Yanmar 3jh3e oil level high advice?

[Captain Bill]

Originally Posted by Sailingallday

I am no expert at all. Just some dude who sails and tries to fix things on his own. Buuut - a different way of looking at it

Coolant gone/low. Oil high/overflowing. Why do these two facts not have you looking more intently at the coolant side of things leaking into the oil? Yes, I accept the pressure test showed no loss of pressure, but all the tests on the fuel side of things have also been clear. To me, the high temp alarm, loss of fluid and gain of oil would have me much more closely examining the potential of coolant into the oil

In the morning I will add a 50/50 coolant mixture to a about a quart of the contaminated oil in a glass jar and see if it stays mixed or separates. I believe it will separate, but it won't cost me much to test it. I know Rotella is supposed to have some good detergent in it, but I don't see how it could mix with equal parts coolant and not look like a chocolate milkshake.

[skipperpete]

Well spotted, its a supply line and the oil level needs to be higher than the cam but I guess yanmar doesn’t trust the bearing to maintain that level so they provided a seal to stop outward flow through the bearing but put 3 drain holes above the camshaft and into the timing case so that the front bearing is fully submerged
If it transpires that the coolant has mixed happily with the lube oil and not formed an emulsified mess with water droplets on the inside of the tappet cover…… that will be yet another “ today I learned” moment. I have however seen a lot of clean green coolant come out via the sump plug, usually from liner oring leaks, blown head gaskets and leaky gear driven circulating pumps.

[Captain Bill]

I was discussing this with a friend and he consulted Chatgpt. It said that on rare occasion an injector will get stuck open while the engine is off and will gravity feed from the fuel tank through the cylinder and into the oil. I had checked the oil on Thursday because I new we were going out the next day. I also started the engines and let them run for about 30 seconds just to make sure they would start as we were planning for friends to join us. The oil in that engine was a little less than half way between the add and full marks. Our engine sits level and the marks seem accurate. I guess it's possible this was just a freak occurrence and the injector was stuck open. Is it actually possible for dripping injector to leak 4 liters in 24 hours? If it had I would have expected something really bad or strange to have happened on startup. The was no smoke, no oil sheen on the water and no hesitation at startup. It started in less than a second and idled smoothly.



None of the injectors have had any unexpected leaks during testing. I changed the oil absorbent pad under them after the test and there are no drips on it. I may just have to put the injectors back in and let it run for an hour and see what happens. I make it a practice when the engine hasn't been run for a few weeks to pull out the fuel cutoff, engage the starter and wait for a 10 count before pushing the fuel cutoff back in and letting the engine start. Is this a bad practice that could have contributed to the problem? I did not check the oil the actual day of the incident because I had just checked it the day before and only started the engines briefly.

[skipperpete]

Chat GPT is partly correct, this injector leak event is very engine specific, Cummins PT injectors do it, GM series yanmars don’t, common rail injectors do, regardless of engine type but yes it is a thing …..but injectors rarely heal, once buggered they stay buggered so it’s a long shot but not totally impossible. Is your fuel tank higher than your engine?

[Captain Bill]

Well at the moment I'm am a bit more concerned than I was an hour ago. As recommended by Sailingallday I did the test of antifreeze/water in the oil. I poured about 250 ml of the oil I pumped out of the engine and added about equal amounts of 50/50 antifreeze to the bottle. While the total volume went up as expected there was only about a cm of green coolant in the bottom of the bottle yet the total volume went up about 6 cm. I then shook the bottle for about 30 seconds and the layer of green coolant disappeared. Now 40 minutes later and no sign of separation, whats more while the oil appears dirtier it does not at all look like a chocolate milk shake. I'm gobsmacked. It looks like I've been chasing a wild goose. Clearly mixing ethylene glycol with Rotella is not the same as mixing seawater with it. It's been 45 years since I took organic chemistry, but after thinking about the structure of the molecule it makes sense. The ethylene side mixes with non-polar solvents(oil) and the glycol side is attracted to polar solvents(water). This would allow it to mix freely with the oil and water at the same time thus holding the water in suspension in the oil. That is however bad news for me. The usual culprits that allow coolant into the oil are a cracked block, a cracked head, or a blown head gasket. Of course I'm hoping for the latter. In any case this means a major project. I'm still confused as to how the cooling system would hold pressure for 5 hours. Has anyone ever heard of head gasket resealing itself after an overheat event, or for that matter a crack that only leaks when the engine is hot?


