Mystery of the Vanishing Coolant

[Alpine 1789]

My SV regularly looses coolant. But, here is the mystery: I really don't see evidence if it spewing out anywhere and there is no water in the oil (nor oil in the water). Plus, the radiator holds pressure for weeks; removing the cap almost always results in pressure escaping. I thought maybe it was coming out of the overflow tube, so I rigged a temporary overflow tank up. Not only isn't there more fluid in it than when I started, there is less, which I assume means it has been sucked back into the system.

(Just for completeness for those who remember my earlier post about a leak in the upper outlet, soldering it didn't help. I will probably end up having a spare one boiled out and tested, but for now I just sealed it up with JB Weld, which has done the trick as far as stopping that particular leak.)

For the most part, this isn't too big a problem and I rarely drive far enough that I have to deal with it, as long as I check the fluid level frequently. However, I am trying to get the car ready for the Invasion and want to take care of this. Unless I am missing something, a new rad and hoses won't be the solution, since the current one does hold pressure. Any thoughts on where to look next?

Thanks.

 

[Bill Blue]

Jim, a slow leak around the water pump shaft can be devilishly hard to detect. They sometimes leak only under certain conditions and if the engine is running, the coolant is spread hither and yon, leaving not a drip or a drop to be found. Does the inside of the water pump pulley feel damp or oily?

Bill

 

[Alpine 1789]

Just checked: dry as a bone.

 

[Alpine 1789]

Just thinking: am I making too much of the fact that the rad holds pressure? If I had a slow leak in the lower hose or in the lower part of the radiator, wouldn't the cap still hold pressure? It would have to be somehow intermittent or only happen when I was driving, else I'd have puddles under the car in the garage, but would this be possible?

 

[Nickodell]

The same thing happened to me years ago with Matilda. A friend suggested lifting the head, and sure enough there was a tiny leak in the gasket between the water passage and #3 (as I recall) cylinder. It was just big enough to allow a few ccs of coolant to be drawn in, either on each inlet stroke or when the car was standing with the engine off. But it wasn't big enough to pressurize the system or show an anomoly on a cylinder compression test.

I never found any water in the oil, either. Seems the small amount finding its way in the cylinder was either just blown out the exhaust or turned into steam, and a few ccs each day can amount to a quart or so in a month.

 

[Alpine 1789]

Thanks Nick. I didn't know it was possible to have a head gasket leak without seeing oil in the water or vice versa. Still, I'd like to eliminate other potential causes before pulling the head. It is also worth noting that a few years ago (this has been going on for a long time ) Bill Blue brought a head gasket leak tester to the Invasion. Bill can probably fill in the details on this, but it replaced the radiator cap and tested for minute traces of exhaust in the radiator. We tested mine and there were no traces. I don't know if that eliminates the kind of leak you had, but it is another data point.

 

[RootesRooter]

I hadn't heard of a leak detector that sniffs for exhaust gases before. The one I use, which also attaches in place of the radiator cap, simply tests the pressure. You pump the handle til the pressure guage reads about 7lbs, then sit back and watch the needle. If the pressure doesn't drop over the course of a minute (usually 10 seconds is enough), then your head gasket is fine.

If the pressure does go down, then you inspect the outside of the block, hoses etc for seeping. Still can't find any leaks? Uh-oh. It's pull-the-head time.

Any chance the heater core or its hose connections are leaking? (Do run enough anti-freeze that you could smell it?) Also, slow seeping from a bad gasket between the engine and thermostat housing can be difficult to detect. The housing itself can become distorted, leading to gasket failure.

 

[RootesRacer]

Thanks Nick. I didn't know it was possible to have a head gasket leak without seeing oil in the water or vice versa. Still, I'd like to eliminate other potential causes before pulling the head. It is also worth noting that a few years ago (this has been going on for a long time ) Bill Blue brought a head gasket leak tester to the Invasion. Bill can probably fill in the details on this, but it replaced the radiator cap and tested for minute traces of exhaust in the radiator. We tested mine and there were no traces. I don't know if that eliminates the kind of leak you had, but it is another data point.

I have one of these and they do work, though I havent yet used one to affirm an HG failure (you can blow through the test liquid and the CO2 in your breath with change the color).

I bought mine on ebay.

 

[George Farrell]

Jim

I think it was me you were thinking of with the Lisle testor at the Invasion. I have used it a few times and it does seem to work well. No false positives and as far as I know no false negatives. I can't remember, right now if the blue fuluid changes to yellow or yellow to blue if there any exhaust gasses in the coolant.

George

 

[alpine_64]

Jim,

Back in 2003 i was having a similar issue with the SII when i was driving it as a daily driver. I had a return system, i couldnt find any leaks.. it was just slowly loose coolant without a trace.. This went on for several months. In the end it turned out to be what nick explained, a slow small leak in the HG between the water passage and the cylinder. The only thing i can suggest is also check to see if the exhaust feels a little cooler than you would expect.. though this is unlikely to be that noticeable till it gets worse.

You could put that HG treatment in the cooling system, the ones that plug leaks (though i always figure they make a mess of everything else in there surely?)

This is what lead me to making the custom HG and doign the head mods to eliminate the cresent recess water passages.

 

[V6 JOSE]

Hi Jim,

I would suggest pulling the spark plugs to see if one is cleaner than the others, which would indicate that there is a leak into that cylinder.

