Perkins loses prime overnight

[Lazerbrains]

Pulling my hair out working on my Perkins 4.108.

Fairly new (to me) boat has 4.108 which started developing air in the lines. I can bleed the system and it will run fine all day, but after sitting overnight will need to be rebled again. Symptom is it starts and runs for a few seconds then dies - after bleed procedure runs fine all day. Tank is below engine. Fuel goes from pickup line to Racor, from Racor to electric fuel pump, then to lift pump, then to secondary filter, then to injector pump.

A few questions:

A) I rebuilt Racor, and it has a pressure gauge installed where the t-handle should go. Shows pressure when running and after shutting down, but next day shows no pressure - is this normal with these gauges, or should it hold the pressure here? (I also am a bit suspicious that pressure gauge is cause of air in system)

B) Could a leak in the injector pump cause this, or should I be limiting my search to the vacuum side (pre lift pump)?

Thanks so much for any advice - at my wits end trying to figure this one out. Motor runs fantastic once bled, but I cannot go on bleeding it every time I want to use it.

[SVQuestOfPaget]

I've done probably the same amount of bleeding fuel lines as most cruisers, but am no expert. If you're losing prime overnight, then I'd check every fuel fitting for cleanliness and tightness. The fact that it runs briefly suggests to me that you're getting fuel at least to the injector pump.

When you bleed the system, does it take some time to get fuel to the engine?

I think you might be correct in your suspicions regarding the Racor, and if all is tightened at the engine, there's few places air can get in there. I'd focus on the lines pre-lift pump. Let us know what you find!
-Eric

[Cheechako]

You have an air leak somewhere.
Also, I would put the electric pump between the tank and the Racor. Right now it's inviting air into the system via any micro leak as it sucks through the Racor ...right?
Your vacuum gauge shows pressure? It should show vacuum... They are intended to show the filter condition by how much vacuum there is right?

[Mike OReilly]

I feel your pain. You've got a leak somewhere. Very hard to find small ones. We chased air and fuel leaks for a whole season a few years ago. The last one confounded me for weeks. Turned out to be a fitting on the engine-mounted (secondary) fuel filter. Found this only after changing virtually every olive connection, bronze gasket and a few high pressure lines.

[acem]

You have an air leak. The electric fuel pump may patch the problem for a while but it could cause a fuel leak while running. It's easier to find a fuel leak than a vacuum leak but leaking fuel is flammable.

Check your fuel lines and all connections. Check the line in the tank for cracks. The injection pump can leak. Lift pumps can also leak.

Rubber fuel lines get hard or soft with age and need replaced. Check them closely.

Thx-Ace

[Lazerbrains]

Originally Posted by SVQuestOfPaget

I've done probably the same amount of bleeding fuel lines as most cruisers, but am no expert. If you're losing prime overnight, then I'd check every fuel fitting for cleanliness and tightness. The fact that it runs briefly suggests to me that you're getting fuel at least to the injector pump.

That is what I did yesterday - will check this morning to see if it worked. Thanks!

Originally Posted by Cheechako

You have an air leak somewhere.
Also, I would put the electric pump between the tank and the Racor. Right now it's inviting air into the system via any micro leak as it sucks through the Racor ...right?
Your vacuum gauge shows pressure? It should show vacuum... They are intended to show the filter condition by how much vacuum there is right?

Yes, I have considered moving the electric pump to where you suggest so that it pushes instead of pulls through the Racor - that is always how I have had previous boats setup. Previous owner had installed a fuel polishing loop there (controlled by a valve) which is part of the reason for the location. I already removed the fuel polishing loop hoping it was the cause of the air in system. Funny thing is I thought I had found the problem as before I removed the fuel polishing loop the fuel level in the Racor would go down to the tube overnight, after removing it the fuel stayed full in Racor, but still same problem of needing to be bled to start. Yesterday I simply double checked all connections again. If no start today I will likely remove the fuel pump out of the system to isolate it.

[Lazerbrains]

Originally Posted by acem

Check your fuel lines and all connections. Check the line in the tank for cracks. The injection pump can leak. Lift pumps can also leak.

Here is the question I am grappling with - could a leak somewhere in the injection pump cause the overnight symptoms I am describing? I would think even if it had a leak there it would not cause this situation, but I am second guessing myself at this point.

[Lazerbrains]

Also to be clear, engine will start and restart for hours after being bled. It takes overnight for it to develop air in system and does not do so at all while running (tested by running motor for 12 hours straight).

[wsmurdoch]

The electric fuel pump is usually after the initial filter (Racor). From the Walbro electric pump installation instructions, "A fuel filter must be installed between fuel tank and pump...". The filter keeps junk in the fuel from getting to the small check valves in the pump or (in some pumps) to the small protective screen in the pump inlet.

