Gut Check: Attempting to Fix a Seeping Head Gasket Tomorrow by Replacing Head Bolts

[SlimKlim]

I've been dealing with a head gasket letting coolant seep to the outside of my engine. You can see it right under the rear-most cylinder where the block and head meet. I probably lose ~1in of coolant in the reservoir every time I heat cycle it. It isn't mixing with the oil and I've driven it like this for maybe 10k miles. I've just assumed that a full HG job was needed, or just a motor replacement, since mine has 184k on it.

However, I was at a C&C Meet last weekend, and ran into a really nice older guy with an '88 M3 who told me just to replace the head bolts one at a time to see if it seals it back up and squeezes a bit more life out of it. He said they can stretch/loosen over time, and when he ran a shop they replaced them at 100k miles just for good measure.

I did some googling, and it seems like some people actually have had success with it, especially in a situation like mine where it's only an external seep and isn't blowing oil into the coolant and vice versa. The OEM head bolt set was $15, so I figured what the hell, it's worth a shot. I also ordered some spare washers just in case, and a new VCG set because I can't remember how old mine is.

My plan is to replace the bolts one at a time, but I will follow the 3-stage procedure in the Bentley to properly torque them. So that means I'll remove them one at a time and install the new ones to 22ft-lbs. Then I'll tighten them, in order, another 180* in two more passes. Here are the Bentley pages I'm referencing:









I've spoken to a few different people about how to approach this, and I am NOT planning on draining the oil or coolant before I do it. My logic is that the overall torque of the head will never be below 22ft-lbs, which should be enough to hold the seal, and I don't think there's a ton of water or pressure in the head when the motor is off, otherwise I would have huge puddles in my garage after letting it sit for a week. Unfortunately I didn't think to get Mr. E30's contact information to ask him exactly how he did it.

Anyway, I'm probably going to do this first thing tomorrow morning, and just thought I'd post a thread to see if anyone had any last minute words of wisdom or advice about my plan of action, particularly about not draining the coolant beforehand. I'm trying not to get my hopes up, I think there's a pretty slim chance of it actually working, I'm just hoping that the worst case scenario is that nothing changes, and I don't end up getting involved in a full HG job. If things do go south I'm going to have to park it for a while, I'm moving in May and I'm not going to risk having to move a disassembled car. I'd rather tow it over there broken but at least in one piece.

[CMG]

You need a headgasket. The bolts are not loose the headgasket is going. Rechecking torwue on hg bolts is only something you would do after a headgasket has been replaced and using certain bolts/studs or hg that might compress.

[JDStrickland]

I'm not sure your plan will work, but allow me to throw in my $0.02 ...

The three-phase tightening is for replacing the gasket. It assumes the head is removed and reinstalled, and the idea is that you want to load the head evenly and in steps to preclude the incidence of warpage that comes from tightening a long surface down all at once.

If you are right, and the bolts are loose, then why not just double check that they are at the final torque value? Perhaps retighten using the three-phase sequence that is recommended, but just use the bolts that are already there. I think you might just find the rear bolt(s) to be loose. If your hunch is true, then all you really need to do is tighten the rear most bolts, and you might find them to only be finger tight at his point, which is why the gasket is not sealing.

If you have a box full of new bolts, then try putting your torque wrench on first to see if there are loose bolts where the leak is. If yes, then save the new bolts for when the head gasket finally fails.

- - - Updated - - -

There is zero water pressure when the engine is off. You can remove the radiator cap to be sure this is true.

You have a '95 M3, so you have the S50 engine which has a cast iron block so there is little (almost zero) chance that the bolt holes will be stripped. There's a small chance that you can snap a bolt off, but this should be a near-zero chance also. And, if you do snap a bolt off, you might snap it off removing it or it might be snapped already -- which is why there is a leak -- and you have to pull the head to replace the head gasket anyhow, so the broken bolt is the least of your trouble.

