Likely CRD coolant leak locations?

dgeist · **Posted:** Tue Jul 01, 2014 12:00 pm

So, I've been tracking a slow leak in my cooling system for the better part of a year now (about 1-2 cups per month, or so). I've had the water pump changed with the TB service, I've checked on all the indicators of HG leakage (and found none) and am not having any other drive-ability, economy, or other issues. I've also cleaned the whole top-end of the engine fairly well to see if I can find the leak visibly, to no avail.

Where you y'all find is the most common place to find a leak (specific joints, locations if you have them)? I'd like to add some UV tracer fluid and start tearing into things over the long holiday weekend to locate the issue. Anyone know of a UV leak detection additive that's safe for HOAT? I'm not able to find any that list it explicitly.

Dan

papaindigo · **Posted:** Tue Jul 01, 2014 12:45 pm

Any coolant system UV dye should work just fine if a) you make sure you get coolant system UV dye as there are very similar products for oil and transmission fluid leaks, b) you do a pretty decent job of cleaning the engine bay before hand (apply an engine degreaser to warm not hot engine, let sit a bit, rinse don't blast off, run engine again to start the drying process), and c) after the engine is completely dry and cool put the dye into the coolant tank and just drive normally for like a week or at least say 50 miles to be sure enough of the dye escapes to be seen with a UV light.

I've dealt with this before for a very slow leak, like what you have, and the dye can be real hard to see especially if there is lots of outside light so do the light check in a garage or after sunset or nearly so. In my case the leak was a tiny one, not my CRD, at the lower radiator hose connection to the radiator due to a weak factory spring clamp.

General leak locations:
1. obvious ones - upper and lower radiator hose connections, heater hose connections, thermostat housing base and hoses running from there to water pump via the bypass/viscous heater/across top of engine to EGR
2. less obvious ones that are hard to get to - EGR cooler and associated hoses, oil cooler, heater core (if drip from HVAC drain on passenger side of firewall tastes sweet the heater core is leaking/alternatively you can loop the heater hoses so the heater core is out of the system and see if leak stops or do a pressure test of the heater core independent of the rest of the system and then of the rest of the system), coolant bottle and hoses (don't forget where the coolant level sensor goes into the tank on the bottom.

Good luck.

65Corvair · **Posted:** Tue Jul 01, 2014 1:31 pm

A difficult spot to see could bebetwwn the thermostat and cylinder head. Probably pretty rare to leak there, but the coolant would evaporate pretty quick.

Image from Squeeto's t-stat replacement walkthrough.

papaindigo · **Posted:** Tue Jul 01, 2014 2:30 pm

I sort of mentioned that spot in passing. That said the post reminds me of something I should have stressed more clearly. A small coolant leak in any of a variety of areas often results in no visible wet spot even at the point of the leak as the adjacent area is often hot enough for leaked coolant to evaporate instantly. That's why I recommend driving a bit to allow time for a deposit of UV dye from the evaporated coolant to build up enough to see.

dgeist · **Posted:** Tue Jul 01, 2014 3:19 pm

This is the only UV dye I can find that actually discusses coolant compatibility:
http://www.amazon.com/Dye-Lite-Rite-Blend-Extended-Life-Coolant/dp/B000K25U0I

It doesn't look to be available retail anywhere near me, so I'll be waiting a few days on an amazon order

Dan

racertracer · **Posted:** Tue Jul 01, 2014 3:51 pm

Have you checked the cooling system with a presssure tester? Autozone will loan the tester.

Separate the heater core by disconnecting the heater hoses leading to the core and looping them back.

Isolate the cooling system so that only the engine part is observed.

If you still have a leak then isolate the egr valve and pressurise the system.

Not sure how to isolate the oil cooler but I am sure there's a way.

Look for dried up trails/streaks of white coolant on the block.

Process of elimination.

When using the coolant pressure tester, to prevent any damage to the oil cooler or other parts, make sure not to over pressurize the system. Keep it just below 15lbs if that.

flman · **Posted:** Tue Jul 01, 2014 4:30 pm

If you changed any of your rad hoses and got one from ID parts, I found the clamps did not hold as well as the OE hose clamps, so I had to put SS screw clamps on both ends.

The upper rad hose only cost me about $12, the price was right, but the clamps were a little too soft.

dgeist · **Posted:** Wed Nov 11, 2015 6:50 pm

I eventually found the leak in question. It was actually the gasket for the thermostat right where it shows above. Now, more than a year later, I have a much worse leak (from full to "needs coolant" in two days). I still have dye in the system and it's all over everything on the bottom passenger side, but I can't see anything up higher than the motor mounts from beneath. From the top, there's a slight glow around the head seam, which I think is inconsequential. The big spots are all beneath the turbo and thermostat area on the oil cooler (it's soaked). Any good advice for getting at the cooler with minimal disassembly?

Dan

racertracer · **Posted:** Wed Nov 11, 2015 9:27 pm

Do you have any pressure under radiator cap....on top of the expansion tank?

Mike92104 · **Posted:** Thu Nov 12, 2015 3:17 am

dgeist wrote:

I eventually found the leak in question. It was actually the gasket for the thermostat right where it shows above. Now, more than a year later, I have a much worse leak (from full to "needs coolant" in two days). I still have dye in the system and it's all over everything on the bottom passenger side, but I can't see anything up higher than the motor mounts from beneath. From the top, there's a slight glow around the head seam, which I think is inconsequential. The big spots are all beneath the turbo and thermostat area on the oil cooler (it's soaked). Any good advice for getting at the cooler with minimal disassembly?

Dan

If it's leaking that fast, clean it all off, and go for a short drive. Then look for new dye. I had a bad oil leak on my nissan, and I made the mistake of driving it too long. The dye just coated the entire underside of the engine. I cleaned it up, and ran it around the block a couple of times and spotted the leak instantly.

Mike

dgeist · **Posted:** Thu Nov 12, 2015 8:35 am

Mike92104 wrote:

If it's leaking that fast, clean it all off, and go for a short drive. Then look for new dye. I had a bad oil leak on my nissan, and I made the mistake of driving it too long. The dye just coated the entire underside of the engine. I cleaned it up, and ran it around the block a couple of times and spotted the leak instantly.

Mike

Thanks, Mike. I tried the "spray it off with a good rinse of the garden hose then go for a drive" test the other day. There was still a ton of florescent splatter all around the oil cooler and below (and just a smidge on the area above it).

There doesn't appear to be pressure in the coolant, but thanks for reminding me. I'll check that again today to see if the head seal is intact. Mostly, I need a description of how the oil cooler assembly looks/works and approximately how to tackle addressing the leak (i.e. is it a bad seal or to I need to shell out the coin for a new one).

Dan

flash7210 · **Posted:** Thu Nov 12, 2015 9:39 am

dgeist wrote:

Mike92104 wrote:

If it's leaking that fast, clean it all off, and go for a short drive. Then look for new dye. I had a bad oil leak on my nissan, and I made the mistake of driving it too long. The dye just coated the entire underside of the engine. I cleaned it up, and ran it around the block a couple of times and spotted the leak instantly.

Mike

Thanks, Mike. I tried the "spray it off with a good rinse of the garden hose then go for a drive" test the other day. There was still a ton of florescent splatter all around the oil cooler and below (and just a smidge on the area above it).

There doesn't appear to be pressure in the coolant, but thanks for reminding me. I'll check that again today to see if the head seal is intact. Mostly, I need a description of how the oil cooler assembly looks/works and approximately how to tackle addressing the leak (i.e. is it a bad seal or to I need to shell out the coin for a new one).

Dan

If your oil cooler were leaking externally, you should be able to see that and would find a small puddle or dripping from it.
Most of the time the oil cooler will leak internally, mixing the coolant with the oil.

There is only one external connection to the oil cooler and it feeds coolant down to the water pump inlet. There are three seals where the cooler fits against the block. Two for oil and one for coolant.
I recommend either buying or borrowing a cooling system pressure tester. Pump it up to about 16 psi and wait for the drips to appear. And if you haven't already, remove the engine skid plate. This will make it easier to find the puddle on the ground and look up from underneath.
I would want to be absolutely sure that the oil cooler is leaking before I went through the trouble of trying to replace it.

racertracer · **Posted:** Thu Nov 12, 2015 11:37 am

Dan,

Coolant leaks in our CRDs is almost always due to high cooling system pressure. I never heard of a hose leaking due to deterioration. A hole in the radiator yes but never a hose leaking out of the blue due to a hose failure.

I have heard of high cooling system pressures being the cause of busted oil coolers... mixing oil in the coolant.

This condition is also caused by over-pressuring the cooling system with a coolant pressure tester (pump) going past the 16lbs mark.

I have experienced coolant leaks at the egr valve due to pressure in the system. The 5/8" dia. coolant hoses began to leak around the spring clamps, so I replaced the OEM spring clamps with new screw clamps and I tightened them down good. This stopped the leak at that spot but due to the pressure that was continually building up in the system, more leaks sprang up elsewhere. Chasing them was endless.

Till the known head gasket leak condition sprang up. The head gasket was blown and I didn't realize it until it manifested itself when coolant began to spew out of the coolant reservoir leaving a puddle on the pavement just under the passenger side. I eventually used an exhaust gas in coolant tester to determine that the head gasket was blown.

If you're chasing coolant leaks in the engine compartment, then I would immediately check for a head gasket leak. Because the sooner you do the less stress your system will be under. Especially if it begins to re leave pressure on the oil cooler.

If there is a weak deteriorated hose somewhere, you can test the system for leaks using a coolant pressure tester (http://www.amazon.com/Stant-12270-Cooli ... 03V9YSRW01) and Autozone has them for rent. Just remember not to over pressurize the system past 16lbs, going over that may blow the oil cooler and it has been done before.

And for your info, I wouldn't use the dye. I used it to check for leaks and the amount of time it takes to remove it from your system after the repair is complete, isn't worth it.

