Fuel systems leak detection

I have a new boat I built. I tested the fuel tanks to 2 psi and both passed, then sealed them in under the sole..they can be cut out, if needed.

They are the older style; not diurnal tanks, but rotomolded Moeller tanks. 2x 64 gallons, one per hull in a catamaran.

The boat has 3" conduits running from the tank lockers to the engine bays and then to the helm station which is attached to the main cabin/salon.

I keep getting a hydrocarbon odor in my salon. The helm station has an enclosure fan (cpu style) that turns on when it gets above 95F in the enclosure and I get a fuel odor there as well. The active cpu fan is pulling air from the fuel tank area.

I have a very good nose and can smell a cigarette four cars ahead on the road and when they put it out; so part of me wonders if it is me picking up some low rate of fuel.

I tightened up all the tank connections, but not sure it'll matter.

So, I have a bunch of questions.

1. Can anyone recommend a meter for gasoline detection so I can determine whether there is a variation from hull to hull? And is there a 'safe' level of hydrocarbon odor?

2. Is there a standard in boats for stopping the flow of air from a fuel locker to other areas that I have botched? I ordered a couple innertubes and plan to seal off the wires and 3" conduits for hydraulic steer, etc, but this may not be enough.

3. The fuel tank lockers are not really vented, per se. There are a couple places where conduits exist for the fuel supply lines and wiring that exit the fuel tank lockers; should those holes be foamed in and would I then fume detect??

4. The fuel fill and vent lines are open to a rope locker on each hull. When I stick my nose in there; no notable odors are present, but the active cooling fan in the helm station would draw from there; across the tank locker and through the console rear, so it makes sense to not smell gas there.

Any advice on the tank lockers and airflow would be appreciated. I am also planning to run a humidistat and two enclosure fans to help with minor condensation in the area between the engine bay and the fuel tank lockers.

I did not expect odors at all, but perhaps some are expected and I need to manage them differently is all. Thanks.

The boat is an outboard and the rules on them are a bit looser as many of you will know. This does not help with odor mitigation.

 

All the basics, like two clamps per connection, retirquing of them and the sender mounts have been done, etc.


I think this is a design error..

 

For leaks there is dye you add to the fuel then look for it with an UV lamp. Kits are available from several manufacturers, specific to the medium (gasoline, diesel, water, etc.)

Your tanks and hoses are probably not perfectly gas tight, a sensible nose might detect the smell, altough this normally happens with old plastics not new ones.

Ideally the fuel tank lockers should be isolated and vented separately using an explosion proof fan. Appropriate blowers are made for gasoline powered inboard boats, fit one of these extraction fans to each compartment and run them on a timer several times a day, and continuously when the engines are running. This should ventilate the compartments well enough to remove smells.

Change the path for the helm ventilation fan, it should suck air from a different area. This fans are not spark proofed, if you really have a leak it could be a disaster.

To explode the fuel air ratio must be in a certain range, usually the human nose can detect gasoline before the mixture is right, but that's hypothetical because a human isn't always present, and gasoline vapors can accumulate over your absence. Standard procedure is to always asume danger is present when smelling gasoline.

Given your description of the fan sucking air from a locker in wich the fuel vent terminates, it's no wonder you can smell fuel. Even if the locker is properly vented at the bottom, your fan is creating a venturi effect with the tank vent. It literally sucks air mixed with gasoline vapors right out of the tank. You can't smell it because it only happens when the fan is running and the lid closed, the moment you open the lid even if the fan is running it probably isn't powerful enough to create the small underpressure needed to extract the fumes from the vent.

 

thank you; embedded replies

I am going to create a new intake source for the helm fan and use innertubes to close off the other pathways a bit.

 

Did you use Marine grade hose on the fill, draw and vent lines?
Automotive hose has a higher permeability allowance ABYC states "permeation of hoses within the limits of SAE J1527" Perhaps there is some permeation occurring

H-2 24.5.2 The entire system shall be liquid and vapor tight to the hull interior spaces

Permeation of Fuel Tanks within limits of CFR33 SupartK, 183.620 I would expect that Moeller tanks should meet these specs

 

well, I botched making the spaces vapor tight to the cabin...sort of

How does one run fuel sender data to a helm station then? I suppose foaming off the wire and fuel chases...but the boat was originally to be demountable and I wanted access to wire chases, and the helm access panels I figured were good enough to meet vapor tight...but they were off for awhile this year...

everything is marine grade

I'm nervous there is a pinhole leak...or such...I did not test the lines..just the tank with terminations

 

Does anyone know if there is an acceptable limit of gasoline vapor? So, if I buy a meter, I can probe the locker to determine ppm?

 

Testing for leaks is easy. Plug all the hoses but one, install a pressure gauge and pressurize the tank. Wait a few hours and check if the pressure goes down.

 

Never had an old moeller that didn't stink a might on a hot day, always assumed they were somewhat semi vapor permeable. Installed 4 with that in mind, all enclosed away from liveable space.

 

How long did you test it for?
Fuel tanks should best held at the pressure, you require, for a minimum of 12 hours.

 

The tests were done for 24 hours. I only had one leak and that was on the test gauge; so one tank was tested twice.

The boat did sit for about two years after the tests, but before launching. So, I did go back and retorque all connections are the tank.

 

Would you remove the fuel for such a test?

I would not call it easy. The fill hose would be a pita, unless I could test it with the cap on. It is all bedded in, but I used butyl; so can be done easier than other sealants.

I tested the original tests at 2 psi, but a test with fuel I would probably only go to maybe 1/2 psi. There is no reason to puff the tank way up on a recheck.

Let me know on the if filled. This might really be the way to go. And if I do some isolation; it'd be the quickest way to leak detect, but I'd need to be careful to avoid any spark from the air system...

 

I smell gas vapors in both tank lockers; and the tanks are both vented; so I don't understand it.

My problem is likely me and the Moeller tsnk combined with the vapor leaking into the cabin a bit via chases. I can smell a covey of grouse in the field before I see em.

I just need a good way to block off the chases and maybe a ppm detector to get an idea of the amount. I'm going to try innertubes, and then create a new pathway for the helm fan. The helm is angled up to about a 2.5 inch dimension from about 12" and so I installed a fan with temp controller that turns the fan on and I smell fuel vapors outputting there. I need to change the direction/source of the ventilation system and add a couple small bilge blowers.

 

I read online where a guy had a sewer gas leak enter his home via conduit and was instructed to use duct seal putty; so I could try that, but I also need to revise the negative pressure issue.

Thanks for all the comments.

I still want to buy a meter or gas sniffer for the hulls.

 

I bought this meter to help quantify the issue.

Chaoos Gas Leak Detector, Combustible Gas Detector 11-In Gooseneck, Natural Gas Detector 0.1%LEL/1PPM Resolution, Locate Sources Of Methane, Ethane, Propane, Butane, LNG, LPG And Combustible Gas Leaks https://a.co/d/drO3vxa