Pressure testing integral fiberglass diesel tanks

It is mentioned in a posting from 2006 that fuel tanks should be tested at 3psi. This is the number used in the ABYC standards for metal tanks but the document is mute when it comes to fiberglass tanks.

I have used a high precision pressure gauge (multi-rotations, like an altimeter, one rotation equals 2psi, reads to the 1/100 of a PSI) and have pressurized the 95 gallon integral fiberglass diesel tank in my sailboat to 1.5PSI where it starts creeking and is well domed. I am afraid to go much higher in pressure lest my tombstone reads "killed by fiberglass shards in a boat construction accident..."

Does anyone know of any authoritative recommendations for pressure testing fiberglass integral fuel tanks?

Thanks!

 

A pressure test is always the same setup on tanks, no matter which material.

And I assume you are using water.

Regards
Richard

 

No, I am testing with air so I can use soap solution to see any leaks.

 

Do´nt do that!!!

Water is the common medium and leaks would become impressively obvious.

But it seems you are already at the end of the range (pressure) the tanks can stand.
Probably a reinforcement will be the next step.

 

If you are reading your gauge correctly, your tank is too weak to test at three PSI. If you want it to pass ABYS, it will need reinforcement. I am concerned that you may have damaged the tank. Is there a sticker on the tank that says "This tank has been tested under 33 CFR 183.580" ? If not, who built it? I cannot recommend that you leave the test at 1.5 PSI, but why are you testing, anyway? It's not a gasoline tank, is it? No integral tanks are allowed with gasoline. Generally, an integral tank will be of the same thickness as the hull. Does it seem that way? Typically, a hand pump is used so that an accidental overload is less likely to happen.

 

Right Mark, a serious accident cannot happen on a pressure test with water and a simple handpump, as it is common. With any gas as medium (like air) that of course is not valid! But that is a insane way to test a tank!

Regards
Richard

 

Can any of your folks point to a definitive document regarding tank testing? (internet opinions are not what I am seeking).

This is a new tank that I fabricated using standards for thickness and materials that I found published and tied back to authoritative sources. I am testing it so I can label it appropriately, per ABYC standards. Again, ABYC says nothing about fiberglass tanks. (I have the standards book).

 

There have been no "internet opinons".

ABS might be the better source for standards? Have no access to any (GL, DNV or Lloyds, I am sailing at present), and do´nt have ABYC and ABS anyway.

And sure air is not the way to perform a tank test! You mentioned the tombstone already!

 

Air is suggested by ASTM documents for testing new tanks, including specifically fuel storage tanks. I agree that one does not want to test a tank with fuel in it via air pressure testing. The government of Australia publishes a document regarding pressure testing marine fuel tanks with air pressure: http://www.dtei.sa.gov.au/marine/survey/media/pdf/Pressure_Testing_of_Fuel_Tanks.pdf
So, yup, sounds like "internet opinions" and not reference documents to me.

 

Okay to test for leaks with air and soapy water, you dink. If you test for fail capacity with just air, you are as dumb as you are ungrateful. 3 PSI - go ahead and take it up to 3 bars to be sure - I don't care. I repeat, CFR § 183.514(b)(5). I'm NOT going to wipe for you, as well -here; http://www.uscgboating.org/assets/pdf/downloads/FUELSYSTEM.pdf . Now, get outa here.

 

there is only one way survey societys use and I have used it for years, the manometer test
I posted a dwg years ago
basically you have a clear tube in a loop, the tube needs be 8 ft high above the bottom of the loop, you half fill with water, , you have the tube attached to tank top with a valve at the outlet, you have air into the tank also with valve
you admit air until the water in the tube rises to top but do not let it pass the bottom of loop our it,ll blow out
when the water is near top shut off air to tank, and wait , if that water drops you will have a leak
the pressure is equivelent to 8 ft head or around 2 psi

I leave it there overnight, the level will drop when the ambient temp cools, and rise in heat of day But this will show even the tiniest leak, which can be found with soapy water
this is the only test I am aware of that they make on liferafts, if you need drawing let me know, there is one here somewhere I did years back
you should know that if you have PIN PRICK in a 200 gal tank with 2-3 psi in it, there will still be pressure there in 48 hrs, so you may think you have no leak, but with the above test any pin prick will show a drop in water level

http://www.boatdesign.net/forums/boat-building/tank-testing-22057.html

 

Tank testing, as boiler tests, are made safe ONLY by filling them with water and applying the required (very low) force.

Liquids are non compressible, therefore you need only low forces to do the test. And a leak only drips a few drops where a gas filled tank would burst completely! Air test is Insane!

A gas in a tested tank is under high pressure and is a danger when the tank bursts.

Simple physics, not "internet opinions" .............bigmouth.

And water tests are the the only accepted method by all classification societies for ages.

Regards
Richard

 

Any testing involving compressed gases is hazardous because the compression builds up a source of energy that will explode on release. Using fluids is much safer as the pressure will dissipate quickly if a failure occurs, although there is still the danger that trapped air could still be present.

I was told by my old boss many years ago that air will show a leak at one seventh of the pressure that water will. I assumed he maintained this because of the size of the molecules or something like that.

By testing with air you are not putting fluid in the tank as it would contain in its normal state, so your test is not reflective of the in-service state of the tank.

If you need to proceed with an air test, 3 psi is equivalent to a head of 7 ft of water so to find leaks you could reduce the head to 1 ft of water in a manometer. You can set one up with a clear plastic hose.
The other thing you need to do is set up your air supply through a double block and bleed valve arrangement. This means that you would have two valves in line on your air supply with a teed-off bleed valve between them. When you get to the desired 1 ft on your manometer, lock shut the two valves and lock open the bleed. If you can't lock the valves you must disconnect the air supply, you must have full control of the pressure source.

From what you say, either your tank is too weak and needs reinforcing as Apex 1 has already said or else you are not reading your complicated pressure gauge correctly. You could test your gauge reading against a manometer (U-tube) reading, 2 ft on your manometer should be 0.87 psi on your gauge.

Unless you are familier with testing procedures, I would advise you to stick with water for your testing,

 

Mac when I was building under lloyds this was the only one we used
naturally you can not fill an alloy tank with water as if you do have to fix a leak you can not weld it Naturally the tube in OPEN, so that anything over 3ps blows out naturally it is safe and naturally you se t you pressure IN reg to such a very lo pressure
I did see a tank once that was left to fill, the guy forgot to turn off the ait, it blew at 120psi , was luckily only a skin cooler for oil
testing with liquid, , unless the tank was deep then you would have no pressure, and none at top, then in anything apart from steel you would have to dry the thing each time Also water will not leak through somethng that diesel will So you may think your take is tight but it is not
We are TALKING STORAGE TANKS HERE, WE ARE NOT TALKING PRESSURE VESSELS SUCH AS WATER

 

Can anyone explain the difference between 1 psi of water pressure versus 1 psi of air pressure?