Chasing air in fuel line two years

IF only the resistance of the filter is needed a DP, differential pressure gauge can be dash mounted.


It picks up info just before and just after the filter .OTS

I recently watched a sail video of a guy in Tonga who spent a long time trying to diagnose a chronic air leak jn his fuel system. He must have bled the system 20 times.

In the end, he figured out his secondary filter (on-engine one) had been mislabeled by a previous owner. It was slightly different.

Reminiscent of a problem I had on an old BMW motorcycle. I was many miles into the Henry Mountains in Southern Utah in dirt roads. I had an after market oil cooler with a housing that attached at the lower side of the engine replacing the triangular shaped cover where the filter was. I hit a rock just right and broke housing. My buddy towed me to the nearest town (Hanksville for those who care). I came down the following weekend with the original cover. It leaked. A lot. I tried all sorts of gasket cement, and could not get it to stop leaking. Turns out the 10mm bolts that came with the oil cooler housing were just a tad too long and bottomed out preventing the cover from being fully tightened.

Peter

Air can be entering from return. With whole injection system above day tank, fuel can slowly drain back with air coming from return path. It sits under a slight vacuum. If that is the problem, return to tank can be set up with dip tubes that go below fuel level instead of dribbling at top of tank.

Ski in NC said:

If that is the problem, return to tank can be set up with dip tubes that go below fuel level instead of dribbling at top of tank.

Click to expand...

I have that setup on my generator. Made for a real head scratcher when changing the secondary filter after the lift pump. Where the hell is this fuel coming from with the Racor bowl empty and the fuel valve closed?

Ted

Do you have a shutoff at the tank or day tank. If you do hook up a temporary vacuum pump at the engine interface and then pull a vacuum and watch the vacuum gage. You can easily pull 15 to 20 inches of vacuum. It the line is tight the gage should remain steady, if not you have a vacuum leak in the line. Some ball valves are rated for vacuum operation, most are not so check you system. The vacuum gages can be purchased at most auto parts stores. It gives you a way to check, in minutes or a few hours not waiting weeks to find success. Old copper washer can be reused if you anneal them with a propane torch and flatten on a sharping stone. New ones are the best route but the copper needs to be annealed to work. Copper work hardens when clamped in a fitting.
Good luck chasing the leak.

Hi Nick, the vacuum gauge is intended to show a restriction in the filter media. It would also measure any restrictions back to the tank. If the day tank is the lowest point in the system, it should always be full, therefore there shouldn't be any siphon effect, perhaps what you are describing. The day tank should have a float mechanism to shut off the fuel when the tank is full otherwise it might overflow. If it's shut off, no vacuum.

If the fuel tank has its outlet at the bottom and your shutoff valves are leaking in air, you have a restriction in the tank, probably a screen that's blocked. With the head pressure of a tank, it should flow very well. If it flows uphill such that it needs a pump, the vacuum should be negligible unless there is a restriction. Once the pump stops, so should the vacuum. Most of us just have an engine driven pump, ymmv. My conclusion is that if you are showing a vacuum with the engine stopped, you have an issue.

What am I missing?

By the way, the ball valve in the Racor is intended to shut off the fuel flow if the bowl fills with water. I am looking at the ball, sitting on my window sill, that I forgot to reinstall last time I changed the filter in my 1000s, over 3 years ago. Oops. I can assure you that it does not compromise the filter at all unless you have water...

Many tanks have screens. You need to open the tank and remove them. Get a cable camera (lapiroscope?) and see if you have them. There is no advantage to screens, only grief! That's why we all have comprehensive filtering but a screen is not serviceable.

Hi Xsbank - I am assuming that the liquid surface in the "day tank" is below the height of the vacuum gage. Perhaps the OP can confirm this.

Regards,
Nick

Hi Nick, I still see no way there can be a vacuum unless there is some motive flow. The fuel system is open to the tank, even with the engine running at cruise, there would be no vacuum unless there is a restriction; if not a filter or a blocked tank screen, or the tank valve is closed. With a vacuum indicated there is a risk of air being introduced, which circles back around to the air leak, a poor mechanical connection somewhere.

