Narrative:

My aircraft; a 1978 grumman american GA-7 light twin GA aircraft; made an off-airport landing on a gravel road as a result of fuel starvation. No injuries to the 3 occupants. All 3 exited the plane upon landing. The plane; however; suffered extensive damage on landing. Aircraft was fueled prior to departing on this flight. At the refueling; I selected 30 gallons as the purchased amount from the self-service credit card payment automated fuel pump. I installed 15 gallons; watching the meter; then filled the other tank until the pump automatically shut off. (Subsequent review of my credit card statement confirms a charge of $127 which is consistent with 30 gallons at $4.25/gallon.) after adding the 30 gallons; and prior to engine start; I checked the fuel level in each tank with a self-calibrated dipstick. The indications were that I had 32 gallons in the left tank and 25 gallons in the right tank prior to engine start. Fuel planning called for a 1 hour 45 min flight to my destination. The poh for the GA-7; borne out by my 120 hours in the plane; calls for 15 gph total fuel burn at cruise power settings of 24 inches/2400 RPM. Including taxi; run-up and climb; I was estimating 30 gallons of fuel burn for the leg and would have about 27 gallons at landing. That would be sufficient for a diversion if need be. I did not review the fuel gauges on the panel prior to takeoff. I depended on the dipstick; which is my regular method. Once airborne; the fuel gauges on this aircraft are 'bouncy.' fuel movement in the tanks makes the gauges move as much as 25%. You don't know what is accurate with that much displacement. While I do routinely review the gauges when airborne; that is more trend monitoring than absolute indication. On this flight; the trend downward was happening faster than anticipated but I did not begin to become cognizant or concerned of that until perhaps 30 mins prior to scheduled landing. At that point in the trip; turning back to my diversion airports were about as far a distance as continuing to the destination. I was not aware of any other alternative airports. I reduced power and decreased fuel richness in order to conserve and then lost power in the left engine about 7 mi from the destination. Shut fuel flow to that engine and feathered the propeller immediately. Flew another 2 mi and lost power in the right engine. We were at pattern altitude of 900 ft AGL; 5.2 mi from the airport; and about 1 mi south of the runway centerline. I announced to ZZZZ radio my intention to land on a gravel road that was within gliding distance and gave my approximately location. Don't recall if I feathered the right propeller. Lowered the landing gear. Forgot about using flaps. Reached the 'threshold' of the road at about 100 ft and 115 KIAS. Used up the length of the road. Probably would have made a good landing if I had remembered to deploy the flaps or not had feathered one/both engines. At the 'end' of the road was an approximately 90 degree left bend in the road. Continuing straight would have carried us into a stand of tall trees. Didn't care or that option. Was over the bend at about 15 ft AGL and 50 KIAS. Figured that trying the turn the bend; which would provide another long length of straight road; would cause a wing stall. Seemed like the preferred course of action; as opposed to hitting trees. Initiated the turn (perhaps 60 degree bank) and didn't stall; but did bleed off most of the remaining airspeed. The plane settled into a ditch adjacent to the road. Left wingtip contacted first on the shoulder of the road and caused a 90 degree left rotation of the plane. The plane settled right side up; 90 degrees rotated to the left after about 50 ft of skidding. We all immediately exited and called 911. Another interesting thought is that with one or both propellers feathered; approach speeds are much higher than any training flight with engine(south) idled and propellers windmilling. In my training; no flight instructor ever feathered a propeller on final approach. They all wanted the safety of a quick engine start. When in real life; the propellers are feathered; the wind resistance is down dramatically; and the approach speeds are higher than practiced. So all of the 'muscle memory' is incorrectly calibrated.

Related reports:

Google
 

Original NASA ASRS Text

Title: PLT OF GA7 LANDS SHORT OF DESTINATION DUE TO FUEL STARVATION TO BOTH ENGINES.