I say, put your wrench on 7, 12, and 13, and see if they are torqued to the final spec. If they are not and you can torq them, then you simply inspect to see if you solved the problem. If they are to spec, then putting new bolts in and putting them to spec isn't gonna change anything. If you want to take the extra steps of starting at 1 and hitting all of the bolts, then go for it, but the leak is at 7, 12, and 13. Mostly 13, by the way. (these numbers are the tightening sequence numbers on the diagram you posted.)

[SlimKlim]

You need a headgasket. The bolts are not loose the headgasket is going. Rechecking torwue on hg bolts is only something you would do after a headgasket has been replaced and using certain bolts/studs or hg that might compress.

I've been if the motor overheats the bolts can stretch out or loosen. My car has been overheated twice since I've owned it, never got that hot and I shut it down immediately, but it started seeping after the second time.

I'm not sure your plan will work, but allow me to throw in my $0.02 ...

The three-phase tightening is for replacing the gasket. It assumes the head is removed and reinstalled, and the idea is that you want to load the head evenly and in steps to preclude the incidence of warpage that comes from tightening a long surface down all at once.

If you are right, and the bolts are loose, then why not just double check that they are at the final torque value? Perhaps retighten using the three-phase sequence that is recommended, but just use the bolts that are already there. I think you might just find the rear bolt(s) to be loose. If your hunch is true, then all you really need to do is tighten the rear most bolts, and you might find them to only be finger tight at his point, which is why the gasket is not sealing.

If you have a box full of new bolts, then try putting your torque wrench on first to see if there are loose bolts where the leak is. If yes, then save the new bolts for when the head gasket finally fails.

- - - Updated - - -

There is zero water pressure when the engine is off. You can remove the radiator cap to be sure this is true.

You have a '95 M3, so you have the S50 engine which has a cast iron block so there is little (almost zero) chance that the bolt holes will be stripped. There's a small chance that you can snap a bolt off, but this should be a near-zero chance also. And, if you do snap a bolt off, you might snap it off removing it or it might be snapped already -- which is why there is a leak -- and you have to pull the head to replace the head gasket anyhow, so the broken bolt is the least of your trouble.

I say, put your wrench on 7, 12, and 13, and see if they are torqued to the final spec. If they are not and you can torq them, then you simply inspect to see if you solved the problem. If they are to spec, then putting new bolts in and putting them to spec isn't gonna change anything. If you want to take the extra steps of starting at 1 and hitting all of the bolts, then go for it, but the leak is at 7, 12, and 13. Mostly 13, by the way. (these numbers are the tightening sequence numbers on the diagram you posted.)

Really good points, thanks for the advice! I've been told not to retorque the original bolts because they are designed to stretch the first time you torque them, but I can check the torque of 7, 12, and 13 and see if they are loose and replace them if they are, the head bolts are only like $.90/each. I'm not sure what the final torque value is because its 22ft-lbs+90*+90*, but I can check the other cylinders for comparison.

The guy who suggested I tried this was a retired BMW shop owner and said they always replaced them on customer's cars at 100k miles because they have a tendency to loosen, and aren't reusable. Seems pretty uncommon but maybe he's onto something. It's cheap and will only take an hour or two, worth a shot!

[JDStrickland]

Personally, I've not heard that advice before about replacing head bolts, but I can't think of a reason not to if there is an issue you are trying to resolve.

Is the spec 22 ft lb, plus 90 degrees, plus another 90 degrees? (I've not done a head in a very long time, and I don't recall the specific procedure.)

If the spec is what you said, then I guess you have to check that the bolts (7, 12 & 13) are at 22, then give them each another 180-ish degrees. There's no way to know what the number to make the torque wrench click should look like. I guess, but have no way to know, that 22 + 90 should be about 50-ish ft. lb. If you don't get a click from your wrench at 50, then giving another 90 should not break anything.

Food for thought though, the head is aluminum, so it should be soft enough that the bolts should not break easily.

I like the Cheapest Thing is the Main Thing school of repair. I say, twist a bolt and hope it doesn't snap off, and if it does then it should be ready to snap anyhow. If the block was aluminum, then I'd worry about stripping it. But it's cast iron so you will be okay.

