I took the opportunity last Sunday to enjoy the sunny weather from aloft, and went flying in Norbert the Champ.
I put the dipstick in the tanks, and decided that I had about 12 gallons of fuel. Plenty for the ~1 hour flight I had in mind, and it would put me in reasonable territory to drain the tanks and re-calibrate the dipstick, which had seemed a little off when I was trying to work out the math between fill-ups and gallons per hour on recent trips.
The flight went as planned, and I set up in the hangar to drain the fuel from the tanks, eventually (holy buckets does it take a long time to drain a fuel tank by the sump drain) getting about 4 gallons out of the two tanks. Given that I had just flown 1.4 hours at ~4 gph, that already suggested I was on the right track: I should have had about 6 gallons left, not 4.
The tools I had at my disposal were two 6-gallon jugs, and a "2 gallon" jug, which I decided would be my measuring container. It turned out, half way through, that it's actually a "2 gallons and 8 ounces" jug, intended for mixing 2-stroke oil with gasoline, so my measurements were on the imprecise side. I also had the magic hydrophobic funnel, which allows me to do all this fuel pouring without transferring water or other crud.
The fuel system on the Champ is very simple: two 13 gallon tanks reside in the wings, and are connected together above the fuel shutoff valve. This means that they (very slowly) cross-feed at all times, so any fuel additions would have to be to both tanks before taking a reading, to ensure I wasn't seeing one tank slowly creep up the level of the other.
The plan was this (and it worked reasonably well): go to the pumps and fill the two 6-gallon jugs with exactly 10 gallons of fuel, 5 per jug. Fill the 2-gallon jug twice, and fill each tank on the plane with 2 gallons of fuel. Dip the tanks, and mark where the fuel hit, having previously erased the old marks. Lather rinse repeat until the 10 gallons are in the tanks, then go fill the two 6-gallon jugs again and repeat the whole process.
Since I started this exercise with about four gallons of fuel, I first poured this amount into the tanks, and got my first surprise. The right tank registered about an inch of fuel on the stick. 2 gallons, marked and done. The left tank left the stick completely dry. No 2 gallon mark. (The landing gear on the left is always extended slightly further than the gear on the right, so that the left wing is several inches further from the ground at rest.)
Keep going: 4 gallon marks for each tank -- one side of the dipstick is marked RT, and one is marked LFT, but previously, the levels had been exactly the same on each side; don't ask me why there were two sides before. Interestingly, the 4 gallon mark for the left side was only about 4mm below the 4 gallon mark for the right side, suggesting that at 2 gallons, the left tank was just barely shy of hitting the stick.
So I kept going, marking 2, 4, and 6 gallons, then marking 7 when I split the final 2-gallon jug between the two tanks. I went back to the pump, and refilled with another 10 gallons, figuring I was already half-way there, so I might as well finish the job. Marked 9, then 11 gallons, although the right tank overfilled and spilled down the wing by some amount, so the final count was not terribly accurate. Presumably, the tilted wings mean that it would be possible to fill the right tank so far that it comes out the vent, when filling the left tank to the very top.
In all, I put in around 24 gallons of fuel (exactly 20 from the pumps, and about 4 that had already been there, minus the overfill), and completely filled the right tank, while the left tank got near the filler neck but didn't quite touch it. The dipstick reads about a quarter inch difference between the right and left tanks, and had been reading almost two gallons too full on the right tank, and about one gallon too full on the left.
That means I had previously thought I had half an hour more fuel than I actually did. That's a sobering error, and one that could have bitten me badly. It's a more than 10% error.
Since the 2 gallon jug was actually more than 2 gallons, and I didn't fill it to exactly the same amount each time, I'm still only going to trust the dipstick as an approximation of the fuel left in the tanks. It's a more pessimistic approximation now, though, and that suits me just fine.
Until I can redo the job with a properly calibrated filling rig, I'll live with the knowledge that I may have a little bit more fuel than the dipstick says I do, but do the math as if I don't. It's always a better surprise to find out you have more fuel than you think you do, rather than the other way around.