This is my first major project with the car and honestly I have no idea the state of the system before doing this. Have driven the car many months without issue. Had the cooling system apart for the front end service. Fully drained rad, replaced coolant. Replaced coolant tank and rad cap. In the process of running the car for the first time and filling the coolant, I ran the reservoir dry for a bit before I could get more coolant in there, probably pumped a ton of air into the system. After that I followed the normal bleed procedure in the forum threads and youtube videos I've watched. Rad cap off, idle for 10-15 minutes, fill as necessary, hit the bleed screw.

Car runs fine, no overheating, drove it for 30 minutes yesterday and the coolant level dropped by half an inch after I thought I had bled it fully (5 bleed screw openings, 15m idle, some with rad cap off). I drove the car 1 hour since then, coolant level hasn't changed every time I've checked it after driving.

Symptoms:

• Coolant bubbling sound when you shut the car off, sometimes. Twice it didn't, the last few times it has. Moves around. It's been the radiator, and around the turbo coolant lines.
• Air in the system. I see air bubbles, sometimes, coming through the rad return line. Sometimes it will be fine for 5-10 minutes. Often changing heat\sport mode settings cause it to spit some out. Sometimes very tiny bubbles, but I've had a few whammies come through just tonight.
• The coolant level never changes, not as much as I would expect as the amount of fluid that I'm letting out of the bleed screw nor from how much air is coming from the return line from the radiator.
• I am not getting overheating. Scanner confirms 105-108 degree temp limit max, then the fan kicks in. Sport mode sends it straight to 90c. Car doesn't overheat while driving as far as I can tell.
• I keep getting spurts of air coming out of the bleeder screw hole. I'll open it, a 0.25 second spurt of air will come out, I'll close it, a few seconds later I'll repeat and more air will come out. I've opened the bleeder screw probably like 40 or 50 times now and I'm still getting air coming out.
• I snapped the bleed screw b\c I had to open it so many times, replaced with Dorman brass screw w\ o-ring. Happened at opening #25 or so, didn't change symptoms at all when I replaced.
• Radiator cap is NOT hissing\venting pressure, even after idling & revving car for 1+ hour with system fully sealed.

I've spent maybe two or three hours with the car in every configuration of: idling, driving hard, in sport mode, heat full blast, or normal mode, A\C on, driving the car.

No coolant leaks, no codes, no drips, no weird smells. When I drained the oil, it looked perfect with no visual amounts of coolant intrusion - tho I did not get it tested.

I just spent an hour tonight hitting the bleed screw every minute or two, revving the engine, going on short drives. Coolant level barely dropped enough to be noticable, even though there was enough coolant to start burning on the engine block and dripping onto the ground. I'm still getting air out of the bleed screw.

There are no leaks so I can't imagine any of the hoses or connections are letting air into the system? Where the hell is all this air coming from. Is there really THAT much that not using a vacuum bleeder really is this difficult?

 

Personally, I like squeezing both hoses together just above the shift linkage as hard as I can, loosen the cap, let the air out, tighten the cap and then let go of the hoses. I will "squeeze pulse" both hoses repeatedly, then try that again. Then afterwards do the bleed screw while squeezing the hoses. This has always worked for me.

 

If you can get your hands on a vacuum coolant fill tool, you’ll save yourself further headaches. I have one that I bought the first time I had to change out my thermostat. With the vacuum tool, it’s a full fill with no air first time every time. I have used it three times, including that first time, and had no air in the cooling system each time.

This is the kit that I have:
https://www.ecstuning.com/b-schwaben...l/003466sch01/

 

This. Most OEMs spec a service tool such as this to prevent water pump cavitation and trapped air.

You might also want to try jacking the front end of the car up to allow the air to travel to the highest point. It’s an old school technique, but it does work.

To be honest, I would overfill slightly above the full mark and drive the car while keeping an eye on ECTs. It will likely work itself out of the system.

 

Njaremka is right. I got this tool.
It's more expensive, but the quality is much better that the knock off version my brother purchased.

I've attempted to attach the BMW TIS bleed proceedure ... let's see if this works.

 

MFB, it is possible you have a head gasket or oil filter manifold gasket issue. Remove your coolant recovery tank cap, and install the pressure test gauge from the kit onto the coolant recovery tank. Start the engine, if the head gasket is blown, you're going to see an immediate pressure rise on the gauge.

Other options for you are to collect an engine oil sample and send it to the lab; they test for coolant.

Local auto parts stores also have combustion gas testers that you can connect to the top of the coolant recovery tank. If you have combustion gasses in your coolant, the fluid in the tester will change color and confirm you have a head gasket problem.

 

Ok folks thanks a bunch I have a ton of things to look into now. Will report back.

 

so I am just re-reading things - I did re-seal the oil filter housing as part of my service. Replaced all the seals as the old unit was leaking really badly. Is it possible I messed up that somehow in my reattachment? I cleaned the mating surfaces as best as I could and took pretty good care of the unit when I was putting it back on. I had the front of the car off so I think it was the best case scenario for having a successful operation.

