Narrative:

I think the left tank vent iced up. After a 4 hour flight from DCU, decatur, al, to ID2 at plymouth, mi the engine quit on the downwind leg for landing when I went through the gump check. I switched (stupidly!) to the left tank (gauge read higher than the right), and about the time I completed the gump check the engine quit at about 800 AGL. I quickly switched tanks, pumped the throttle, and pulled up to clear some trees and wires as I glided into and gently landed on a snow covered field just south of runway 36 an ID2, on a heading of about 070. After landing, and checking the tanks, I switched to the right tank, started the engine and taxied up to the runway and then to a fuel truck. The left tank took 40 gallons, the right tank 13. Each tank holds 40 gallons. How it was discovered: it is fairly obvious when an engine quits. Contributing factors: I did not religiously stick with a 30 min time limit on each tank as I usually do, starting with the left tank. The engine fuel system bypass returns fuel into the left tank. If it overfills, fuel can be pumped overboard. This flight was estimated at 4 hours, and it was exactly that. I left the selector on the left tank 1 hour to start. Then an hour on the right, but before the flight was up it looked like the left tank was filling again, so I switched to it 'for a while'. And here my recollection gets fuzzy. The first hour and half of the flight was in VMC. About then I was in and out of clouds, but below the freezing level. Around cincinnati I was in solid cloud at 6000. Shortly after entering toledo approach airspace I picked up some mixed ice, maybe 1/4 and requested a higher altitude. I was immediately given 8000. I was in and out of snow showers at 8000 but picked up no more ice. I completed the VOR approach to ID2 (on the right tank) and broke out of the clouds at about 900 AGL/1600 MSL, had the airport in sight, cancelled my IFR flight plan with detroit approach and switched to unicom frequency. Unicom advised that 36 was the active runway, no traffic. I usually fly a fairly close-in downwind leg, and did this time, and was about abeam the touchdown point when I went through the gump check. However, the gear was already down as I had lowered them to expedite my descent to minimize the time in the clouds. After I had switched to the left tank, I then lowered the flaps, pushed in the mixture control and advanced the propeller to high RPM. Corrective actions: 1. Always, always, go by time and fuel burn rate in fuel management. 2. Check that the fuel tank vents have been 'anti-ice' modified. When the vent ices over, and the fuel is drawn out, the tank bladder can pull up, or collapse, the fuel sensor is affected and the fuel gauge will read erroneously. 3. The final corrective action is to also check the O rings or seals on the filler caps. If they leak (usually can see a fine mist of fuel after takeoff with full tanks) the fuel will feed down a little and not come out, but the air flowing over the operation of the cap reduces the pressure inside the tank and can pull up or draw in the fuel bladder nd the gauge will again not be accurate. Be sharp, and quick, on emergency procedure. The engine might have restarted if I had also flipped on the electric boost pump after switching tanks (not a bad habit on any case). However, I was busy turning toward the only open area available and losing altitude fast. (Feeble excuse). Finally, I was luckier'n a blind dog in a meat house on this one.

Google
 

Original NASA ASRS Text

Title: LIGHT ACFT SUFFERS FORCED LNDG DUE TO FUEL STARVATION. FUEL STARVATION DUE TO A COMBINATION OF ACFT EQUIP PROBLEM AND PLT ERROR.