Narrative: MY ACFT; A 1978 GRUMMAN AMERICAN GA-7 LIGHT TWIN GA ACFT; MADE AN OFF-ARPT LNDG ON A GRAVEL ROAD AS A RESULT OF FUEL STARVATION. NO INJURIES TO THE 3 OCCUPANTS. ALL 3 EXITED THE PLANE UPON LNDG. THE PLANE; HOWEVER; SUFFERED EXTENSIVE DAMAGE ON LNDG. ACFT WAS FUELED PRIOR TO DEPARTING ON THIS FLT. AT THE REFUELING; I SELECTED 30 GALLONS AS THE PURCHASED AMOUNT FROM THE SELF-SVC CREDIT CARD PAYMENT AUTOMATED FUEL PUMP. I INSTALLED 15 GALLONS; WATCHING THE METER; THEN FILLED THE OTHER TANK UNTIL THE PUMP AUTOMATICALLY SHUT OFF. (SUBSEQUENT REVIEW OF MY CREDIT CARD STATEMENT CONFIRMS A CHARGE OF $127 WHICH IS CONSISTENT WITH 30 GALLONS AT $4.25/GALLON.) AFTER ADDING THE 30 GALLONS; AND PRIOR TO ENG START; I CHKED THE FUEL LEVEL IN EACH TANK WITH A SELF-CALIBRATED DIPSTICK. THE INDICATIONS WERE THAT I HAD 32 GALLONS IN THE L TANK AND 25 GALLONS IN THE R TANK PRIOR TO ENG START. FUEL PLANNING CALLED FOR A 1 HR 45 MIN FLT TO MY DEST. THE POH FOR THE GA-7; BORNE OUT BY MY 120 HRS IN THE PLANE; CALLS FOR 15 GPH TOTAL FUEL BURN AT CRUISE PWR SETTINGS OF 24 INCHES/2400 RPM. INCLUDING TAXI; RUN-UP AND CLB; I WAS ESTIMATING 30 GALLONS OF FUEL BURN FOR THE LEG AND WOULD HAVE ABOUT 27 GALLONS AT LNDG. THAT WOULD BE SUFFICIENT FOR A DIVERSION IF NEED BE. I DID NOT REVIEW THE FUEL GAUGES ON THE PANEL PRIOR TO TKOF. I DEPENDED ON THE DIPSTICK; WHICH IS MY REGULAR METHOD. ONCE AIRBORNE; THE FUEL GAUGES ON THIS ACFT ARE 'BOUNCY.' FUEL MOVEMENT IN THE TANKS MAKES THE GAUGES MOVE AS MUCH AS 25%. YOU DON'T KNOW WHAT IS ACCURATE WITH THAT MUCH DISPLACEMENT. WHILE I DO ROUTINELY REVIEW THE GAUGES WHEN AIRBORNE; THAT IS MORE TREND MONITORING THAN ABSOLUTE INDICATION. ON THIS FLT; THE TREND DOWNWARD WAS HAPPENING FASTER THAN ANTICIPATED BUT I DID NOT BEGIN TO BECOME COGNIZANT OR CONCERNED OF THAT UNTIL PERHAPS 30 MINS PRIOR TO SCHEDULED LNDG. AT THAT POINT IN THE TRIP; TURNING BACK TO MY DIVERSION ARPTS WERE ABOUT AS FAR A DISTANCE AS CONTINUING TO THE DEST. I WAS NOT AWARE OF ANY OTHER ALTERNATIVE ARPTS. I REDUCED PWR AND DECREASED FUEL RICHNESS IN ORDER TO CONSERVE AND THEN LOST PWR IN THE L ENG ABOUT 7 MI FROM THE DEST. SHUT FUEL FLOW TO THAT ENG AND FEATHERED THE PROP IMMEDIATELY. FLEW ANOTHER 2 MI AND LOST PWR IN THE R ENG. WE WERE AT PATTERN ALT OF 900 FT AGL; 5.2 MI FROM THE ARPT; AND ABOUT 1 MI S OF THE RWY CTRLINE. I ANNOUNCED TO ZZZZ RADIO MY INTENTION TO LAND ON A GRAVEL ROAD THAT WAS WITHIN GLIDING DISTANCE AND GAVE MY APPROX LOCATION. DON'T RECALL IF I FEATHERED THE R PROP. LOWERED THE LNDG GEAR. FORGOT ABOUT USING FLAPS. REACHED THE 'THRESHOLD' OF THE ROAD AT ABOUT 100 FT AND 115 KIAS. USED UP THE LENGTH OF THE ROAD. PROBABLY WOULD HAVE MADE A GOOD LNDG IF I HAD REMEMBERED TO DEPLOY THE FLAPS OR NOT HAD FEATHERED ONE/BOTH ENGS. AT THE 'END' OF THE ROAD WAS AN APPROX 90 DEG L BEND IN THE ROAD. CONTINUING STRAIGHT WOULD HAVE CARRIED US INTO A STAND OF TALL TREES. DIDN'T CARE OR THAT OPTION. WAS OVER THE BEND AT ABOUT 15 FT AGL AND 50 KIAS. FIGURED THAT TRYING THE TURN THE BEND; WHICH WOULD PROVIDE ANOTHER LONG LENGTH OF STRAIGHT ROAD; WOULD CAUSE A WING STALL. SEEMED LIKE THE PREFERRED COURSE OF ACTION; AS OPPOSED TO HITTING TREES. INITIATED THE TURN (PERHAPS 60 DEG BANK) AND DIDN'T STALL; BUT DID BLEED OFF MOST OF THE REMAINING AIRSPD. THE PLANE SETTLED INTO A DITCH ADJACENT TO THE ROAD. L WINGTIP CONTACTED FIRST ON THE SHOULDER OF THE ROAD AND CAUSED A 90 DEG L ROTATION OF THE PLANE. THE PLANE SETTLED RIGHT SIDE UP; 90 DEGS ROTATED TO THE L AFTER ABOUT 50 FT OF SKIDDING. WE ALL IMMEDIATELY EXITED AND CALLED 911. ANOTHER INTERESTING THOUGHT IS THAT WITH ONE OR BOTH PROPS FEATHERED; APCH SPDS ARE MUCH HIGHER THAN ANY TRAINING FLT WITH ENG(S) IDLED AND PROPS WINDMILLING. IN MY TRAINING; NO FLT INSTRUCTOR EVER FEATHERED A PROP ON FINAL APCH. THEY ALL WANTED THE SAFETY OF A QUICK ENG START. WHEN IN REAL LIFE; THE PROPS ARE FEATHERED; THE WIND RESISTANCE IS DOWN DRAMATICALLY; AND THE APCH SPDS ARE HIGHER THAN PRACTICED. SO ALL OF THE 'MUSCLE MEMORY' IS INCORRECTLY CALIBRATED.

Data retrieved from NASA's ASRS site as of May 2009 and automatically converted to unabbreviated mixed upper/lowercase text. This report is for informational purposes with no guarantee of accuracy. See NASA's ASRS site for official report.