Perkins 4.108 - oil in coolant

[sf_pearl]

I put a rebuilt long-block from Transatlantic Diesels into Pearl (Islander 36) in 2013. It only has about 200 hours on it to date.


I pulled the "radiator" cap early in the month to find oil in the coolant. Not a lot, but enough to make the coolant (antifreeze) look pretty milky.


My first thought was the oil / trans oil cooler, which was cleaned and tested in 2013 but not new. Pulled it out (real fun), had it tested -- no leaks


Put it back in, changed all fluids. Ran it for an hour at the dock (some of the time with the trans in gear to add load), and changed the coolant from the header tank about three times.


At this point, it looks like whatever leak there is must be fairly small -- not a lot of oil circulating in the coolant after an hour or so of running. But there is a leak, of course. Engine starts and runs as normal, no smoke. Also, the oil level hasn't changed and looks as crystal-clear as it always does after an oil change; at least on the dipstick.



So. Head gasket? Are there any seals elsewhere on the engine that could be suspect? Lastly, could it possibly be just some torquing of head bolts that could fix it? I've never re-torqued them since installation; TAD said the engines are "fully broken-in" after rebuild.


Anything beyond re-torquing head bolts is going to require pulling the engine, unfortunately.

[model 10]

There is a small hole in the block and head for the oil passage to the rocker arms. It is on the right side in the middle. There is a possibility that you could tighten the head bolts, but you probably need a new head gasket. I don't see how they could break in an engine without running it for the 25hr it takes before the re- toque.

[sf_pearl]

Originally Posted by model 10

There is a small hole in the block and head for the oil passage to the rocker arms. It is on the right side in the middle. There is a possibility that you could tighten the head bolts, but you probably need a new head gasket. I don't see how they could break in an engine without running it for the 25hr it takes before the re- toque.

Thanks, model 10. My understanding was that they do run them for 25hrs, but perhaps my understanding was sadly wrong.


In any case, I'll look at re-torquing. Somewhere I have a 4.108 manual with the torque specs.

[sf_pearl]

Re-torqued the head bolts today. Took a long time; have to remove the coolant header tank to get to the front two bolts. And remove the rocker arm assembly of course, but that was simple enough.


Several of the bolts were a good 1/2 turn from the 60-lb/ft spec, ug. So I should have done that long ago, dammit. The one in the middle, closest to the oil tube to the rocker arm was loosest. Fooey.



So we'll see -- the good news is that I'm confident now that I wouldn't have to remove the motor to get the head off, if I do have to replace the head gasket.


So now I'm re-flushing the coolant a couple of times and we'll see. Gapped the rockers while I was at it; only two slightly off.


The 4.108 manual doesn't have a torque spec for the nuts on the rocker assembly -- they're small, so I just went to what it seemed like to get them off, which is probably 10-15 lb/ft.


Crossing my fingers. Tomorrow I'll flush the coolant a couple of times and hopefully take her out for a spin to get her good and warm and see what floats to the top.

[rbk]

Originally Posted by sf_pearl

Re-torqued the head bolts today. Took a long time; have to remove the coolant header tank to get to the front two bolts. And remove the rocker arm assembly of course, but that was simple enough.


Several of the bolts were a good 1/2 turn from the 60-lb/ft spec, ug. So I should have done that long ago, dammit. The one in the middle, closest to the oil tube to the rocker arm was loosest. Fooey.



So we'll see -- the good news is that I'm confident now that I wouldn't have to remove the motor to get the head off, if I do have to replace the head gasket.


So now I'm re-flushing the coolant a couple of times and we'll see. Gapped the rockers while I was at it; only two slightly off.


The 4.108 manual doesn't have a torque spec for the nuts on the rocker assembly -- they're small, so I just went to what it seemed like to get them off, which is probably 10-15 lb/ft.


Crossing my fingers. Tomorrow I'll flush the coolant a couple of times and hopefully take her out for a spin to get her good and warm and see what floats to the top.

Pretty sure (without actually checking) the shop manual say 15lb/ft for the rocker assembly bolts.

[Squanderbucks]

Some will tell you to just torque it down and good to go. Mistake after all these hours. Gasket is burned in places and contaminated with oil and water. That would have been a Choice right after first few hours when all was still clean. Don’t waste your time.
Bite the bullet and do it the way you really know in your heart that it should. Sometimes we ask for validation of shortcuts to make us feel better. It seldom works.