- - - Updated - - -

I would not use the same head bolts over again, but if they are loose then they were never tight so making them tight now should not be a problem.

[tufasttt]

I have the same HG coolant seeping issue with my car, you can see my post HERE. I've also read on the forums/google that some install new bolts using the 3 step torque sequence and put a band aid on the leak, however, it slowly comes back after a few thousand miles. I actually spoke with a mechanic yesterday and he recommended doing the same for a temporary fix (as the E30 guy recommended to you), replacing with new bolts one at a time, starting with the last cylinder. I personally don't think it's a good idea to re-torque the original bolts because if you snap one, then you will have to take off the head for sure, and deal with a broken bolt. Due to high heat and age, the old head bolts most likely lost their original qualities and will be more brittle and easier to snap if they were to be torqued. Plus, unless you know the actual final torque (of 22ft lb + 90deg +90deg=?ft lb), you will be just applying torque in the blind (most likely over torquing the recommendations).

If you do decide to keep the original bolts, a better idea might be just to use a regular ratchet and without twisting the bolt more, 'hand feel' how much pressure you can apply to some of the center bolts that have no leaks, and then a check around cylinder 6, etc.

In my situation, I'm collecting info on the HG DIY (since I've never done it before) and will most likely be ordering ARP bolts, which can be reused more than once. My car has only 100k miles, never overheated (that I know of) and already HG issues. Just seems to be a weak point in the e36 engines.

By the way, when I checked the coolant leak behind my engine, I didn't see any cracks in the gasket or places where the coolant would be coming from. Visually, it looks like the head lost it's clamping ability and the coolant just seeps through the top and bottom. If you end up having head bolts that are loose or snapped around cylinder 6, please keep us updated, as that might be great info for all others for a very cheap solution/cheap test before taking on the HG repair.

Good luck!

[sjpgoalie]

You're going to be torquing bolts to spec on a pinched gasket. Pound for pound, yes, but not pounded for pound whilst seated correctly. Not going to go well with new or stretched bolts. There is a reason you torque to x spec, then go forward x degrees, etc. You're putting more stress on the bolt and both sets of threads without a thicker gasket, providing tension. The less stress on the gasket, the more on the threads. That is why bolts back out with old gaskets. The threads jangle loose from one another. Plus the stretching from compression over time.

Tl;dr Your torquing things incorrectly with a pinched gasket (using new or stretched bolts) and may warp your head beyond repair.

JD, the bolts being loose does not mean they were torqued incorrectly in then first place. At all. It means they have stretched and the thread contact has altered. And when they stretch, the head lifts on a pinched gasket. If it lifts, it has vertical play and that is why the bolts end up loose. They're not contact anything correctly at that point.

This is a terrible idea. Deal with the seeping gasket or change it. Of the three solutions, these two will cost you less in the long run.

[epmedia]

What about just one table spoon of black pepper?

[SlimKlim]

Well I'll be god. dammed. I think it actually worked.

I checked the torque on most of the bolts, and 5, 7, and 13 were noticeably looser than the others, started to turn at ~30ft-lbs. I replaced those three and torqued them to 22ft-lbs, then did an additional 90*, which was as far as they were going to turn, probably because of what sjpgoalie said about the gasket already being compressed.

I put it back together and took it for a quick 10 minute test drive to get it to temp and run it through the revs a couple of times, and then let it cool down for the rest of the day. I just opened the reservoir to check the fluid and heard that *pchsss* sound I haven't heard in a loooooong time! It was still a little warm and the coolant level was just a click over the cold mark. Usually it loses about an inch of coolant in the reservoir every time you heat cycle it.

I'll report back when I put some miles on it to see how it holds up, even if it just gives me a few thousand miles I'll be a pretty happy camper, I'm only putting a couple thousand on it a year these days.

[sjpgoalie]

Well I'll be god. dammed. I think it actually worked.

