Pressurising expansion tank - airlock, cap or head gasket?

[cheesegas]

Hi all. I changed all the coolant hoses, gear box oil cooler end caps and coolant on my boat about a year ago. All has been fine since then, including pushing the engine fairly hard on the Stort. The skin tank is big for the engine, it happily sits at maximum power for as long as required without overheating.

 

However, recently it's started to pressurise the expansion tank to the point where it leaks out the cap. Initially I thought the coolant was overfilled, but with the engine cold there's about half an inch of coolant in the bottom and it still rises up and leaks out. Before, I used to fill it about an inch short of the top and it never rose enough to leak. The engine isn't overheating and it runs fine.

 

The min/max markings on the tank don't mean a whole lot as the tank appears to be from a Ford car or van. Might there be a chance the cap is faulty, or is the rise too much to indicate a cap issue? Or might it be an airlock somewhere which has only just sprung to light after the system was drained to change the hoses?  The tank bottom hose goes to the water pump where the calorifier is usually connected (unsure if it's the outlet or inlet) and the top vent hose is tee'd off the highest point on the engine, just above the thermostat. 

 

The coolant system takes one hell of a lot of fluid and it's a mess to drain due to the location of the lower hose - is it worth draining and refilling slowly?

 

Possibly head gasket? This is also a bit more of a pain to change than usual as it's an overhead cam engine, and genuine parts are £££. Considering getting it leakdown tested first!

 

Isuzu 3KC1, skin tank cooled, air cooled exhaust manifold, no calorifier.

Edited March 11 by cheesegas

[MtB]

Does the skin tank self-vent? 

 

I had this on a boat with a manual air vent on the skin tank and opening the vent revealed that gas was accumulating in the skin tank. Venting it made the problem go away for a while but it would return after a few hours as more gas buitl up. I concluded there was a cylinder head crack as the box of spares for the engine contained several used head gaskets. 

 

I fixed it the quick and pragmatic way - an auto air vent on the skin tank. No more coolant loss!

[cheesegas]

13 hours ago, MtB said:

Does the skin tank self-vent? 

It doesn't, and there's no vent on the tank at all. I've changed the coolant twice before though, and it's been fine. The hose routing is exactly the same with the new hoses too, so I don't think that's caused the airlock directly...if that's the case.

[MtB]

1 hour ago, cheesegas said:

It doesn't, and there's no vent on the tank at all.

 

These two statements seem incompatible.

 

Either it self-vents, or it needs a manual (or automatic) air vent. Otherwise it would never fill up with coolant! 

 

Point is, if air/gas cannot escape from the skin tank, that air pocket in the skin tank will get ever bigger if combustion gas is leaking into the coolant system, and more gas in the skin tank expands more when it heats up. And this seems consistent with the symptoms you report. 

 

As DMR occasionally points out, most engine faults can be diagnosed by considering the symptoms and thinking about it. In fact should be, before taking stuff to bits. 

 

 

 

Edited March 11 by MtB
Add a bit.

[Tony Brooks]

8 hours ago, cheesegas said:

It doesn't, and there's no vent on the tank at all. I've changed the coolant twice before though, and it's been fine. The hose routing is exactly the same with the new hoses too, so I don't think that's caused the airlock directly...if that's the case.

 

The routing on the highest hose on the skin tank is vital in that case. I must have upward or downward bows hat can trap air, and it must run up hill all the way to the header tank so air and gas can rise up through the coolant, but as many skin tanks have the hot in connection at the top, once the engine is running any air or gas is trying to move against the water flow. This is why many automotive header tanks have a second connection, specifically to make venting air or gas into the tank easier.

 

By all means try a new pressure cap, they won't cost much from automotive sources, but it is not there to keep the running system under pressure and the coolant in, unless there is too much coolant (as you imply) and on a canal boat at canal speeds it is not necessary to do that. If you are sure the pipework is exactly as it has always been, the alternator belt is tight, and the skin tank is large enough, then the symptoms do suggest a probable head gasket. However, you can but test liquid that tests the coolant for exhaust gas content to help confirm such a diagnosis.

[cheesegas]

9 hours ago, MtB said:

These two statements seem incompatible.

Ah that's my bad wording - there's no vent, manual or automatic on the skin tank, but at the highest point of the skin tank inlet (which is at the top of the tank) there's a T piece which vents into the expansion tank. This makes it self venting I guess.

 

1 hour ago, Tony Brooks said:

The routing on the highest hose on the skin tank is vital in that case. I must have upward or downward bows hat can trap air, and it must run up hill all the way to the header tank so air and gas can rise up through the coolant, but as many skin tanks have the hot in connection at the top, once the engine is running any air or gas is trying to move against the water flow. This is why many automotive header tanks have a second connection, specifically to make venting air or gas into the tank easier.

 

By all means try a new pressure cap, they won't cost much from automotive sources, but it is not there to keep the running system under pressure and the coolant in, unless there is too much coolant (as you imply) and on a canal boat at canal speeds it is not necessary to do that. If you are sure the pipework is exactly as it has always been, the alternator belt is tight, and the skin tank is large enough, then the symptoms do suggest a probable head gasket. However, you can but test liquid that tests the coolant for exhaust gas content to help confirm such a diagnosis.

