Narrative:

The flight profile was a routine training flight to review takeoff and landing procedures for a pre-solo student pilot with 20 hours flight time. The flight consisted of normal pattern work at el monte and brackett/lavern airports. During the preflight the amount of fuel was checked independently by my student and by myself both visually and by the gauge. It was determined that the aircraft held 1/2 tank of fuel. Approximately 9- 10 gals. Contributing factor: it is a difficult process to check the fuel level. The fuel tank is located in the center of the fuselage behind the pilots' seats. The inspecting pilot must sight down the 1 1/2 to 2 ft long filler neck to try and see the fuel. If the light is too dark or too bright, then this becomes even more difficult, if not impossible. A factory calibrated dip- stick was not yet available. Fuel loads above 1/2 tank cannot be visually verified. For fuel loads below 1/2 tank, it was necessary to compare the fuel level to 3 welds in the side of the tank. It was recommended that the 'top weld' was equivalent to 1/2 tank or 10.15 gals. The DA20-A1 burns on average 5 gph. Thus the 20.3 usable gals of fuel is approximately 4 hours of flight time. It is also often required that the DA20-A1 be operated with 1/2 fuel load due to weight limitations and or density altitude restrs. Problem discovery: having agreed that we had sufficient fuel for the flight, (EST 1 hour plus 50 min) we departed. Upon returning to el monte from brackett, I evaluated the remaining fuel remaining. We had been flying for 1 hour and 06 min, 1.1 hours, thus should have 45 mins of fuel remaining. It was decided to accomplish 3 more patterns. On the up-wind leg of the 3RD pattern (approximately 300 ft AGL) the aircraft lost power. It was noted at this time that the gauge still indicated 1/4 tank. For several seconds the engine power surged in and out providing partial power to climb an additional 100 ft and maintain altitude long enough to turn an early left crosswind. This was the start of an attempt to return to the field. Shortly after this turn the engine lost considerable power and I made the decision to make an off airport emergency landing. An emergency was declared with el monte tower. The best available landing area was the I-10 freeway eastbound due to the traffic moving at the approximately landing speed of the aircraft. An area of the freeway was chosen free of overhead wires. Full flaps were applied to match the speed of traffic and fly to the chosen area. A safe and successful landing was made with no damage or injury to people, property or aircraft. After we were on the ground, the fuel gauge still read 1/8 of a tank full. Corrective actions: since the incident, a dip-stick or pipette has been made available from diamond aircraft. With this pipette, we have made a careful study of the other katanas in our fleet. 7 in all. We have found the same errors in every gauge. The average error is 1/4 tank high on the gauge. Some of the katanas can have as much as 1/2 tank error at times. Each aircraft also seems to register more fuel in-flight than when on the ground. The pipette has also revealed that the prior visual estimations of fuel available below 1/2 tank have been low in error as much as 2 gals. There also seems to be a temperature factor with the gauge errors (warmer the more error). There is also a small error in the pipette if it is used incorrectly. It may show a 1/2 inch error if dipped at an angle due to the strange geometry of the fuel tank. Our policy now is to have each pilot dip their tank twice before each flight to verify fuel loads if the tank is not full. It has also been suggested to diamond aircraft that a fuel low level light be installed independent of the fuel gauge. This would be helpful since there is no mixture control on this aircraft. There are 2 bing altitude sensing carbs. Fuel burn might vary from aircraft to aircraft and many pilots that have new certificates may not be cognizant of how long they have been flying. Callback conversation with reporter revealed the following information: the reporter stated he now personally will not take off in a katana with less than 3/4 tank. With a 200 pound flight examiner and a 200 pound pilot the katanawould be over gross with full fuel. There have been numerous incidents of pilots running the katana out of fuel. The fuel gauges have been off as much as 1/2 tank. There are 3 weld lines in the side of the tank. The top weld was said to be at the 1/2 tank level or 10 gals. What was not known was that 1/4 inch below that weld line is 1/4 tank. The geometry of the tank is deceptive. The operator has 7 katanas. The success of his emergency landing on the freeway he attributes partly to the preplanning he had done in the past. The flying school no longer allows pilots to fly the katana solo departing with less than the full fuel.