Even if the day tank is below there would have to be a restriction to measure vacuum and at the same time the fuel would be trying to flow to the tank? Then it would flow until the tank was full then equalize, no vacuum.

It's difficult to understand all this I'm finding as the subject boat is somewhere else and it has an unique system but the principle is the same, restrictions anywhere in a fuel system are a fault.

Bloody snow.

Since you have no fuel leaks reported when running and pressurized the question becomes where is the air coming from?. The most obvious source is the fuel return. I suspect a bad check valve is allowing fuel to drain out of the IP side of the system back to the tank and drawing air from the tank.

Dear Bloody Snow,
I don't know how I can explain it better - difficult without being in person. The day tank is lower than the engine (and lower than the point in the system where the vacuum gage is located). The engine has to suck the fuel up against this height difference. Even when the flow stops the suction holds - the fuel is trying to run back down into the tank.

To further discuss other people's suggestions - the air has to be able to enter the system somewhere. Possibilities are:
1 return line to the tank - would require reverse leakage of the check valves in both the injector pump and in the lift pump (a total of at least 3 checks). Seems unlikely, but is possible. In the industry I work in we are told never to rely on a check valve totally sealing.

2 a "vacuum" leak between the day tank and the injector pump. This is interesting because from the day tank to the lift pump is always under a (slight) vacuum when the engine is running. This means that the engine would be pulling in air with its fuel. Perhaps this tiny amount of air is able to pass through the engine during running but the engine cannot handle the arrival of the accumulation of air which builds up during long periods when stopped.

3 a "one-way" leak between the lift pump and the injector pump, which only allows air in but not fuel out.

I think that option 2 is the most likely. I have personally seen air leaking into a filter set-up when subjected to a mild vacuum. If the leak is indeed between the day tank and the lift pump it could even be a bidirectional leak and no fuel dripping is seen because this system is always under a slight vacuum. You could try putting pressure on this part of the system and look for fuel leaks. It would be nice to pull a vacuum on the fuel line and watch the vacuum gage, but this would probably take a very long time to show anything. The transparent tubing suggestion sounds good as well but, once again, it would probably take a long time to see. Perhaps put the clear tube between the filter and the lift pump and watch for a slug of air at start up. If you see the air, the problem is between the day tank and the filter.

Nick (suffering from the sun)

I should have posted this earlier, but attached is a pic of my problem fuel piping. The day tank is at the very bottom of the pic and you can see various lines and ball valves going in and out. The engine filters and engine pump is about the same height as the Racors. All in all this seems like a pretty standard arrangement.

The Racor pressure gauges are about 3' +- above the fuel level in the day tank. My math says that with the engines off and feed valves open there should be about a 2 inch vacuum at the gauges (the fuel in piping wants to return to tank creating vacuum). Indeed, that is what the gauges are showing in the picture.

During normal operation the added vacuum from the clean filters add another 1-2 inches of vacuum, which seems normal.

Bruce. If that manifold with 6 outlets plus drain is involved then a search for a recent thread is warranted as in the end if I recall correctly the manifold was the source of air.

Its very simple. If the pressure in the line is lower than the outside ambient (14.7 at sea level) air has the potential to leak into the line. If the fuel tank is below the engine, or there are restrictions in the line the engine fuel pump creates a suction which lowers the pressure line pressure further. A column of fuel of about 30 inches creates about 1 psi pressure at the bottom of the column. To lift a column of fuel above the free surface of the fuel tank takes about 1 psi to lift 30 inches. Flow velocity or screen restrictions will add further to lower the pressure. Without flow the column of is at zero velocity, but there is still a vacuum, so a very small air leak can cause the line to drain back to the tank, which makes starting difficult if you do not have fuel in the line. The other problem over time air will come out of solution in the fuel if there significant vacuum. (pop bottle effect) It is difficult to find these small leaks, every thing in the fuel line is a potential source and requires carefull review of the installation and review the condition of the components in the line.

Thanks Dennis and Soo Valley - very much in line with what I was trying to explain - in fact, explained better!

As regards a workaround - There should be a valve right at the outlet from the day tank. Closing this valve whenever the engine is stopped should prevent the drain-back problem.

Nick