Yep, the highest hose on the skin tank goes up at a 20 degree ish angle with no bends, straight into the thermostat housing. Just before it gets into the thermostat housing, there's a T piece with the T facing up that vents into the expansion tank. There's no calorifier.

 

The skin tank is large enough, it's the entire height and length of the swim plate, and I've run the engine at full power for a long time on the lower Thames quite a few times with no overheating issues. The pipework is exactly as it should be, perhaps with less bends in the skin tank top hose as it used to sag between the tank and engine. Alternator belt is tight and I've had the water pump out last year to change the seals and bearings, the impeller's not corroded.

 

The temperature is the same as always. It doesn't have a temperature gauge, only a warning light, so I take it with an infrared thermometer aimed at the top hose - 85 degrees C. When running at lower revs, the thermostat cycles open and closed.

 

I didn't know that was a test for exhaust in coolant, thanks! I'll do that, easier than a leak down test. 

 

First I'll change the coolant and fill it very slowly. I've always done it in the same way however, via the tank, making sure that the vent spigot on the tank is clear. The system takes around 40 litres of coolant so it's not cheap! 

[Tony Brooks]

Before changing the coolant, which I suspect is unlikely to make any difference, it is certainly worth testing the coolant for exhaust gasses and I would consider changing the T piece to expansion tank hose for clear plastic so you can see any air/gas bubbles in it with the engine running.

 

Further thought - where is the pressure cap that is leaking, on a remote expansion bottle or on the exhaust manifold? If there is a pressure cap on the expansion tank, then any cap on the engine needs to be airtight or set at a higher pressure than the one on the expansion tank. Often the one on the engine is a plain cap with a rubber seal under the cap to seal on the filler neck. If what is the pressure cap on the engine is opening at  a pressure close to or at the expansion tank one then a new higher pressure one may stop the leaking.

Edited March 12 by Tony Brooks

[cheesegas]

Just now, Tony Brooks said:

Before changing the coolant, which I suspect is unlikely to make any difference, it is certainly worth testing the coolant for exhaust gasses and I would consider changing the T piece to expansion tank hose for clear plastic so you can see any air/gas bubbles in it with the engine running.

Will do. The only reason I was going to change the coolant is that I faintly remember rushing the filling process last time for some reason - the vent line from the T piece is only 6mm hose so it may have not have let the air out fast enough in hindsight.

 

I think have a piece of clear hose so I'll put that in after doing the test. Thanks!

[Tony Brooks]

2 minutes ago, cheesegas said:

 

I think have a piece of clear hose so I'll put that in after doing the test. Thanks!

 

Don't leave it on for long, I doubt it will be sufficiently heat and pressure resistant.

 

I add to my last reply.

[cheesegas]

8 minutes ago, Tony Brooks said:

Further thought - where is the pressure cap that is leaking, on a remote expansion bottle or on the exhaust manifold? If there is a pressure cap on the expansion tank, then any cap on the engine needs to be airtight or set at a higher pressure than the one on the expansion tank. Often the one on the engine is a plain cap with a rubber seal under the cap to seal on the filler neck. If what is the pressure cap on the engine is opening at  a pressure close to or at the expansion tank one then a new higher pressure one may stop the leaking.

The pressure cap that's leaking is the one on the remote expansion bottle.

 

The exhaust manifold is air cooled, there's no coolant tank around the stubs and the only cap on the system is on the expansion bottle. It's a pre-HMI-blue Isuzu so as far as I can tell, the only marinising that's been done to the industrial base engine is an oil change pump and a heavier flywheel, unlike the HMI Isuzus which have a water cooled exhaust manifold.

[Tony Brooks]

Just now, cheesegas said:

The pressure cap that's leaking is the one on the remote expansion bottle.

 

The exhaust manifold is air cooled, there's no coolant tank around the stubs and the only cap on the system is on the expansion bottle. It's a pre-HMI-blue Isuzu so as far as I can tell, the only marinising that's been done to the industrial base engine is an oil change pump and a heavier flywheel, unlike the HMI Isuzus which have a water cooled exhaust manifold.

 

Thanks, do the tests and decide form there.

[cheesegas]

22 hours ago, Tony Brooks said:

Thanks, do the tests and decide form there.

Testing kit arrived yesterday evening (Amazon are evil but oh so convenient!), as I'm not moored near anyone I ran the engine up this morning, then put the little demi-john airlock looking thing on the neck of the tank with the special fluid inside. A key bit of info I forgot to add was that the coolant rises a little on a cold start, and then under load it rises further.

 

Did it twice, showed positive for combustion gases, I'll get a leakdown test done too however just to be sure.

 

Looks like Engines Plus might be making me a little bit poorer with the most expensive head gasket possible!

 

Thanks @Tony Brooks for the recommendation of the combustion gas test kit!