Newbie with mysterious coolant loss

Hi, all.

My first post here and am hoping for a little help with coolant issue that has me stumped. The truck is an 03’ Dakota. It has a 4.7L with 263K on it.

For the past few months, I seem to be losing a little coolant every few days. So far I’ve replaced the coolant filler neck and radiator cap. The radiator is only a few years old. I ran a pressure test which found a minor drip in a heater bypass hose, which I’ve fixed. It seems the drip only showed up with the pressure test because the area always dry before that. Current pressure tests show less than 1 PSI drop, after leaving it pressured for 20 minutes at 21 psi. I’m guessing that’s acceptable.

Now here is the thing. I can fill up to the neck with coolant while cold and leave the overflow at half capacity. After a day of driving, and full cool down, the overflow bottle ends up full and the neck is empty. In other words, it seems to be pushing coolant out to the overflow tank, but not sucking it back in when cool, as it’s supposed to. Overall, I’m still losing coolant somewhere though, as I end up adding here and there. I found no further leaks, none in the heater core either. If this was a head gasket issue, I guess I’d have coolant in the oil and vice versa… I see nothing, plus again, it passes pressure tests.

Maybe you guys have come across this before. It’s a minor annoyance now, but I want to figure it out before it gets worse.

Thanks.

Oil and coolant clean.




Overflow started out with half capacity, now full after a day or two


Coolant neck shows no coolant after day or two. Takes about an ouce or two to refill.


20 min pressure test start

20 min pressure test end

Any air bubbles in the radiator when running? how old is the water pump? any wetness at the seep hole.

Water pump is two years old. If I run with cap off, I may get a single air bubble every now and then. No leaks anywhere that I can see. I've spent time under truck with strong light while running and while pressure testing. Saw no leak.

first I'd solve why the coolant ain't being drawn back into the rad: bad recovery cap or leak in the hose or container (above coolant level being burped out) that is keeping vacuum from drawing the coolant back to the rad when the system cools off. IF (I ain't sure) recovery tanks are vented (very well might be as the hose comes in thru the bottom) then the hot coolant could simply be evaporating out of it as it sits.

New cap. I checked the tank and hose... no leaks. Indeed, the tank is vented. I closed off vent and applied air pressure with my mouth, finger over end of hose. It held air tight. I feel like I've covered all the bases, but no dice.

My '02 Durango 4.7 does the exact same thing. I also can't find a cause. My thoughts are possible headgasket. Although there are no signs of that. My system holds pressure in hoses well after it cools.

Radiator level should be about 1" down rather than full to the top.

I've had the truck for a few years and in the past, it was always up to the neck when I checked. The issue is relatively new. I've never run it low on coolant or oil and the temp needle is always just below half. Never overheated while in my possession. I can't see the head gasket just going bad on it's own and even if it was, pressure test would fail, I think?

on a tough case with a slow leak that ain't showing anything with normal diagnostics (like yours) I would get it hot at the end of the day & pump it up to 15 lbs with the pressure tester & let it set overnight & see if it holds. A person could have an emissions testing station stick their probe into the air space above the coolant under the rad cap & see if the machine shows any hydrocarbons present (power brake it to stress it) or they have that blok check kit with the yellow fluid & you suck some of that same air up into one of those mini antifreeze testers with the balls in it & if it turns blue then you know what comes next. (dealing with that myself with the cooling system from Hell in the stock car)

The rad cap on these engines are inline with upper hose. Rad doesn't have a cap. They use a large res mounted on shroud. You can't see rad level.

True... this is how it looks

That pipe plug on housing by clamps is to bleed engine. Should be removed while filling till a/f comes out of it.

That bleeder was tightened by the devil himself. I've never been able to crack it. I put the front end on ramps and run with the cap off till bubbles stop.

compression test may reveal whether it is a head gasket or not.

Your not losing coolant..more like pushing ot out..most likely a start of a head gasket

There is no real reason to remove that plug because rad cap itself is at same level., and thermostat is at engine lower hose. I as well, would think a head gasket may be starting to fail, especially at that mileage.

A slow coolant leak can be sneaky. My/my daughter's Neon was loosing coolant and I couldn't find it. Had to top it up every week for several months. Then as she was pulling into the garage one day, the radiator just let go. There was a small leak in the corner of the core hidden by the shroud. It evaporated as is leaked so it was not dripping and I couldn't see the white telltale residue from it evaporating. New radiator fixed it.

In your case, suggest you get it good and hot running, shut down and inspect carefully. Also, may be worth trying another cap... I know you replaced it but it sure sounds like the cap may not be doing its job.