I found a leak in my Racor filter by temporarily installing a length of clear vinyl hose in place of the fuel hose after the filter. I could see the small air bubbles coming down the line when the engine was running. The hose, which was above the Racor filter, slowly filled with air when the engine was shut down. The leak was between the metal case and the plastic bowl. The white paint on the metal case had separated from the metal, and the air was leaking in between the paint and the metal when the Racor was under vacuum. No fuel leaked out when the engine was off.

You can also close the valve at the fuel tank with the engine running. The engine will eventually stop. Read the vacuum gauge to see the highest vacuum (dirty filter) that will still let the engine run at that rpm and load. If the vacuum holds, you have no air leak. If the vacuum falls off, you have an air leak.

I have also had a small leak on the high pressure side of my injection pump at one of the three discharge valves. It leaked fuel out when the engine was running and air in when the engine was shut down.

Bill

[acem]

A leaking injector pump, depending on locatio,can allow the air into the system and to loose prime.

Ideally the pump should have a filter to keep trash from damaging it. However a screen will suffice. On larger mechanical diesel engines there is normally a coarse filter or screen between the tank and lift pump. Then fine filtration between the pump and injection pump.

If you shut off the fuel at the tank does it loose prime?

Thx-Ace

[Cheechako]

Originally Posted by wsmurdoch

The electric fuel pump is usually after the initial filter (Racor). From the Walbro electric pump installation instructions, "A fuel filter must be installed between fuel tank and pump...". The filter keeps junk in the fuel from getting to the small check valves in the pump or (in some pumps) to the small protective screen in the pump inlet.

I found a leak in my Racor filter by temporarily installing a length of clear vinyl hose in place of the fuel hose after the filter. I could see the small air bubbles coming down the line when the engine was running. The hose, which was above the Racor filter, slowly filled with air when the engine was shut down. The leak was between the metal case and the plastic bowl. The white paint on the metal case had separated from the metal, and the air was leaking in between the paint and the metal when the Racor was under vacuum. No fuel leaked out when the engine was off.

You can also close the valve at the fuel tank with the engine running. The engine will eventually stop. Read the vacuum gauge to see the highest vacuum (dirty filter) that will still let the engine run at that rpm and load. If the vacuum holds, you have no air leak. If the vacuum falls off, you have an air leak.

I have also had a small leak on the high pressure side of my injection pump at one of the three discharge valves. It leaked fuel out when the engine was running and air in when the engine was shut down.

Bill

Yeah, they like to keep their pumps clean, in their own interest, but that may not always be in your best interest if you keep clean fuel. Those little pumps, well really all pumps, push better than they suck IME !

[jpendoley]

If the electric fuel pump does not have a check valve it can lose prime. If the fuel pump was recently replaced I would check the pump specs.

[Lazerbrains]

Originally Posted by jpendoley

If the electric fuel pump does not have a check valve it can lose prime. If the fuel pump was recently replaced I would check the pump specs.

That's next on my checklist. Today's attempt was to completely remove the Racor from the fuel path. Will have to wait till tomorrow to see if that does it. If not, the next step will be to remove the electric fuel pump from the path.

BTW - thanks for all the suggestions and advice so far. Keep them coming as I have yet to isolate the problem yet.

[jimbunyard]

It's crude, but one could always just buy a cheap 1/4" ball valve and two appropriately-sized hose barbs, and insert it (consecutively and individually) before every appliance downstream from where the engine rubber hose begins, and shut the valve off every night.

If the negine starts and runs the next next morning, move it downstream to the next position till it doesn't, and you've found your problem 'appliance'.

For most small boats it'd probably be easier just to redo the whole system (though more expensive).

I'm betting on the Racor, but have you made sure you have clear, unobstructed flow from the tank, including the vent and the anti-siphon tank valve (if fitted)?

[Lazerbrains]

Still troubleshooting this problem - takes a long time as I have to re-bleed the Perkins and wait overnight every time I make a change. At this point I have isolated each component on the vacuum side of the pump and am satisfied that all is good on that side, which leaves me to the pressure side. The problem is there is no visible leakage after sitting overnight. Is it possible that it is leaking back into the return line and down into the tank? My tank is lower than the engine, FWIW. When engine is running there is evidence of a small leak in the injector pump - about one drop per minute or two. My theory is that when shut off, air comes in at the same spot and it leaks down the return into the tank. Does this make sense? I do have a small cutoff valve on order to put in the return line which should test this theory. Again, the symptom is that after sitting overnight, when started the motor runs for 5 secs and then dies.