Coolant overflowing from the expansion tank.
This is not for the faint-hearted. My apologies for the length of this. I didn't want to miss out any of the observed symptoms in case they might help. I know that you may just see the description in the title and respond with "Cylinder Head Gasket" but I am interested in acquiring understanding. So here are the details:
Kia Sedona 2.9TDi 2000
The first symptom was a loss of warm air from the heater and the engine temperature was fluctuating from the middle of the gauge down to cold. So, I replaced the thermostat and that sorted out the fluctuation and cold temperatures. I filled it with water and that gave me heat inside the car again. But then, I was losing lots of water. I would travel for half an hour and lose about 1 litre of water. It was very difficult to locate the source of the "leak" but eventually I tied a little plastic bag over the overflow pipe coming from the expansion tank and I found that it filled with coolant. Initially, when the engine is cool nothing goes into the bag but it does blow up like a balloon. It only starts to fill with coolant after the engine has come up to the normal operating temperature.
So, I felt that this must be a blown cylinder head gasket, or a cracked head (woe betide).
Now, I actually have some (limited) repair insurance with the AA, so I called out the AA and the patrolman immediately suggested that it was probably a head gasket. I then had to take it to a garage to get a definitive diagnosis so that the AA could authorise a repair. They tested the air-tightness of the water system: they removed the radiator (not actually on the radiator) cap and used a device to pump up the water system to about 15psi and then the pressure gauge on the device showed it losing pressure very quickly. They found that the pipe (about an inch in diameter) that goes from the (radiator) pressure cap housing (containing the thermostat) off towards the heater matrix, had sprung a small leak which was resulting in some hissing. They tightened the jubilee clip around the end of that pipe where it met the pressure cap housing. When they tested again, the coolant system was holding steady at 15psi with no leakage.
After that, I didn't manage to get the coolant to overflow from the expansion tank. However, there were bubbles coming through into the expansion tank, so they tested these gases by removing the pressure cap again and putting a device into that housing that allowed these bubbles to pass up through a purple liquid. This was supposed to detect any carbon monoxide by turning yellow. Despite having gas bubble through it for ages, it did not detect anything. So they concluded that the cyclinder head gasket must still be intact. So presumably we must have been looking at air pockets gradually passing out. When we revved the engine the bubbles slowed down. So, my only problem with this conclusion is that I have never seen coolant coming out of the expansion tank while the car is stationary. This gas test was done while the car was stationary. But would we not expect some of these trapped gases to have contained some exhaust gases?
So, I bombed up and down the motorway for about half an hour and I still didn't get any further coolant overflowing. That really surprised me.
So, why would that jubilee clip have resolved that problem. I developed a theory whereby, the coolant would expand when hot into the expansion tank and then when it cooled, it would suck that coolant back into the system, but if it had an air leak then maybe it would just fill with air instead, so that when it heated up next time it would push coolant out of the expansion tank because the volume of coolant and air would now be larger.
Unfortunately, I have now come to use it again and this time coolant has overflowed out of the expansion tank again.
Finally, here is my general reasoning. I have had the experience of topping up my coolant at the top of the engine where the pressure cap is and making sure that my expansion tank is below the maximum line, and then going for a drive and finding that it had overflowed. Now, the only reason I can see for it overflowing is if gas is entering the system. The expansion tank should have enough space for the coolant. Considering the coolant should be under pressure, I don't think that the additional gas could be air because that is at lower pressure, so I can't see how that would get into the pressurised system while it is running. So I can only see high pressure gas entering from the cylinders.
Can anybody help me with their greater understanding?
Last edited by nowthenkia; 11-18-2012 at 06:18 PM.