M54 head gasket? Combustion gases in coolant.

[oldskewel]

EDIT - SOLVED. It turned out to be a warped head. Head gasket looked OK, but was replaced anyway of course. Head was not cracked, just needed to be re-surfaced. I did a ton of other stuff too while I had it all apart.

Here's the story chronologically.

2001 E53 3.0i, 185k miles, owned for 3+ years now, running GREAT.

2 weeks ago my son was driving it home from a school graduation party in the middle of the night. Upper radiator hose sprung a leak (the plastic at the bend where it goes into the radiator) and a lot of coolant drained. He now knows he should have just stopped right there. But he pulled over, walked to Safeway and got a gallon of coolant, added it and drove the ~3 miles home from there. He says he did not notice overheating, and I don't know how bad it was before or after this incident. I think there's a good chance it overheated and he did not notice it.

Once home, it took 6 quarts of water to fill. I replaced upper and lower radiator hoses and a new radiator for good measure, all OEM (Behr, Rein, BMW).

It tested OK, I thought, then my daughter needed it for a 400 mile round trip to Lake Tahoe. She had issues starting about half way there with low coolant. She added coolant as needed, and made it home without too much issue. I think she was burning through about one quart of coolant per hundred miles.

Testing all day today ...
Engine cold, I can pressurize the system to 23 psi with no leaks at all. Bone dry, holds pressure for a really long time. I did not initially go that high, but after the other tests, I went that high to check if the reservoir cap might be opening too early (nope).

After a quick 0.5 mile test drive around the 'hood, with the hood up and engine idling (and warmed up), coolant flowing out of the reservoir cap (at the 2-bar pressure release point). Cap is Rein, a year or two old, works fine as far as I can tell. 2 bar release pressure, tested to 23 psi as noted above.

Lisle 75500 Combustion Leak Detector gave a positive result after about 5-10 minutes. The blue turned to yellowish green.

No oil noticed in coolant. I have not checked the oil for coolant yet.

The SUV still runs PERFECTLY other than this. Revs great, idles smooth, no codes, etc.

EDIT - it does NOT overheat. My daughter, on her 400 mile trip was told to pay very close attention to the gauge. And I can confirm I've never seen it go past 12 o-clock. So the damage seems to have come from that one-time event.

----------------

Does this sound like a head gasket? Worse? Anything else I can do to test things before tearing into it?

Any other ideas for a problem that would give these symptoms (no coolant leaks other than radiator cap release, combustion gases in coolant, coolant overpressures when driven)?

I'm thinking to do a compression test (at least) before tearing into it. If I find problems I may do a leakdown test too. After that, it's probably time to start taking things apart until I can find the actual failure point. I've got a driveway, a garage full of tools, but no lift.

I really don't need another project, but if needed, I would do everything to fix this myself. I've rebuilt simpler engines back in the day, and have plenty of good general purpose tools. No special BMW tools, though.

I spent a couple of hours searching for info, and there is not much on this that I found. I guess I'd follow the Bentley manual.

If it ends up being a bunch of "work" taking stuff apart, cleaning it, replacing things, sending the head out for machining, etc. I'm very happy to do that sort of stuff. But if it ends up being a logistical optimization to figure out what to do, what special tools to buy vs. hack something and take a risk, etc., maybe spending $2,000+ and wondering if I should just punt. That's not so much fun for me.

Any advice would be appreciated.

[andrewwynn]

It does sound like head gasket.

I just helped swap a few heads on M62 motor. It's no doubt a big DIY project but very worthwhile. I have a brother that can machine the head flat but if you take it out and measure and the flatness is within spec you can just replace the head gaskets.

Don't make my mistake an miss the phase three of tightening: on M62 you pre-tension to 30 N·m then two separate passes of 80° turns. Read the bently to determine the exact value for that car.

I wouldn't go as far as to say the car overheated it's just as likely the symptom followed the cause. Eg. If the head gasket gave way and is letting combustion gas into the coolant it will over press the system and vent out the cap while not hot.

It should be on the order of $200-400 in parts and a solid weekend of work to do the head gasket. Iirc you said M54 gas motor so just one head half the work.

When I did the M62 motor I was able to remove the fuel rail then the spark hardware and flip it up over the windshield leaving all the Electronics connected save two connectors that go down the front. One is the oil level sensor the other is possibly the crank sensor, or another block temp sensor it's right on the bottom of the M62.

My thought on that repair is that like paying for a class, it would be worth the $400 even if you discover mid-way or after reassembling that something bigger inside killed the engine.

On the M62s we rebuilt we had to swap valves that were bent from broken timing chain and yet they came back together and ran afterwords.

[andrewwynn]

A thought: pressurize the coolant for a good while like 2 hours then look in the cylinders with inspection camera for some florescent green drops to confirm but it has to be head gasket or crack

[sgrice]

Do a google search for YouTube 50skid m54 rebuild. There are about 30 videos in a series covering head gasket repair on the M54 engine. He discusses compression testing, what tools to buy, what the cost was, etc. I’m not an expert, but it sure seemed to me that he covered it well.

Looks like about 10 hours or more of videos, so go ahead and get started watching!

Good luck - let us know how it turns out. As Andrewwynn pointed out, it’s possible the head gasket went first, then high pressure popped a weak spot in the hose. I don’t know how one can tell if the chicken or the egg came first. From a practical standpoint it doesn’t matter - at this point the repair is the same.

