I realized yesterday that my means of measuring usable fuel was somewhat flawed, in that measuring leftover capacity in tanks drained is not valid. It’s nice to know that there is .0312 gallons left in each tank after a draining, but what we really need to know is how much is left after the tanks are pumped dry.
So I went back up there and raised the tail up to flight attitude again. This time it was a little easier, because I found a weird old pallet lift at the back of the hangar, which allowed me to precisely lift the tailwheel to a desired height. I drained out both tanks again, then starting with the right tank, I added exactly one gallon of fuel. That’s when the fun started.
I’d made a fitting and tube to allow me to pump fuel from the firewall fitting into a measured pitcher below the airplane, and yes, I did make a bonding cable from a piece of 10ga wire and a couple of cleco clamps. One end was clipped to a steel rod sitting in the pitcher and the other was clamped to the airframe to mitigate static buildup. With this setup, I should have been able to pump the tank dry and measure the difference.
Wrong. The pump buzzed and blew some air but no fuel. Hmm.. Won’t prime. I attached a rubber tube to the vent fitting and blew through it, thinking maybe a little more pressure would help the pump get started. Oh yeah, that… sort of works. But then, I saw fuel dripping from the skin, outboard where it meets the spar. WTF?
In the cockpit, there was a nice puddle of fuel on the floorboards. Oh f—. Stop everything, grab paper towels, mop up the spill. This is where the lesson in finishing comes in. When I reinstalled the fuel pump and covers, the intention was to get the covers and such back on so I could fit the interior, carpet, etc. An actual functioning fuel system was not in the scope of the job. It should have been.
There were three loose fittings at the fuel selector end, which were wet enough to feel. Idiot. After mopping up the fuel and letting the compartment ventilate for a while, I tightened those up and whaddya know, fuel starts shooting into the container when the pump runs. I backed up and drained the fuel again, repeating the measured pour into the tank. This time, I got 8 oz pumped into the waiting pitcher. Add the 4 ounces of fuel left in the tank after the drain test from yesterday and that’s 12 oz of unusable fuel in the right tank.
The left tank went fine, only this time I got 11.5 oz of unusable fuel.
At this point, what the hell, let’s see if the fuel system will charge. I reconnected the fuel pump feed line to the firewall fitting and hit the boost pump again. And got a spray of fuel coming from the fuel pressure sensor hose fitting. This was my second warning. At this point I started to clue in and checked all the fuel fittings downstream of the firewall, which were plenty tight. Fire up boost pump again, and look for leaks. None firewarll forward. But I got another puddle of fuel on the floorboards for my trouble. Huh?
I removed the fuel fitting panel on the carb heat cover and yep, the fuel feed line and the purge valve return line were both loose. How the hell did that happen? Did I just not bother to tighten anything at all in there?
With those tight, the only leak came out of the sniffle valve on the sump, which is fine. It was also just a few drops, probably because I’d left the throttle open when I ran the boost pump.
Finally. I poured the fuel back into the tanks and buttoned everything back up. Stress levels down. A lemonade pitcher full of avgas is a scary thing and you can’t take your eyes off it for a second, because some fuel dribbling in the cockpit is one thing (on the ground at least). Half a gallon of avgas spilling across the hangar floor is another.
So the takeaway from this lesson? Check all the dependent systems before performing tests. Ensure proper installation of dependent equipment before running live fuel. That means pull the covers and check the fittings. You can bet I’m not going to be nearly as complacent when it comes time to check the ignition system.