Event ERA13LA143

Factual narrative

On February 24, 2013, about 1142 eastern standard time, a Cessna T337G, N8594M, registered to and operated by Exec Aviation LLC, was ditched in Biscayne Bay, near Homestead, Florida. Visual meteorological conditions prevailed at the time and no flight plan was filed for the 14 Code of Federal Regulations (CFR) Part 91 personal flight from North Perry Airport (HWO), Hollywood, Florida, to Key West International Airport (EYW), Key West, Florida. The airplane sustained substantial damage and the private pilot and three passengers sustained minor injuries. The flight originated about HWO about 20 minutes earlier. The pilot stated that before departure 78 gallons of 100 low lead fuel were added; after fueling the left fuel tank had about 56 gallons and the right fuel tank had about 52 gallons. After fueling he checked the fuel tanks and fuel strainers for contaminants and water; none were found. He departed with the left fuel tank supplying fuel to the front engine and the right fuel tank supplying fuel to the rear engine. After takeoff he proceeded eastbound to the coast then flew south along the coast while obtaining VFR flight following with air traffic control. When the flight was near Key Biscayne with the landing gear retracted flying at 900 feet, the rear engine quit suddenly. Because of the low altitude he didn’t have time to look at the engine gauges, but did glance enough to confirm the rear engine was not developing power. He added full power to the front engine and it did respond and also restarted the rear engine but during the process of restarting the rear engine, the flight lost about 500 feet in altitude. He reported that the rear engine manifold pressure after restart was between 15 and 17 inches, and the engine rpm was between 2,000 and 2,200. Because of the low altitude, and his concern about stalling the airplane, he did not want to continue to trouble shoot the rear engine. He realized in that configuration he could not make it to land and elected to ditch the airplane. He looked around and found a clear area and lowered the landing gear before ditching. He flared for landing and after touchdown the airplane nosed over. He reported having trouble with the cabin door latch but was able to kick the upper door open. All occupants exited the airplane within 20 seconds and while in the water, one passenger used a cell phone to call 911. All occupants were rescued by a boat then taken to land where they were treated by EMT’s. Following recovery of the airplane it was taken to a Federal Aviation Administration (FAA) certified repair station for an examination and operational testing of the rear engine. Also, the on-board JPI Fuel Scan 450 Twin Fuel Flow Gauge secured in the instrument panel was retained for examination by the NTSB Vehicle Recorder Laboratory located in Washington, D.C. Since the wings had been removed to recover and transport the airplane, an alternate fuel supply was plumbed into the right wing fuel supply line at the wing root. All spark plugs were removed, examined, and then tested as received at 80 psi. All spark plugs exhibited normal wear and tested satisfactory. With the spark plugs removed, the engine was motored several times using the starter in an effort to remove water and debris from the cylinders. Also, compressed air was used in an effort to remove moisture from both magnetos which were properly timed to the engine. The engine was started but only found to be operating on the left magneto. A slave magneto was installed on the right position and the engine was started and operated to 2,780 rpm and 34 inches manifold pressure (the airplane type certificate data sheet indicates maximum red line rpm and manifold pressure are 2,800 and 37 inches of Mercury, respectively). Before shutting down the engine a feather check was performed and the propeller fully feathered. The right magneto was retained for further examination. Inspection of the right magneto was performed at a FAA certified repair station with NTSB and FAA oversight. The magneto was placed on a test bench as received and noted to produce spark at all test leads; however, one test lead stopped firing. The magneto was removed from the test bench, and disassembled which revealed moisture and corrosion inside the magneto and distributor housings. While disassembled, the coil primary and secondary resistance was checked. The primary resistance was 14,000 Ohms (specification is 12,000 to 16,000 Ohms), while the secondary resistance was .3K Ohms (specification is .2 to .6K Ohms). No carbon tracking was noted on the distributor block. The condenser tested .420 microfarads (specification is .375 microfarads plus or minus 10 percent microfarads). The batch code on the condenser was 0834, and it did not pass most likely due to moisture. The point gap measured 0.021 inch (specification is 0.012 to 0.024 inch). The magneto was air dried, reassembled, and returned to the test bench where it sparked consistently at all leads between 150 and 4,700 rpm. The impulse coupling was noted to operate satisfactory. The JPI Fuel Scan 450 is a fuel flow indicator and totalizer, which in this particular model provides fuel information from both engines such as, fuel flow, fuel used and fuel remaining to the pilot. A chip level recovery was attempted after the unit did not power on after receiving external power. The binary file image created was decoded to provide the pilots last fuel used and fuel remaining for each engine. The left engine fuel used was 183.9 gallons, the right engine fuel used was 64.0 gallons, and the unit was powered up for a total of 83.6 hours. The airplane type certificate data sheet indicates a 62.5 gallon fuel tank is installed in each wing. Review of the airplane Owner’s Manual revealed that the procedures for engine-out during flight call to identify inoperative engine, open the cowl flaps on the operative engine, adjust the mixture on the operative engine and before securing the inoperative engine, check the fuel flow. If fuel flow is deficient, the checklist calls to advance the mixture control to full rich and turn on the auxiliary fuel of the inoperative engine. The checklist also indicates that if the inoperative engine does not start, to place the mixture control to idle cut-off, move the propeller control to feather, and turn off the auxiliary fuel pump, alternator, ignition switches, and fuel selector. The final step calls for closing the cowl flaps on the inoperative engine. Excerpts from the Owner’s Manual are included in the NTSB Public Docket for this case. According to the owner of the facility that fueled the airplane last, Truck 1 was utilized. He reported that they have not had a complaint pertaining to the quality of their fuel. Additionally, every day personnel inspect in part the condition of the truck, and check the compartment and filter sumps and record the findings. On February 24, 2013, the records indicate that with respect to Truck 1, the fuel sample taken from the compartment and filter sumps were both recorded to be clear and bright. The airplane was fueled before departure, and after fueling, the pilot reported that he checked the fuel tanks and fuel strainers and no contaminants were found. The facility that fueled the airplane reported no complaints or discrepancies related to the quality of fuel. The pilot departed and flew south along the coast at 900 feet above ground level when the rear engine suddenly lost power. The pilot was able to restart the rear engine, which regained only partial power, but not before losing 500 feet of altitude. Because of the low altitude, he did not want to troubleshoot the rear engine any further and opted to ditch the airplane with the landing gear extended. After touchdown, the airplane nosed over and all occupants exited the airplane and were rescued. The pilot further reported there was no issue related to the front engine. Following recovery of the airplane, rear engine spark plugs tested satisfactorily. An alternate fuel supply was plumbed into the right wing root, and, after water and contaminants were removed from the internal portion of the rear engine, it was started but found to be only operating on the left magneto. The right magneto was replaced, and the engine was restarted and operated satisfactorily. A feather check of the rear propeller was performed during the engine run and no issues were noted. The right magneto likely did not operate during the postaccident engine run because of internal moisture and corrosion due to salt water submersion. After cleaning the magneto, only the condenser tested outside its limits, most likely due to internal moisture. The magneto was placed on a test bench and it operated normally. Postaccident testing of the rear engine did not reveal the reason for the reported loss of power. Source: NTSB Aviation Accident Database Retrieved: 2026-02-12

NTSB Findings

Hierarchical cause / factor breakdown from the FAA bulk avdata database. Each finding tagged C (Cause) or F (Factor).

• C Not determined-Not determined-(general)-(general)-Unknown/Not determined - C