Pin Hole leaks on jacketed pessure vessel 2

[koshyeng]

All,

We notice pin hole leak as shown in the attachment. This is a jacketed pressure vessel with 304 shell and 517 CS jacket with internal baffles for chilled water. Could you please shed some light on damage cause?

Chilled water is DI water at 47 F which is treated with molybdate. pH of chilled water is 8 to 9.

When grinding the vessel, we did not find general thinning of the vessel wall (1/4" thick). Upon grinding, we noticed the pin hole is located right above the baffle (Stitch welded to the shell). During grinding, the wall thickness was close 1/4" indicating there is no general thinning of the wall.

Appreciate for your help.

 

[EdStainless]

Do these line up with the toe of the welds for the baffles?
Is this in service? How long?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[koshyeng]

Ed,

The crack is on jacket wall. The baffles are stitch welded to the shell wall and there is about 1/8" gap between the baffle circumference and jacket i.d .

This vessel has been in service for 25 years and has never been taken out of service except for shutdowns.

I assumed under deposit corrosion as i have seen red sludge on the surface of the baffle as well as on the shell wall. However, the baffles which are made of same material (SA517), have no signs of pitting corrosion.

Appreciate for your help.

Thanks

 

[EdStainless]

There is a crevice between the baffle and jacket. Between the fit and corrosion product I would presume that in some places it is a tight crevice.
Do these leak?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[koshyeng]

Ed,

I agree with you that there could be tight fit in certain areas that could contribute for crevice corrosion. I did not see any corrosion signs on the baffles. See attached picture showing the baffle.

 

[EdStainless]

Can you make a quick sketch, I am not sure of what I am looking at there

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[koshyeng]

Please see the attached description about the cracks and pictures.

From the pictures, I ruled out crevice corrosion. Appreciate your help.

 

[koshyeng]

Any ideas on corrosion mechanism for the issue addressed above?

thanks

 

[metengr]

From the pictures, confirmed as crevice corrosion and the cracks are located where the corrosion has thinned the wall under the baffle attachment where it could no longer provide any structural support.

 

[koshyeng]

Metengr,

Thanks for your response. There is no crevice between the baffle and jacket wall. The last picture of the plate is upside down.

The baffle is stitch welded to the jacket wall on the top side (3" weld). The 1/4" baffle is around the read area. Right above the baffle you see a ridge which seem to be weld and the crack is right on top of the weld location.

The crack is as long as the weld section is. We identified 8 such cracks and all are loacted right above the weld. We cut a section off as shown in the attached pictures. For the rest, we ground the cracked area and welded it back. The grinder went through the stitch welds for all 7 cracks.

 

[EdStainless]

Then they are either cracks from the original welds, or there was enough stress from the welds, the metal was hard enough, and corrosion generated enough hydrogen (after all that is what it does) to cause hydrogen embrittlement cracking. With the SS in the system you know that the CS will be corroding. Sounds like you need sound welds and a PWHT.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[koshyeng]

metengr,

I have attached some studies done on the cracked plate along with the pictures. It seems more like a thermal fatigue at the welds. There is 90 degree temperature swing at least 3 times/day.

Appreciate your thoughts.

 

[SwinnyGG]

Thermal fatigue would not be my first, second or third guess at an explanation.

What is the mean operating temperature of the jacket? 90 degrees (C??) of temperature change, once every 8 hours, is a relatively low level of thermal loading.

 

[koshyeng]

Chilled water is 50 F and hot water is 140 F. We use hot water for one hour during charging. At the end of polymer chanrging in to this vessel, we will start agitation and switch to chilled water.

 

[SwinnyGG]

If you're talking 90 degrees F, now you're talking an even smaller thermal change.

Given the amount of sludge, and surface corrosion, visible on the baffles in your photos, the other posters pointing out corrosion issues look to be correct.

In any event, a failure after 25 years of continuous service seems to me to be a pretty reasonable track record. The hard part will be cleaning all of the corrosion out of the jacket to avoid this problem popping up in other areas that you haven't yet inspected. My suspicion would be that the conditions that caused this pinhole in this area are present in other locations as well, and if they aren't leaking yet they will be eventually.

 

[koshyeng]

One reason for me to suspect thermal stresses is I have noticed bulged jacket wall at 5 locations. This bulging is not all around the circumference of the jacket but only in few location. These points are inline with where the baffle is inside the jacket.

It is hard to see the bulges with naked eye for 5 ft away. I notice when I was looking at the defects about a feet away from the wall and felt it.

Thanks

 

[SwinnyGG]

If the numbers you are quoting to us are accurate, there is simply not enough energy flux through that part for the root cause of this failure to be thermal loading. 90 degrees of temperature change is meaningless in the grand scheme, especially given the large thermal mass you are dealing with.

All of your photos, explanation, and data point directly at a corrosion problem. The two other posters in this thread are posters who are well known on this board as excellent technical resources, and they are all saying the same thing.

 

[EdStainless]

The bulging of the jacket means that the stress are lower than if it stayed flat. With thermal expansion things either move or they fail.
This still looks like primarily a corrosion issue to me.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[koshyeng]

Ed a& Metengr,

Thanks. 4 years ago, the chilled water is switched from regular industrial water (clarified water with pH 8 to 9) that was treated with nitrite type chemical to Demineralized water (pH 7.0) treated with Molybdate ( >100 ppm with water pH at 8 to 9).

Last week we inspected another vessel with same design, service and operating conditions as the failed vessel in discussion and found out that the vessel jacket has lost entire 40% of wall thickness. Shear wave indicated pitting corrosion.

I would like to know if the DI water is causing this corrosion problem even with molybdate treatment.

 

[EdStainless]

If the molybdate is not coating the surface then the DI could be the issue. Many inhibitors require and initial high dosage rate to achieve coverage, and the lower dosage maintenance rate.
I trust that lab samples were run, giving them a long term exposure in the old water, and then into the new and checking to see that they remained passive?
There is also the issue of dissolved gasses in the water. This could be very different between the two waters and could be playing a role.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube