I think the internals are ok. I ran it at idle for a few minutes and no leaking. A bit of white smoke at first but nothing crazy. I kept it revved at 3k rpm for a few minutes, no fluid came out, and no white smoke. The exhaust smells a bit "sweet" but that could just be coolant burning off the pipes. I didn't get it all the way to temp but Im not gonna try until the new TStat gets here. If it happens at that point I'll swap in the new one, if not I'll hold on to it and see what happens.

Sent from my POCOPHONE F1 using Tapatalk

Exhaust smells sweet? Is that at the tailpipe or just overall around the exhaust?

Think you just had an air pocket or you over filled. I would get it up to temp at idle and check level...on a level ground that is or with only the rear lifted.

You could also rent a coolant pressure tester and see if it holds pressure. You can do this with the engine off at least.

Not to scare you even more, but that's a common method of failure in Subarus: Failing HG causing pressure into the coolant system. You can often see the bubbles coming up in the expansion tank. No oil/coolant intermix, generally. I don't know if the same happens in the M96. And like you said, M96 HGs are rock solid.

I knew about the ring land failures for the subies, didn't know about the bad head gaskets. My test today certainly isn't conclusive, I need to see what happens when I run it under load. If it leaks then, compression and leak down tests will be in order. I appreciate the heads up though, I will do some reading in them tonight.

Sent from my POCOPHONE F1 using Tapatalk

Pull a vacuum on the cooling system with the airlift system and see if any air comes out. Sometimes when one of these has air pockets, you get a surge in the system that can annoy the plastic tanks and create sudden leaks.

When did you last open up the cooling loop? I assumed from your comments on just getting the car back together that it was recently.

Oh now I get what you were saying. No, I dont think this is a tank issue. I'm nearly certain all the flow was coming from the overflow tube itself.

The last time I did anything to the cooling system was around 2 years ago - system drain, 04 oem cap, and LN TStat install. No problems at all since then and had the usual "success" with lower running temps up until this happened.

I did a bit more poking around in the trunk while running it at 3k rpm (no load obviously) - mild vacuum on the oil tank, absolutely no positive pressure, so I think my rings are good. No signs of water/oil on the dipstick and no signs of oil in the coolant resivoir. However my oil stinks of gasoline. It's been 6k miles since last changed but that was a bit over a year ago. I figure it's from cylinder washdown since it's a 2.7 with a 3.4 tune, but I haven't driven enough to check afrs yet.

It can never be easy...but I suppose that's half the fun!

Sent from my POCOPHONE F1 using Tapatalk

OP said clear smooth liquid...could this be something other than coolant which is colored...or at least every antifreeze I've seen is colored not clear.

Pressure test the cooling system?

New BW TStat came today. I'm juggling a bunch of stuff right now but as soon as I get a chance I will drive the **************** out of it and see if it happens again. If it does, new TStat is going in.

Unfortunately I don't have a way to pressure test the system. I don't usually like just replacing parts but at $50 with the housing included this just seemed like the easiest routes to take.

Sent from my POCOPHONE F1 using Tapatalk

Update - I ran her hard for 20 miles today, needle straight up and kept it there for probably 15 minutes or so. No more coolant at that point coming from the overflow tube! Hooray!

However, when I popped the trunk there's coolant around the cap (.04 model, changed 2 years ago) and also the burp valve was closed but pushed open! I was able to push it back down but it just popped right back up.

Something is definitely up here. Im on the verge of going nuclear and replacing all of it - the TStat, cap, and burp valve but I know that's irrational and what I really need is more information. I think I will go ahead and flush the system tomorrow and swap in the new TStat just to make sure there are no signs of intermixing or sludge/debris in the system.

I'm hoping this doesn't have anything to do with the white chunks of debris I found in my oil filter tonight but I think that after the flush I'm gonna have to start doing some pressure testing of the coolant system and cylinders.

Any other thoughts?

BTW, feel free to chime in on the white chunks in the oil filter if you know what they are (pics posted on other thread)

Sent from my POCOPHONE F1 using Tapatalk

Maybe combustion gases are getting into your coolant. Try this, it's cheap and quick:

https://www.harborfreight.com/combustion-leak-detector-64814.html?_br_psugg_q=combustion+leak+detector

Hey thanks piper! I haven't seen such a thing before but looks simple enough. I appreciate the help.

Sent from my POCOPHONE F1 using Tapatalk

If you use a combustion gas tester, test the tester fluid before relying on the results. I have had a kit not work before, it wouldn't change colors even when exposed to exhaust straight from the tail pipe. I think the one I got sat on the shelve for so long it lost its effectiveness maybe. Idk. Probably an isolated incident, but thought it worth mentioning.

Sent from my SM-G991U using Tapatalk

Thanks GL, that's a great idea. Book knowledge is important, but experience is importanter ;)

Sent from my POCOPHONE F1 using Tapatalk

Quick update -

Changed the oil today (reeked of gasoline) and put new injectors in her. I used castrol gtx 20w50 for a few reasons, I wantes to go up to 50 weight and it was the only one walmart had, and I was just looking for something to use as a cheap flush for a few hundred miles.

But ho-lee-schnikes - when I fired up the car I honestly though something was wrong because it was SO QUIET (keep in mind that she is straight piped!)

I'm starting to think that a bad injector leaked a ton of fuel into the oil while I was working on her this winter, and that oil was so dilute and had such a low viscosity that it allowed a massive amount of blow by into the crank case, then causing my coolant puking.

I should know more this weekend though, just got my combustion gas tester and plan on doing that, along with a coolant flush and new BW TStat install this weekend. I'll keep everyone posted!

Sent from my POCOPHONE F1 using Tapatalk

Flushed coolant and swapped tstats last night. Comparing the BW and LN side by side, the only noticeable difference I could tell was the spring (with flange) that is on the water pump side is much more stiff on the BW then the LN. I killed my battery while flushing the system though so hopefully I can't finish up tonight or this weekend.

Btw, I found that I drained an extra gallon of coolant from the front radiators by jacking the front up about 4" higher. No need to disconnect the front hoses (or the AC compressor lines, which I did in an exceptionally bone head move lol).

On a side note, I cleaned my radiators. I've seen some videos where guys don't remove the surrounding shroud and separate the two but you have gotta do this - my compressor fins looked pretty clean but when I moved them out of the way I was shocked to see how much **************** made it through those fins and into the radiator fins. Pic below, just for funsies. https://uploads.tapatalk-cdn.com/202...a6516f8b8f.jpg

Sent from my POCOPHONE F1 using Tapatalk

BW TStat installed, system filled and bled. A bit of coolant came out the overflow during the initial burping process but I got it up to 180 and kept it there without any further spillage. Interestingly my radiator fans did not kick on at that point so I'll have to use the piwis to interrogate those. At the end of it I used my leak tester kit and it stayed blue! No head leak! I did double check to make sure the fluid was reactive by stocking the tester in front of the exhaust and it did indeed turn yellow at that point, so I think I'm in the clear.


On a side note, I thought long and hard about replacing the water pump while I was there. I have no leak from the housing though and there is no bearing play when the serpentine belt was off. I kept trying to figure out a better way of interrogating it (for peace of mind) and then I realized that with the TStat housing off, I was able to stick my flex scope up in the water pump and visualize the turbine blades, which appeared to be in perfect health. Once I could see the guts of the thing I stuck my finger up there and tried to see if there was any internal play in the bearing, which there was not. So in the end I left it in place and slept well after that.

Anyway, I now have a questionable LN TStat for sale, with the housing and a slightly used gasket included! ;). Thanks to everyone for the help.

Sent from my POCOPHONE F1 using Tapatalk