Jose

 

[bernd_st]

Check all the core plugs for proper sealing . Especially the rearmost small one just above the clutch. Very common failure failure that it leaks...

 

[Alpine 1789]

So, after a nice drive today, I have a couple of new data points that might help diagnose the problem. It doesn't do anything for me, but hopefully it will help someone here. This will take a bit of 'splaining, so please bear with me.

I drove the car to the golf course today, about a 30 minute highway drive. I checked under the hood when I arrived and noticed two things. First, there was some moisture in the area around the hood hinges. Not enough to really explain the problem, but I wondered if it was possible that my radiator cap was the problem, especially given problem number two: there was pressure escaping from the cap. I could easily hear it. I've heard that before, although it is not always there. It went away when I pressed on top of the cap. I then went away and played 9 holes, about two hours, letting everything cool down.

When I came back I decided to tighten up the cap a bit. It never felt right to me and can easily be turned all the way around; meaning that the filler neck doesn't stop the cap when it should if you turn too hard. All I did was use a vice grip to slightly lower the lower edge of the cap's flange. Hopefully, that makes sense. The flange on the cap is slightly angled so the leading edge can slip under the flange on the radiator while the trailing edge catches at the right spot. I simply increased the angle by squeezing the trailing edge just a hair. I then drove home, this time taking country roads and engaging in some pretty spirited (and immensely enjoyable) country driving.

About 15 minutes into my drive I got a very strong coolant smell. (Note: there is normally a slight smell, one of the things that makes me think the problem is radiator or hoses, not engine.) Since there was no good place to stop and the engine temp was normal, I kept going until I found a good place to pull over. When I did, I discovered that the engine compartment was pretty wet and there was a fair amount of coolant coming out of the front of the engine. I was unable to find a specific location for the leak, but slowly removed the cap and was surprised to find that the level in the radiator was not that different from when I'd left the golf course.

I remembered that I had a spare cap in the trunk and decided to put that on for the remainder of my drive. That was when I noticed that the two caps were pretty different. The one on the car was the correct 7lbs, while the spare was 9lbs. (I don't remember when I used it, but do remember putting the current one on in an attempt to stop the leakage.) Also, the spare was a 9lb cap and shorter than the current one. The one I have been using is about 1 1/4" from the top of the cap to the bottom of the rubber gasket, while the spare was a half inch less.

Despite the difference in the two caps, it felt like the spare was still sealing so I decided to give it a try. I topped off the radiator and drove the rest of the way home - about 15 minutes on the highway - without incident and without overheating. I checked the cap when I arrived and it was also leaking pressure, more so than the longer 7 lb cap. And, unlike the 7 lb cap, the leaking didn't stop when I pressed down on the cap. But, I don't think I'd lost any more coolant on the rest of the drive, although it is hard to be 100% certain, as the engine compartment was still pretty wet from the first incident.

So, again, I apologize for the long-winded explanation, but does any of this suggest a further diagnosis to anyone?

 

[RootesRacer]

The difference in length suggests a mix of series caps.

IIRC the earlier ones are longer, and there are even differences between the SII models that had separate releif valves vs. the ones that had only the cap.

You have one of two possibilities, either condition seems to show you are going over pressure to whatever the cap is actually relieving at.

Possibility 1, coolant temp and its requisite vapor pressure is greater than your cap is rated.
You can determine the vapor pressure of the coolant by getting a coolant/water % vs temperature (vapor pressure) chart.
Get infrared thermo and measure the tank temp, which is a good indicator of coolant temp inside.
Get a radiator test setup (HF sells an affordable one) and measure the actual coolant pressure (and temp) and see if the pressure correlates with temp.
Additionally, some repair shops have test kits to measure rad cap releif pressures.

Possibility 2, you have a blown head gasket and have exhaust gasses pressurizing the cooling system. Taking the procedure from above, if the vapor pressure and temp dont correlate and the pressure rises and rises, then likely its a failed fire ring near one of the crescent cooling ports.
You can confirm this condition with a combustion gas leak tester from Lisle.

I went through this last year with my alpine and ended up convincing myself to install a higher pressure rad cap from a "new" application which required me to change out the filler to match the new cap.
I went to 9 PSI and now it holds very nicely.
The stock 7 PSI cap showed leaking at only 4 psi, which is below the vapor pressure of 50/50 glycol at 200 deg F.


So there you have it, an equally verbose response.

 

[tony perrett]

You could try mixing a concentrated water-based dye with the coolant and pressurising the system for several minutes. Any external leak should then be evident.

 

[RootesRacer]

You could try mixing a concentrated water-based dye with the coolant and pressurising the system for several minutes. Any external leak should then be evident.

No dye should be needed. Automotive coolant is florescent with a UV lamp.
You would just need to clean the area where the coolant may have leaked well before the test.

 

[tony perrett]

No dye should be needed. Automotive coolant is florescent with a UV lamp.
You would just need to clean the area where the coolant may have leaked well before the test.

Not everyone owns a UV lamp.

 

[RootesRich]

Not everyone owns a UV lamp.

True, but here in the States you can buy a UV bulb at Home Depot for $4 that will work in an inspection light.

 

[Nickodell]

Cool for 60s - theme parties, too.

 

[RootesRich]

Cool for 60s - theme parties, too.

Jim,
Perhaps you'd have more success in finding the coolant leak with the blacklight if you played "Age of Aquarius/ let the Sunshine In" in the background...