Narrative: I THINK THE L TANK VENT ICED UP. AFTER A 4 HR FLT FROM DCU, DECATUR, AL, TO ID2 AT PLYMOUTH, MI THE ENG QUIT ON THE DOWNWIND LEG FOR LNDG WHEN I WENT THROUGH THE GUMP CHK. I SWITCHED (STUPIDLY!) TO THE L TANK (GAUGE READ HIGHER THAN THE RIGHT), AND ABOUT THE TIME I COMPLETED THE GUMP CHECK THE ENG QUIT AT ABOUT 800 AGL. I QUICKLY SWITCHED TANKS, PUMPED THE THROTTLE, AND PULLED UP TO CLR SOME TREES AND WIRES AS I GLIDED INTO AND GENTLY LANDED ON A SNOW COVERED FIELD JUST S OF RWY 36 AN ID2, ON A HDG OF ABOUT 070. AFTER LNDG, AND CHKING THE TANKS, I SWITCHED TO THE R TANK, STARTED THE ENG AND TAXIED UP TO THE RWY AND THEN TO A FUEL TRUCK. THE L TANK TOOK 40 GALLONS, THE R TANK 13. EACH TANK HOLDS 40 GALLONS. HOW IT WAS DISCOVERED: IT IS FAIRLY OBVIOUS WHEN AN ENG QUITS. CONTRIBUTING FACTORS: I DID NOT RELIGIOUSLY STICK WITH A 30 MIN TIME LIMIT ON EACH TANK AS I USUALLY DO, STARTING WITH THE L TANK. THE ENG FUEL SYS BYPASS RETURNS FUEL INTO THE L TANK. IF IT OVERFILLS, FUEL CAN BE PUMPED OVERBOARD. THIS FLT WAS ESTIMATED AT 4 HRS, AND IT WAS EXACTLY THAT. I L THE SELECTOR ON THE L TANK 1 HR TO START. THEN AN HR ON THE R, BUT BEFORE THE FLT WAS UP IT LOOKED LIKE THE L TANK WAS FILLING AGAIN, SO I SWITCHED TO IT 'FOR A WHILE'. AND HERE MY RECOLLECTION GETS FUZZY. THE FIRST HR AND HALF OF THE FLT WAS IN VMC. ABOUT THEN I WAS IN AND OUT OF CLOUDS, BUT BELOW THE FREEZING LEVEL. AROUND CINCINNATI I WAS IN SOLID CLOUD AT 6000. SHORTLY AFTER ENTERING TOLEDO APCH AIRSPACE I PICKED UP SOME MIXED ICE, MAYBE 1/4 AND REQUESTED A HIGHER ALT. I WAS IMMEDIATELY GIVEN 8000. I WAS IN AND OUT OF SNOW SHOWERS AT 8000 BUT PICKED UP NO MORE ICE. I COMPLETED THE VOR APCH TO ID2 (ON THE R TANK) AND BROKE OUT OF THE CLOUDS AT ABOUT 900 AGL/1600 MSL, HAD THE ARPT IN SIGHT, CANCELLED MY IFR FLT PLAN WITH DETROIT APCH AND SWITCHED TO UNICOM FREQ. UNICOM ADVISED THAT 36 WAS THE ACTIVE RWY, NO TFC. I USUALLY FLY A FAIRLY CLOSE-IN DOWNWIND LEG, AND DID THIS TIME, AND WAS ABOUT ABEAM THE TOUCHDOWN POINT WHEN I WENT THROUGH THE GUMP CHK. HOWEVER, THE GEAR WAS ALREADY DOWN AS I HAD LOWERED THEM TO EXPEDITE MY DSCNT TO MINIMIZE THE TIME IN THE CLOUDS. AFTER I HAD SWITCHED TO THE L TANK, I THEN LOWERED THE FLAPS, PUSHED IN THE MIXTURE CTL AND ADVANCED THE PROP TO HIGH RPM. CORRECTIVE ACTIONS: 1. ALWAYS, ALWAYS, GO BY TIME AND FUEL BURN RATE IN FUEL MGMNT. 2. CHK THAT THE FUEL TANK VENTS HAVE BEEN 'ANTI-ICE' MODIFIED. WHEN THE VENT ICES OVER, AND THE FUEL IS DRAWN OUT, THE TANK BLADDER CAN PULL UP, OR COLLAPSE, THE FUEL SENSOR IS AFFECTED AND THE FUEL GAUGE WILL READ ERRONEOUSLY. 3. THE FINAL CORRECTIVE ACTION IS TO ALSO CHK THE O RINGS OR SEALS ON THE FILLER CAPS. IF THEY LEAK (USUALLY CAN SEE A FINE MIST OF FUEL AFTER TKOF WITH FULL TANKS) THE FUEL WILL FEED DOWN A LITTLE AND NOT COME OUT, BUT THE AIR FLOWING OVER THE OP OF THE CAP REDUCES THE PRESSURE INSIDE THE TANK AND CAN PULL UP OR DRAW IN THE FUEL BLADDER ND THE GAUGE WILL AGAIN NOT BE ACCURATE. BE SHARP, AND QUICK, ON EMER PROC. THE ENG MIGHT HAVE RESTARTED IF I HAD ALSO FLIPPED ON THE ELECTRIC BOOST PUMP AFTER SWITCHING TANKS (NOT A BAD HABIT ON ANY CASE). HOWEVER, I WAS BUSY TURNING TOWARD THE ONLY OPEN AREA AVAILABLE AND LOSING ALT FAST. (FEEBLE EXCUSE). FINALLY, I WAS LUCKIER'N A BLIND DOG IN A MEAT HOUSE ON THIS ONE.

Data retrieved from NASA's ASRS site as of July 2007 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.