I checked the torque on most of the bolts, and 5, 7, and 13 were noticeably looser than the others, started to turn at ~30ft-lbs. I replaced those three and torqued them to 22ft-lbs, then did an additional 90*, which was as far as they were going to turn, probably because of what sjpgoalie said about the gasket already being compressed.

I put it back together and took it for a quick 10 minute test drive to get it to temp and run it through the revs a couple of times, and then let it cool down for the rest of the day. I just opened the reservoir to check the fluid and heard that *pchsss* sound I haven't heard in a loooooong time! It was still a little warm and the coolant level was just a click over the cold mark. Usually it loses about an inch of coolant in the reservoir every time you heat cycle it.

I'll report back when I put some miles on it to see how it holds up, even if it just gives me a few thousand miles I'll be a pretty happy camper, I'm only putting a couple thousand on it a year these days.

Lol dude so you are using a mix of new and stretched bolts?? For real? Additionally, the new bolts that were getting tight - that head is not mounted straight on the block and you're pressing the head to the block with unequal pressure along the mating surfaces.

[snaponbob]

To the OP, if it "worked", you were very lucky. Ultimately, you are going to have to address the real issue - the HG is failing. Why? Who knows. Maybe it has been off before and the owner or mechanic cheaped out and reused the bolts or did not torque them correctly. The correct torque is all bolts to 22 ft/lb, then all bolts 90 degrees, and then all bolts to 90 degrees. This assumes a new head gasket.

[SlimKlim]

I know the full job is unavoidable, for the time and money involved i figured this was worth a shot, best case it buys me a little time, worst case it doesn't and I still need a HG.

As long as it seals I don't see the problem with this as a quick fix, my ideal solution is to build up a good S52 on the side and then drop it in the car, so if this gives me a few thousand miles of not having to top up the coolant every time I drive it then I'm a happy camper. I'm not advising anyone to do this, just sharing my own experiences.

Took it on a longer drive this morning and will check the coolant level once it cools back down.

[sjpgoalie]

As long as it seals I don't see the problem, my ideal solution is to build up a good S52 on the side and then drop it in the car, so if this gives me a few thousand miles of not having to top up the coolant every time I drive it then I'm a happy camper. I'm not telling anyone to do this, just sharing my own experiences.

Took it on a longer drive this morning and will check the coolant level once it cools back down.

Who cares about coolant. The head isn't seated correctly, man. I don't understand how you give a damn about pissing coolant away, but over torquing a few of the head bolts on purpose is absolutely fine. Not sure what you are basing that on...have you dealt with this before? This affects chain tension and stuff like that too, dude. Not much, but it does. You're ruining parts that you could reuse. But hey, as long as you don't lose coolant right? I know I sound like a dick, but it is because this is foolish and I would want someone to tell me I was making a mistake. You shouldn't be putting motors together with this sort of logic.

[SlimKlim]

Was it not seated improperly after it overheated and stretched the bolts, posing the same risks? I didn't do the full 180* after 22ft-lbs, I did my best to match the torque of the bolts on the opposite side of the motor.

Regardless, the deed is done, I can't go back and undo it. I've made my bed and now I'm going to lie in it. It's an S50 with a fuckton of miles on it and I'd rather have the 3.2 anyway, I'm hoping to score a deal on a lowsish mileage S52 to convert to OBD-I this upcoming winter and I'll sell the S50 for parts.

I haven't dealt with this before, but as I said an older retired BMW shop owner suggested I try it, and the techs at race shop I use (they specialize in Ferrari track cars but do all euro stuff, I only go there because I'm friends with the service advisor) couldn't think of a reason not to try it. Worst case I still have a broken S50 that isn't worth much. I'm not DDing the car anymore so if it breaks I'm not completely screwed, I'll just park it until I can get a new motor in it.

- - - Updated - - -

Just checked it again now that it's all the way cold, level is still right where it's supposed to be. I've done three heat cycles and maybe 20 miles so far, I'll report back after I have a chance to put some miles on it.

[Eric93se]

Are you 100% sure its the headgasket leaking and not the water pipe at the back of the engine that is known to fail??

