Coolant loss mystery

Pilchard

Member
My 1.4 diesel is losing coolant somehow.

It was serviced and topped up 11 days and 110 miles ago. Following the warning alarm today it took 0.85 litres to bring it to the max line.
The same happened 2000 before the service, and again just before the service, so I asked the garage to check. He had it on a pressure test while he did the service and he reported no pressure loss.
A bit of a mystery.

The temp-gauge is very lazy and never moves from its 60 degree start point unless on a long run... and then never hits the middle point where you'd expect it to be. There could be a connection... but maybe not. Either way... the lazy gauge is another mystery I've had in mind to look into.

There is no engine-bay staining visible from just casual inspection and, fortunately, there is no sign of oil contamination.

I did notice a little windscreen misting the other day, but assumed it was condensation from the water ingress I've been getting in the rear battery/underfloor area from the (I presume) vents behind the bumper.
Water-pump was done along with the cam-belt early last year.

So... can anyone advise on classic or well-known coolant loss locations?
 
Does seam to be a long time for trapped air but you never know. They may have just chucked the coolant in and not burped the car then. Repeat the test and let the car start to over heat a little so the fan cut in. Also repeating the the with the cap off will prevent the system from pressurising but may get any air out.
 
Can you rotate the radiator fan by hand? If not fan bearings could be seized. The fan controller on the top rear of the radiator could have damaged or corroded wiring so they need to be carefully checked and cleaned or repaired. Controller itself could also have failed.

If I remember correctly the fan is 3 speed with climate and 2 speed without. The first slow speed is only on cars we climate and the fan turns slowly with the climate on. Second speed is for the first engine hot and is 95 Degrees. The third speed is for the engine is really too hot and that is about 102 Degrees. As the engine runs cool there is a good chance that you never reach the first overheat let along the second higher overheat unless say in slow traffic on a hot day. The opposite is true slow traffic cold day can see the gauge drop from 90 to 60 very quickly.
 
Does seam to be a long time for trapped air but you never know. They may have just chucked the coolant in and not burped the car then. Repeat the test and let the car start to over heat a little so the fan cut in. Also repeating the the with the cap off will prevent the system from pressurising but may get any air out.
I'll repeat the test... both ways... as you advise.
Why the interest in the fan cutting in? Just to see if the coolant IS eventually overheating, despite the dash gauge saying not.... to see if the temp fan sensor works... to see if fan works... or some other reason?
Presumably no harm can occur from extended fast idle without the cap on, so how about a little throttle action to get the coolant really racing through and dragging through any air that may be there.
I'm assuming that that bleed point on the top hose is the highest point in the system... after the expansion tank... but are there other points (eg. Webasto?) where air could become trapped?
 
There is supposed to be a bleed point on the Webasto hose. If the fan is not cutting in you risk the chance of the engine cooking itself. Say there is air or sludge on the sensor it may read below the actual temperature and that then moves past the sensor the temperature would increase and the fan needs to cut in to cool the radiator. Ideally while running we get the actual temperature over 90 on the gauge and the fan kicks in. Not overly worried about the second speed but 95 cut in would be nice to confirm.
As for the extended fast idle obviously is you are blowing coolant out or the level is climbing up the reservoir then back off. Want to make sure there is not other issues with the water pump, head gasket or partial blockage causing a localised hot spot that could turn the coolant to steam and erupt from the top of the tank, very dangerous.
Frequently a 20-30 minute drive with the cap fitted and a stop half way through to visually check levels leaks etc is enough to force air locks out. If it is something causing a blockage well that can be fun .
I know it is the wrong time of year but do you get full heat from the cabin heater? I would also do the runs with the heater fully hot.
 
There is supposed to be a bleed point on the Webasto hose. If the fan is not cutting in you risk the chance of the engine cooking itself. Say there is air or sludge on the sensor it may read below the actual temperature and that then moves past the sensor the temperature would increase and the fan needs to cut in to cool the radiator. Ideally while running we get the actual temperature over 90 on the gauge and the fan kicks in. Not overly worried about the second speed but 95 cut in would be nice to confirm.
As for the extended fast idle obviously is you are blowing coolant out or the level is climbing up the reservoir then back off. Want to make sure there is not other issues with the water pump, head gasket or partial blockage causing a localised hot spot that could turn the coolant to steam and erupt from the top of the tank, very dangerous.
Frequently a 20-30 minute drive with the cap fitted and a stop half way through to visually check levels leaks etc is enough to force air locks out. If it is something causing a blockage well that can be fun .
I know it is the wrong time of year but do you get full heat from the cabin heater? I would also do the runs with the heater fully hot.
Much to check out... but all advice sounds pretty logical, so will do that tomorrow and get back. Will also have to have another look for that Webasto hose bleed point.
Have also ordered a Sniff Test as it's a bit of a worry.
 
Last edited:
This has been thus far a very informative thread, I've noticed my temp gauge not quite reaching 90, it has had to be on a really long run to get to full temp, that said I've not had any loss of coolant or any issues arise from it. After having my cambelts and waterpump replacing on Monday they renewed my coolant so may see a change. I do hope you get this mystery solved so good luck your in safe hands with all the very knowledgeable folks here.
 
Hi, when my 1.4 petrol had coolant loss it was from the junction of the rear coolant pipe to the thermostat housing. Had an MOT shortly after buying so was having a good look under while it was on the ramps, tester noticed some pink staining on the bottom of the pipe (to the housing) but otherwise no major issues. Then I changed the cambelt which involves some engine movment and after that it dumped a lot of coolant so that movement must have sent an iffy seal over the edge.

