Pipe fitting sealant question

[MVCalypso]

Hi TF denizens,

I'm experiencing a very small diesel weep from a pipe fitting assembly (as in maybe a drop a week) & I'm looking for advice on how to stop it.

Here is the pipe fitting sequence (all are NPT):

Al 1" female port welded into an Al tank,
1" SS close nipple inserted into the tank fitting,
SS shut off ball valve,
SS adapter to fuel hose.
​

The joints were assembled with Rector Seal #5.

I think the weeping is between the AL female and SS close nipple. It's hard to tell, as after a week the bottom of the assembly will be damp to the touch making it hard to localize the source joint.

Side bar: I'm told that SS is used as "close to Al" to avoid (minimize) galvanic action at the tank fitting. Frankly, I don't know why that matters in this case as the joint is not in an electrolyte (fluid is diesel). But I understand it's the "standard practice" and the above is the part sequence I'm dealing with. As there are no Al body ball valves, I think there has to be an Al to SS transition somewhere.

Goal: Understand why this is happening & stop it.

Theory 1: I think NPT threads seal (at least partially) as the result of deformation of the thread crests. I perceive the SS threads as much harder than the mating Al threads. I'm wondering if the SS male thread is simply re-threading the Al female fitting instead of deforming and making a seal.

If this is happening, I'm loath to tighten the close nipple into the Al fitting, as it won't make the joint tighter, and the valve handle could end up in a position (as the valve rotates with the nipple) where is would be hard or impossible to access the handle, or the handle could not rotate a full 90 degrees for on/off.

If this is a valid concern? If so, what I could I do differently...?

Theory 2: The rector seal stays soft forever. I'm not a pipe fitter, but I'm having trouble believing in a "soft sealant" as being very good at sealing. I observe it hasn't "sealed" the joint so far. Diesel is getting past the rector seal goop.

Is there a better goop to use for this?

Any words of wisdom appreciated!

 

[DennisB1]

I have had difficulty with this the Rector seal on occasions. I switched to a different product which worked well on stopping the small seeps on pipe threads. I have been using Permatex high temp thread sealant which requires a Permatex surface prep spray as a cleaner and degreaser. Start by thread cleaning with a wire brush, then the spray then the sealant following the directions. Then engage hand tight plus couple of turns. Don’t over torque. You will need to empty the tank to below the thread fitting. I bought the sealant and spray at a local auto parts store. Good luck

 

[DDW]

The thread crests do not touch in a properly cut NPT thread. That leaves a gap between the thread crest and corresponding valley which leaks and must be sealed. There is a variation of the NPT spec, NPTF (for "Fluid") in which the thread crest is fully formed and intended to seal fluids when assembled dry. Almost no routinely used marine plumbing is NPTF.

I'd use Loctite 567. It is a low strength anaerobic methacrylate sealant that will cure and harden, but still allow removal. It is rated for diesel. The surfaces need to be clean, that might be a problem in your application. Also it is best to let it cure for 24 hours before using it. I've used this successfully for gasoline fittings where nothing else worked. I now use it routinely for gas and diesel.

 

[Datenight]

I'd use Loctite 567. It is a low strength anaerobic methacrylate sealant that will cure and harden, but still allow removal. It is rated for diesel. The surfaces need to be clean, that might be a problem in your application. Also it is best to let it cure for 24 hours before using it. I've used this successfully for gasoline fittings where nothing else worked. I now use it routinely for gas and diesel.

Loctite 745 was also recommended to me by the Research Lab in Rock Hill, CT.

Rob

 

[mvweebles]

Similar use for me - I asked Steve D'Antonio if he had a recommendation. He said the yard he worked at used "Leak Lock" for years without issue.

Good luck - total PITA.

Peter

 

[Aguyinak]

Stainless steel pipe threads are notoriously difficult to properly seal in a fuel system. Unfortunately Rector Seal #5 is about the worst type of pipe dope to use in this application, in fact in my company we stopped using it over 30 years ago. Two excellent products have already been suggested, we have found both the Loctite 567 and the Permatex to work very well with new, clean piping. We also use two other products that I believe will work better in your situation, AST-Seal and Leak Lock, with Leak Lock being the best for old pipe fittings. The only downside to Leak Lock is that it takes time to cure completely.

https://www.antiseize.com/ast-seal-hyd-hydraulic-anaerobic-thread-sealants

https://www.amazon.com/Highside-Chemicals-10004-Brush-Top-Plastic/dp/B00009W5HX?th=1

 

[boatnbump]

I had a problem with a tank valve in a Jupiter that I have. I called the company that makes the tanks for Jupiter. Best Fab. They told me blue Lock tight only. I used it and will never use anything else. The interesting thing is when you take a fitting back part and see how it looks on the threads you know it was sealed.

 

[Nick F]

As you can see, there are almost as many recommendations as there are replies!

So here is one more:

For joints which are upstream of a filter, I use Petro-tape. It was recommended to me by a fabricator of aluminum tanks and has consistently worked for me.