[SlimKlim]

Well, presumably if it was the water fitting, and not the HG, I would still be losing coolant.

If that logic doesn't sit well with you, then yes, I've given the car a full visual inspection and recently replaced all the water hoses and radiator, and I'm 100% sure it *was* the headgasket leaking.

[sjpgoalie]

Was it not seated improperly after it overheated and stretched the bolts, posing the same risks? I didn't do the full 180* after 22ft-lbs, I did my best to match the torque of the bolts on the opposite side of the motor.

Regardless, the deed is done, I can't go back and undo it. I've made my bed and now I'm going to lie in it. It's an S50 with a fuckton of miles on it and I'd rather have the 3.2 anyway, I'm hoping to score a deal on a lowsish mileage S52 to convert to OBD-I this upcoming winter and I'll sell the S50 for parts.

I haven't dealt with this before, but as I said an older retired BMW shop owner suggested I try it, and the techs at race shop I use (they specialize in Ferrari track cars but do all euro stuff, I only go there because I'm friends with the service advisor) couldn't think of a reason not to try it. Worst case I still have a broken S50 that isn't worth much. I'm not DDing the car anymore so if it breaks I'm not completely screwed, I'll just park it until I can get a new motor in it.

- - - Updated - - -

Just checked it again now that it's all the way cold, level is still right where it's supposed to be. I've done three heat cycles and maybe 20 miles so far, I'll report back after I have a chance to put some miles on it.

I'll make things clearer:

YOU CAN'T GET STRETCHED BOLTS AND A FEW NEW BOLTS TO TORQUE DOWN SIMILARLY. YOU HAVE NO IDEA HOW TORQUING A THREAD BOLT WORKS. STOP. IT IS PHYSICALLY IMPOSSIBLE TO MATCH TORQUE. YOU HAVE NO IDEA WHAT THEY ARE TORQUED TO RIGHT NOW, AS MUCH AS YOU THINK YOU DO. THIS WOULD BE LIKE ATTEMPTING TO PRELOAD WORN BEARINGS. IT IS A SHOT IN THE DARK. 20LBFT ON A STRETCHED BOLT IS NOT THE SAME AS 20LBFT ON A NEW BOLT.

No, it wasn't seated correctly beforehand... But it was not being compressed downwards at various points, then not others. There is a reason there are very specific torque instructions to bolting a cylinder head down. You just bowed the head. It is going to have a crescent shape to the underside when you remove it and put a straight edge to it.

Think of how much pressure is inside each cylinder that drives a power stroke. You are actually lifting the head at points that are more loose than the three bolts you tightened. That is why it will come off warped like a bastard. Heads warp at correct torque specs as gaskets settle, burn and pinch over time...let alone mismatched torque specs.

You have stretched bolts that aren't "loose"...then you have a few new bolts at the "correct" torque spec.

You're throwing a head away. But, hey man. You no better than us. Not sure why you bothered asking for advice, if you were going to ignore anything logical. This is like replacing three of four rotors and no pad change. It is equally stupid. Sorry, but you seem pretty confident that it will be fine and I have no idea what your basing it on, but this is either super-hick or super ghetto, depending on where you live. No one in their right mind does this crap.

If it isn't leaking...just wait. You're going to be dealing with compression loss eventually. Every time you run that motor, the head is bowing more and more.

[JonTheGreek]

Good job brother on fixing the issue and doing a follow up. Honestly, I don't see anything getting messed up in the long run. Sjpgoalie, I really don't see how this can possibly effect his chain tension. He only did 3 bolts and they were all the way in the back. As long as he did them 1-by-1 there shouldn't be an issue warping the head, at all. Also there is no possible way that the chain is THAT sensitive to change. Is it sensitive to change, yes it is. But the way you are describing it is quite frankly preposterous.

[sjpgoalie]

Good job brother on fixing the issue and doing a follow up. Honestly, I don't see anything getting messed up in the long run. Sjpgoalie, I really don't see how this can possibly effect his chain tension. He only did 3 bolts and they were all the way in the back. As long as he did them 1-by-1 there shouldn't be an issue warping the head, at all. Also there is no possible way that the chain is THAT sensitive to change. Is it sensitive to change, yes it is. But the way you are describing it is quite frankly preposterous.