The main points are a) you wouldn't see this leak point unless it was on ramps and you had a very good look deep into the engine from underneath and b) it was intermittent.

From your posts this looks like a bit of a long shot, but if your under the car do have a look here.
 
This has been thus far a very informative thread, I've noticed my temp gauge not quite reaching 90, it has had to be on a really long run to get to full temp, that said I've not had any loss of coolant or any issues arise from it. After having my cambelts and waterpump replacing on Monday they renewed my coolant so may see a change. I do hope you get this mystery solved so good luck your in safe hands with all the very knowledgeable folks here.
If you have just had your water-pump replaced it might be worthwhile opening that bleed-screw on the top hose where it connects to the radiator. See my earlier photo. I found a quantity of air at the top of the radiator which was displaced by coolant when I slackened off the expansion tank cap. It's too early to say if it has made any difference at all in my case, but considering it is so easy to do, it'd be interesting hearing from you if you do it.... and/or others.

It has occurred to me that although water has very low compressibility and expands little when heated, air does both compress significantly and expands significantly when heated. Maybe my chain of logic is flawed, but the air COULD be the cause of the strangely fluctuating levels and losses through the expansion tank overflow.
Another thought... I have a feeling the water-temp sensor is located at the top near-side corner of the radiator... closely above that bleed screw. If that sensor has been measuring air temp rather than water temp, then it might explain my temp gauge reading low.

Will be running through Audifan's instructions and report.
 
Hi, when my 1.4 petrol had coolant loss it was from the junction of the rear coolant pipe to the thermostat housing. Had an MOT shortly after buying so was having a good look under while it was on the ramps, tester noticed some pink staining on the bottom of the pipe (to the housing) but otherwise no major issues. Then I changed the cambelt which involves some engine movment and after that it dumped a lot of coolant so that movement must have sent an iffy seal over the edge.

The main points are a) you wouldn't see this leak point unless it was on ramps and you had a very good look deep into the engine from underneath and b) it was intermittent.

From your posts this looks like a bit of a long shot, but if your under the car do have a look here.
Although I've looked to see if there are any leaks, the obvious dumping of coolant from the expansion tank was a no-brainer as to the source of the loss. I'd be pretty pleased if it was an easy fix like a split or rotten hose. Having said that, our drive is gravel so impossible to spot any spillage on it. I could always park the car on an old bed-sheet, but my uptight neighbours already see me as.... challengingly different.
 
No coolant sensor is located on the right side of the engine not on the radiator...

1619011250190.jpeg
 
No coolant sensor is located on the right side of the engine not on the radiator...

View attachment 79334
Thank you again Audifan. I sometimes have an infinite capacity for talking through my ill-informed arse. Even before I saw your correction I took a quick peek today and that doofer.... whatever it is... has no physical connection with the radiator. Will plot a course to that sensor using your photo, and file it's position for future reference.
 
Just got back after a quick round trip jaunt to the shops in town. 5 miles there, 20 minutes max in the shops, 5 miles back. Fairly swift but not flooring it. Tried the heater on the way out, as Audifan had recommended, and that blew hot enough considering the gauge was still only just creeping up to 70 degrees at the time. And I had also spun the radiator fan before setting off as recommended... no probs there. before setting off the expansion tank level was a bit below minimum, but when I got home it had risen to well above max and had dumped a few drops of coolant through the overflow vent.
Having released air the other day from the top of the radiator, which dropped the level from well above max to below minimum... I suspect the same thing will occur if I do it again. I'll do it tomorrow.
This sounds to me very much like exhaust gasses blowing into the cooling system. I can't see what else it can be.
Can you?
 
Having just bled the radiator... which reminds me its been a while since I did the C/H rads at home... there was air there... but not as much as before. The level dropped from the well over max level noted before, but this time down to the max line instead of below the min line.
Still a mystery. Still suspecting the worse. Now just following Audifan's last bit of advice and waiting for the Sniff-Test to arrive.
 
A head gasket seal isn’t too expensive so maybe it’s an easy repair if you like the idea of decapitating your engine block. Sounds fun right?

Wait for the sniff test to check though.


Sent from my iPhone using Tapatalk
 
Pretty conclusive I reckon.
This was also the first time I ran it with the cap off. Had I done so before, I might have guessed that the little stream of froth/bubbles running down the inside of the expansion tank from the upper tank hose (what I can only assume is a breather) strongly indicates leaky head gasket causing exhaust gases to enter cooling system.

As this certainly isn't a job for me, I now need to find a garage. Although I have no reason to doubt competence of the local garage I've only just started using, I think on balance I'd be better off with an A2 or Audi or VAG specialist.

Balls...balls... balls... balls... balls!
 

Attachments

  • Head Gasket Test.jpg
    Head Gasket Test.jpg
    877.7 KB · Views: 151
Out of interest, how long did the test fluid take to change colour?

Andy
About 5 minutes. After a few minutes nothing seemed to happen and I thought my luck was in. But then, very gently and carefully in order to not cause a surge or spurt in the coolant, I raised the revs 200 or 300rpm above tick-over for about 30 seconds, then after that noticed the gradual change. A cheap, simple, and useful test.

Anyone got any ball-park figures as to cost?
Anyone got any advice re jobs to be done at same time?

I'm really regretting buying a cheap high miler now, but it was supposed to be an experiment, and... I suppose... head gasket failure can happen to expensive cars too.
I'd thought about just scrapping it but... it's had cam-belt, pump, wheels tyres, battery... I'm in too deep now. God forbid the clutch or turbo dies in the next six months!
 
Back
Top