Thoughts about this weird microscopic corrosion

[BoxOhJelly]

Hello forum, interested to hear what thoughts y'all may have about this really weird looking corrosion I found.

Parts are thin wall 304L stainless tubing, annealed. They get filled with good ole tap water but failures are very random, sometimes they'll last a few days, sometimes a few months or even a few years. Some don't fail at all.

I've seen a good number of failures so far, but I have only seen this phenomenon once before back in October and now again this week so here I am asking about it.

First particular instance, it was filled for 3 hours before leak was discovered.

Second particular instance, only info I have is that it leaked 19 days after installation.


My first thought was preferential dissolution of cementite/ferrite in pearlite because of the lamellar kind of look to it but then I remembered this is austenitic. My other idea was slip bands or twinning, something along the lines of higher stressed regions, corrosion starts on areas with the highest energy, but I just dont think that's right (too small of a pattern to be twinning, and the tubes are not under any external stresses besides simply existing, and they're annealed so shouldn't be any residual stresses). The tubes can vary in length though these were both 72 inches long, different spots along the tube, not near the weld seam, not the same water between these two occurrences.

Appreciate any and all thoughts you peeps may have!

-BoJ

 

[mfgenggear]

OP
test the water for chemical, or has it been done. is this a DI system

how is the integrity of the wall thickness measured.
are the tube contaminated prior to installation.
inquire with the manufacturer of his process.
are the tube been passivated., clean process.

from previous experts chloride or salt attacks SS.

 

[BoxOhJelly]

OP
test the water for chemical, or has it been done. is this a DI system

how is the integrity of the wall thickness measured.
are the tube contaminated prior to installation.
inquire with the manufacturer of his process.
are the tube been passivated., clean process.

from previous experts chloride or salt attacks SS.

Testing the water is harder than it sounds, I've tried to get them to do it. These were actually from 2 different systems on different continents as well. Not DI, just straight tap water/city water.

I'm not quite sure what you mean by "how is the integrity of the wall thickness measured."

Well they certainly shouldn't be contaminated prior to install. Several tubes are installed per system, and as I said different continents, I find it unlikely they would both get contaminated in the same exact way. I've also looked at many many other failures of these tubes its just these 2 that jumped out since the corrosion looks so weird. "Normal" failures are pretty standard pitting with just a single big pinhole, maybe some small pits around it.

We manufactured the process and it has regular compliance audits at our suppliers, everything looks to be in check there. Unless the suppliers are falsifying their reports but I don't think we're at that stage yet.

I've been told that the annealing process passivates the tubes yes, and that it is a very clean process. Lots of definitions and specs for cleaning and drying, allowed detergents, DI water only (which is tested with every batch with strict requirements).


Someone has suggested that this could be some kind of microbial contamination, city water is clean enough to drink but its still going to have microbes and bacteria and things in it. I'm not sure how much I believe that but I know nothing about bacteria really, I will have to do some reading on it this week. We do sometimes detect sulfur and chlorine in our EDX and as you said salts attack SS so thats not unexpected, but the sulfur doesnt exactly make sense

 

[mfgenggear]

City water depending on the area is loaded with chemicals. But mostly chlorine.

Annealing is not passivation, and annealing causes scale, and if not in a protective coating or atmosphere will damage the surface from oxidation.
Once damage there is possibility of the chemicals are attacking the SS.

Passivation for SS , ref AMS 2700

 

[EdStainless]

Break these open and look under the SEM.
What is on the surface of the 'cracks'?
As a tubing guy this looks like contamination that was baked in during the anneal.
Do you have any idea who made the tubes and how they do it?
Sorry but a detailed water analysis is step #1.
You can't begin to troubleshoot this without it.

 

[mfgenggear]

OP
Integrity of wall thickness.
Depending on how the tubing was manufactured and how it was bent.
If drawn the material is cold worked to obtain the outside diameter. Wall thickness could have indication , defects or under size wall thickness.
Since it is difficult to measure an ultrasonic gage is required to measure thickness.
And after manufacture , after passivate an NDT operation is required to look for defects.

 

[Stick.]

Since it's tap/city water, can you at least get water quality reports from the water supplier? That should help at least get a general sense of what you may be dealing with.

 

[dik]

Since it's tap/city water, can you at least get water quality reports from the water supplier?

They may not provide a full, detailed analysis.

 

[BoxOhJelly]

Since it's tap/city water, can you at least get water quality reports from the water supplier? That should help at least get a general sense of what you may be dealing with.

Unfortunately I cannot. We are aware that tap/city water will contain ions, especially chlorine, and in most cases they eventually corrode. I've tried to push higher ups to focus on the water more but for whatever reason everyone refuses to do anything about the water. Apparently customers are warned that crappy water will have adverse effects but they still act surprised when it leaks.