Right, so now the back of the head is pulled down more than it should, which stretches the front bolts even more, allowing the front of the head to lift...even more. Chain tension? Not a whole lot. Warped? He already warped it - 100%, guaranteed. OP will never post back with pictures of the bottom of this head.

Jon, read what I just posted. You don't know anything about torquing thread bolts either.

In fact, the two of you should go read about torquing threaded bolts before you destroy something.

This is flat out retarded. No one does this to an aluminum head and for very good reason. Let alone one fastened to a block that is leaps and bounds more rigid.

[Eric93se]

He's not going to leverage the front of the head by replacing the rear bolts. I wouldn't torque the rear bolts that are already there though.

SlimKlim, just go ahead and try it, your not going to blow anything up. Let us know how it goes.

[sjpgoalie]

If the front bolts don't stretch more as he cranks the back of the head down (incorrectly), then he is just warping the head. One or the other, there is no way around those two options. He very well may need the block plained. It happens with abused iron blocks as well.

Picture an oreo, as silly as it sounds. What happens to the opposite side of the end you press down on? It lifts, unless the cookie is stale. Then it warps in the latter case. Oreo or not - same exact physics. In this case, the sticky filling is acting as the front bolts/mating point and over torqued rear bolts act as your finger pressing down. Same exact concept.

[SlimKlim]

SJP, let it go man. I understand where you're coming from but you're making an assumption that the head is/was warped. I believe if the head had warped, installing new bolts would not have corrected it, and it'd still be leaking from the original spot. Installing new bolts on the rear of the motor is not going to affect anything on the opposite side. If your theory was correct and the head was warped, I'm sure I would still be leaking coolant, but I'm not. If I decide to rebuild this motor instead of swapping to a 3.2L I'll send the head out to have it rebuilt, which I would do regardless of whether I attempted this fix.

You can explain your theory seven ways to Sunday. However, the fact of this particular situation is that I had a car that was leaking coolant and now I don't anymore. I don't understand how you can sit there and tell me it didn't when I'm the one looking at the car.

I WOULD NOT SUGGEST ATTEMPTING THIS REPAIR IF YOU ARE UNCOMFORTABLE WITH IT. There are a lot of different ways a headgasket can fail. You could have a full on blow-through and be mixing coolant with your oil, or you could have a warped head. I don't believe this fix would work in either of those situations, I think if the head was warped the new bolts would've just stretched without closing the gap, and it would still be leaking from the same place. This leak started after I swapped a radiator and got distracted while bleeding the coolant, and the car sat at 2/3rds temp for a few minutes. A couple years prior to that (2010ish), I blew an idler pulley on the highway and the car went all the way into the red while I limped it off the road, which may have contributed to the problem, but the leak didn't start until the 2nd overheating in 2012. It lost about ~1" or a little less of coolant level in the reservoir every time the car heat cycled.

Especially after the success of this fix, I'm led to believe that air was trapped in the rear corner of the head the 2nd time it overheated, and the aluminum head expanded a little, slightly stretching the bolts on that corner. Which is why replacing them was able to reseal it. I think this is the only situation in which this repair could work. However, an external seep from that corner of the motor seems like a fairly common issue, so people who are also dealing with this, and willing to accept some risk, may be able to buy some time with this fix, at the very least.

Eric, I already did it, replaced the 3 bolts closest to the leak and it is holding up so far, but I've only had a chance to heat cycle it 3 times and put about 20 miles on it. However, before doing this, I would already need to top up the reservoir again.

SJP, just for you, if, I decide to rebuild this S50 instead of swapping an S52, I will come back and post photos of the head when it's off the car. It might be a couple years though, I got a replacement daily driver last year so I'm only putting a couple thousand miles on this car per year. I also promise, hand to heart, if/when the leak reappears, I'll update this thread. I have nothing to prove here man, I'm just trying to share my experiences for others that may be able to benefit from them.

