How to Find Source of Fuel Smell?

[Piper J3]

I just did the Fuel Tank Attach Modification Service Bulletin 13-12-19. I accomplished the modification by carefully removing the tank lid to gain better access to the lower front edges of the tank. All went well and tank was resealed with ProSeal. Tank was filled with 20 gallons 93E10 to calibrate the fuel gauge. The airplane sat for a full week with completely full fuel tank and no fuel odor was noticed when canopy was raised. I even left the fuel shut off valve in the open position to static test the fuel lines up to and including the gascolator. Very good.

Now I have flown the plane three times for roughly an hour each flight and have 5 gallons remaining in the tank. On the first flight I noticed fuel smell occasionally on climb out and occasionally in cruise flight with a fair amount of turbulence. Second flight was a little less fuel smell and the third flight was still less but smell was occasionally still there. I suspect the leak to be very small, and of course, difficult to find.

I see no evidence of a leak at the tank. It might be carb float bowl overflow but I don?t see evidence of residue in the drip pans. I have searched VAF and found a thread for RV-12 Fuel Tank Leak that?s not much help.

My question is? has anyone used a dye solution specifically made for detecting fuel leaks? I found a UV dye additive for auto fuel that fluoresces under black light. Anybody try anything like this?

[KALEWIS]

Run a tank of 100LL .... it has blue dye in it

A few gallons of avgas is ok for the Rotax

[Mark Albery]

I'm not sure about the -12, but with the other RVs completely filling the tank can allow fuel to enter the vent tubes with their opening near the filler neck. Then when you fly the expanding air in the tank (not being used) pushes that fuel out of the fuselage vent outlet and you can often smell it in the cockpit.

Possibly the turbulence could allow fuel to splash down the vent in a similar manner.

When the tank is less full there is less likelihood of fuel getting into the vent.

Whether this applies for a single tank set up will depend if the air expands in the climb at a greater rate than fuel is being drawn from the tank.

[Piper J3]

I forgot to mention that tank is an early design with visual sight gauge on the side and vented fuel cap. I elected not to install the new vent line and instead stay with the vented fuel cap. The vent holes in the cap are fully open.

Also, as for carb bowl overflow possibility, during last condition inspection I weighed the floats and they are within spec. Both carbs were overhauled by Lockwood prior to my purchasing the plane.

Another data point - while doing the fuel tank modification the carbs sat for five months and most likely the fuel evaporated from the float bowls. I'm wondering if gaskets may have dried out.

[Pat Stewart]

Check the carb floats especially if you get this smell on climb out.

[John-G]

Jim-

Five months is a long time for auto fuel to sit in a carb bowl compared to aviation fuel. You just may want to pop off the carb bowls and have a good look for gunk.

If you look carefully at the flight data recorder you may see a small indication of what is going on ... I would think if you are smelling fuel and it is coming from the float bowls, there should also be a small change in RPM, EGT, fuel pressure and fuel flow at the time of the "event" when you are smelling the fuel. If you can see the "event" reflected in the flight data, it would indicate you are not getting the fuel smell from your repaired tank ... its coming from under the cowl from a carburetor(s) overflowing.

As another data point, you may want to place 30% or more 100LL in the tank and make a test flight. You just may be seeing the results of winter blend auto fuel being used in the spring when the temperatures are warming up.

Keep us posted as to what your findings are.

[DHeal]

The only time I get a whiff of fuel odor in my RV-12 cockpit is during climb-out with what was later to be determined were over-weight fuel bowl floats. You may or may not see any evidence of fuel residue in the carb overflow trays.

[Piper J3]

I checked my engine log book and carb floats were weighed 8 months ago as follows - 6.1g (R) 6.2g (L). So the floats are less than the 7.0 gram spec set by Rotax for SB-912-065UL. I guess I'll have to re-weigh the floats now that I'm having a problem.

[Dave12]

In almost 600 hrs. of 12 time, the only time I have smelled fuel was when the floats were heavy.k

[snopercod]

Amazon sells Combustible Gas Detectors.