Narrative:

The purpose of the flight was to transport a used experimental glasair on behalf of a friend who had just purchased the airplane from its point of purchase (ZZZ) to its new home base (ZZZ1).the original owner and builder of the aircraft was deceased. Prior to picking up the aircraft; myself and the new owner reviewed all the pertinent documentation and ensured it was correct and legal. Another friend; an ia; had previously performed a 3-4 hour pre-purchase inspection at the point of purchase. He noted no discrepancies effecting airworthiness. I reviewed the cockpit layout pictures and the poh prior to picking up the airplane. I also spoke; via phone; with the airplane builder's pilot-friend who was a frequent pilot of the airplane; although not recently. I specifically queried him about the airplane's fuel system and gear hydraulic system. I noted that the fuel selector had three positions: on; off and aerobatic. I confirmed with the pilot-friend that on was the main wing tank and aerobatic was a flop tube in the header tank. The ia indicated he was confident this was the case as well; based on other glasairs he was familiar with.myself and the new owner (in the passenger seat) picked up the airplane in ZZZ. I did several run-ups and other systems checks and fueled both tanks (single wing tank and header) within an inch of the top. There were multiple checklists in the aircraft and procedures listed in the poh; most of which did not agree with one another. We drained and tested the fuel from each drain point; did one final run up and associated checks. We departed with the fuel selector in the 'on' position. Both the ia and the pilot-friend along with all available checklists indicated that this was the correct procedure. Initially; I had planned to stop in ZZZ2 to practice some landings and check the aircraft for any anomalies. Prior to overflying ZZZ2; I elected to continue to ZZZ1; our final destination; as the airplane was running flawlessly and I did not want to get the airplane home at dusk. I informed ATC of my plan and turned towards the VOR. I climbed to 10;500 [feet] from 8;500 [feet] for extra margin as we would be overflying mountainous terrain. About 7 miles northwest of ZZZ2; about 25 mins into the flight; the engine coughed and quit cold. A scan of the instruments didn't reveal any anomalies (that I could see). The main fuel tank gauge between my legs indicated 3/4 full which seemed about right and turning on the fuel pump and switching to 'aerobatic' did not bring the engine back; so returned it to 'on'. With plenty of altitude to make ZZZ2; I began a turn towards the airport and [advised ATC]. At this point; I was committed to a power off landing; and was focused on getting the plane safely on the runway. However; just as a final attempt; I turned on the main fuel pump and two pumps labelled 'lt x-fer pump' and 'rt x-fer pump' which I had been told transferred fuel to the header tank from the main. I proceeded to approach the airport for landing. At approximately 5500 feet; the engine came to life. All vitals appeared good. I continued my approach. At about 3500 feet; I ran the engine up to full power for about 10 seconds to test it; since I wanted more information about the power loss; but it ran flawlessly; and I reduced power for the precautionary landing. After a normal landing; I removed the main tank fuel cap and noted that some air appeared to enter the tank. I assumed I had a venting issue. Later that day; we blew air through and confirmed the vent lines were now clear. I was not convinced there had been a blockage though. We left the airplane; returned to ZZZ1 in another aircraft and I spoke to several friends about the incident. One idea was that I might have actually been burning fuel from the header tank and run it dry. The timing and the aircraft's fuel burn seemed about right to exhaust a 5-6 gal tank in 25 mins. That would also explain why the engine came back and ran flawlesslya couple mins after turning on the two transfer pumps; as the header tank would have refilled. A call to the pilot-friend and sending him cockpit pictures to refresh his memory had him confirming; again; that the main tank was the one feeding in the 'on' position; not the header. I told him I suspected that was incorrect; but he held to his opinion. The new owner; who was riding right seat with me during the flight; also confirmed that he saw two lights light up in succession - one stated 'header tank 70% full' and the other 'header tank 25% full'. He did not tell me he saw these; and from my angle in the cockpit; it was not obvious. He stated he wasn't concerned as we weren't running on the header; and there were numerous other warnings and false alarms generated by several signal problems related to non-critical auxiliary EFIS and EMS functions; so he assumed he should ignore all warnings.myself; the owner and another glasair owner returned to ZZZ2 three days later to investigate and determine the cause. Upon arrival; we noted several things: 1) the main tank had less fuel than on departure from ZZZ. 2) the header tank was full; likely from the transfer pumps running during the emergency. 3) it wasn't possible to track all the fuel lines inside the cockpit due to the aircraft's build. We ran the transfer pumps and visually and auditory confirmed that each pumped fuel into the header tank. We also confirmed that when the header tank reached about one inch from the top; the fuel was siphoning back out of the tank; maintaining that level. We put the fuel selector 'on' and I ran the engine for 10 mins on the ground; with no pumps activated. I was able to ascertain that one page of the EMS system showed a bar graph indicating 'C fuel' which appeared to read '5' when full and represented the header tank. We measured and confirmed that the fuel burned was from the header; not the main tank and that the quantity roughly matched that bar graph on the EMS. One of the checklists indicated to turn on a single transfer pump at engine start and never indicated if/when we should turn it off. We read some builders' hand-scribbled notes and determined this single pump would flow upwards of 28gph - this was enough to feed the header tank and keep it full for the trip home. We then topped the tanks; checked the fuel and departed ZZZ2 for ZZZ1. The airplane flew flawlessly. The header tank remained full and the main tank drained as expected as the transfer pump ran. It was now obvious that the aircraft's non-standard fuel system; where both settings for the selector drained from the single; small header tank; and the undocumented procedures required to keep it full; along with our lack of knowledge about the aircraft systems and incorrect information from the pilot-friend; had caused the header fuel tank to run dry temporarily until power was restored by running the transfer pumps during our prior engine stoppage. There are some obvious human factors things here - like a better briefing between the owner-passenger and me; with the emphasis that any indication of a problem is to be taken seriously and communicated. In addition; positively testing all aircraft systems in an experimental aircraft where they are not clearly documented is essential. Several things we did do well: flying high; adding full fuel to both tanks and turning on the transfer pumps after the initial failure all helped to prevent a more serious incident or accident.

