Narrative:

We departed sua on an IFR flight plan, destination hagerstown, md. I was flying in an small aircraft and based on winds and WX information obtained from miami a time of filing, we estimated 4:50 en route with 5:50 fuel aboard. The flight was routine except that winds were not as helpful as predicted, and after passing sdz omni, we were approximately 30 mins behind flight plan estimates. We then changed our destination with dca center to winchester, va. As we approached the general area of winchester, the WX was poorer then forecast with multiple circuit breaker buildups and visibility below the haze level of 3-4 mi. Flight deviations based on stormscope readouts and approved by center added another 20 mins delay, and as we came up 20 or so mi south of winchester, it became obvious that it would be close on fuel. I asked dulles approach to expedite our descent, and they were unable. I reduced power to conserve remaining fuel and decided it should be sufficient. In the small aircraft aircraft return fuel is to the left tank only, so it is necessary to use 'left' frequently to keep tanks balanced. I had run the right tank down (on the 'both position'), to where the needle read east but was still bouncing, then switched to left tank. The operations manual is specific that for landing the fuel selector be on 'both', so as soon as were cleared for down for the VFR approach into winchester, I began the landing checklist so that I could keep a good look out for traffic in the 3-4 mi visibility. Approximately 8 mi out at about 4800' the engine died, and I set up the aircraft for best glide airspeed, richened the mixture and hit the boost pump. The left tank needle was still bouncing (in the turbulence), but the right was not. At his point I had been on 'both' tanks for about 5 mins. When after 90 seconds or so the engine did not fire up, I declared an emergency with dulles approach. I continued looking for the winchester airport (with which I am familiar), continuing descent at about 700 FPM. I then reviewed the situation and decided that since there obviously was fuel in the left tank, I'd switch the selector from 'both' to left. When I did, the engine came back up very smoothly. By then I was on winchester unicom (monitoring dulles approach on the #2 communication), so 30 seconds later when I saw the field I called dulles approach back and cancelled the emergency, explaining that the engine was now running ok. We proceeded in and landed west/O incident. We refueled and took approximately 67 gallons, 31 in the left, 36 in the right indicating at least 3 usable remaining in the left tank. I believe the cause of this incident is the poh in that nothing is said about the engine failing on the 'both' position if one tank is dry and sucking air.

Related reports:

Google
 

Original NASA ASRS Text

Title: ENGINE FAILURE. PLT OF SMA LOW ON FUEL FOLLOWING ACFT MANUAL INSTRUCTION, BOTH TANKS ON, SHUT OFF RIGHT TANK AND WAS ABLE TO RESTART ENGINE WITH LEFT TANK ONLY FOR LNDG. RIGHT TANK DRY, LEFT TANK HAD 3 GALLONS REMAINING.

Narrative: WE DEPARTED SUA ON AN IFR FLT PLAN, DEST HAGERSTOWN, MD. I WAS FLYING IN AN SMA AND BASED ON WINDS AND WX INFO OBTAINED FROM MIAMI A TIME OF FILING, WE ESTIMATED 4:50 ENRTE WITH 5:50 FUEL ABOARD. THE FLT WAS ROUTINE EXCEPT THAT WINDS WERE NOT AS HELPFUL AS PREDICTED, AND AFTER PASSING SDZ OMNI, WE WERE APPROX 30 MINS BEHIND FLT PLAN ESTIMATES. WE THEN CHANGED OUR DEST WITH DCA CTR TO WINCHESTER, VA. AS WE APCHED THE GENERAL AREA OF WINCHESTER, THE WX WAS POORER THEN FORECAST WITH MULTIPLE CB BUILDUPS AND VISIBILITY BELOW THE HAZE LEVEL OF 3-4 MI. FLT DEVS BASED ON STORMSCOPE READOUTS AND APPROVED BY CTR ADDED ANOTHER 20 MINS DELAY, AND AS WE CAME UP 20 OR SO MI S OF WINCHESTER, IT BECAME OBVIOUS THAT IT WOULD BE CLOSE ON FUEL. I ASKED DULLES APCH TO EXPEDITE OUR DSCNT, AND THEY WERE UNABLE. I REDUCED PWR TO CONSERVE REMAINING FUEL AND DECIDED IT SHOULD BE SUFFICIENT. IN THE SMA ACFT RETURN FUEL IS TO THE L TANK ONLY, SO IT IS NECESSARY TO USE 'L' FREQUENTLY TO KEEP TANKS BALANCED. I HAD RUN THE R TANK DOWN (ON THE 'BOTH POS'), TO WHERE THE NEEDLE READ E BUT WAS STILL BOUNCING, THEN SWITCHED TO L TANK. THE OPS MANUAL IS SPECIFIC THAT FOR LNDG THE FUEL SELECTOR BE ON 'BOTH', SO AS SOON AS WERE CLRED FOR DOWN FOR THE VFR APCH INTO WINCHESTER, I BEGAN THE LNDG CHKLIST SO THAT I COULD KEEP A GOOD LOOK OUT FOR TFC IN THE 3-4 MI VISIBILITY. APPROX 8 MI OUT AT ABOUT 4800' THE ENG DIED, AND I SET UP THE ACFT FOR BEST GLIDE AIRSPD, RICHENED THE MIXTURE AND HIT THE BOOST PUMP. THE L TANK NEEDLE WAS STILL BOUNCING (IN THE TURB), BUT THE R WAS NOT. AT HIS POINT I HAD BEEN ON 'BOTH' TANKS FOR ABOUT 5 MINS. WHEN AFTER 90 SECS OR SO THE ENG DID NOT FIRE UP, I DECLARED AN EMER WITH DULLES APCH. I CONTINUED LOOKING FOR THE WINCHESTER ARPT (WITH WHICH I AM FAMILIAR), CONTINUING DSNT AT ABOUT 700 FPM. I THEN REVIEWED THE SITUATION AND DECIDED THAT SINCE THERE OBVIOUSLY WAS FUEL IN THE L TANK, I'D SWITCH THE SELECTOR FROM 'BOTH' TO L. WHEN I DID, THE ENG CAME BACK UP VERY SMOOTHLY. BY THEN I WAS ON WINCHESTER UNICOM (MONITORING DULLES APCH ON THE #2 COM), SO 30 SECS LATER WHEN I SAW THE FIELD I CALLED DULLES APCH BACK AND CANCELLED THE EMER, EXPLAINING THAT THE ENG WAS NOW RUNNING OK. WE PROCEEDED IN AND LANDED W/O INCIDENT. WE REFUELED AND TOOK APPROX 67 GALLONS, 31 IN THE L, 36 IN THE R INDICATING AT LEAST 3 USABLE REMAINING IN THE L TANK. I BELIEVE THE CAUSE OF THIS INCIDENT IS THE POH IN THAT NOTHING IS SAID ABOUT THE ENG FAILING ON THE 'BOTH' POS IF ONE TANK IS DRY AND SUCKING AIR.

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.