 

[MVCalypso]

Similar use for me - I asked Steve D'Antonio if he had a recommendation. He said the yard he worked at used "Leak Lock" for years without issue.

Good luck - total PITA.

Peter

Thanks for the tip. Do you happen to know if joints done with Leak Lock can be taken apart?
The stuff seems to be resin based with fibers which sounds somewhat like FRP for threads.

 

[twistedtree]

Thanks for the tip. Do you happen to know if joints done with Leak Lock can be taken apart?
The stuff seems to be resin based with fibers which sounds somewhat like FRP for threads.

It can definitely be taken apart.

 

[MVCalypso]

The thread crests do not touch in a properly cut NPT thread. That leaves a gap between the thread crest and corresponding valley which leaks and must be sealed. There is a variation of the NPT spec, NPTF (for "Fluid") in which the thread crest is fully formed and intended to seal fluids when assembled dry. Almost no routinely used marine plumbing is NPTF.

Thanks for that info. I looked NPTF threads up in machinery's handbook to learn more. It's just amazing how much info is packed into that one book.

I'd use Loctite 567. It is a low strength anaerobic methacrylate sealant that will cure and harden, but still allow removal. It is rated for diesel. The surfaces need to be clean, that might be a problem in your application. Also it is best to let it cure for 24 hours before using it. I've used this successfully for gasoline fittings where nothing else worked. I now use it routinely for gas and diesel.

This looks promising. Being removable is a feature I like (I've avoided considering some epoxy type sealers) just in case this ever needs to come apart down the road.

I think I can get the threads clean- I have a fuel polish system that lets me pump from tank to tank, so I can empty the tank, disassemble, clean the threads and reassemble. Cure time is OK as the task doesn't have to fit into a single day.

 

[MVCalypso]

DennisB1: Thanks, I'm off to read the permatex data sheets.

 

[KnotYet]

Obviously double check that the leak isn't from the ball valve packing which can
be tightened in some if not most ball valves.
Pipe fitting quality varies with manufacturers so I always try to buy domestic made.

 

[twistedtree]

Obviously double check that the leak isn't from the ball valve packing which can
be tightened in some if not most ball valves.
Pipe fitting quality varies with manufacturers so I always try to buy domestic made.

Good point about the fittings themselves. I have encountered more than my fair share of threaded parts with all sorts of defects. The worst seems to be when there is a casting seam that the threads haven’t cut fully through, creating a straight line of defect right through the joint.

 

[HopCar]

I only keep one sealant on my boat or home. It works with any liquid or gas I need it too, Leak Lock.
https://highsidechem.com/leaklock/

 

[DDW]

Good point about the fittings themselves. I have encountered more than my fair share of threaded parts with all sorts of defects. The worst seems to be when there is a casting seam that the threads haven’t cut fully through, creating a straight line of defect right through the joint.

A "feature" of die cut NPT threads is that there is always a discontinuity in the thread form. The thread has a specified taper to the major and minor diameter, which is why it tightens up as you turn it. A tap used to cut an internal thread (or a die used for external) has a matching taper, with each successive tooth set cutting a little deeper than the last. Therefore, when you stop the tap or die and back it off, the point at which you stopped has a step in the thread form a few thousandth of an inch high, on every thread at every tooth. This can look like a parting line in a cast part, though usually more subtle and usually four such spaced at 90 degrees around the diameter (matching the teeth on the cutting tool).

The only way to cut NPT threads (that I know of) that does not have this step is to thread mill them which can form a perfect thread. Or use a thread forming tool that deforms the material to the required shape (these are very difficult to use and set up correctly). (Or mold them in plastic, but then you may have a parting line for the mold). Almost everything you run into on a boat is tap or die cut.

 

[Southern Boater]

I recently repaired multiple, long term diesel fuel leaks at the threads of the inspection covers of the steel fuel tanks. The fuel level was two thirds up the inspection covers and the lower four threads on each 12” plate were oozing consistently.
I was advised to use Loxeal 58-11, but was sceptical any sealant could actually stop flowing diesel.
I removed the nuts, one at a time, a quick wipe with a clean rag, applied the Loxeal at the plate end of the thread, into the oozing diesel, and re-tightened. It stopped the flow immediately and so far, four weeks later, still dry. I’m certainly impressed with this products performance.

 

[twistedtree]

A "feature" of die cut NPT threads is that there is always a discontinuity in the thread form. The thread has a specified taper to the major and minor diameter, which is why it tightens up as you turn it. A tap used to cut an internal thread (or a die used for external) has a matching taper, with each successive tooth set cutting a little deeper than the last. Therefore, when you stop the tap or die and back it off, the point at which you stopped has a step in the thread form a few thousandth of an inch high, on every thread at every tooth. This can look like a parting line in a cast part, though usually more subtle and usually four such spaced at 90 degrees around the diameter (matching the teeth on the cutting tool).

The only way to cut NPT threads (that I know of) that does not have this step is to thread mill them which can form a perfect thread. Or use a thread forming tool that deforms the material to the required shape (these are very difficult to use and set up correctly). (Or mold them in plastic, but then you may have a parting line for the mold). Almost everything you run into on a boat is tap or die cut.

