[newsance]

Utterly befuddled with this car. It has 143k on it. Headgasket failed at 95K miles.

I suspected another failure. Car temp spikes up to 235* ish when run for more than 50 miles at a time, acting like it develops an air pocket in cooling system. I changed radiator in case it was bad cooling, no change in condition. It is consuming coolant and I can't see an external leak anywhere. Closest I can come is that the turbo banjo fitting has some dried coolant on it bit no sign of an active leak.

I checked oil, and found the oil level high on the dipstick by maybe 1/2 quart. I immediately stopped driving the car, assuming coolant was getting into the oil. I didn't see signs of water in the crankcase (like gunk under the oil cap or milkshake oil) but unless I am losing it I didn't overfill oil.

Here is where it gets weird. I drained the oil and sent it to Blackstone for analysis. I wanted to see if the oil showed bad engine wear to make sure that the engine was in condition to do another HG.

The results state they found NO coolant in the oil.


Quote:

Jeff, You're concerns about a bad head gasket leaving coolant in the oil, but things look just fine on that front. Potassium and sodium shown coolant in our testing and both are very low. Wear metals are in good shape too, so there are no mechanical worries on the radar. Universal averages show typical wear levels for this type of engine on the oil. Silicon is a tad on the high side. Check air filtration and the intake for leaks/cracks, but this looks more like harmless additive or maybe sealer from recent work performed. Looking good at 144,132 miles.

Potassium shows 0 (universal average 3) and sodium 11 (average 27).

Calculated water % and Coolant % are both 0. Fuel % Shows .5, should be less the 2%.

As for Silicon, I live 6 miles up a dirt road. I am sure I'm getting more dust past the filter than most.

I'm at a loss for how to proceed. Change the banjo fitting crush washers and see what happens?

 

[ForesterCub]

If you checked the oil at anything other than stone cold, it would have read high. Unless you made allowances for the engiune being hot?

How much coolant are you losing / over what time period?

 

[newsance]

I have a 30 mile drive to work, I lose a few cups of coolant over that period.

 

[ForesterCub]

OK so quite a lot. I'd have a comprsssion test done of the cooling system which is usually the best way of identifying a leak if it's there and have a second eye look over the system to ensure it's completely purged. Followed by compression test of cylinders. Good luck.

 

[ForesterCub]

These are 2 of the more helpful vids out there about this issue -

 

[TMX]

+1 Check for combustion gases in your cooling system

 

[newsance]

Thanks. Looks Like Autozone loans out the cooling pressure testers, so I will go that route.

Last go around, I had real trouble getting a good read of combustion gases in the coolant, even with what presented as a much larger leak.

I would love for this not to be another headgasket.

 

[The Tinman Rides Again]

From everything you’ve said: certainly sounds like you are burning coolant in the combustion chamber to me. When the last head gasket repair was performed do you know if/how heads were resurfaced, if overheating had occurred, if the block was tested for trueness and what gaskets were used and who did it? Best of luck! Hopefully you have a tiny coolant leak somewhere and can find it.

 

[newsance]

Minor overheating had occured- above 235 but less than 250.

They resurfaced the heads and checked for flatness. I am not 100% sure they checked the block for flatness as well.

I'm still also having a hard time finding another likely culprit, but not finding ANY water or coolant in the oil confuses the hell out of me. I was expecting that a given, I wanted to see how badly it had affected lubrication...

 

[ForesterCub]

Coolant in the oil is but one of several possible symptoms of HG issues. If there is a HG issue, the coolant may be disappearing via another route across the head.

Seriously, get a cylinder compression test done, if you haven't already. It will tell you a lot about what you need to know right now.

 

[ForesterCub]

It's a long read but if you haven't already seen this I can highly recommend this article.

https://allwheeldriveauto.com/subaru-head-gaskets-problems-explained-part-ii/

 

[The Tinman Rides Again]

Minor overheating had occured- above 235 but less than 250.

They resurfaced the heads and checked for flatness. I am not 100% sure they checked the block for flatness as well.

I'm still also having a hard time finding another likely culprit, but not finding ANY water or coolant in the oil confuses the hell out of me. I was expecting that a given, I wanted to see how badly it had affected lubrication...

How is it that you know the exact temperatures that the motor hit? Is this stored in the ECU for retrieval at a later date?

Was the head resurfacing done by a subaru dealership or service center, or were the heads sent out to a dedicated machine shop. My inderstanding is that if this issue is handled correctly: it should be a very long term fix. If not: not so much. I’m wondering if the head was. “Resurfaced” using a wire wheel or the like.

Head gasket issues can present themselves in any number of ways. Leading up to when I had to do it both oil and coolant looked A-OK but like you: I was slowly losing coolant.

 

[TMX]

Did they reuse the old head bolts during the head gasket job? You must use new ones if you're sticking to OEM or they will never seal properly after the first use.

ARP make reusable head bolts

 

[newsance]

Well, I know the exact temps because Accessport, but also because the temp light flashes at 235+ and stays lit at 250+. The temp sensor maxes at 250.

Yes, new head bolts, heads sent out to a machine shop.

That said, the cooling system pressure test shows a leak at the banjo fitting that provides coolant for the turbo. Only leaks with system under pressure, but then it drips quite a bit. I am thrilled to have a culprit that does not involve pulling the engine, but I may need to yank turbo to get it.

 

[The Tinman Rides Again]

Well, I know the exact temps because Accessport, but also because the temp light flashes at 235+ and stays lit at 250+. The temp sensor maxes at 250.

Interesting, I have an accessport but I don't leave it plugged in. I have a Greddy Informeter Touch that monitors engine temps but I figure most people are just relying on the built-in temp monitor in the cluster. Was not aware of the flashing light @ 235f.

Glad to see your problem turned out to be a simple coolant leak!

 

[newsance]

Follow-up-

What mimics a failed headgasket is a failed headgasket. Nothing I tried could resolve the issue and the problem was getting worse, and eventually I got a positive result on a combustion gas test.

Yanked the engine and did the job myself. I found headgaskets had failed on both sides, one very minor, one NOT minor. Both heads had a fair amount of warp. I found nothing else to hint at why the repeated failure happened.

The head gaskets installed by a shop were felpro. I feel I had requested OEM gaskets. Repair bill doesn't represent they put OEM on, so it would just be my word. I'm also not sure whether they truly resurfaced the heads, I would have no real way to tell.

The felpro gaskets have a coating on them (similar to the OEM non turbo gaskets) this was deteriorating all over the place. I WILL NOT be putting felpro head gaskets into any other Subaru.

 

[TMX]

Ouch

At least if you do the job this time round, you will know it is done right.

With the failure rates of OEM head gaskets on the turbo, I would be tempted to use a Cosworth, RCM or Tomei head gasket - Bit more money but if it means a better seal that doesn't fail.....Might be worth it

I don't with to alarm you but:
Did you pressure test the cooling system in the end? It might be a case of a crack in a radiator or pipework allowing air to enter the system, causing an over heat, taking the gaskets with it...
Have you checked the bores are perfectly circular and not warped/oval-ised etc? It would be a shame to do the hg job and find the block wasn't straight and you get a leak past the rings.