If coolant is getting into a combustion chamber, it will tend to steam clean the plug. It would be worth pulling and inspecting the plugs... if one looks really clean good chance it is a leaking head gasket or crack.

See pics of oil and coolant in first post. No smoke and exhaust smells clean. Been driving it a few years with no overheat, so if the gasket failed, it would be for no real reason. If it was the head gasket, I think it would lose pressure during the pressure test, right?

Yes but you need to get it hot (duplicates actual conditions when it acts up) and let the tester set on it/pressurized overnight cuz it is a slow enough leak that it needs time to show itself

Must say, wow! The rate of response in this forum is incredible. I'd have to wait 5 days in the other forums for all these responses. I'm impressed and humbled. Thanks, guys.

I tried this last night. I ran the truck with the pressure tester attached. From cold, the pressure started slowly from zero and rose steadily as it heated up. I killed the engine when it hit 23 PSI because the tester's cap was not going to vent into the overflow tank and it was rising rapidly at that point. Since the pressure was from thermal expansion and not me pumping air pressure, it dropped slowly as it cooled, back to zero. Saw no leaks. I'm not ruling out head gasket, but I saw nothing unstable as I watched it. It was all very linear. I will be pulling plugs next.

Thanks again.


------------------------------

RTH, my understanding is that it shouldn't hold pressure when it cools. It should actually be a vacuum, which ends up being zeroed by return flow from the expansion tank. So, there should be no hiss when you open the cap, perfectly cold.

Hey We're the best, F the rest!. You know what I'd do now is take out the stat housing/stat/fan belt & fill the housing neck up to the brim with coolant & start it & power brake it & see if you see bubbles or the coolant starts to rise. You only have a brief window of several minutes give or take till heat expands the coolant the same way (wouldn't be bubbles tho) but if it bubbles or rises right off the bat within several minutes then that is a dead giveaway for you know what. Good point missed earlier & that is to put on a non recovery cap then pump it up to 15 then you will have a sealed system that ain't going nowhere unless there is a leak in the rad, in the hoses, in the pump (all external) or internal/inside in the head gasket (BB) or head gasket and or intake gasket (SB) & alot of those recovery systems dont work right in the first place (they dont end up drawing the coolant back to the rad) so pumping up a recovery cap system overnight ends being a valid test by default

Unfortunately, the 4.7 thermostat and housing sits at the bottom where the lower rad hose meets the motor. If I removed it, all coolant would drain out. The non recovery cap comment, I don't understand. Sorry, but this is where the newbiness shows. LOL Wouldn't the pressure tester's cap be considered a non-recovery, since it has no valves? Currently I'm pumping it up to 21 PSI, since that's the max pressure on the cap.

Yes you are correct! A brain fart on my part.

Were the best, F the rest.....LOL

You guys are a trip... I think I'm going to like this forum.

A few pics of the old battle axe that won't quit.

Well, guys.

I can definitely rule out the overflow tank and hose as having a leak or being faulty. The only thing left is head gasket, but I think I can almost rule it out too. I went through some steps from the service manual, including hooking up the tester, pumping to 16 psi and revving in place. Manual says if gasket was perforated, the revving would have caused pressure fluctuation. The needle was rock steady, and rising slowly with heat expansion. Furthermore, I smelled the exhaust and not a hint of coolant. If anything, a little water vapor with the cold weather. Heater core is dry too and there is not a single drip anywhere. Still, the coolant does not seem to be returning from bottle, so it must be losing the vacuum effect somewhere.

I'm totally baffled at this problem. W-T-F? LOL

headgaskets can leak into the oil pan and not the combustion chamber, how's the oil level look? See any milky residue at the fill cap?

Does you new cap work properly when the system cools downn, creates a vacuum and tries to suck coolant back into the system?

I'd get it hot & pump it up to 15 & let it set overnight. If it ain't leaking externally then it has to be leaking inside the eng

Cap is a week old. Nothing changed when I replaced it. I just took this....

I pumped it up hot just now. Pressure drops as it cools, which seems normal. If I pump it up cold, pressure holds for a long time. I will pump it up and leave overnight.

actually not normal. when pumping it up when hot then the pressure starts to drop not because it is cooling off but because the HG is leaking. pumping it up when cold would highly likely leak if let set up overnight (just takes longer to develop when the eng is cold). Supercuda has you covered in his last post

And no oil in coolant...

Oil and coolant look good, so that leaves the new cap. Does your tester also have a vacuum function to test the cap and make sure it'll draw in coolant under a vacuum?

If so test it, if not I would suspect the cap of not doing it's job even if it is new.