Google
 

Original NASA ASRS Text

Title: Experimental aircraft pilot reported a loss of engine power due to a non-standard fuel system and lack of system knowledge.

Narrative: The purpose of the flight was to transport a used experimental Glasair on behalf of a friend who had just purchased the airplane from its point of purchase (ZZZ) to its new home base (ZZZ1).The original owner and builder of the aircraft was deceased. Prior to picking up the aircraft; myself and the new owner reviewed all the pertinent documentation and ensured it was correct and legal. Another friend; an IA; had previously performed a 3-4 hour pre-purchase inspection at the point of purchase. He noted no discrepancies effecting airworthiness. I reviewed the cockpit layout pictures and the POH prior to picking up the airplane. I also spoke; via phone; with the airplane builder's pilot-friend who was a frequent pilot of the airplane; although not recently. I specifically queried him about the airplane's fuel system and gear hydraulic system. I noted that the fuel selector had three positions: On; Off and Aerobatic. I confirmed with the pilot-friend that On was the main wing tank and Aerobatic was a flop tube in the header tank. The IA indicated he was confident this was the case as well; based on other Glasairs he was familiar with.Myself and the new owner (in the passenger seat) picked up the airplane in ZZZ. I did several run-ups and other systems checks and fueled both tanks (single wing tank and header) within an inch of the top. There were multiple checklists in the aircraft and procedures listed in the POH; most of which did not agree with one another. We drained and tested the fuel from each drain point; did one final run up and associated checks. We departed with the fuel selector in the 'on' position. Both the IA and the pilot-friend along with all available checklists indicated that this was the correct procedure. Initially; I had planned to stop in ZZZ2 to practice some landings and check the aircraft for any anomalies. Prior to overflying ZZZ2; I elected to continue to ZZZ1; our final destination; as the airplane was running flawlessly and I did not want to get the airplane home at dusk. I informed ATC of my plan and turned towards the VOR. I climbed to 10;500 [feet] from 8;500 [feet] for extra margin as we would be overflying mountainous terrain. About 7 miles northwest of ZZZ2; about 25 mins into the flight; the engine coughed and quit cold. A scan of the instruments didn't reveal any anomalies (that I could see). The main fuel tank gauge between my legs indicated 3/4 full which seemed about right and turning on the fuel pump and switching to 'aerobatic' did not bring the engine back; so returned it to 'on'. With plenty of altitude to make ZZZ2; I began a turn towards the airport and [advised ATC]. At this point; I was committed to a power off landing; and was focused on getting the plane safely on the runway. However; just as a final attempt; I turned on the main fuel pump and two pumps labelled 'LT X-fer Pump' and 'RT X-fer pump' which I had been told transferred fuel to the header tank from the main. I proceeded to approach the airport for landing. At approximately 5500 feet; the engine came to life. All vitals appeared good. I continued my approach. At about 3500 feet; I ran the engine up to full power for about 10 seconds to test it; since I wanted more information about the power loss; but it ran flawlessly; and I reduced power for the precautionary landing. After a normal landing; I removed the main tank fuel cap and noted that some air appeared to enter the tank. I assumed I had a venting issue. Later that day; we blew air through and confirmed the vent lines were now clear. I was not convinced there had been a blockage though. We left the airplane; returned to ZZZ1 in another aircraft and I spoke to