Now 55 minutes and no separation. Given that the contaminated oil was already 50% fuel or antifreeze/water it's apparent that you can mix a lot of coolant in the oil, at least for a while.

[Sailingallday]

Sounds like it takes temperature to make the leak between oil and coolant happens. That would be my next focus. You are already on it with the possibilities listed. Hope it’s the cheaper one

[Captain Bill]

I forgot to answer Skipperpete's question The tops of the tanks are above the engines and the tanks were full, so the fuel level was above the engines by a few inches. I also found the starter rpm rating which is 200. That puts the calculated leakage for one minute of cranking at 12 ml.



I just reached three hours on my oil/coolant test and separation is starting to happen. At about 1.75 hours the mixture started getting cloudy. I suspect that this is because the miscibles are coalescing into larger droplets. There is a top layer of what appears to be oil about 4mm thick that is not cloudy.


Just for fun I got some clean Rotella and mixed it with equal parts coolant. When I poured them together I got an immediate separation and a of green coolant and a layer of clean oil. I shook the mixture for about 30 seconds and immediately got the green chocolate milkshake effect I would have expected. Within 15 minute there was a clear separation with about 75% of the coolant accounted for in the green layer. The upper layer still looked like thin chocolate milk. It is continuing to slowly separate. There is clearly a difference in the behavior of clean oil versus used oil. There is still no green layer in the used oil test.

[Captain Bill]

I now know why it overheated. I opened the heat exchanger and found impeller fragments blocking tubes. along with what looked like hair over the bottom row of tubes. I guess the full throttle blast was more than the impeller could handle. I replaced it last November, but the impeller I put in was one of may spares that had been sitting in the drawer for several years. I should have used a fresh one. I guess there are enough vanes left to pump some water as there was water coming out of the exhaust . Getting the heat exchanger apart let me see the actual coolant level. It was probably only a pint low and accounting for what I added it was probably only a quart or so low. Definitely not enough to account for a 4+ quart rise in the oil level. I guess it's back to looking for the fuel leak.


I consulted the oracle for a diesel in the oil test and found one. Definitely diesel in the oil.

[Captain Bill]

Well it looks like Skipperpete called it right originally. I decided to connect the lift pump to the boat fuel system rather than test it by submerging in fuel and pumping air through it and watching for bubbles according to the manual. I did not put it back on the injector pump but exercised the plunger by hand. After priming with the squeeze bulb I started pushing the plunger. I figured there might be some air in there that would take a while to clear so kept it up for a while. At some point I got a squirt of fuel out of the oil weep hole on the plunger. I pushed it a few more times and nothing. After about 10 pushes I got another squirt so I started counting pushes. Every 8-10 pushes I got a drop of fuel or a small squirt. Also the outer surface of the plunger wept enough fuel to make the surface wet. While I never saw any bubble in the submerged test I only pushed the plunger about 5 times. I don't know how a diaphragm pumps leaks intermittently but It's enough evidence for me to order a lift pump. If I continue to have the problem after installing the new lift pump I will have the injector pump rebuilt. I would replace the lift pump on an injector pump rebuild anyway so it's not wasted money.

[jcapo]

Originally Posted by Captain Bill

I consulted the oracle for a diesel in the oil test and found one. Definitely diesel in the oil.

What did the oracle say? All I find is to have the oil tested.

Not to hijack a thread but this is why I ask.

My 4JHDTE used to use oil and now it doesn't. It has always emitted a bit of smoke but its whitish smoke and not blueish. I had the oil tested and the report was that there was oil in the diesel but the report did not give a percentage.