I'm sure if I leave it overnight, it will lose some pressure, but according to everything I've read, the system is not expected to hold perfect pressure with a tester attached all night. I've left it for an hour and lost less than a PSI (cold). Most say if the pressure does not drop much in 20 minutes, there is no leak. Remember, cooling overnight should create pressure drop enough to create a vacuum.

If I pump it up while hot, I'm combining thermal expansion with air pressure from the pump. As it cools, pressure will drop due reducing thermal expansion, till the expansion is no longer a factor and that would leave only the air pressure...so there would be a considerable pressure drop.

With the tester attached, I found a minute leak which I never saw under normal operation (last week). I fixed those with heightened hopes, only to be disappointed.

Nope...wish it did.

Alright how about this, pump it up stone cold to 15 & let it set overnight & if it drops to say 10 then we have lost 5 lbs of psi somewhere & now that might be 5lbs of air at the top of the system or coolant in the rest of the system or a combo of both but we ARE losing coolant & tho you and me neither want it to be a HG issue, that 5 psi loss (& some if not all is coolant cuz we are losing it) then it has to be going somewhere

Hate to leave a thread hanging so I thought I'd come back and update and to let you know you guys were right. I finally ran a chemical test on the block and it failed. Video below.

I ran a second chemical test yesterday with the truck hot, and the fluid quickly changed to yellow without me having to raise the RPMS like I did in the video.

I still don't understand what caused the issue though, as I never overheated this truck and checked fluids anally. Again, truck runs and idles perfectly. No smoke and no overheat. No oil in coolant or vice-versa and pressure tests were passed. No bubbles in coolant, etc. Obviously though, exhaust gases are in the coolant. So it might be a cracked block, cracked head or bad HG.

So there it is... "head gasket" failure. I was driving the truck daily till diagnosis. Now, I drive it only a few miles on weekends to keep it from sitting too long. Waiting until the weather gets warm to work on it. Don't want to get worse. Hate to say this, but I'm going to try using block sealer first. I think with such a mild condition it may work. Will bypass heater core when I do this. Got this truck cheap and the miles are pretty high. Don't know if digging into it and possibly finding a cracked head or block is worth the investment. Plus I'd have to change the timing chains too.

I just saw this thread. As soon as the pressure tester went to 23psi there was your answer. How can a 15lb cap seal 23psi? It can't. Thus the coolant fills the overflow bottle. After working in dealerships over 27 years I can tell you head gaskets fail on a regular basis w/o being overheated. All types of engines.
Doug

I don't understand your theory. The pressure tester's cap replaces the vented cap. The pressure tester has no vent to expansion tank. The pressure can rise till the hoses (or something) bursts by either pumping up enough air pressure, or coolant thermal expansion when running engine. BTW, the cap is 21 psi on this engine so 23 is just a tad over.

More pics of the oil and coolant. I also tried a sizzle test. I dripped the oil onto a hot skillet and there was zero sizzle, so there is no water in the oil. Coolant, you can see for yourself.

Your cap or any cap for that matter is designed to bleed at a given pressure. In your case 21 psi. If your coolant system runs more than 21 psi it will bleed coolant into the over flow until the pressure is below the target amount. You've been driving around over 21 psi, thus the coolant loss. If there is no vacuum on the engine side coolant is never drawn back into the engine. The only way coolant will back feed into the engine is if the pressure in the closed side of the system is below atmospheric.
Doug

Doug, everything you just said is correct, except

"You've been driving around over 21 psi, thus the coolant loss."

The expansion tank should be above max line then. It's not. It was actually running a bit low when I last checked it. The coolant is being burnt in the combustion chamber, I believe. The same perforation that expels the coolant, prevents the vacuum. Overall, I'm experiencing loss.

I agree the gasket leak is why it doesn't recover. In your 1st post you stated that the over flow jug filled up. So it's either it can't recover, filled more than it should have due to excessive coolant system pressure, or a combination of the two. Now time for a gasket.
Doug

I just wanted to come back and say thanks for all the help and to update you on my final results, almost a year later. Thought you guys might like to know how the fix went down.

As earlier confirmed, with a chemical tester, there was a small HG leak, or a cracked block. I think the former, given the mileage. I scoped the cylinders and it turned out that cylinder 5 had a small amount of coolant in it. The plug also looked a little cleaner than the rest.

Anyway, I used this K&W block seal, making sure I isolated the heater core first. The first time, I followed the instructions and idled with it for 20 minutes, but it didn't do squat. The leak wasn't big enough for the product to make it's way into the fault area so quickly. I tried another bottle, this time I ran it in the motor for 500 miles and that did the trick. Almost a year later and the coolant stays filled to the neck.