Maybe I should also add this as it might be important... these weird spots were both found on the OD of their tubes.
I did some EDX and the ID shows varying amounts of Na, Ca, Cl, K, Mg, S on both occurrences, on the first occurrence one spot far away had a big Al peak, another different spot had a little bit of Zn, its pretty obvious they're just using crappy tap water but I still want to rule out something on our end that might make these tubes more susceptible.
Additionally I have been told that "other companies with SS tubes don't need to treat their water or do anything besides hook em up, so if we want to compete we have to be just as resilient".

Break these open and look under the SEM.
What is on the surface of the 'cracks'?
As a tubing guy this looks like contamination that was baked in during the anneal.
Do you have any idea who made the tubes and how they do it?
Sorry but a detailed water analysis is step #1.
You can't begin to troubleshoot this without it.

The only direct EDX I did on the 'cracks' was this came back with some C, O, 0.5% Cl and 0.5% Ca, 2.3% S. Sulfur shows up on every suspect spot between 2.3-3.9%. I will have to go back in a little bit when my coworker gets done to look at the other occurrence.
I do want to ask what you mean by "break these open" since it is so small. I can attempt to pull the ends apart like a tensile test to stretch it out?
I do know who the supplier is, the process I do not fully know about but I do have our annealing spec and a cleaning spec in front of me. Not sure how much info I can give away though. I don't see a spec or process for how the tube is actually manufactured, it might be the supplier's trade secret or something. I'll ask around to see what I can find out.

City water depending on the area is loaded with chemicals. But mostly chlorine.

Annealing is not passivation, and annealing causes scale, and if not in a protective coating or atmosphere will damage the surface from oxidation.
Once damage there is possibility of the chemicals are attacking the SS.

Passivation for SS , ref AMS 2700

The annealing is done under a protective argon atmosphere, batch annealed. We also have a similar product that does an in-line anneal instead that does not have these same corrosion issues. I will try to find out some more info about the whole annealing and manufacturing process, nowhere am I seeing anything about passivating but you're right it wouldn't make sense to skip passivation...
We are looking into the differences between the annealing, but right now I'm more focused on these weird looking 'cracks' as Ed called them, since they look so unusual.

 

[EdStainless]

You can cut into a sample from each side so that it will tear where you want it to.
After you cut into the edges of coupon you either bend it or pull on the ends.
C, Al, Cu, and Zn will all cause significant intergranular issues.
Look for carbide stringers in the grain boundaries.
If you keep finding a lot of S then biological activity is highly likely.
Ar does not provide a bright anneal and the surface will be covered with chrome oxide that is too thick and porous to provide corrosion resistance.
Passivation will do nothing about this, it would require and actual pickling treatment.

 

[mfgenggear]

ED
my experience with Heat Treat Ar atmosphere has been .0005 inch max.
have you seen more than that?

Op
may be a light bead blast not aluminum oxide unless 60 max PSI. after annealing, then blow with compressed air to remove debri.
then passivate. I would not recommend to pickle. darn operators some time leave in the tank to long.

prior to annealing run metlab samples see if the grain and metlab meet the specification
could be the oven for annealing might have an air leak as well.

 

[BoxOhJelly]

You can cut into a sample from each side so that it will tear where you want it to.
After you cut into the edges of coupon you either bend it or pull on the ends.
C, Al, Cu, and Zn will all cause significant intergranular issues.
Look for carbide stringers in the grain boundaries.
If you keep finding a lot of S then biological activity is highly likely.
Ar does not provide a bright anneal and the surface will be covered with chrome oxide that is too thick and porous to provide corrosion resistance.
Passivation will do nothing about this, it would require and actual pickling treatment.

Interesting that you say argon doesn't provide a bright anneal... we switched to argon because we had sensitization issues in the past, it has since been solved, no GB carbides or sensitization, tubes are shiny and bright to me (I'm young and dont have any experience with heat treating so I'm not really sure what a non-bright anneal would look like in comparison but my boss says its bright sooo I'll see what he says about this, since he is the guy who decided on switching to argon) I do have some mounts made so I can check the grain boundaries just need to get them polished.
We did just send some tubes out to do some Auger analysis, they reported the passivation thickness of ~150 angstrom on an unannealed tube vs ~50 angstrom after it was annealed. 150 angstroms is about 0.0005 inch as @mfgenggear said...

In other failures we aren't seeing any intergranular issues, but once there is a big pit that spot is corroded and gone, so maybe that's hiding the issue from us? I guess my first thought is that if it was an intergranular issue we'd be seeing it all over the tube not just in one spot, but maybe that one spot just had the most susceptibility.
It is rare that we see Al, Cu, or Zn show up. And I slightly misunderstood. We actually do see carbon and sulfur in pretty much every scan of a suspect area, I thought that it was just the corrosion (rust) as they also usually have a decent oxygen peak, but I think it's actually a biofilm from bacteria

(Photo is from a different tube I didn't do EDX on but its similar to what I've seen before)
And now that I'm looking at this again I'm noticing the visible grain boundaries which makes me think there might be an intergranular problem or maybe sensitization has popped up again.

