I have a '91 7.3 with an after market ATS turbo. When I first start the engine the coolant is slowly forced into the overflow reservoir. It appears to leak about 1 1/2 gallons but stops leaking when the temperature gauge gets about half way to the "Normal" range. When the engine reaches operating temperature I can add the coolant back to the radiator and run all day without noticing any problems. Milage and power seem normal. The truck mechanic at the local Ford shop I visited didn't have any answers but the person at the Schucks parts counter suggested the head gasket. I just got back from a 600 mile trip and haven't drained the pan but I can't detect any water in the oil when I pull the dip stick. The truck started easily even though the temperatures were in the low teens but runs really rough with lots of white smoke for a couple of minutes. I figured I might have some glow plugs out but searching these threads for clues to the pressurization of the cooling system led me to think the parts man may have been right about the head gasket, however, I didn't think a blown head gasket would seal up when the engine warms up. I had the engine sleeved about 45,000 miles ago and have about 155,000 miles on the truck. Thanks for any suggestions.

 

Were all 8 cylinders sleeved. What PSI is the rad cap, it should be 13 psi test it.

Next time with a cold engine try squeezing the upper hose and releasing several times, listen for a tinkling sound that would be the check ball in the upper manifold.

This is a WAG but a cracked area by an exhaust valve could cause the cooling system to presurize, when it heats up it could seal. Again if the crack was at an intake valve the coolant would enter the cylinder and smoke would be seen.

Are you losing or using any A/F, is the hose on the recovery bottle tight at the rad and clean/clear. So if the rad cap was off then it would push out coolant.

Could be Cavitation, head gasket, cracked head???..........

 

Well, when these heads crack, they crack between the two valves. They don't really cause any problems until the crack opens up to water. Then the water goes everywhere but the radiator! This is usually caused by putting cold water in an overheated engine (or just an overheated engine ).

I would think head gasket, but with a 7.3, the pinholes are always a problem.

 

Thanks for the suggestions. I'll test the cap but it holds pressure fine at operating temperatures. I had the upper radiator hose off when I replaced the thermostat but it stayed connected at the thermostat housing. Is the check ball located above the thermostat? It only appears to lose antifreeze when the engine in cold. It fills the recovery bottle to the top and then overflows about 1 1/2 gal. it takes a minute or so for the bottle to fill but if I open the cap the antifreeze gushes out and the one time I did this I lost about 2 gallons of antifreeze. Am I correct in assuming that cavitation can be ruled out since all 8 cylinders were sleeved?

 

I lost a radiator hose in the Oregon desert about 3 years ago and the engine got hot. Since then it has warmed up pretty good a couple of times climbing grades under load but has stayed well withing the operating range. On this trip it ran normal with the usual increase climbing the mountain passes with a heavy camper on the back. It has a 4:10 differential and I normally run about 2200 to 2400 rpm in 5th (70 mph) but drop a gear and run about 2400 - 2800 rpm (60 to 65 mph in 4th) when climbing. If I loose that gear and drop to 3rd I'll run 2700 to 3000 rpm for the rest of the climb. (50 to 55 mph) Am I pushing the engine too hard? It has never gotten hot on a climb unless I let it lug down too far before shifting.

 

Well the only thing I can see is when cold the air/compression intrusion is there. So to isolate the cylinder you may have to do a pressure test on the rad with the GPs out. This should allow the coolant to find the leak so that when the engine is rotated the coolant should come out the offending side/GP hole.

Or you can do this test to try and find the compression leaker......

To accurately pinpoint the cylinder leaking the full steps are as follows:

Warm up the engine
Remove all glow plugs
Remove valve covers, rocker arms, and pushrods
Using an adapter that threads into the glow plug hole, attach compressed air (100-175psi)
Wait 3 mins for each cyl being tested.
A failed cylinder will make the Coolant level slowly rise and then overflow from top of radiator. It won't blow out of the radiator, but it will be obvious.
On Power Stroke Diesels a false failure may be indicated due to a leaky bottom injector o-ring. So, before you run off in a panic, consider that the o-ring may be defective, effect repair of the o-ring in that cylinder, then re-test.
NOTES:
Valve train should be removed to allow the piston, in the cylinder to be tested, to drop to BDC when shop air pressure is applied. This insures the valves will be closed and the entire cylinder surface can be tested under pressure.
You can purchase an air pressure adapter at most tool shops (they normally are sold to adapt a compression tester to the glow plug hole)
The cylinders that are most prone to cavitation are listed in order. #8, #7, and #4. Be sure to test all cylinders as I have seen some of the other cylinders occasionally deviate from this norm.
Most coolant leaks into the cylinder will show as the glow plug tip will be wet on the bad cylinder.

If you do remove rockers etc ensure they are replaced to the same position removed from.

 

I think I would do PLC7.3's test with the air, but since your leak is only with a cold motor that seals it's self I would do it when it is cold.

I would be guessing either a seeping head gasket or cracked head.

I recently blew a head gasket between #8 cylinder and the adjoining water passage. In 15 seconds of running it built 30 plus pounds of pressure in the radiator.
When just the regular cap was on it, it blew the coolant recovery tank cap off and shot coolant 4 feet in the air.

This issue needs to be adressed real soon. If your leak gets any worse, the pressure in the radiator will force the coolant into the cylinder. When you go to start the engine it will hydraulic from the coolant in the cylinder not compressing and you may bend a rod.
At best you will not be able to start the engine.

 

I think I would do PLC7.3's test with the air, but since your leak is only with a cold motor that seals it's self I would do it when it is cold.

I would be guessing either a seeping head gasket or cracked head.

I recently blew a head gasket between #8 cylinder and the adjoining water passage. In 15 seconds of running it built 30 plus pounds of pressure in the radiator.
When just the regular cap was on it, it blew the coolant recovery tank cap off and shot coolant 4 feet in the air.

This issue needs to be adressed real soon. If your leak gets any worse, the pressure in the radiator will force the coolant into the cylinder. When you go to start the engine it will hydraulic from the coolant in the cylinder not compressing and you may bend a rod.
At best you will not be able to start the engine.

Having all eight cylinders sleeved should rule out cavitation.

When it is running rough does the white smoke smell like fuel or is it sweet like antifreeze?

The leak is putting compression in the cooling system.
Untill it seals, I am betting the rough running is antifreeze seeping into the cylinder causing the white sweet smoke and that cylinder is probably not firing or not firing correctly which is the roughness.

If it happens to fail completely and you suck a cylinder full of antifreeze with the engine running, you are looking at major damage to the engine.

 

Sorry for the double post, I hit post instead of edit.