A trick I was taught is to put in the head gasket. Tourque it all down. Then run the engine for a few minutes without the water pump">raw water pump hooked up to get the engine up a little over normal running temp. Say 200F. Shut it down and let it cool enough to work on and re-tighten to final specs. Some suggest to do the initial burn in of the gasket even before putting the block coolant.

[Scubaseas]

12 to 15 ft lbs is fine for 8mm/5/16 nuts/bolts.. There is an engine oil cooler in the heat exchanger but do you also one on the transmission? While head gasket is a possibility any place where the oil (high pressure) meets coolant (low pressure) can be your source. Most common is oil coolers from my auto experience. Maybe try bypassing the oil cooler as a proof or do your own leak test overnight with compressed air. Possible if you had a radiator shop test it they only went to 18 PSI or so instead of oil line pressure of 40 to 60 PSI.

[model 10]

You might put some dye in in the coolant that detects combustion gas too.

[foothillsailor]

Don't be too quick to blame a head gasket ---- not so sure on a Perkins, but on my Westerbeke, the seal on the water pump">raw water pump shaft failed letting salt water into the oil chambers of the engine ----(caught and fixed before any damage). A bunch of amateurs and mechanics had all incorrectly 'diagnosed' the problem as a failed head gasket (ie. time and $$$$)



However in your case because its is oil in your coolant then I would at least first take a look at an internal parts diagram of your coolant pump and see if there are any internal seals or gaskets which if they failed would permit the passage of oil into the coolant... just something I'd check before I pulled the head ... good luck

[sf_pearl]

Originally Posted by Scubaseas

12 to 15 ft lbs is fine for 8mm/5/16 nuts/bolts.. There is an engine oil cooler in the heat exchanger but do you also one on the transmission? While head gasket is a possibility any place where the oil (high pressure) meets coolant (low pressure) can be your source. Most common is oil coolers from my auto experience. Maybe try bypassing the oil cooler as a proof or do your own leak test overnight with compressed air. Possible if you had a radiator shop test it they only went to 18 PSI or so instead of oil line pressure of 40 to 60 PSI.

I like the idea of bypassing the oil side of the cooler (yes, it does the transmission too) to really be double-sure it's not the cooler. I'm still somewhat suspicious of it, your comment is relevant that they may have only tested to 18-20 psi at the radiator shop.


I got the engine good and hot today after a pretty reasonable 2g coolant replacement, and there's definitely still a leak. But I also pumped out the trans fluid, and it was way browner than I would have thought given that I had just changed that as well -- no obvious motor oil or coolant in it, but I was surprised at how not-pink it was given that I just changed the fluid. So that gives me a nice (and probably false) hope that it's actually the cooler...


Thanks for all the feedback, folks.

[sf_pearl]

Well.


I bypassed the oil cooler with a straight piece of copper tubing (with two flare fittings on the ends), rode her hard after changing out 3 of 4g of coolant, and... oil in the coolant.


So now I've pulled the head.


Certainly nothing obvious / visibly wrong: The oil hole (is there only one?) for the passage to the rocker arm connection doesn't seem to show any leakage. Doesn't seem to be any obvious leakage anywhere, really.


I'm attaching a picture of the oil hole on the head, fwiw.


I haven't experienced that many leaking head gaskets in my time, but for the couple I have, it was pretty obvious. This one sure isn't.


It can't be the coolant water pump (no seal with oil there), so that leaves... what else?


I have the new head gasket from England. Any suggestions for cleaning up the head / block surfaces before reinstalling (and crossing my fingers)?

[Lepke]

Since you have a suspected head gasket failure, I'd use a head gasket sealant. I'd clean the head and block surface carefully. You don't want scratches to lead to another failure. I usually wipe the surfaces down with acetone or something similar. One thing to consider, you may not be getting all the oil out of the cooling system. Oil is on top of the water, so if the water isn't completely drained, you still have the oil. When I'm chasing/repairing that problem, I drain the coolant, add some dish soap to water and run the engine until the thermostat is open. The soap causes the oil to blend with the water and when draining, it more thoroughly gets the oil out.

[Scubaseas]

Too bad the cooler didn't proof out as bad. If your new gasket has red sealing areas you need a spotless surface. I use a razor blade like a cabinet scraper (about 5* from vertical and carefully so as not to dig in and scratch) and clean surfaces with carb or brake klean. Acetone or MEK would also work fine.

