On March 16, 2012, at 1100 eastern daylight time, a Flight Design GmbH CTSW, N379CT, was substantially damaged when it impacted an airport perimeter fence following a loss of engine power on takeoff from Venice Municipal Airport (VNC), Venice, Florida. The certificated commercial pilot and the pilot-rated passenger were not injured. Visual meteorological conditions prevailed, and no flight plan was filed for the flight, which was destined for Punta Gorda Airport (PGD), Punta Gorda, Florida. The personal flight was conducted under the provisions of Title 14 Code of Federal Regulations Part 91. Prior to departing from his home airport, the pilot had the airplane serviced so that about 20 gallons of fuel were onboard. He then departed for VNC, where he picked up the pilot-rated passenger, and after about 5 minutes on the ground, they boarded the airplane and departed on the accident flight. The pilot performed a brief run-up check of the engine, and then taxied onto runway 4 after another airplane had departed ahead. The pilot initiated the takeoff roll normally, observed normal indications on the electronic engine monitor, and noted that the tachometer was reading greater than 5,500 rpm. Upon reaching the intersection of runway 4 and runway 13/31, the airplane lifted off, then the warning annunciator "control" illuminated briefly and extinguished. Simultaneously, the primary flight display and the electronic engine monitor displayed the message, "Warning – Low Fuel Pressure." The engine then immediately lost power and the propeller ceased rotating. The pilot responded by lowering the airplane's nose, and was subsequently able to land the airplane within the grass overrun area beyond the departure end of the runway. During the rollout, the airplane struck the airport perimeter fence, resulting in substantial damage to the left wing. A representative of the engine manufacturer examined the engine under the supervision of a Federal Aviation Administration (FAA) inspector following the accident. Two of the three propeller blades were fractured from about their mid-points to their respective blade tips. The crankshaft was free to rotate, and thumb compression was confirmed on all cylinders. The magnetic chip plug was removed and a normal amount of metallic accumulation was observed. The top four spark plugs were removed and exhibited normal wear with some lead deposits, and the electrodes were white in color. All of the ignition components appeared to be in serviceable condition. The gascolator was disassembled and the fuel screen was found to be contaminated with an undetermined particulate, similar in appearance to sand. The fuel pump was removed from the engine and a clear fluid consistent in appearance and smell with automotive gasoline was observed at the inlet and outlet side. When actuated, the pump appeared to function normally. Both carburetors were disassembled and the right carburetor contained a small amount of particulate contamination in the fuel bowl, along with a normal fuel level. The left carburetor contained a significant amount of particulate contamination, which was sand-like in appearance, along with a lower-than-normal fuel level. Testing of the fuel recovered from both carburetors was negative for the presence of water, and the fuel had a blue tint consistent with 100LL aviation fuel. The engine was subsequently reassembled and a test run was performed on the airframe. The engine started and ran normally and throttle response appeared to be normal. A check of the ignition system showed that both the left and right side systems were functioning normally. The fuel pump was subsequently removed from the engine and forwarded to the engine manufacturer in Austria in order to confirm that it was operating within its required parameters. The pump was subsequently tested on January 26, 2013 under the supervision of the Austrian Civil Aviation Safety Investigation Authority. After being installed on a test bench, a fuel flow and pressure test was conducted. The test engine was subsequently started normally and ran smoothly. The fuel pressure remained within the prescribed limits throughout the engine test rpm range, which was between 1,800 and 5,800 rpm. During the test, a small drip of fuel was noted to be discharging from a weep hole on the pump, which did not affect its operation. Following the test, the fuel pump was disassembled and its internal components examined. Detailed examination of the pump’s rubber diaphragm showed that superficial cracks were present on both the fuel and dry sides. The diaphragm was then fitted into a test fixture and a vacuum was applied to test for leaks. No leakage of air was observed during the test. The accident airplane was equipped with a Dynon EMS-D120 engine monitoring system. In addition to providing a digital display of numerous engine parameters to the pilot, the system also recorded those parameters to a file. Following the accident, the system was recovered from the airplane and forwarded to the NTSB Vehicle Recorders laboratory, where its memory contents were downloaded. The downloaded data contained records that were recorded at a rate of once per second, and 1,798 records were present. The records appeared to capture the entirety of the accident flight, along with the conclusion of the preceding flight. The data began recording the accident flight when the system was powered on at 1048:38, and the engine was started about 1 minute later. At 1058:52, the engine rpm began increasing above the previously established average of about 1,700 rpm and reached about 4,800 rpm about 15 seconds later. About 4 seconds after the engine rpm began its increase, the fuel pressure began a rapid decrease from the previously established average of about 4.2 psi. At 1059:08, as the engine rpm reached nearly 4,800, the fuel pressure had fallen below 2 psi, and the fuel pressure low alarm triggered. The engine rpm continued to increase, reaching about 4,900 rpm at 1059:27, while the fuel pressure continued falling, reaching a low of 0.25 psi. The engine rpm then began dropping rapidly from about 4,900 to about 2,000 over the next 10 seconds, while the fuel pressure remained at 0.25 psi. As the engine rpm began to stabilize around 2,000, the fuel pressure increased briefly, reaching nearly 5.0 psi, but the engine rpm then continued to fall until it reached 0 at 1059:52. At that time, the fuel pressure had fallen to 1.94 psi.