The oil level in the engine does not change. The oil smells, feels and looks normal. The engine does not loose any coolant.

I would gladly have the injector pump rebuilt but removing a 4JH pump is difficult, even more so when the engine is below the cabin sole.

The engine has over 5000 hours but runs fine.

[Shaneesprit]

As you had the engine running for an hour a month, for how many months? Or how many hours would you have accumulated?
Assume you were doing this to charge batteries?
Also just had injectors done 100hrs previously- I would be getting them out and testing them.
Low load running puts diesel in the oil, although in your case this is obviously excessive, but if you also had a leaking injector at that time, without the load it’s possible you didn’t notice.
Still lift pump is the most likely, will it run without it?

I would change the oil and filter and run it again. To see what happens while.

Also worth considering your fuel tank location is the top of the tank below the height of the injectors?

[Cheechako]

If there's water in your oil you will see it; bubbles or anti freeze color. It will separate in a jar readily.
If no water then you know it's diesel if your crankcase is filling. There's nothing else for it to be. (well, maybe dark matter from space?)
Did you ever check to see if a push rod etc is off?

[Jon Hacking]

I haven't read through all of these, but an issue we've had with both of our 4JH2s is coolant leaking into the exhaust manifold. Remove the wet-elbow & check for corrosion on the heat-exchanger/exhaust-manifold, right at the flange for the wet-elbow.

Ours tends to corrode around the inside of the flange. If that corrosion is allowed to continue, it can work its way back to the coolant jacket around the exhaust manifold. It can look solid, but poking at it with a screwdriver can reveal that it's not very strong & can allow coolant in.

If you catch it early, the fix is pretty easy. Just find a machine-shop that can weld aluminum to build the lost alloy back up, then machine the flange flat again.

While running, you lose coolant. But if that coolant is allowed to leak while the engine's still, it can leak past an open exhaust-valve & into the cylinder, resulting in much carnage.

Just a thought, & it doesn't explain the oil situation at all, but the 2 might not be related.

[Captain Bill]

Jcapo
The test for diesel in the oil was pretty simple and is essentially a twist on the paper chromatography procedure I learned in high school chemistry. Get a couple of small containers that are the same size and put fresh oil in one and the suspect oil in the other. Cut a paper towel into a couple of strips that will easily fit into the containers you are using. Dip the two strips into the containers at the same time. If the diesel is contaminated with diesel the liquid will wick up the paper towel in the contaminated oil container faster than in the clean oil container. A simpler method is to take an eye dropper of the suspect oil and one of the clean oil and put a drop of each on a clean paper towel. The contaminated oil will spread much faster and farther than clean oil. The difference is really obvious. FWIW my oil did not feel normal, it was substantially less viscous.



Shaneesprit;
I run the engine to circulate the oil and keep things lubricated. I checked the oil so whatever happened was sudden. The injectors are out, still connected to the high pressure pump, and none are leaking. I turned over the engine with the starter for a full minute and the injectors were spraying normally and none leaked after the test. Unless the injector fixed itself I think I've eliminated them as the problem.


Jon Hacking;
I didn't mention it but when I took the back cover off the heat exchanger the seal between the raw water side and the coolant side came off in two pieces. A piece about an inch long remained stuck to the case and the rest came off with the cap. It was also near the top of the heat exchanger so an air leak there would explain why coolant from the overflow tank was not being sucked into the coolant side. I suspect my missing coolant when out of the raw water side when the system pressurized.


Cheechako, I have not pulled the valve cover yet but will do so this weekend.

[sinnerman]

My two cents:

If you DO have the pump rebuilt in the future, remove the injectors as well and have the shop clean and test the injectors at the same time.

Tests will be: Spray pattern, firing (cracking) pressure and leakage (no dripping after the injector fires.)

Periodically removing and servicing the injectors is maintenance item which often gets neglected. (along with periodic valvce clearance adjustments)


Cheers