I have had the car running for total of probably 4-5 hours now, I feel like if there were air leaking into the system from a head gasket (or coolant coming out) we would have seen over-pressure on the coolant system by now? I have spent lots of time revving the car parked with my head sticking out the door to watch the coolant tank, I'd have noticed.

Will be getting help from a friend with these tools next week to give concrete answers on these questions, but just want to throw this out there for now.

 

If you think it may be getting into the engine, a simple check of the color of your oil will tell you if it is or not. Dont be surprised if under the cap has some "milkshake" on it, it often will if the engine doesnt get hot enough to burn out the moisture in it - but if you pull up your dipstick and there is anything other than normal looking oil, then the coolant is leaking into the engine.

 

Keep in mind that oil pressure will be higher than cooling system pressure. Even at idle My R56 is running 11 PSI oil pressure. I have seen engines where oil worked its way into the cooling system, but the engine oil looked normal; I even sent an oil sample out for analysis and there was no coolant in the oil.

 

Engine oil looks normal, coolant is totally normal too. I did an oil change as part of the service, so 100 miles ago. Oil looks fine, smells fine. Coolant looks fine, smells fine. I checked both yesterday.

So, I put another hour into the car this week. Still getting boiling fluid from the front end when I shut the car off (it's back in the radiator now), still getting a little *psst* of air every time I hit the bleeder screw. I can't tell if I'm making overall progress on getting air out of the system yet or not.

Temps are fine, I monitored it last week for maybe 30 minutes while driving. 108 degrees fan comes on and it comes right down. Sport mode shot it directly to 90 degrees in less than a minute. I've got maybe 100 miles and 2 idle hours on the car and all the driving I have done lately does not change the coolant level in any way. Rock solid. The only change in coolant levels feels like the trickle of coolant that comes out whenever I hit the bleed screw.

I've opened the cap a few times when the car is fully cold and its always under pressure whenever I crack it. I get air escaping and the coolant level in the tank rises by maybe 1-1.5cm.

Plans changed, I've pushed to this weekend for borrowing a vacuum bleeder w\ a friend to give that a shot. Hopefully that will get me to where it needs to be and I can be more confident in the car. I don't want to drive it hard in the canyons until I get this cooling system bled properly, I get the car pretty hot at altitude, it needs all the cooling it can get. That, plus my front strut mount bearing (it's binding) giving up due to being in the air for a month - I've got some more work to do before the car is ready.

Thanks everyone for your input and help so far!

 

There a simple chemical combustion test for testing for head gasket leak:

https://a.co/d/6tw6Sas

https://a.co/d/f7BJZvi

https://a.co/d/hSDhWQ3

Jack the car up or park on a hill so the fluid res is the highest point. Start from cold and open the res cap. When the coolant starts to come to the top, put the cap on and tighten. If you want, you can squeeze the hoses while warming it up. Be careful reaching in.

 

I have been beginning to suspect that as well. I also have a misfire on cold starts. Low idle it misfires until its warm, and then if you put too much load on the motor when it's extremely cold it has a pretty good misfire. Once warm, the motor feels perfect even under full throttle across all the RPM range.

Ran the HG test. Now, this only tests to see if there is combustion gas in the coolant. It doesn't measure if the HG has broken between the coolant and the oil. But I've checked the oil and the coolant - visually, they both continue to look fine.

But I've passed the test with totally flying colors. Ran the car for about 3-4 minutes, spend the last 90 seconds revving the motor. Saw many bubbles go through the test fluid (big bubbles once a second). The fluid did not change color, even slightly.

 

Well, not the head gasket then. This is just a weird one: I'd think any air would've worked its way out by now. And it's not like the system would "suck in" air anywhere but the expansion tank cap, where it's supposed to (it has vacuum relief).

I'd suggest pressure testing the system to verify no leaks: you should be able to rent a pressure tester from your local auto parts store, and it's an easy test. If that looks okay then drain the coolant again and do a refill from scratch using the vacuum tool.

 

Ok sounds good. Memorial day today, I'll hit up the local parts stores to see if I can rent what I need. My neighbor down the street didn't have one, and was the one to actually suggest the head gasket test (along with the other gentleman here).

Will go back to square 1 trying to get it bled then.

Is it possible to have a crack somewhere in the cooling system, maybe at a high point, that can let air in, but not liquid out? If that were the case, at some point it should leak fluid right. Hopefully you're right and the vacuum test will yield good results here.

By the way the coolant tank and res cap are both brand new. Replaced them as part of the overhaul I did. I think the tank is Lemforter, and the cap is OE.



And even if it did let air in, it wouldn't be in the cooling system. It would just be on the air bubble at the top of the tank. IDK how it's getting into the loop.

 

Pressure test kit didn't have a cap which fits on the mini, had to use the universal adaptor. Highest test I was able to run is ~8PSI.

Rock solid did not change pressure at all after several minutes.

Should I keep trying to see if I can get a seal which will let me test at the full ~15 PSI?

Will source a vacuum bleeder next. Problem that I have is that I don't have air. I might just end up going to a shop with this one.