If it starts leaking again or causes other issues, you can publicly shame me to your heart's desire and I'll stand here and take it on the chin. Like I said, I've got nothing to prove, I'm just trying something and sharing my experiences, whether it succeeds or fails.

[JonTheGreek]

OMG, do you people think that engines are made from glass? LOL Listen, if he used a beam typed torque wrench, which it sounds like he did, he should be able to keep the same clamping force on the head. Personally, I think he should have changed all of them.

P.S. Thank you for the recommendation on my need to further educate myself on threaded bolts, but seeing as that I am not getting a mechanical/materials engineering degree my education level stops at the plastic state of materials when they reach their yield strength. But who knows, maybe I should read a little more

[sjpgoalie]

SJP, let it go man. I understand where you're coming from but you're making an assumption that the head is/was warped. I believe if the head had warped, installing new bolts would not have corrected it, and it'd still be leaking from the original spot. Installing new bolts on the rear of the motor is not going to affect anything on the opposite side. If your theory was correct and the head was warped, I'm sure I would still be leaking coolant, but I'm not. If I decide to rebuild this motor instead of swapping to a 3.2L I'll send the head out to have it rebuilt, which I would do regardless of whether I attempted this fix.

You can explain your theory seven ways to Sunday. However, the fact of this particular situation is that I had a car that was leaking coolant and now I don't anymore. I don't understand how you can sit there and tell me it didn't when I'm the one looking at the car.

I WOULD NOT SUGGEST ATTEMPTING THIS REPAIR IF YOU ARE UNCOMFORTABLE WITH IT. There are a lot of different ways a headgasket can fail. You could have a full on blow-through and be mixing coolant with your oil, or you could have a warped head. I don't believe this fix would work in either of those situations, I think if the head was warped the new bolts would've just stretched without closing the gap, and it would still be leaking from the same place. This leak started after I swapped a radiator and got distracted while bleeding the coolant, and the car sat at 2/3rds temp for a few minutes. A couple years prior to that (2010ish), I blew an idler pulley on the highway and the car went all the way into the red while I limped it off the road, which may have contributed to the problem, but the leak didn't start until the 2nd overheating in 2012. It lost about ~1" or a little less of coolant level in the reservoir every time the car heat cycled.

Especially after the success of this fix, I'm led to believe that air was trapped in the rear corner of the head the 2nd time it overheated, and the aluminum head expanded a little, slightly stretching the bolts on that corner. Which is why replacing them was able to reseal it. I think this is the only situation in which this repair could work. However, an external seep from that corner of the motor seems like a fairly common issue, so people who are also dealing with this, and willing to accept some risk, may be able to buy some time with this fix, at the very least.

Eric, I already did it, replaced the 3 bolts closest to the leak and it is holding up so far, but I've only had a chance to heat cycle it 3 times and put about 20 miles on it. However, before doing this, I would already need to top up the reservoir again.

SJP, just for you, if, I decide to rebuild this S50 instead of swapping an S52, I will come back and post photos of the head when it's off the car. It might be a couple years though, I got a replacement daily driver last year so I'm only putting a couple thousand miles on this car per year. I also promise, hand to heart, if/when the leak reappears, I'll update this thread. I have nothing to prove here man, I'm just trying to share my experiences for others that may be able to benefit from them.

If it starts leaking again or causes other issues, you can publicly shame me to your heart's desire and I'll stand here and take it on the chin. Like I said, I've got nothing to prove, I'm just trying something and sharing my experiences, whether it succeeds or fails.

You don't understand what I am telling you. You think I am saying it is still leaking, amongst other insane claims in there. Good luck.

[SlimKlim]

Yeah I used a 3/8ths beam type, not a click type, I wanted to see where they started to turn. I flipped back and forth on whether to replace all of them, but once I got in there and realized how much looser 5,7, and 13 were than the others, decided just to replace those. Not sure that was the best decision but it's holding up so far. I'll report back once this awful winter finally blows over and I can get some more miles on it.