If you want to use a sealant I would go with Spray High Tack on both sides of gasket. Not liking the pitting on the head around the oil gallery. Might just be the picture though. If it doesn't clean up definitely use a sealant in that area. Spray on head gasket and spray around the oil gallery passage on head itself

https://www.permatex.com/products/ga...asket-sealant/

Did you notice if the engine oil level was dropping? Or possibly the transmission level dropped? If your oil leak into coolant is coming from the engine you should see a drop in the engine oil level, no? Your comments on color of trans fluid are concerning. When the cooler was bypassed was any oil coming out of the cooler itself?

It can take a long time to get all the oil out of the cooling system. Dawn soap (2 teaspoons) works pretty well as does radiator cleaners.

[sf_pearl]

Thanks for the advice, Scubaseas and Lepke.


That "pitting" is actually just a buildup from the old head gasket; stuff I have to clean off yet (some on both sides). The new head gasket has red sealing areas, it doesn't look like the old one did. Maybe that's good.


The engine oil wasn't dropping enough to really tell, but the oil in the coolant didn't amount to much in volume either. I tried mixing a little Dexron with coolant in a jar to see what the result looked like, and it seemed like it did not look like the black/brown stuff I found in the coolant in the header tank. I had hoped it would so that it would point to the oil/trans HE again but after emptying / refilling the trans, I wasn't seeing a drop in level there either.


Re: getting the oil out of the cooling system, 2 tsp of Dawn sounds like a good idea -- but what ratio of antifreeze to water do you think might be safe while doing flushes? I can only get about a gallon (out of ~4) out of the header tank and a few pulled hoses around the engine, and went through about 8g of 50/50 mix just trying to test the oil HE. I know the ceramic bearings in the (fresh) water pump are sensitive, so I don't know if it's a good idea to use a mix that approaches 100% water (and a little dish soap) to flush?

[Scubaseas]

Originally Posted by sf_pearl

Thanks for the advice, Scubaseas and Lepke.


That "pitting" is actually just a buildup from the old head gasket; stuff I have to clean off yet (some on both sides). The new head gasket has red sealing areas, it doesn't look like the old one did. Maybe that's good.


The engine oil wasn't dropping enough to really tell, but the oil in the coolant didn't amount to much in volume either. I tried mixing a little Dexron with coolant in a jar to see what the result looked like, and it seemed like it did not look like the black/brown stuff I found in the coolant in the header tank. I had hoped it would so that it would point to the oil/trans HE again but after emptying / refilling the trans, I wasn't seeing a drop in level there either.


Re: getting the oil out of the cooling system, 2 tsp of Dawn sounds like a good idea -- but what ratio of antifreeze to water do you think might be safe while doing flushes? I can only get about a gallon (out of ~4) out of the header tank and a few pulled hoses around the engine, and went through about 8g of 50/50 mix just trying to test the oil HE. I know the ceramic bearings in the (fresh) water pump are sensitive, so I don't know if it's a good idea to use a mix that approaches 100% water (and a little dish soap) to flush?

You can use straight water to flush and run. The ceramic seal on the coolant circulator pump won't care if it's under say 20 hours. I would get as much oil as possible out by using paper towels or rags into the expansion tank/filler port. Or use a turkey baster or suction pump of some sort. You
can reuse your mixed coolant. Just skim the oil off the top.

When I do a flush to remove oil I try and mechanically get as much oil out as possible, then remove the thermostat. You want no t-stat so you have full flow. Then put in Dawn or radiator cleaner, then let motor run for 1/2 hour or so, then slowly fill the cooling system while the engine is running by using a garden hose on about a 1/4 gallon per minute flow. This helps flush the oil out of the reservoir as the oil will still float to the top although the soap helps emulsify it. It takes a while to get it all out. I also pay attention to all the rubber hoses as motor oil can make them go soft or disintegrate. When flushed and no more oil, drain system, put the thermostat back in and fill with 50/50 mix antifreeze to water. Re check daily for oil as there may still be a small amount still in the system. You may still have to pull out small amounts over the next week or so. As long as it's not increasing day by day you are probably fine.

The red sealing material is a good way to solve design issues or poor sealing I've quite a few engines that had upgraded/revised head gaskets that looked like they just added the red sealing stuff. Just make sure you have a clean surfaces and follow the torque pattern when installing. Put a little oil on the bolts when installing to get accurate torque values.

You could always introduce a dye into either oil to rule out which one it is. Tracer dye is available online and your Friendly Local Auto Parts Store.