Question about failing bolts in chilled water pipe connectors (Victaulic) 13

[Lnewqban]

Dear all,
I am involved in an engineering problem related to pipe connectors that have been failing around every five years.
This has happened in Florida, in an office building, and for a chilled water piping system serving individual coils and fed by VFD pumps.
The bolts of the couplings have been failing, only weeks apart, and flooding the corridors of the building.
That happened about five years ago, time at which all the original bolts were replaced, and is currently happening again.
The manager of the building has contacted my company about a permanent solution to this problem.
Please, see attached picture of broken bolts.

There is a rooftop chiller, from which two main insulated pipes go down and branch out in two opposite directions for about 60 yards each way, making two closed loops.
The broken pipes (branches closer to the roof chiller and vertical pipes) should be 2-1/2 inches diameter, according to specified in the drawings.

A local AC contractor has estimated a cost for replacement of all the 375 couplings and rubber seals, installing stainless steel bolts rather than regular steel, as original.
The bolts that have failed show signs of normal corrosion and all have broke by the section where the nut meets the casing.
There are no expansion loops in the original system, and the broken sections show areas of corrosion and new metal, which make me believe that the failure may be caused by metal fatigue.
The system rests at night and moves chilled water during the day.

I don’t have any experience with this type of connectors.
I would love to learn your first impression about the probable causes or failure, and possible remediation.
Perhaps a type of coupling able to absorb the bending stress of thermal dilatation-contraction cycles?
Thank you for your time.

"Where the spirit does not work with the hand, there is no art." - Leonardo da Vinci

 

[dik]

Have you talked to Victaulic to see if they have any suggestions? or can possibly recommend fastener types.
The failure appears to be corrosion related. What is the environment. The lack of expansion loops can be a problem, too. Can you introduce a 'bellows' type of expansion, else create an expansion loop.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[EdStainless]

So these bolts are wet all of the time.
And being hammered by fatigue.
How near the coast are you? Any chance for chlorides to be present?
It would be worth having some of these looked at by a good lab you see if fatigue really is the root issue or does it set in after the corrosion has weakened the bolt.
I ask about chlorides because if so then 304 or 316 bolts might not be the best option.
You would be left with going to duplex SS hardware (2205) which would get pricy.
Or maybe a better option would be Xylan coated steel bolts.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[EnergyProfessional]

I don't have experience with Victaulic, but we typically don't allow them. The reason is the gaskets don't have an infinite design life and you can count on having to replace them after 20 years. Same is true for similar systems for copper pipe.

From your description it doesn't sound like a gasket problem and 5 years is awfully short. You should talk to Victaulic and review the corrosion situation as others have recommended. Obviously something went wrong installing them. You don't want the contractor repeat the same error.

I don't know if this is an option or possible at all. But maybe they can weld in connectors and then you have welded pipe and don't have to deal with gaskets anymore.

 

[GD2]

Lnewqban,

Tell us more about the joint configuration:
1. What pipe material the clamps joins?
2. Are the clamps insulated too?
3. Any specific area or joints it fails repeatedly?
4. Had there been leak at the failed joints?

GDD
Canada

 

[TugboatEng]

Why do you consider fatigue to be a contributing factor? Is this joint vibrating?

When the bolts fail do the pipes remain in place or do they dislodge? If they dislodge this indicates excessive pipe strain from misalignment.

I don't see anything from your picture that would indicate corrosion to be a contributing factor. Yes, corrosion is present but it appears superficial.

Stainless fasteners are a terrible solution. Ed mentions chlorides for good reason. The problems with SS start from the very beginning. The tensile strength is low, 300 series stainless bolts are only about 70ksi. Even worse, they gall when tightening which gives incredibly inconsistent tensions.

 

[MJCronin]

Like TugboatEng, I also agree with the sage advice of EdStainless .... EdStainless still makes massive contributions to these fora

Chlorides are always a prime suspect in this kind of system-wide bolting failure.

I suggest that you again contact VICTUALIC and not talk with the "twenty-something" newbie in sales ...

Find the engineering gray-hairs and start askling about experience with thier systems in corrosion prone environs. I am sure that someone somewhere has encountered a similar problem with a VICTUALIC system.

Ask the geezer about his experience with premium coated bolting and hardware, such as XYLAN (as mentioned above)

I also suggest that you invite VICTUALIC to your plant site to do a system review and evaluation. They may be VERY interested in finding out if you somehow ended up with a bad batch of bolts.

