloration on them. The Champion Spark Plug "Aviation Check-A-Plug" trouble shooting guide states that the following conditions typically occur when a spark plug is carbon fouled: "sooty deposits from extensive ground idling, idle mixture too rich or plug type too cold."

The fuel lines leading to the electric boost pump, and from the strainer to the engine fuel pump were intact. All fuel lines were secure and intact between the engine driven fuel pump and carburetor. No evidence of leakage was observed between or on these two components. The carburetor's fuel inlet screen and bowl were found clean and free of contamination. The composite floats and needle valve exhibited no evidence of contamination or distortion and operated properly when tested.

The review of the aircraft and engine logbooks revealed that both magnetos were replaced on August 30, 1994, (90 hours before the accident). Both were replaced a second time on October 11, 1994, (58 hours before the accident). The #2 cylinder's spark plugs were replaced on August 11, 1994, (101 hours before the accident). The carburetor bowl gasket was replaced on February 3, 1992, when the aircraft had accumulated 1,631 total hours (247 hours before the accident).

Based on the reported temperature and dew point (30 and 12 degrees Fahrenheit respectively) carburetor icing was not a possibility. An FAA carburetor icing probability chart is appended to this report that illustrates this. FAA publication AC 61-21A, "Flight Training Handbook," says that when carburetor ice forms it reduces the power output. N694SK's pilot operating handbook (POH) states, "Carburetor heat should not be applied unless there is an indication of carburetor icing....."

The starting sequence described by the pilot closely resembles the flooded engine start described in the airplane's POH. A copy of this procedure is appended to this report. The engine start checklist directs the pilot to "Turn the electric fuel pump ON" when starting a cold or hot engine. The use of the engine fuel primer is not mentioned in the POH's starting procedures. A copy of the starting procedure is appended to this report.

According to the FAA PMI, a sooty carburetor venturi shows a backfire during an engine start. Spark plugs with carbon-like deposits and sooty coloration show a rich mixture during engine operations. The engine primer assembly was found bent to the right about 20 degrees during the on-scene investigation. The distance from the instrument panel and end of the knob is about 1 1/4 inches. This corresponds with a stowed position. It was not determined if the knob had been locked in the stowed position before the accident.

The primer on N694SK's engine is connected to 3 of the 4 cylinders. The primer is connected directly to the fuel strainer and engine cylinders. An open primer handle will, according to the engines manufacturer representative, allow additional fuel to enter the cylinders. He said an activated electric boost pump will add more fuel to the cylinders. An excerpt from N964SK's POH describing the fuel system is appended to this report.

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