Hi everyone,
I'm currently battling some kind of a mystery on my car and I need your help in two ways.

First if someone with a M278 (M278.922 would be perfect but I don't think it matters that much) could do an experiment for me: let the car sit for a couple days and then turn on the contact (without starting the engine) so that the coolant circulation pump runs. After that, do you hear the pump running through some air? Is it a normal behavior?

Second, if someone has a "procedure" I can go through to find where that coolant is going, I'd be glad because so far, it's been guesswork by everybody involved.

Here is the sum up of what is happening :
- Car is a 2012 CLS500 (x218) with a M278.922 with an ETH reprog and no other mods.
- Coolant is disappearing whether the car runs or not. I'm not sure it's worse when the car runs, I'd need to experiment and conditions are not the best for that.
- I have white smoke in the morning and I'm not sure it's normal or not given that I'm surrounded by cars with engines a quarter the size of mine and with half as many cylinders, it's difficult to compare. I have very little to no smoke when the engine is hot in the afternoon but quite a lot in a morning commute.
- I don't see milkshake in the oil and the dipstick looks like it's only oil there both in cold engine and cooling down engine conditions.
- Coolant appears to be clean

What has been done so far:
+ Garage 1
- Pressurizing + visual inspection: No leak found and the circuit holds the pressure well.

+ Garage 2
- Pressurizing + visual inspection: No leak found, circuit holds pressure
- Chemical block test: no trace of exhaust fumes in the expansion tank

+ Garage 3, trustworthy mercedes garage
- Pressurizing: Gone above normal working pressure, it held properly for 15 minutes
- Visual inspection: nothing to report
- Changed coolant -> did not solve the issue
- Re did the pressure check on second visit, same result.

Garage 1 did not give any recommendation due to lack of tools.
Garage 2 recommended changing a seal behind the timing plate (can't remember which one, it was a while ago)
Garage 3 is now recommending to change tank and cap but I don't see how it would explain coolant loss without running the engine
Note that the car is the perfect example of the mechanic syndrome as in both time to the Merc dealership, she stayed there a couple days and did not loose coolant...

If it were your car, what steps would you take to identify where the coolant is going?
Is it possible for a leak to be inside the intercooler? The turbo? an EGR or other stuff?

Thanks for reading through.

 

M278 ... an interesting case of disappearing coolant.
Multiple Pressure tests returned no active leak.

How much coolant do you need to add?

At this stage it's likely your coolant tank cap. It is removed to pressure test the car thus never tested.

 

I added about 6 quarts (6 litres) in 8 month and 4000 miles.

What I don't understand is why I don't have any trace around it and that coolant disappears even when I don't run the car (about 2oz a day from what I measured).

 

That's a lot of coolant !!

There's another place the coolant can got when it's not going into engine oil sump... the tranny but your system was tested and is not leaking !!!

The tranny would not work long on mayo mix.

So I am going to leave it at replace tank CAP.

 

Yes, I wouldn't be worried if it wasn't constantly asking for me to top up.
I'm scheduled to change the tank and cap (and AC condenser) before may so I'll see if it fixes it.

 

The white smoke would have me concerned if the chemical block test had not been performed.

Strangely, a few months ago, when it was colder, I would get a low coolant warning in the morning. I topped it off and no more problems. But, where did it go to begin with? This was not my first winter with the car.

Check hoses. Out of nowhere, a permanently banded coolant hose came off of the plastic T. Bought a new hose ($120) and then realized a good hose clamp ($7) would work just as well. The parts guy had no explanation.

 

And your car ran without leaking coolant everywhere despite that hose being disconnected ?

I'll try to go and see if I can spot anything obvious but that problem has me more worried each day. I started a datasheet to try to corellate liquid loss but I can't really see any pattern so far but that it is worse with the car running.
I'll try to see if I can redo the tests but that liquid is going somewhere, just not anywhere I can see and it's worrysome.

 

Oh, no. I don't know when exactly it happened, but there was no steam or anything when I parked my car and went into a store. When I came out, I saw all the coolant on the ground. I thought, "Hmmm, that ain't good!"

 

it's indeed quite visible!
I'll collect some more data over the weekend to see if I can muster an idea of what is happening to my engine and that coolant, I wish it were on the ground or that the leak was identified at least.

 

CaliBenzDriver has the best suggestion ref reservoir cap. Check for residue on the threads. The cap is cheap to purchase; I've had the same issue on both of my E55s. New cap fixed issue both times.

 

The bad CAP is right but too simplistic. We're dealing with luxury here.

The root cause of extreme coolant pressures is though to be heat soak after engine stop.
A LOT of heat may gets trapped while running the cylinders dry at normal driving speed.
Cooling circuit may relief pressure out and leak some coolant.

Extreme heat > high pressure > relief vent > coolant loose even with a fresh cap.

There's a set of experimental techniques to normalize coolant heat in M278 engines by spraying cooling dry-lubed pistons with oil. This is unwarranted departure from factory procedures.

 

Other-side-of-the-coin thinking:
What is the longest period of time not running the car that you lost coolant?
What is the shortest period of time not running the car that you lost coolant?
Engine should be cold in both scenarios. Is there a loss/days ratio?

 

Hey, sorry for not updating for so long.
I changed the cap a few days ago and it didn't solve the issue.
I'll give the data I gathered at the bottom of the post.

For when I checked I've seen coolant continuing to be lost after 5 days (checking every day)
Hard to say but sometimes feels a bit random. I've sometimes spent 2 days without loss and then loss on day 3. But sometimes it looks like it doesn't stop disappearing.
During the last time I tried looking at it, I was somewhere around 100ml (3.5ounces) mark per day.

Here is the data I gathered lately, the miles are done between previous and current line. There was some data loss due to chair-keyboard interface between 16 and 20th of April (data points lost but average is still accurate):


I'm currently stuck having to use that car very frequently until I put the Crossfire back on the road properly so I won't be able to have her sit for more than a few days at most.

 

Well, it's now official blown head gasket(s) as there are combustion gases in the expansion tank when the engine's warm.
Minimum price to repair 8200€ (8800$) and could be more depending on what they found upon opening the engine: scoring, porous head, cracked cylinders etc.

Time to see what the seller offers but I have to say I'd rather them take the car back at that point and move towards a more reliable engine (M273 E-Class?).

 

Ok so I've got the epilogue of that story for anyone interested.
Turns out it was not the head gasket(s) in the end, coolant was slowly going in oil. So slowly in fact that it wasn't detectable (probably evaporated before it turned into milkshake).
The garage that sold me the car took it back to investigate and repair, they put the coolant circuit under pressure overnight and found a gallon of coolant in oil the next day, which ruled out the head gaskets as no coolant was in combustion chambers.
Culprit was an O-ring somewhere near the timing plate, they found it cracked in several places.

Engine appears to not drink the coolant anymore, I'm still monitoring to make sure it's perfectly fine.

 

GLAD you got this sorted at once - Thank you for the heads up!

Normally the Timing Cover seal failure kills as many engines it affects.
You got away, lucky