Thanks, very helpful. My "machinist" experience is limited to one summer job making custom parts on a lathe & milling machine. It was a lot of fun and I learned a ton, but surely just scratched the surface.


Let me try to find a picture because I don't think I'm explaining it correctly. The casting seam I'm trying to describe runs perpendicular to the threads, and parallel to the flow through the fitting. I assume the threads were then cut using a die (this was a male fitting). The issue was that you could still see the casting seam after the threads were cut, leaving a thread defect each time the thread crossed the seam. If the threads had been cut a little bit deeper the seam would have been cleaned up, but then of course the threads would have been out of spec. I guess another way to describe it is that the surface defect left behind by the casting mold was deeper than the threads, so left a defect in each thread.


Does that make more sense?

 

[Snapdragon III]

All of the above advice is good. Another thing that helps a lot, especially on stainless to stainless connections is to put a little bit of the pipe dope on the inside threads of the female fitting. But in this case make sure the pipe dope is not going to dissolve in the diesel and mess up your engine.

 

[MVCalypso]

I want to thank everyone for their inputs, it's been very helpful!

The use of Loxeal to stop a leak on an inspection plate without needing to empty the tank was impressive. We wanted to try Loxeal 58-11 but it appears hard to find a vendor in the US (where I am). Lots of suppliers in Europe, Aus & NZ. Price with shipping and the lead time required, resulted in us going to another choice. Another thought was that a couple of videos said that Loxeal required heat to disassemble things.

The Loctite 567 appeared to be another good choice but the tech data sheets really stressed having very clean (or new) fittings. Leak Lock's instructions appeared a tad more forgiving.

I've ordered Leak Lock and we're going to dissemble the fittings, clean them and redo the assembly with leak lock.

It's a case of pick your poison and give it a try. I'll report back after we get the job done.

 

[JDCAVE]

I’m not sure where you are at with this, but I had to redo my sight gauges on the 4 fuel tanks as the surveyor didn’t like them. You know the whole insurance thingy. Transport Canada would “accept” self closing ball valves. I had petcocks. So I replaced with self closing ball valves.

After discussion with machinists and pipe fitters, I used a combination of Teflon pipe dope and Teflon tape on all joints. Pipe dope on the male threads first and Teflon second. Job done 5-6 yers ago. No leaks.

Jim

 

[JDCAVE]

See attached.

 

[DDW]

Thanks, very helpful. My "machinist" experience is limited to one summer job making custom parts on a lathe & milling machine. It was a lot of fun and I learned a ton, but surely just scratched the surface.


Let me try to find a picture because I don't think I'm explaining it correctly. The casting seam I'm trying to describe runs perpendicular to the threads, and parallel to the flow through the fitting. I assume the threads were then cut using a die (this was a male fitting). The issue was that you could still see the casting seam after the threads were cut, leaving a thread defect each time the thread crossed the seam. If the threads had been cut a little bit deeper the seam would have been cleaned up, but then of course the threads would have been out of spec. I guess another way to describe it is that the surface defect left behind by the casting mold was deeper than the threads, so left a defect in each thread.


Does that make more sense?

Of course that could happen, you hope not in a fitting made with any quality. Was it a cast fitting or a pipe nipple? A lot of pipe is seamed, rolled from plate and then welded to form the tube. That seam can be imperfect, but also is likely to create a HAZ, which will cut differently than the unaffected metal. When I drill seamed steel tube or pipe, I try not to drill though the seam, as it is noticeably harder than the rest of it (not true of SS or alum which don't harden with heat). Fittings are generally cast, and steel is usually sand cast so there will be a parting line (also, it is an imprecise process). Stainless fittings are often investment cast, you could still see a parting line if their investment mold was crap, but otherwise it should be cleaner and tighter controlled.

I should have added to my prior post that an external or internal tapered thread can be correctly cut by single point threading on a lathe. Too slow for most production fittings, only practical for parts than can be fed through the stock with a bar puller (or small runs).

 

[MOJO]

I'm not wild screwing a male stainless steel nipple into a female Aluminum fitting. Stainless steel is so much harder than aluminum that I'd be worried about deforming the aluminum threads inside the fitting. Bronze (not brass) might be a better choice. I believe bronze is also closer on the galvanic scale to aluminum although you said that's not really a concern in your application.

 

[diver dave]

….

After discussion with machinists and pipe fitters, I used a combination of Teflon pipe dope and Teflon tape on all joints. Pipe dope on the male threads first and Teflon second. Job done 5-6 yers ago. No leaks.

Jim

I have seen both used as well. On an industrial mineral oil system. But the tape was put down first.

 

[MVCalypso]

Hi All -

An update:
first, sorry for the delay in posting an update. The project slipped out as I was down with a bug for a couple of weeks.

Anyhow, the job is now all done. All the joints were taken apart, cleaned and redone using Leak Lock (blue). I gave all the joints a full 24+ hour cure time. It's been a few days now & I'm a happy camper!

Not even the smallest sign of a diesel weep!