Water has gotten into crankcase and Oil

[ndutton]

Be prepared to do multiple pressure tests:

1. Test the entire system, block, pump and manifold all included. If it holds air, the exhaust as the water source is indicated.
2. If the first test leaks air, test the system with the water pump disconnected. No leak, the water pump is the problem.
3. Still leaking air? Test the manifold by itself, inlet and outlet ports on the top. No air leak points to the block as the problem.

A full complement of tests should pinpoint the water incursion source in maybe an hour.

[RobH2]

Ok Neil, thanks. I appreciate the additional detail.

[roadnsky]

FOUND IT (Well, you actually did)

Decided to re-read the whole thread from the beginning and realized in your POST #4 you'd actually linked to Don's post that I was referring to.

...see if a Sherwood or Jabsco water pump is installed on the engine. Both of these brands have the potential of passing a bit of water into the crankcase if their water seals leak...
Oberdorfer and MMI flexible impeller pumps don't have this same potential.

That same post also give good instructions for doing a "quick" cooling jacket test as well as pressure testing the block.

[RobH2]

10-4. Yes, that's what I was using as my reference for the test.

Update this evening...

[Don Moyer]

Water pumps and water in oil

I need to make a small correction to the record. While MMI and Oberdorfer flexible impeller pumps to not have the same potential for leaking water into the crankcase as Sherwood and Jabsco pumps do, due to their larger weep holes, we do have several documented cases where a leaky seal in both an MMI and an Oberdorfer pump did result in a small amount of water getting into the crankcase. Water is apparently able to cling to the rotating shaft of the pump and work its way past the oil seal in the flange end of the pump. The key here is the amount of water getting into the oil. If there is only a very small amount of water getting into the oil, the water pump should be considered suspect (even MMI and Oberdorfer pumps) in cases where no leak shows up in a pressure test of the water jackets. Don

[RobH2]

Did the water jacket pressure test. Put 15lbs on the system and 1 hour later still had 10lbs. That was comforting.

Then did the 3 oil changes. Ran the engine for one minute after each refill. Had the raw water inlet turned off till engine was running and then opened it for the last 30-seconds of that one minute just so some water would flow in.

After each change, the apparent water was disappearing from the oil. That was encouraging. After the third change, I started the engine, opened the raw water valve and let it run for 5-minutes.

I shut it off and went to check the oil. LOTS of water. About a quart if I use the marks on the dipstick. The level was about the same distance above "Full" as "Low" is below the Full mark.

So, what can that mean? The only other thing I did in all of this was pull my impeller. It's on a shaft. I pulled it out, removed the C-clip, changed the impeller and reinserted the shaft like I've done the past 9 years. It interfered once so I turned it and then it seated. I put the cover and gasket back on.

Could I have pushed some seal out when I reinserted the impeller shaft?

[ndutton]

Rob, does your pump have a hole in the bottom of the casting?

[RobH2]

I don't know. I need to pull it tomorrow. Here Is a photo.. Can you tell what pump it is from this?

[ndutton]

That is an Oberdorfer N202M-3 or M-7 depending on the internal cam shoe. It has the drain hole I pictured so I'm having a difficult time wrapping my head around the pump as the source of the crankcase water, especially the volume you report. Water that makes it past the pump cavity seal dribbles out of that hole before it gets to the second seal at the mounting flange end. That's why the hole is there. Some water can make it through if both seals have failed but the bulk of it should be dribbling out. If the pump cavity seal failed you would expect a failed pressure test too.

If that reasoning is correct and we accept the pressure test as accurate, we are left with water coming from the exhaust. And yet, I'm not completely convinced of that either.

[roadnsky]

Agree with Neil and I didn't mean to send us down the wrong rabbit hole with addressing the pump.
I think it's too much water.

Going with Neil's exhaust thought... can we look closer at the "steam" on the exhaust?
Maybe disconnect the exhaust and run the engine to see if water is still getting in?
It would eliminate or point to the exhaust.

[RobH2]

How do I go about disconnecting the exhaust and running the engine? Do you mean disconnect the pipe at the back of the manifold. Wouldn't water spew all over in the boat?

[RobH2]

What malfunction in manifold would cause this do you think? If it had a crack internally, wouldn't the pressure test have failed? Just trying to learn here.

If I disconnect the exhaust and run the engine are you saying to pull the pipe off of the back of the manifold and run the engine? Wouldn't that spew water all over the inside of the boat? Not that it would really hurt anything, but is that what you are saying to do?

Edit: Sorry, duplicate posts of a sort. My tablet said the post failed when in fact it appears to have gone though.

[ndutton]

Your successful pressure test leads us away from casting cracks which is a good thing.

The exhaust coming out of the manifold is dry exhaust. Water is injected after the high point in the wrapped pipe. The test Jerry is describing is loud and dangerous with exhaust gases inside the boat. Please take all available precautions.

[RobH2]

Understood, but before I attempt this what is it that we are seeking to discern from this test? What's the thought behind doing it? I don't want to go into it not knowing what it's trying to determine.

I am eager to try anything that will help solve the issue without starting to tear the engine apart. It's crammed in a very tight space that's very hard to work in.

Of course I'll do whatever is needed but the more I understand and know about what a particular test is trying to accomplish the better I can diagnose it as I'm performing it. Make sense?

Edit:
I'm back now and can use a real computer. Here are some photos. The top shows how little pressure dissipated over one hour. The middle shows water in oil on the dipsick. The bottom photo shows how the oil change looked coming out. The last time it was literally full of water.

[ndutton]

Understood, but before I attempt this what is it that we are seeking to discern from this test? What's the thought behind doing it? I don't want to go into it not knowing what it's trying to determine

Jerry covered it pretty well:

Maybe disconnect the exhaust and run the engine to see if water is still getting in? It would eliminate or point to the exhaust.

I'm back now and can use a real computer. Here are some photos. The top shows how little pressure dissipated over one hour. The middle shows water in oil on the dipstick. The bottom photo shows how the oil change looked coming out. The last time it was literally full of water.

As I said earlier, I'm doubtful pump seal damage could ship this volume of water on its own and now I'm also doubting the same with the exhaust although the wetness of the wrap certainly needs to be chased down. Following the exhaust test I expect to be asking for some more but isolated pressure testing. It goes against the grain to follow up a successful test with more testing but the volume of water you are reporting keeps drawing me back.