Hundreds of gallons of clear water via a garden hose went through the cooling system trying to get rid of the dye and it didn't work, it's still in there mixed with the new coolant.

flash7210 · **Posted:** Thu Nov 12, 2015 12:28 pm

Quote:

Coolant leaks in our CRDs is almost always due to high cooling system pressure...

This is one reason why I switched to a 7 psi radiator cap.
Ran all summer like this, using a 195F t-stat, and it never boiled and I never lost a drop of coolant.

dgeist · **Posted:** Thu Nov 12, 2015 5:54 pm

flash7210 wrote:

If your oil cooler were leaking externally, you should be able to see that and would find a small puddle or dripping from it.
Most of the time the oil cooler will leak internally, mixing the coolant with the oil.

There is only one external connection to the oil cooler and it feeds coolant down to the water pump inlet. There are three seals where the cooler fits against the block. Two for oil and one for coolant.
I recommend either buying or borrowing a cooling system pressure tester. Pump it up to about 16 psi and wait for the drips to appear. And if you haven't already, remove the engine skid plate. This will make it easier to find the puddle on the ground and look up from underneath.
I would want to be absolutely sure that the oil cooler is leaking before I went through the trouble of trying to replace it.

I did check the cap this morning after about a 30 second run cycle. There was a slight suction formed when I opened the cap (at least I'm pretty sure it was suction, hard to tell with that little).

The cooler is definitely covered in coolant. I can shine a UV light down there and see it from over the turbo area. The hose coming down from the water pump doesn't look to even be wet, though. I guess it could still be coming behind that area and I just can't see it. I'm just running out of places it COULD be sourced from that aren't already identified as not leaking.

I suppose the cooling system pressure test is the best way to go.

Dan

racertracer · **Posted:** Thu Nov 12, 2015 10:31 pm

dgeist wrote:

The cooler is definitely covered in coolant.
Dan

Was the OEM thermostat replaced recently?

dgeist · **Posted:** Fri Nov 13, 2015 12:01 am

racertracer wrote:

dgeist wrote:

The cooler is definitely covered in coolant.
Dan

Was the OEM thermostat replaced recently?

I replaced it with one of the manufactured/machined units from our friend kapalczynski in in Jan '15 after a long period of slow leaking and finding this at the gasket:

It's been rock-solid for 9 months. Just started leaking again (from somewhere) about 2/3 weeks ago.

Dan

racertracer · **Posted:** Fri Nov 13, 2015 4:42 am

It looks like the leak may be flowing down from the thermostat above.

I would remove the thermostat and inspect it for leaks. Remove the old gasket material and replace it with a new gasket. But before reinstalling the new gasket, coat it with a thin amount of RTV, both sides.

Also check that one of the bolt threads in the block is not stripped.

The shorter length bolt at the bottom of the stat is the one that can strip the threads in the block if it is over torqued.

The thermostat has 3 bolts that hold it tight to the block, 2 longer bolts and 1 shorter bolt that is approximately 3/4" long .

And then there is an even shorter bolt (7/16" long) at the top of the stat that connects it to an L shaped bracket.

A total of 4 bolts that hold the thermostat in place and all the bolts have the same size threads.

If you mistakenly swap the short 3/4" long bolt with the shorter 7/16" long bolt, the threads in the block can and will strip when torqued. There just isn't enough threads to bite into.

This is the spot of a significant coolant leak... the stripped bolt area at the thermostat.

The leak is difficult to detect at that spot because the coolant tends to escape when the cooling system builds up pressure and begins to push the thermostat away from the head. This only happens after the engine has reached operating temperature and pressure builds up in the system. The coolant begins to evaporate when it comes in contact with the hot block metal.

This is the reason why a coolant leak at that location is difficult to detect.

I experienced this condition....... I caused the problem when I incorrectly installed the thermostat, it was the cause of my head gasket failure.

I experienced 3 separate head gasket failures on two separate Jeep CRDs, and this was the cause of just one of them.

flash7210 · **Posted:** Fri Nov 13, 2015 8:44 am

dgeist wrote:

racertracer wrote:

dgeist wrote:

The cooler is definitely covered in coolant.
Dan

Was the OEM thermostat replaced recently?

I replaced it with one of the manufactured/machined units from our friend kapalczynski in in Jan '15 after a long period of slow leaking and finding this at the gasket:

It's been rock-solid for 9 months. Just started leaking again (from somewhere) about 2/3 weeks ago.

Dan

Why is there a green puddle there?
Are you using the correct type of coolant?

dgeist · **Posted:** Fri Nov 13, 2015 11:13 am

flash7210 wrote:

Why is there a green puddle there?
Are you using the correct type of coolant?

It's G-05 (or the OEM Mopar HOAT, not sure which at that point. Those two are all I use). The green is from the fluorescing additive I put in. I had the UV light pointed at it for the shot. Made identifying the location (vs other wet things and old crusty spots) very easy. It also makes the spot on my garage floor that has appeared in the last couple weeks glow in a really interesting creepy way.

I rented the pressure tester this morning at the auto parts store. 15 PSI (and hopefully more insight) here I come!

Dan

Likely CRD coolant leak locations?

Who is online