 

I mean, ideally you'd want to test up to at least the full working pressure. Per the service manual system pressure is 1.4 bar (20 PSI) and test pressure is 2 bar (29 PSI) so I don't think 8 tells you much.

Yes, you'd need a compressor to use the vacuum fill tool.

 

Oh there is no way that test kit I rented would work then. I'll need to find one with a proper adapter that actually screws on, instead of the universal adapter I had to use.

I tried searching around forever and couldn't find the pressures. The closest I could find was the standard 15PSI. I had no idea we're running at 20\29.

 

Yeah, here's what I found:

 

So I drove the car for 100 miles today. This doesn't make any sense to me. I'm still getting overpressure from the system, but it's not blowing by the cap. I'm still getting a *psst* from the bleed screw, still getting air bubbles coming out of the radiator overflow, still getting boiling in\around the T-junction and\or thermostat area on the car shutdown.

Yet the coolant level didn't change. There's air continually going into the system, but, it's not blowing the radiator cap. The exterior of the expansion tank is completely dry, always. There is never any evidence of the overpressure getting by. I even drove the car hard and rapidly pulled over and in 5 seconds was looking under the hood. Nothing.

I even drove the car for those 100 miles without shutting it off once, as maybe the boiling is continually re-adding air into the system without ever getting fully bled. And so it's acting somewhat as a thermosiphon, as the steam moves elsewhere into the system and recondenses back into water. And would explain why there is still pressure in the coolant tank even after sitting all night. Enough that cracking the cap is a pretty big event and it splashes to the top of the tank, though not very much volume of air escapes.

I checked all the auto parts stores for a pressure test kit that might work, but they all either carry the same kit or don't have any adapters. Might just have to capitulate and go to a shop. Oh well.

I am going to try the shop vac trick to apply vacuum first though. May as well before I pay the $200\hr labor rates.

 

I dunno what to tell you man, it seems pretty weird. Anyway I don't think the shop vac is going to work. The way the vacuum tool works is you drain all the coolant first, then pull vacuum, and then let the vacuum pull the coolant in to fill the system. A shop vac isn't going to pull enough vacuum, nor is it going to do anything significant on a filled system.

Given shop rates it might be worth just buying a pressure test kit (or adapter), a vacuum fill kit, and a small compressor if you think you're going to use them again. Like, most modern cars benefit from a vacuum coolant fill tool, so you'll almost certainly need it again. Buy once, cry once I always say. Or if you have a buddy with a small compressor see if you can borrow it for a bit.

I've got a small 1 HP, 3 gallon compressor (2.4 SCFM @ 90 PSI) and it can do the job no problem. You can probably get away with a smaller compressor but you'll need to shut the valve a few times to let the compressor catch up.

I cannot vouch for quality, but this kit has both the vacuum tool and claims to have the Mini pressure test cap for $61:

 

Ok cool thanks for the input. I will speak with my neighbor this weekend again and get their input. I actually don't know if they have air or not, so I'm going to ask if they do as well (they're also a mechanic) so I can actually run these tests properly. And I'll make a game plan on how to continue testing when I know whether or not they have air.

In my digging I did find something to be worried about though. This car has the same bubbling noise my car does, though it's leaking\blowing out coolant from the expansion tank like mine isn't. The Car Ninja pretty quickly called it a failed head gasket. Skip to about 7 minutes.

I wonder if mine has a very very minor version of this and its only a matter of time.


It's so weird because my car isn't overheating, at all. And not over-pressurizing, at all. I dunno. The mystery deepens! I really hope it's not a head gasket. I thought the major project was behind me, and a HG is beyond my skill level probably.

 

2 months later update time. I just got the car back today, I've put about 500 miles on it since I last updated this thread. Had no issues during this time, though, I did not drive the car hard either. No full throttle, no lugging the motor, lots of highway.

Took the car to a shop, vacuum bled the system, no change. Their opinion is that it is the aux coolant pump which continues to circulate coolant for a period of time after the motor shuts off.

This is difficult for me to personally confirm because the boiling seems to stop before the pump turns off. But they're confident that's the case.

What do we think here?


I only just got back the car today so I haven't seen if the other symptoms have changed post vacuum bleed.

 

It's not the aux pump making the noise, the boiling stops while the pump is still running (put screwdriver to the engine block to my ear). Once the boiling stops the pump moves water almost silently. And there is a linear taper for the violence and loudness of said water noise that linearly tapers off over time. There's still air coming back to the system through the return line. Car sometimes has a slight "smell", a whiff of coolant if I come to a stop after driving the car hard.

The only thing is that I think I just caught it early, because it must be only leaking under boost. And the shop I brought it to said the car is kosher, even after me pushing back and asking followup questions, when it very clearly is not.

**** my life, welp I guess that's that.

 

So it's not quite clear to me: beyond the noise are you still having any issues? Did you or the mechanic ever do a full pressure test up to 29 PSI? If not you really need to do that.

Also I know the cap is new but it's entirely possible it's not working right. If any question get the cap tested or try a new cap.

If the system isn't holding rated pressure for any reason then you're very likely going to have issues with the coolant boiling. This can be the cap itself or any leak anywhere. That's why the pressure test is so important.