Another option might be to gather a bunch of failed bolts and contract with a bolting specialist to perform a PAID forensic analysis..

OH .... Wait ... what was that ? ..... Your boss told you that there was no money available in the budget to hire a specialist ...

MJCronin
Sr. Process Engineer

 

[Lnewqban]

Thank you all.
Answering some of your questions:
I have contacted Victaulic, but I have not received any response so far.
I could recommend one or two expansion loops for each problematic branch.

This office building is located about 5 miles from the Atlantic Ocean, but I have no idea how much chlorine could be present; outdoor steel on the roof does not seem to be exceptionally rusted.

1. What pipe material the clamps joins?
Carbon steel pipes.
2. Are the clamps insulated too?
Yes, although, I don't know how well air humidity was able to reach the cold metals.
3. Any specific area or joints it fails repeatedly?
Yes, it has been always the bolts of the couplings located closer (about 50 feet each side) to the main vertical pipes coming from the VFD controlled pump located on the roof.
4. Had there been leak at the failed joints?
No, no sign prior to the rupture. One bolt suddenly fails, and a huge leak follows from the affected coupling (nothing prevents the pump from keep pumping until somebody notices the flooding).

Why do you consider fatigue to be a contributing factor? Is this joint vibrating?
Only because the look of the broken sections of the bolts: there is a peripheral area of "old" rust, and a central area of fresh broken metal. I assume that the changes in internal temperature and pressure of the water induces contractions and elongations of each branch, which could induce excessive bending moments at the couplings.

When the bolts fail do the pipes remain in place or do they dislodge? If they dislodge this indicates excessive pipe strain from misalignment.
I have not been informed about that detail.

I don't see anything from your picture that would indicate corrosion to be a contributing factor. Yes, corrosion is present but it appears superficial.
I agree with you, but it is deep at the threated fractured section, which is always located where the nut meets the casing.

Stainless fasteners are a terrible solution. Ed mentions chlorides for good reason. The problems with SS start from the very beginning. The tensile strength is low, 300 series stainless bolts are only about 70ksi. Even worse, they gall when tightening which gives incredibly inconsistent tensions.
I agree with you. I also don't know how both materials compare regarding resistance to metal fatigue. Victaulic may not provide that type of bolts; I have not found those in their literature (which use could void any warranty).

"Where the spirit does not work with the hand, there is no art." - Leonardo da Vinci

 

[FacEngrPE]

I have used these couplings in some really awful conditions, and have not observed the sort of failure depicted in your photographs. If correctly installed the clamshell halfs have a hard metal seat that is directly lined up with the bolt and nut line of action. The properly tightened (not over tightened) bolt should only see minimal alternating stress even with significant vibration applied.

Posting a photograph of the break with better resolution might be helpful here, the details needed to confirm fatigue failure are smaller then the pixle size in the photo.

If the groove is cut incorrectly the clamshells will not seat correctly. This might result in undesirable alternating stress in the bolts. I suggest that you disassemble the joints where the failure occurred, and measure the groove. We have found that "PI" tape works well, and does not require the same level of skill as a micrometer. Reason for this suggestion is that Victualic never supplies the grooved pipe, it is either grooved by a machineshop, or field grooved by the contractor. We have observed that some of the pipe groovers do not understand keeping the groove in tolerance is important.

 

[MJCronin]

Agree with FactEngrPE ...

A paid one-day walk through with an experienced VICTUALIC rep would, in my humble opinion, be the most cost effective solution ...

.... despite what your butt-hole manager tells you..

Oh, .... and by-the-way..

Out of respect, could you please tell us about your final solution to this problem ...

Thank You ...

MJCronin
Sr. Process Engineer

 

[GD2]

Lnewqban,

Here is my theory of failure:
Given Data:
Building Chiller system, insulated, CS pipes - works on day, off at night
Assumed Summer day temperature: 30 - 40C
Assumed Night Temperature: 20 - 30C
Assumed piping installation temperature: 25C
Assumed Chilled water temperature: 5C

Location: Florida
Weather: Coastal, moist, salty

Bolt failure observation: sheared bolt at nut/clamp interface, general corrosion on the bolt but deeper corrosion at the sheared face meaning there was cracks growing before final failure.

Probable cause of failure: Stress Corrosion Cracking (SCC)due to fatigue.