Thanks again!

I always thought that stuff was snake oil, that's great that it worked good for you & good info that sometimes it takes much longer in there to take care of it

what did that stuff look like out of the bottle ? was it a liquid only, or did it have some chunky, pepper looking [size wise] stuff suspended in it ?

[quote=Dodgevity] I thought I'd see pics of my ex-wife

A few pics of the old battle axe that won't quit.

(slightly) off topic. I remember reading about the pepper in the radiator trick from someone (several times) way back in the Slant six club of America, well one day I had been nursing a fairly small leak in an OE 65 dart radiator & I stopped by the quik shop one day on a lark & bought a tin of pepper & poured it in (some people sure gave me some funny looks) & I drove home & when I got there (a mile or two) the leak had stopped & stayed stopped for a year till another leak sprung in another location. recently I tried it again on another OE copper/brass radiator & mistakenly bought coarse ground pepper (had bigger pieces), I figured they would seal better! & the next day they had swelled so much that it immediately started to overheat & they were embedded securely in the tubes & it was a lost cause (don't remember if the leak stopped or not).

It was watery liquid with what looked like copper dust floating in it. It was added with water after all coolant was drained and motor flushed with water. I did that via a hose nozzle hooked up to the heater hoses and thermostat removed.

Supposedly the stuff has nano particles, but who knows. The idea is that it flows into the crack/leak and the copper melts into the crack. Since the head is the hottest area, it melts there, but won't elsewhere. Additionally, the liquid is supposed to harden like glass, once it's exposed to air and left to dry.

I didn't want any of that crap in the heater core, so I disconnected the heater hoses and looped the heater lines from engine with a single hose. Furthermore, once I was done driving the 500 miles, I drained the radiator and flushed it with a hose, to get it all out of there. Only the motor had it left in. Heater, radiator hoses and thermostat were removed. I left it to dry for a few days. That's when the stuff hardens.

Its not snake oil, but it's probably not meant for a leak as small as mine was. I'm glad I caught it early. If you search on this stuff, you'll see it worked for many, including one in a jeep forum with a very bad HG leak.

Note: There are many brands out there but do NOT use the ones that are mixed with coolant and are not meant to be drained. Those are the ones that will clog things.

PS: I recall reading somewhere that this stuff is good for a final seal after a HG job.

4.7s are notorious for slight head gasket leaks. we see this almost daily at my shop, and your symptoms are classic combustion leak. you may even be able to smell a slight whiff of exhaust in the coolant...

I wasn't able to smell it. The leak was tiny, but the chemical tester sure recognized exhaust fumes in the coolant. I pulled all the plugs and scoped each cylinder. I found out that number 5 cylinder had coolant in it. At first I used a UV light and saw the coolant glowing in that cylinder, then I stuck the scope down in and took this crappy pic. Not a high quality image, but you can see the green coolant pooled from about the center, going diagonally to the right corner of the pic.



To be absolutely sure I was seeing coolant, I rocked the truck back and forth...the movement was enough to see the liquid shaking back and forth in the cylinder.

Anyway, I'm glad that I didn't have to pull the head off. I'm pretty impressed with the 4.7L. This thing is running excellent for having 271K. I drive it daily and it doesn't burn a drop of oil and the coolant stays up to the neck now.

that is quite a success story. I will add that stuff to my list of legitemate out of the bottle fixes, the other one (the only one) I believe in is seafoam 16oz "motor treatment" in every tank of gas.

Yeah, just if you use it, be smart about it. Don't let it flow through, or dry in areas that don't need it. Meaning, disconnect heater core before using and flush radiator before drying.

The first bottle didn't work. I had even run the engine without #5 spark plug, so there would be no back pressure impeding the incoming liquid. None of that crap worked. I just had to drive with it for 500 miles with no thermostat. Of course, you'll want to do that in the summer when you don't need the heater core. I blocked off a part of the radiator with cardboard so it would overheat slightly and keep coolant pressure high, to push it through. I also disconnected the overflow hose (didn't want the stuff in overflow bottle) and ran the overflow hose into a glass bottle (wont melt), which I stuck down near the manifold. Once a day, I'd pour the overflow back into the radiator, so it would keep filled. Even after the 500 miles, it didn't seem to work completely at first. I was still losing a little coolant, albeit, much less. After about a week, it completely sealed up. Weird stuff.

2003 Dodge Dakota Quad Cab SLT
4.7L V8, Automatic (545RFE), RWD
263K

Hey, Guys. Just saying hello. Truck still runs great and losing no coolant 40K miles later. Still my daily driver but a Tesla is calling my name.

From the frying pan into the fire.