Coolant loss
Q
We have a situation with a 2003 MG ZT-T 180+ 2.5-litre KV6 which has covered 60,000 miles, in that it is losing coolant about 200-300ml over about 100 miles. The MG is a one owner car in excellent overall condition. The situation began during an oil service – a small amount of coolant was discovered in the V of the engine, believed to be coming from the rubbish plastic thermostat. It had not previously been losing coolant.
A new thermostat was fitted with new seals, etc., with new inlet manifold gaskets and although there was no further leak into the ‘V’, a loss of coolant then started to occur.
We thought that we may not have put the inlet manifold gaskets back properly, so we removed them again and carefully put them back on with Hylomar to help them seal. We also torqued the inlet manifold bolts to the correct torque and sequence. When we refilled the cooling system although it was not done with a vacuum pump it was done in accordance with MG Rover’s recommendations – the expansion tank was raised as high as possible until coolant started to come out of the bleed screw. We have used this procedure on many previous occasions without any problems.
Unfortunately, the leak persisted. No visible sign of a leak from the engine or water pump. We bypassed the oil cooler which we understand can cause leakage, but there was no change.
We pumped up the cooling system – this showed no leak. We changed the expansion cap for a new one, but the leak persists. Upon removing the expansion tank cap when the engine is hot there is no bubbling with the coolant when the engine is revved.
The engine is running perfectly so we are completely mystified by this. The only things left to change are the head gaskets. We removed the plugs and they look perfectly fine. As there is no evidence of HGF we are reluctant to remove the heads which is a big job on these cars.
Others have indicated that they think it is trapped air in the system, which is not unusual with the KV6 engine, so we drained the cooling system again, removing the plug on the block and carefully refilled the system, though made no difference. There is no coolant on the dipstick and the level has not risen. There is a small amount of mayonnaise in the oil cap, but we have attributed this to the low mileage that the car covers.
We are now at a loss as to how to proceed. Perhaps you will consider and let us know how you think we can resolve this painful situation.
Martin Selwyn
A
You have mentioned pumping up the cooling system but have not said if you have tried doing this whilst the system is hot. If this has not been done, I would try this first. It can often be the case that a leak in the system will only occur when the cooling system is hot. You may find that a small leak from the radiator or another section of the cooling system is the problem, but if the leak is only apparent when hot, the lost coolant may be steaming off leaving no signs of the loss.
You have already checked the water pump and if leaking this does normally leave a stain either from the spindle or the drain hole in the casing. Another contender for the problem may be the coolant pressure cap but again you have renewed this already.
The thermostat housing and the plastic pipes beneath the inlet manifold are troublesome and you do not mention if you also replaced the two pipes PEP103270 and PEP101970 when you carried out the work. If these have not been renewed, then I would consider stripping-off the manifold again to check these pipes for signs of leaks. It would be more likely one of these than one of the head gaskets.
Finally, if no other leaks can be detected it must be considered that a head gasket may be failing. This will not always result in bubbles rising up through the system when revving, but before deciding to remove the cylinder-heads for testing, I would carry-out a chemical test of the cooling system to check for the presence of any gasses.
The Sealey VS0061 Combustion Leak Detector test kit can be bought from here http://bit.ly/cm8cclt and may help solve the mystery. It should be noted that the test should be carried out with a hot engine and over a period of 10-20 minutes, to fully allow any signs of exhaust gasses to be detected.