Justification:
Stress: The system works at one minimum thermal cycle a day. I would expect more cycles a day depending on the start/stop due to weather variations at day and night.

There is no relief of this thermal stress on the CS piping by means of elbows, off-sets or expansion loop. The clamps are located at least 50 ft away from the first T connection (correct me please)

This has added additional load on the bolts due to thermal expansion/contraction of Delta T of at least 15C to 20C on top of the normal bolt torque.

Crack initiation: The on/off start of the system is causing a fatigue stress on the bolt. With 2 start/stop a day (on average), it will be 14,000 cycles a year. With fatigue, all piping have a cycle life. This is the crack initiation process which is accelerated by corrosion.

Corrosion: Evident from the sample. This is due to the humid sea salty environment.

Mitigation:
1. Replace the bolts with Cadmium, xylan coated bolts (very common in offshore). Xylan is preferred. If you are using A193 B7 bolts look for B7M or B16 if available.
2. Reduce stress at the coupling by providing an expansion loop at appropriate location.
3. If you can't provide the expansion loop immediately, replace the bolting, do not over torque and start an maintenance check for cracks, say initially after 4 years and make a maintenance plan to provide the loop.



GDD
Canada

 

[georgeverghese]

Did your installation contractor use moly grease on the bolts? Couplings must be dripping wet with the humid atmosphere in Florida. Got this off the internet:

"When should you not use Moly grease?
Greases containing molybdenum disulfide shouldn't be used in applications that see excessive amounts of water ingress as the molybdenum disulfide can have a corrosive effect of the metal surfaces, and is not suited for bearings and other components that move at a rapid pace.26 Aug 2021"

Am used to seeing cadmium coated nuts and bolts in oil and gas installations.

 

[dik]

GD2... I try to avoid cadmium coatings due to their toxicity.

So strange to see the singularity approaching while the entire planet is rapidly turning into a hellscape. -John Coates

-Dik

 

[LittleInch]

Can you draw a isometric or section view of the system and where these joints are.

It sounds like they are being subjected to a high degree of moment force as the main header expand and contracts so need to show where guides and sorts are.

Initial torque is also crucial to check, but these connectors are not designed to accept forces other than axial/ pressure.

I tried to find out on the victaulic site what the limits were but it doesn't specify it.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[FacEngrPE]

Victaulic I-100 FIELD INSTALLATION HANDBOOK
Contains a section titled: Explanation of Critical Roll Groove and Cut Groove Specifications - Original Groove System (OGS) and EndSeal TM.

The Victualic 26.01 Grooved Piping System – Design Data explains the recommended method of handling expansion, and meeting various pipe design stress requirements (B31.1, B31.9).

I recommend the groove dimensions, and the nearby piping and hanger details be checked against the factory recommendations. For some reason upload is not working, so I have included links to the source.

 

[nuuvox000]

That is some crazy corrosion on bolts that I haven't seen before; it doesn't look superficial at all to me but I only have the one photo. My first thought is that maybe the insulation you are using is not stopping condensation on these pipes (fiberglass is not good to prevent this). If you look at the condition of the bolts near the chiller supply (coldest water in pipe) vs near the return (warmest water in pipe) that may tell you if that's what's happening.

 

[TugboatEng]

You're right, nuuvox. I looked too quickly at the picture and I focused too much on the shoulder area. One bolt has been broken for a while. The other bolt is >50% wasted at the fracture location.

 

[Lnewqban]

Here is the plan view of the second floor pipe layout.

Note that no coupling has ever failed at first floor, only around central area of second floor.
Chiller and pumps are on the roof, just above second floor.




"Where the spirit does not work with the hand, there is no art." - Leonardo da Vinci

 

[Lnewqban]

Estimating around 140 feet of each branch length, I have calculated a max expansion of 0.6 inches, which could create an axial force of about 34,000 lbf, if perfectly restricted.
Would you consider expansion devices for each affected branch?
If so, how many and at locations?

I have just confirmed that original and existing couplings are of rigid type, and that no expansion device was originally specified.
Water working temperature is 45 F, but it could sharply rise when circulation stops during weekends.

Any suggestions are appreciated.

"Where the spirit does not work with the hand, there is no art." - Leonardo da Vinci

 

[nuuvox000]

Lnewqban, see ASHRAE Fundamentals, chapter 22, sections 1.7 and 1.8.
Any and all expansion must be provided for. Usually when it's small, typical hangers and turns take care of it. The definition of "small" can vary widely.