Related reports:

Google
 

Original NASA ASRS Text

Title: THE RPTR, A FLT INSTRUCTOR, AND HIS STUDENT DURING A TRAINING FLT RAN OUT OF FUEL AND MADE AN OFF ARPT LNDG ON A FREEWAY.

Narrative: THE FLT PROFILE WAS A ROUTINE TRAINING FLT TO REVIEW TKOF AND LNDG PROCS FOR A PRE-SOLO STUDENT PLT WITH 20 HRS FLT TIME. THE FLT CONSISTED OF NORMAL PATTERN WORK AT EL MONTE AND BRACKETT/LAVERN ARPTS. DURING THE PREFLT THE AMOUNT OF FUEL WAS CHKED INDEPENDENTLY BY MY STUDENT AND BY MYSELF BOTH VISUALLY AND BY THE GAUGE. IT WAS DETERMINED THAT THE ACFT HELD 1/2 TANK OF FUEL. APPROX 9- 10 GALS. CONTRIBUTING FACTOR: IT IS A DIFFICULT PROCESS TO CHK THE FUEL LEVEL. THE FUEL TANK IS LOCATED IN THE CTR OF THE FUSELAGE BEHIND THE PLTS' SEATS. THE INSPECTING PLT MUST SIGHT DOWN THE 1 1/2 TO 2 FT LONG FILLER NECK TO TRY AND SEE THE FUEL. IF THE LIGHT IS TOO DARK OR TOO BRIGHT, THEN THIS BECOMES EVEN MORE DIFFICULT, IF NOT IMPOSSIBLE. A FACTORY CALIBRATED DIP- STICK WAS NOT YET AVAILABLE. FUEL LOADS ABOVE 1/2 TANK CANNOT BE VISUALLY VERIFIED. FOR FUEL LOADS BELOW 1/2 TANK, IT WAS NECESSARY TO COMPARE THE FUEL LEVEL TO 3 WELDS IN THE SIDE OF THE TANK. IT WAS RECOMMENDED THAT THE 'TOP WELD' WAS EQUIVALENT TO 1/2 TANK OR 10.15 GALS. THE DA20-A1 BURNS ON AVERAGE 5 GPH. THUS THE 20.3 USABLE GALS OF FUEL IS APPROX 4 HRS OF FLT TIME. IT IS ALSO OFTEN REQUIRED THAT THE DA20-A1 BE OPERATED WITH 1/2 FUEL LOAD DUE TO WT LIMITATIONS AND OR DENSITY ALT RESTRS. PROB DISCOVERY: HAVING AGREED THAT WE HAD SUFFICIENT FUEL FOR THE FLT, (EST 1 HR PLUS 50 MIN) WE DEPARTED. UPON RETURNING TO EL MONTE FROM BRACKETT, I EVALUATED THE REMAINING FUEL REMAINING. WE HAD BEEN FLYING FOR 1 HR AND 06 MIN, 1.1 HRS, THUS SHOULD HAVE 45 MINS OF FUEL REMAINING. IT WAS DECIDED TO ACCOMPLISH 3 MORE PATTERNS. ON THE UP-WIND LEG OF THE 3RD PATTERN (APPROX 300 FT AGL) THE ACFT LOST PWR. IT WAS NOTED AT THIS TIME THAT THE GAUGE STILL INDICATED 1/4 TANK. FOR SEVERAL SECONDS THE ENG PWR SURGED IN AND OUT PROVIDING PARTIAL PWR TO CLB AN ADDITIONAL 100 FT AND MAINTAIN ALT LONG ENOUGH TO TURN AN EARLY L XWIND. THIS WAS THE START OF AN ATTEMPT TO RETURN TO THE FIELD. SHORTLY AFTER THIS TURN THE ENG LOST CONSIDERABLE PWR AND I MADE THE DECISION TO MAKE AN OFF ARPT EMER LNDG. AN EMER WAS DECLARED WITH EL MONTE TWR. THE BEST AVAILABLE LNDG AREA WAS THE I-10 FREEWAY EBOUND DUE TO THE TFC MOVING AT THE APPROX LNDG SPD OF THE ACFT. AN AREA OF THE FREEWAY WAS CHOSEN