[oldskewel]

Thanks for the advice so far.

Yes, M54, gasoline, so one head.

I'm thinking basically:
park it in my driveway in a really good spot, where it is likely to be stuck for a while.

some final tests -
pressurize like andrewwynn says (however, I expect it is possible that 20+ psi pushing coolant in on a cold static engine might not succeed, while pushing ~1500 psi combustion gases on a hot, running engine might be causing the problem I've got). Inspect inside.

Compression test. Possibly leakdown test.

Take it all apart to get to the head + gasket. Inspect

Assuming it is worth repairing ...

Have head machined. At 185k miles, valves + seats probably need machining; fly cut the head surface. Will need to buy a new head gasket based on how much machining was done.

I've read / heard that in some overheating conditions the threads in the block are compromised and need to be fixed. Hopefully I will not have that problem since I believe this was a pretty minor disaster.

New gaskets of course on everything else I take apart along the way. Is there a single kit I can buy with everything? Recommended parts lists?

For those that have been down this road, anything here wrong? Any tips, etc.?

[oldskewel]

Originally Posted by sgrice

Do a google search for YouTube 50skid m54 rebuild. There are about 30 videos in a series covering head gasket repair on the M54 engine. He discusses compression testing, what toold to buy, what the cost was, etc. I’m not an expert, but it sure seemed to me that he covered it well.

Looks like about 10 hours or more of videos, so go ahead and get started watching!

Good luck - let us know how it turns out. As Andrewwynn pointed out, it’s possible thr head gasket went first, then high pressure popped a weak spot in the hose. I don’t know how one can tell if the chicken or the egg came first. From a practical standpoint it doesn’t matter - at this point the repair is

Thanks, I'll look that up on the google. 10 hours of videos - wow, I don't know if that is good or bad. I'll definitely take a look.

Thanks for pointing out andrewwynn's comment on maybe the head gasket failure was the first thing to go. I completely missed that on the first reading. Yes, that would give some reassurance, maybe, that I'll find a limited disaster in there when opening it up.

More specifically on the failure point on the upper radiator hose ... it was on the plastic fitting that goes into the radiator. There was a bunch of crusty stuff inside that fitting that would scrape off with a screwdriver. The back end of that hose had no problem, and the lower radiator hose appeared almost like new (and I think it was - one problem with incomplete service records). The plastic fitting makes a bend right past the little opening that connects to the expansion tank, and the hole was there, in a little indentation. Almost like it was designed as a weak point so it would fail there and be easy to spot. I think the crusty residue could have provided some protection for a while, and then at some point it just blew through. So yes, the HG causing the radiator hose to blow could be true. Interesting.

[andrewwynn]

I'm surprised if there is not a thread on xoutpost covering the topic. If not it will be good to get it you may be able to just resurrect an old thread. It's a topic that needs to be here.

It's a big job but many many of the xoutpost members can handle it and can reset the clock on a very fine car and get you many more miles and more years

[andrewwynn]

The fitting on the host is a known weak link.

Internal PSI of a couple engines I've looked up on Google were on the order of 800 so even if you put 20psi (rounding) you could expect on the order of 1/40th the amount of leak going into a cylinder as out. I would run the engine to get it warmed up and especially if it starts to pressurise the cap to vent, then pull the plugs and pressurise the coolant and peer into the cylinders to try to verify before teardown head gasket leak coolant to cylinder.

All indications are that must be the case.

Tip: if working outside invest $80-100 into a 10 ft pop up tent so you can work with minor weather annoyance.


Sent from my iPhone using Tapatalk Pro

[oldskewel]

I took some photos and have some data to post later, but here is the quick update ...

I did apply the cooling system pressure for a few hours, trying to force coolant into the cylinders (this turned out to be a lot easier than I thought it would be). I could not find my endoscope, but I proceeded anyway. Was able to detect coolant on the tip of a screwdriver stuck through the spark plug hole on #3.

Then when compression testing, a significant amount of coolant was blasted out of both #2 and #3 (from the long duration coolant pressurization - not so much after the initial several revs).

Good compression on all. Logged elsewhere, but all were 180-200 psi. This is reassuring that the engine is worth rebuilding, even at high mileage like this.

Another observation from the compression testing that indicates the severity of the leak - if the expansion tank cap was installed tightly during the ~10 revs of compression testing for #2 or #3, coolant would leak out of the cap. Loosening the cap stopped that. This seems to be worse than before all this testing (e.g., it took a half mile test drive to get the leak started before all this) - so somehow maybe the reverse pressurization expanded the leaks??? (confused on this one).

So I'm continuing with the teardown, slowly and carefully. Got the valve cover off so far. Cams look good.

I'll get it taken apart to fully diagnose and then will decide on rebuild (top only) vs. junkyard engine vs. get rid of it.

[andrewwynn]

I think you probably caught the problem at the perfect time to rebuild from the head up. Odds are good that the leak just started when soon had the incident and blew out completely with daughter he the large quantity of coolant blown out.

With any luck out will be just a blown gasket and you won't even need to grind the head flat.

They make a thicker head gasket for post machining I would look into the option of using thicker gasket vs machining to make up for imperfect head. You would lose a tiny bit of compression I'm not sure what other trade off there would be.

Make sure to use new head bolts and follow tightening process properly: helping a guy recently put a new head on and he skipped the last step and naturally coolant got into the oil.