On July 31, 1995, at 1230 central daylight time, a Grumman 164, N949X, registered to Valley Sprayers, Inc., sustained substantial damage following loss of engine power and subsequent ditching in a man made reservoir. The commercial pilot reported no injuries. The 14 CFR Part 137 flight departed from Park River, North Dakota, on a local flight. Visual meteorological conditions prevailed and no flight plan was filed. The pilot reported that when he performed the engine run-up he checked his magnetos several times and cycled the prop about seven times. The takeoff was normal. He reported that just as he was reducing the RPM's back to 2050 RPM's, the engine started to pop and shake. He applied full throttle but the engine failed to produce any additional power. He reported that he began to lose altitude and airspeed. He looked for a place to perform an emergency landing and chose to land in the reservoir. After impact with the water the airplane remained "right side up," and he was able to get free of the aircraft and swim to shore. After the airplane was recovered from the reservoir, an Airworthiness Inspector from the Federal Aviation Administration examined the airplane. He reported that a visual inspection of the engine did not reveal any unusual conditions. The compression ckecked normal, and the engine and supercharger had continuity. (See attachments) The engine was sent to a repair facility for further examination. After the engine had been cleaned and minor repairs performed, which were required as a result of the airplane being submerged for 24 hours, the engine was run on a test stand. The first engine run was unsucessful. It was determined that the left and right magneto breaker point gaps were set at .003 of an inch instead of .009 of an inch specified in the maintenance manuals. Also the primary lead in the left magneto was grounded against the case and distributor cover. After those two discrepancies were corrected, the engine was started and run up to 2050 RPM. The engine ran well but full power was not applied due to the bent crankshaft. The maintenance manager concluded in his report that he felt, "the magnetos were the key problem." He also reported that, "Carburetor adjustments were made to assure a good run because of the cold weather. Keep in mind, because of the submersion, we had little to no cylinder compression and spark plugs tested weak." (See attachments)