FREE OF OVERHEAD WIRES. FULL FLAPS WERE APPLIED TO MATCH THE SPD OF TFC AND FLY TO THE CHOSEN AREA. A SAFE AND SUCCESSFUL LNDG WAS MADE WITH NO DAMAGE OR INJURY TO PEOPLE, PROPERTY OR ACFT. AFTER WE WERE ON THE GND, THE FUEL GAUGE STILL READ 1/8 OF A TANK FULL. CORRECTIVE ACTIONS: SINCE THE INCIDENT, A DIP-STICK OR PIPETTE HAS BEEN MADE AVAILABLE FROM DIAMOND ACFT. WITH THIS PIPETTE, WE HAVE MADE A CAREFUL STUDY OF THE OTHER KATANAS IN OUR FLEET. 7 IN ALL. WE HAVE FOUND THE SAME ERRORS IN EVERY GAUGE. THE AVERAGE ERROR IS 1/4 TANK HIGH ON THE GAUGE. SOME OF THE KATANAS CAN HAVE AS MUCH AS 1/2 TANK ERROR AT TIMES. EACH ACFT ALSO SEEMS TO REGISTER MORE FUEL INFLT THAN WHEN ON THE GND. THE PIPETTE HAS ALSO REVEALED THAT THE PRIOR VISUAL ESTIMATIONS OF FUEL AVAILABLE BELOW 1/2 TANK HAVE BEEN LOW IN ERROR AS MUCH AS 2 GALS. THERE ALSO SEEMS TO BE A TEMP FACTOR WITH THE GAUGE ERRORS (WARMER THE MORE ERROR). THERE IS ALSO A SMALL ERROR IN THE PIPETTE IF IT IS USED INCORRECTLY. IT MAY SHOW A 1/2 INCH ERROR IF DIPPED AT AN ANGLE DUE TO THE STRANGE GEOMETRY OF THE FUEL TANK. OUR POLICY NOW IS TO HAVE EACH PLT DIP THEIR TANK TWICE BEFORE EACH FLT TO VERIFY FUEL LOADS IF THE TANK IS NOT FULL. IT HAS ALSO BEEN SUGGESTED TO DIAMOND ACFT THAT A FUEL LOW LEVEL LIGHT BE INSTALLED INDEPENDENT OF THE FUEL GAUGE. THIS WOULD BE HELPFUL SINCE THERE IS NO MIXTURE CTL ON THIS ACFT. THERE ARE 2 BING ALT SENSING CARBS. FUEL BURN MIGHT VARY FROM ACFT TO ACFT AND MANY PLTS THAT HAVE NEW CERTIFICATES MAY NOT BE COGNIZANT OF HOW LONG THEY HAVE BEEN FLYING. CALLBACK CONVERSATION WITH RPTR REVEALED THE FOLLOWING INFO: THE RPTR STATED HE NOW PERSONALLY WILL NOT TAKE OFF IN A KATANA WITH LESS THAN 3/4 TANK. WITH A 200 LB FLT EXAMINER AND A 200 LB PLT THE KATANAWOULD BE OVER GROSS WITH FULL FUEL. THERE HAVE BEEN NUMEROUS INCIDENTS OF PLTS RUNNING THE KATANA OUT OF FUEL. THE FUEL GAUGES HAVE BEEN OFF AS MUCH AS 1/2 TANK. THERE ARE 3 WELD LINES IN THE SIDE OF THE TANK. THE TOP WELD WAS SAID TO BE AT THE 1/2 TANK LEVEL OR 10 GALS. WHAT WAS NOT KNOWN WAS THAT 1/4 INCH BELOW THAT WELD LINE IS 1/4 TANK. THE GEOMETRY OF THE TANK IS DECEPTIVE. THE OPERATOR HAS 7 KATANAS. THE SUCCESS OF HIS EMER LNDG ON THE FREEWAY HE ATTRIBUTES PARTLY TO THE PREPLANNING HE HAD DONE IN THE PAST. THE FLYING SCHOOL NO LONGER ALLOWS PLTS TO FLY THE KATANA SOLO DEPARTING WITH LESS THAN THE FULL FUEL.

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.