What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Alternate fuel tank leak detection method

claycookiemonster

Well Known Member
Ok, after weeping and gnashing of teeth, and removing and reinstalling both baffles, and then adding a bit more proseal, both my tanks passed the soap bubble test. Good enough.
Or is it?
A friend offered me another way to test the tanks. It's the machine auto repair shops use to find leaks in your car's emission and Evaporation system. You know that "Check Engine" light? Often caused by a leak in the various emission manifolds. So, this machine pumps a relatively low pressure of smoky air into the emission system and then you shut off the incoming flow and it measures how quickly the pressure drops in the sealed system. The smoke allows you to see where the leak might be, and it's calibrated to reveal slowly dropping pressure! Genius!
Now, it's unlikely any of us will pay $1500 for a leak detection machine we'll only use once, but maybe you have a friend who runs a garage? It seriously took only about 10 minutes to re-test (remember, the soap bubble test was a go) both tanks. No smoke, and the pressure gauge was rock solid.
The unit I used (and I have no connection to the company at all) was from Redline Detection. Totally self contained and powered by 12v dc cables like jumper cables. Maybe they're rentable? It was a confidence booster to see the tanks pass a different test with much more precise instrumentation.
 
I suppose you could put a slight vacuum in the tank and watch it to see if it leaks. I have a vacuum pump that I have used to test my pitot/static lines as well as the fuel lines since I built my own hard lines. I would only caution to put very little vacuum so not to claps the tank skin.
But for me, the soap water will be good enough, just waiting for the proseal to cure fully.
 
My tanks passed the soap and water test twice. Yet they still had rivets on the back baffle that were seeping fuel. Not enough to drip, just blue staining. This all transpired during my first 40hrs. Removed the tanks after I finished phase 1 and just out of curiosity, I tested with the soap and water. They passed! No bubbles! I cut the holes in the rear baffle and resealed. No more leaks. I will submit that the soap and water test in not necessarily the end all be all test. At least not in my case.

Regards,
 
He

The best leak test will be a helium leak detection tester. Short of that it is all a guess until avgas goes in.
 
Two advantages of this auto shop smoke tester. 1. Car systems aren't designed for high pressures anymore than our tanks are, so they can only create a little over 1 psid maximum. The guage is calibrated in inches of water, but the real test is to pressurize the vessel (car system or fuel tank) and then very precisely monitor whether the pressure is holding or not. There is a tiny venturi on the machine and so the slightest of leaks is visible instantly.
Anyway, it was interesting to have this machine dropped into my lap, try it out, and to have it confirm the soap bubble test. Barely weeping rivets may indeed show up later, and I'll deal with them then. I think I've done about all I can practically do for now, and that will have to do.
 
Snoop is probably better than dish soap. It has glycerine and is very thin. A bottle from the local air-system supply house is around $5.
 
Smoke test

I've used the shade tree smoke test on my auto. A friend uses a Vape. I just blow cigar smoke into the vacuum line. Comes out where ever there's a leak. Works great. You have to close intentional openings like putting the engine at TDC and taping off the air intake.

Might work on a tank too if you can get enough smoke in the tank.
I would set up a manometer on the tank first. Then invite a bunch of friends for a cigar party. Everyone blow smoke in the tank till it won't take anymore then quickly close the cap and pump it up to 27". Might be hard to see the smoke but maybe a UV light would make it more visible.
 
The little hose puts out a surprising amount of smoke. And, the whole rig is set up to detect leaks in closed spaces. So, after pumping the tank full of smoke to your inches of water, then the incoming smoke is cut off, and the manometer continues to read pressure in the system. Even beyond that, it has a tiny flow gauge that reveals the tiniest of leaks and even quantifies the size of the leak, down to .25mm.
 
I tested mine with fuel (ethanol free Mogas) and UV dye before and after the backplate was installed. Some have mentioned the added risk of using fuel but it wasn't anything that i believed wasn't easily managed. There is no mistaking a leak and/or it's origin. It will light up like a disco.

Note: the "leak" indications in the attached are from the fuel wicking up the aluminum to open backplate rivet holes. Almost p@#$'d myself before realizing what had happened but it ended up being a good control indication for the test. I burned the fuel in my car so the entire test was less than $20.

Before anyone chimes in, I'm just giving the OP another option. It did get me a good amount of confidence regarding fuel tank leak integrity. Hard to beat the surface tension of gasoline for a fuel system leak check.
 

Attachments

  • Tank UV leak check.jpg
    Tank UV leak check.jpg
    291.1 KB · Views: 121
Being the OP, no problem. Having finally gotten through the soap bubbles, and now having passed the smoky auto detector test, I'm also going to put some gas in the tanks and see what happens. Not sure that AVgas is required, so it'll be something I can put in my car after it's done
 
Just remember

Gas being very volatile will not readily reveal a slow leak. The blue dye in Avgas will leave a stain in time. Mogas will leave a stain but it's not as readily noticeable and takes longer. There's lots of dyes available at auto parts stores or amazon. I used the AC Delco based upon it's history (been around for a long time). Just a thought. Good luck.
 
Gas being very volatile will not readily reveal a slow leak. The blue dye in Avgas will leave a stain in time. Mogas will leave a stain but it's not as readily noticeable and takes longer. There's lots of dyes available at auto parts stores or amazon. I used the AC Delco based upon it's history (been around for a long time). Just a thought. Good luck.

That's one of the "problems" of UL - no dye. I tried adding florescent dye but couldn't really see anything. But I could smell it. Took me weeks to find the leak, and amazingly it was a B nut on my fuel return line near the tank that was not seated right - just a bit of crud on the mating surface. Now, no smell at all.

My point is that if you put fuel in and really seal the tank, you should be able to smell leaks.
 
Back
Top