What's new
Van's Air Force

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

What defines a 'sealed' tank

Dave S

Active Member
After some difficulty and several positive soapy water tests, I thought I had my tank sealed.

I have pressurized the tank to less than 1 psi using a homemade water manometer, but the air slowly leaks out. I have tried several times but I am unable to find any bubbles indicating a leak.

Any hints to find where I'm leaking air, or do I consider this a sealed tank?

I pressurized it and when I came back 3 days later the air had escaped.

Thanks
 
leaks

Was the temperature constant?

Also, there is a product that HVAC guys use to check for leaks that works FAR better than soap and water...

Camco 10324 Gas Leak Detector with Sprayer - 8 oz
 
The tank is sitting in my shop which is also where the utilities are (furnace and dehumidifier). We do have our programmable thermostat set to drop about 8-10 degrees during the day.

Even though with the temperature changing as long as the temperature inside and outside the tank are the same shouldn't the differential pressure remain the same?

I didn't check ambient pressure but even as that changes I don't see it getting to the point where the pressure equalizes with the ambient pressure unless there is a leak.

The leak detector, I assume it gets sprayed on just like the soapy water just that you don't have the 'foam' bubble as it comes out like soapy water does from a sprayer.
 
Sealing up a tank 100%, for the purpose of doing the leak testing is very difficult to do. Fortunately it is not important either.

It is common to have very small leaks at locations like the fuel cap, base of the balloon, etc., that will allow air to slowly leak out (even if the pressure is only 1-2 PSI).
As long as the testing method can show you where leaks are located, this doesn't matter. That is why using strictly measured pressure (such as a manometer)is unreliable. With the additional issues of pressure changing with ambient temp, and baro, it is actually possible for you to have an unwanted leak, but have these variables changing in such a way that the gauge reading doesn't change.

It is for all of these reasons that the instructions supplied with the testing kit sold by Van's recommends (at least I am pretty sure it does... haven't had a reason to look at them for a while) that the leak testing be done with a soapy water solution.
This way, if you see a small leak around the perimeter of the fuel cap or the base of the balloon you can ignore it and continue looking. The important thing is to not see indication of a leak at any rivet or seam.

Another good reason to do the testing this way is that even if the pressure reading method accurately tells you there is a leak (entirely possible in your case but not a certainty), you then have to use another method to determine where the leak is anyway. So why not start with a method that is more reliable in the first place.

It seems over the years that the recommendation of using a balloon to evaluate pressure has been mistakenly transposed into a belief that it is to monitor for a change in pressure. The only purpose of the balloon recommendation was for it to act as a visual indicator of how high the pressure is, and to act as a frangible member of sorts so that as you add air, the interior pressure doesn't increase at nearly as high of a rate as it would without it (because the balloon can keep expanding).

My recommendation always is to pressure the tank and then take it to your tub or shower enclosure, stand it on end, and start spraying it down liberally with a soap / water mixture. 1 / 10 soap to water seems to work pretty well if you use something that will bubble easily like dawn dish washing liquid. After really spraying it down well, flip it end for end and do it again to assure you don't miss something. When you are done, spray it down well to rinse it off (if you use a water temp colder than the ambient temp, you will also get a good lesson on how influential the temp can be to just 1 PSI pressure. The balloon will go flat almost instantly).
 
Last edited:
Camco

I found the Camco product to give more consistent results than soap water; you definitely know when there is a leak...
 
Manometer

I used the manometer method as well.
1 PSI is about 28". Every degree change in ambient temp is .8". Mine held pressure fir several weeks. I may test them again before installing the wings.
 
About 24 hours ago I pressurized the tank so that the water levels in the 2 tubes had a difference of 22inches. Right now the water level is down to about an inch difference. I didn't measure the room temperature, but it should be about the same.

I have ordered Camco leak detector, we will see what that shows.
 
Temp change

About 24 hours ago I pressurized the tank so that the water levels in the 2 tubes had a difference of 22inches. Right now the water level is down to about an inch difference. I didn't measure the room temperature, but it should be about the same.

I have ordered Camco leak detector, we will see what that shows.

1 degree will change the level .4" each side or .8" total. Assuming the temp went down. If it went up, that would be the wrong direction. :eek:
 
Back
Top