Transverse Cracking in Carbon steel welds 3

[numberfive]

Hello everyone,

I'm having a problem with transverse cracks in carbon steel welds on a low pressure steam header for a sawmill drying kiln operation.

The header is 20" STD SA-106 Gr. B with 20x20x10 WPB reducing tee's on either end for 10" dirt legs and a 20" WPB pipe cap on either end. The header has (2) ea. 8", 10", and 12" weld-o-let style branch connections with 150# RFWN flanges on each.

The header was built in 2022. Solid wire root passes (Miller RMD PipePro), flux core fill and cap passes. ER70S-6 and E71T-1M respectively.

Operating conditions are 10-12 psi. at around 250 deg. F.

The customer has had (3) leaks to date due to transverse cracks across the welds.

We just stripped all insulation and performed MT examination of every weld today and found a total of (1) transverse cracks.

Weld quality looks really good both on root and cap sides.

Has anyone experienced similar issues with carbon steel? Very strange.

Took multiple hardness readings today and all ranged from upper 150's to lower 170's HB both in weld metal and base metal.

Also took shavings for a chemical analysis of weld wire.

Thank you,
numberfive

 

[r6155]

Thickness, please.

Regards

 

[goutam_freelance]

Any information about fluctuating or reversing loads to check for the possibility of fracture?

Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/

 

[numberfive]

Thickness is 0.375”

No fluctuating or reversing loads. It is a very stable, slow system.

Both of the 8” outlets for kiln steam have approximately a 15’ vertical run then transitioning into long horizontal runs to the kilns. A couple of things I’m a little concerned with are 1) The kiln steam lines are 8” stainless steel per the kiln manufacturers design, 2) All of the pipe supports on the long horizontal runs are welded leaving no room for movement. Typically we would install slide type saddles in this application but the piping contractor used 8” friction clamps welded to a small piece of channel welded directly to the pipe bridge steel.

This may help to explain the cracking on these two o-let welds to the header but doesn’t explain the cracking on the opposite end at (2) 20” butt welds.

 

[Trestala]

Are the orientation of the cracks perpendicular to the SS lines?

Also took shavings for a chemical analysis of weld wire.

Any results on this?

 

[goutam_freelance]

Given your description of lines, there is strong possibility that the lines are partially thermally restrained. Isometric of the lines would have been helpful in understanding the problem.

In any case I suggest you contact a piping design consultant and get the design reviewed.

Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/

 

[numberfive]

On one of the 8" o-lets the cracking is perpendicular to the SS line. On the other, there are (3) transverse cracks equally spaced around the circumference of the 8" o-let to RFWN flange weld on the header.

No results on the chemical analysis for the weld metal yet.

The engineer who designed the piping system will be on site today and I will discuss concerns with the pipe supports.

The thing that really has me puzzled is the type of cracking. In my experience usually when we see cracking due to stress, especially in carbon steel, it is circumferential cracking, usually at the toe of the weld or just off the toe of the weld in the HAZ.

The other thing that I noticed was no use of spring hangers on the 12" steam inlet line. The 12" inlet connection is directly off the top of the header with an isolation valve and then about 15' of vertical pipe which then 90's parallel with the header and travels another 20 or so feet to a pressure reducing station. There are two clevis type hangers in this 20' horizontal run. The 20" circumferential weld in the horizontal header about 12" from the steam inlet has transverse cracking at the 3 and 9 o'clock positions. Could the growth of the steam inlet line in the vertical run be taking the weight off the clevis hangers and transferring it to the header?

 

[goutam_freelance]

Transverse cracks can also develop due to high stress during welding(and also weld and filler metal quality). Please check whether the pipe experienced stress during welding possibly due to restraints or similar reasons.

Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/

 

[weldstan]

Contamination in the joint has proved the most likely cause of transverse cracking in carbon steel weldments in my exprience. The 3 equally spaced cracks might indicate contamination from tacking the pipe in place prior to weld completion with a hot crack in the cotaminated area due to welding over the tack. Similar contamination could come from tooling used to hold the pipe in place for welding. The cracks then propogated during operation. But that's just my experience.

 

[numberfive]

All very good points. Thank you.