several friends about the incident. One idea was that I might have actually been burning fuel from the header tank and run it dry. The timing and the aircraft's fuel burn seemed about right to exhaust a 5-6 gal tank in 25 mins. That would also explain why the engine came back and ran flawlesslya couple mins after turning on the two transfer pumps; as the header tank would have refilled. A call to the pilot-friend and sending him cockpit pictures to refresh his memory had him confirming; again; that the main tank was the one feeding in the 'on' position; not the header. I told him I suspected that was incorrect; but he held to his opinion. The new owner; who was riding right seat with me during the flight; also confirmed that he saw two lights light up in succession - one stated 'header tank 70% full' and the other 'header tank 25% full'. He did not tell me he saw these; and from my angle in the cockpit; it was not obvious. He stated he wasn't concerned as we weren't running on the header; and there were numerous other warnings and false alarms generated by several signal problems related to non-critical auxiliary EFIS and EMS functions; so he assumed he should ignore all warnings.Myself; the owner and another Glasair owner returned to ZZZ2 three days later to investigate and determine the cause. Upon arrival; we noted several things: 1) The main tank had less fuel than on departure from ZZZ. 2) The header tank was full; likely from the transfer pumps running during the emergency. 3) It wasn't possible to track all the fuel lines inside the cockpit due to the aircraft's build. We ran the transfer pumps and visually and auditory confirmed that each pumped fuel into the header tank. We also confirmed that when the header tank reached about one inch from the top; the fuel was siphoning back out of the tank; maintaining that level. We put the fuel selector 'on' and I ran the engine for 10 mins on the ground; with no pumps activated. I was able to ascertain that one page of the EMS system showed a bar graph indicating 'C Fuel' which appeared to read '5' when full and represented the header tank. We measured and confirmed that the fuel burned was from the header; not the main tank and that the quantity roughly matched that bar graph on the EMS. One of the checklists indicated to turn on a single transfer pump at engine start and never indicated if/when we should turn it off. We read some builders' hand-scribbled notes and determined this single pump would flow upwards of 28gph - this was enough to feed the header tank and keep it full for the trip home. We then topped the tanks; checked the fuel and departed ZZZ2 for ZZZ1. The airplane flew flawlessly. The header tank remained full and the main tank drained as expected as the transfer pump ran. It was now obvious that the aircraft's non-standard fuel system; where both settings for the selector drained from the single; small header tank; and the undocumented procedures required to keep it full; along with our lack of knowledge about the aircraft systems and incorrect information from the pilot-friend; had caused the header fuel tank to run dry temporarily until power was restored by running the transfer pumps during our prior engine stoppage. There are some obvious human factors things here - like a better briefing between the owner-passenger and me; with the emphasis that ANY indication of a problem is to be taken seriously and communicated. In addition; positively testing all aircraft systems in an experimental aircraft where they are not clearly documented is essential. Several things we did do well: flying high; adding full fuel to both tanks and turning on the transfer pumps after the initial failure all helped to prevent a more serious incident or accident.

Data retrieved from NASA's ASRS site 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.