 

[EdStainless]

On 304L I am not sure that you can really see the impact of AR over H2 or anneal and pickle.
In higher alloy grades it is easily seen in corrosion testing, especially if you can do electrochemical testing.
Pickling isn't hard.
Control the acid concentrations, residual metals, temperature, and time.
It is like any other process.
We used to anneal in open air and then pickle.
We also hydrogen bright annealed.
For some low grade applications we would anneal in-line (not hot enough and way too short of time) with H2 on the OD and Ar on the ID.
With 304L you couldn't really tell.

 

[mfgenggear]

Interesting that you say argon doesn't provide a bright anneal... we switched to argon because we had sensitization issues in the past, it has since been solved, no GB carbides or sensitization, tubes are shiny and bright to me (I'm young and dont have any experience with heat treating so I'm not really sure what a non-bright anneal would look like in comparison but my boss says its bright sooo I'll see what he says about this, since he is the guy who decided on switching to argon) I do have some mounts made so I can check the grain boundaries just need to get them polished.
We did just send some tubes out to do some Auger analysis, they reported the passivation thickness of ~150 angstrom on an unannealed tube vs ~50 angstrom after it was annealed. 150 angstroms is about 0.0005 inch as @mfgenggear said...

In other failures we aren't seeing any intergranular issues, but once there is a big pit that spot is corroded and gone, so maybe that's hiding the issue from us? I guess my first thought is that if it was an intergranular issue we'd be seeing it all over the tube not just in one spot, but maybe that one spot just had the most susceptibility.
It is rare that we see Al, Cu, or Zn show up. And I slightly misunderstood. We actually do see carbon and sulfur in pretty much every scan of a suspect area, I thought that it was just the corrosion (rust) as they also usually have a decent oxygen peak, but I think it's actually a biofilm from bacteria
View attachment 5690
(Photo is from a different tube I didn't do EDX on but its similar to what I've seen before)
And now that I'm looking at this again I'm noticing the visible grain boundaries which makes me think there might be an intergranular problem or maybe sensitization has popped up again.

It has been my experience if SS parts are zyglo penetrant inspect if material has IGA
It will light up like a Christmas tree..

 

[EdStainless]

or draw a line on them with a Sharpie and see if it bleeds.

I still think that it looks like surface contamination prior to anneal.

 

[weldstan]

What is the flow velocity and does water stay stagnant for somewhat long periods. We had a pitting to failure incidence in a 316L system where the owners engineer allowed the hydrotest water to remain in the piping prior to start up. Leaking began in less than 2 weeks. Water contained over 1000ppm chlorides amongst other organics. We had to replace entire system then hydotest followed by rinsing with deionized water and drying before startup.

Agree with Ed that tubing was possibly improperly cleaned prior to annealing and not pickled and passivated after.

 

[weldstan]

What is the water flow rate and does the water remain stagnant for any prolonged period. We installed a 304L pipe sytsem which was hydrotested with lake water at the direction of the owners engineer and allowed to stay in without cleaning until statup. Water contained over 1000ppm Cl. Leakage occurred within a week. We replaced the entire system, hydrotested and rinsed with deionized water and dryed until startup.

I agree with Ed that the tubing was possibly not properly cleaned prior to annealing and not pickeled and passivated afther. But city water can contain as much as 1000 ppm chloride ion along with other chemicals and metallic ions that would be corrosive to 304L stainless.

 

[BoxOhJelly]

What is the water flow rate and does the water remain stagnant for any prolonged period. We installed a 304L pipe sytsem which was hydrotested with lake water at the direction of the owners engineer and allowed to stay in without cleaning until statup. Water contained over 1000ppm Cl. Leakage occurred within a week. We replaced the entire system, hydrotested and rinsed with deionized water and dryed until startup.

I agree with Ed that the tubing was possibly not properly cleaned prior to annealing and not pickeled and passivated afther. But city water can contain as much as 1000 ppm chloride ion along with other chemicals and metallic ions that would be corrosive to 304L stainless.

Flow rate is zero, system gets filled and stays filled. This is almost always exactly what happens to us. They hydrotest and will usually drain it but no flushing, so some water gets left in the tubes. As these tubes are globally used and I'm stuck in one place all I can do is look at the tubes in the SEM, do some EDX, unfortunately can't do water tests but we know the water most likely will have chlorides and other crap in it. Not uncommon at all for me to find significant concentrations of them and other salts, and recently now I'm seeing sulfur as well.

I've told higher ups that if we just flush it after the hydrotest we most likely will have a significant reduction in leaks but "that's too expensive" as if replacing all these tubes all the time somehow costs less than some DI water.

I'm looking into the possibility that they simply aren't cleaned properly before annealing. Will update if I ever get to the bottom of this

 

[EdStainless]

If you aren't going to flush then hydro test water for 304 or 316 should be better than US Potable water standards.
Both in terms of chlorides and biological control.
I have seen some companies require 50ppm max chlorides and either a specific level of chlorination or a specific hydrogen peroxide addition.
I have seen failures from left over hydro water in a mater of a few days.

 

[weldstan]

Management can sometimes be really stupid.