In the process of excavating (1) transverse crack on an 8" o-let to header weld we discovered a circumferential crack near the root and started chasing it. That revealed multiple small transverse cracks starting at the root and propagating outward toward the surface of the weld (zoom in on attached pic). We are going to completely remove the o-let and take shavings for a chemical analysis on the root pass material. Maybe the wrong weld wire was used for the root pass?

 

[r6155]

Do you have the welding procedure and welder qualification?
I don't like the w-olet for this example.
Excessive welding for design pressure less than 15 psi, and the slip-on flange is easier for this example.

Regards

 

[numberfive]

Yes. Welder was qualified per ASME Section IX using a qualified WPS.

The customer requested design conditions of 200 psi. @ 400F when the unit was built. I think there was talk about changing the use of the header in the future.

It's being significantly under used for what it's designed for that also makes this very puzzling.

 

[r6155]

Are there any x-rays during construction?

Regards

 

[Andrew ONeill]

I'd suggest also not to rule out product induced cracking. When I First read it, my mind when straight to stress corrosion cracking, which usually presents itself as traverse cracks, always initiating from the inside surface. In my industry this is a key damage mechanism, in particular from caustic soda. Although it is a steam service, a small amount of contamination and / or poor water chemistry / dosing could setup conditions for cracking to occur.

As a reference point, trace amounts of caustic soda will cause cracking in carbon will steel with temps above around 90 Deg C. Post Weld Heat Treatment also has a key role to play as well.

It's just another avenue to look at, especially if nothing else is making sense.
Also worth going through API RP 571 Damage Mechanisms Affecting Fixed Equipment in the Refining Industry if you can get a copy, it has a large library of weird and wonderful damage mechanism with photos that might also bring up some things you haven't thought of that you can also investigate.

Andrew O'Neill
Specialist Mechanical Engineer
Australia

 

[Compositepro]

The transverse crack means that the weld is under significant axial stress (circumferential to the pipe). uniform spacing between the cracks is very typical where two materials are bonded together but they shrink or expand differently. The crack relives the stored shrinkage stress of the weld, as it cools from its melting point, in the area near the crack. Because the weld is bonded to the pipe the stress in the weld increases with distance from the crack until the stress is sufficient to cause another crack. The massiveness of the weldolet and the weld contribute to this.

This stress can be mitigated by preheating before welding, and post weld heat treatment.

 

[numberfive]

No x-rays on file however we are thinking about x-raying these welds.

One of my colleagues just brought caustic stress corrosion cracking of carbon steel to my attention this afternoon. I had never heard of SCC in carbon but after researching it seems like what we may be seeing. The mill does use sodium chloride in their boiler water chemistry and each leak has had a chalky white stain surrounding it.

If anyone else has experienced this form of SCC I’m interested in learning more about it.

Will investigate more tomorrow.

Thanks to all for the great insights.

 

[r6155]

Wrong criteria.
Why weld the neck flange (full penetration butt weld) and not perform radiographic testing? What was the designer thinking?
I think it was a welding defect during manufacturing without good inspection and NDE criteria.

 

[Trestala]

r6155, RT is not typically performed on this kind of joint and it is not a butt weld (Category D for Pressure Vessels).

 

[r6155]

@ Trestala
You are wrong.
I say neck flange, look the first picture of (OP) 21 Oct 24 is a welding neck flange.
See ASME VIII Div 1 Figure UW-3 Category C

Regards

 

[Trestala]

r6155, the cracks are present as well on the o-let to header weld which is a category D joint as you can see on another photo sent by the OP. Even with Cat C joint on the o-let to WN flange weld, most likely it is just spot RT.

Anyhow, Caustic SCC seems to be one suspect on the cracks:

3.15 Caustic Stress Corrosion Cracking
3.15.5 Appearance or Morphology of Damage
a) Caustic SCC typically propagates parallel to the weld in adjacent base metal, i.e. in the zone of highest welding residual stress, but can also occur in the weld deposit or HAZ and can be transverse to the weld.
b) The pattern of cracking observed on the steel surface is sometimes described as a spider web of small cracks that often initiate at or interconnect with weld-related flaws that serve as local stress risers.
c) Cracks can be confirmed through metallographic examination. Cracks are typically branched and predominantly intergranular. Cracking that occurs in as-welded carbon steel typically appears as a network of very fine, oxide-filled cracks.

The hardness measurements are 150~170 HB though. Was the line heat treated? Carbon and CE values? Do you have caustic corrosion inside of the pipes?