Laminations found in a Pressure Vessel in Wet H2S service 2

[cespi82]

Good morning Gentlemen,

The situation is as follow:

- Glycol Absorber Column in a Gas Dehydration Unit operating at 1100 psi and 40 Celsius degrees
- Service: Wet Gas with around 400 ppm H2S
- Commissioned in 1979, over 35 years in continue operation
- Material of Construction: SA-516-70
- No History of failures or corrosion issues
- Internal visual inspection revealed no presence of bulging / blisters

During the current Turnaround, a UT scanning was conducted for the 1st time at the “Chimney” Tray level. This Chimney Tray is basically where liquids (Glycol + Water) accumulates. So I presume if there is a potential of Wet H2S damage to occur, it is going to be here.

As a result of the UT Scanning several indications, lamination/inclusion like, were found in this area. Phased Array UT scanning was carried out finding a great extent of indications that were classified by the NDT technicians as laminations and clustered inclusions. After finding this, inspection was extended to other areas of the column (away from the chimney tray) and laminations were also found. Presumably we are talking here about dirty steel with poor control back in 1979.

They reported a total of 7 considerably big laminations. The report included the lamination sizes (they grouped nearby laminations since some of them were very close to each other and classified them as single ones), and the locations and estimated area of the clusters of inclusion.

With the outcomes of the report I ran FFS as per API-579 Part 13 Level 1 and everything is satisfied but the last condition which refer to a maximum lamination size if the service is Hydrogen Charging. I continued further to Level 2 and did not pass either.

If we take this literally, we have to repair the absorber column and eliminate these indications, which obviously implies a huge and long job with a lot of things to consider.

My question is: if this column have been in this service for such a long period of time (35 years) and these laminations have been there since then and there’s no presence of bulging/blisters, why should we consider repair it? Only because we ran FFS and did not pass?

All parameters for susceptibility of hydrogen charging are met (I think), but there hasn’t been any evidence of this to be happening.

Your comments are highly appreciated!!

 

[r6155]

Try with Acoustic Emission Examination (ASME V)

Regards
r6155

 

[TGS4]

You claim that the defects have been there for 35 years, yet you also claim that this is your first inspection. Which is it? Without evidence of the former, other than a hunch, you ought to discard that from your consideration. Accordingly, your vessel does not meet the Level 2 FFS requirements. You have as options: Level 3, repair, rerate, retire.

r6155, what would you expect to discover from AE that the OP hasn't already found with PAUT?

 

[cespi82]

Thanks for your answers.

TGS4, It was the first time that we performed UT scanning on that particular area. What it has been normally done as part of the inspection scope is spot UT gauging on TML's. I presume these indications were not found until now because OP were always gauging at the same point. Just for my knowledge and understanding, what other ways of laminations formation/generation exist, other than during manufacturing process?

I asked up to exhaustion to the NDT technician to assure that there are actual laminations and no the result of another phenomena like HIC damage.

One more point (sorry that I did not mentioned initially), basically the reason for which the component did not pass the FFS was because the service is Hydrogen Charging and the laminations are to big. You've mentioned that my options are repair, rerate, retire. How about applying a HIC Barrier in the area of concern?

If I do so, would I eliminate the Hydrogen Charging issue and theoretically the FFS should pass?

Thanks!

Regards,

 

[TGS4]

What barrier would prevent hydrogen from diffusing through your steel and accumulating in the lamination, thereby preventing additional blistering?

 

[metengr]

I would consider removing a sample plate section (12' x 12") and install a flush patch. Have the sample analyzed at a competent metallurgical lab to check the condition of the microstructure and evidence of hydrogen damage. You may be surprised at the findings.

 

[cespi82]

Hi metengr, thanks for your reply.

I thought about it, but if we plan to do any welding in this vessel that theoretically has been exposed to a Hydrogen Charging service for such a long period of time (35 years), shouldn't we do a preventive hydrogen bake-out procedure before? To do that we might need or either take the column to the workshop (dimensions 59" dia, 25m tall), or make it at site by pulling the column from top with a crane to avoid any deformation/bending during the heating? It is 2.5" thick so probably the baking time will be long.

TGS4, a thick and soundly applied layer of Belzona 1321 wouldn't isolate the base metal from the service? I also thought on applying weld overlay or strip lining but I go back to the point of the previous paragraph: bake-out procedure is required, followed by PWHT after the job is completed?

Waiting for your kind replies!

Thanks in advance

 

[EdStainless]

The other advantage of removing a section is to provide accurate feedback for UT.
The sizing in UT can have significant errors if the reference standards don't closely match the indications found in the field. I have seen people remove a series of 3" diameter plugs for lab evaluation.
I don't know of any coatings that is impervious to hydrogen, only metal overlays.
You will need localized preheat and PWHT for any weld work.


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

 

[cespi82]

Just to clarify my previous post, the coating/metal overlay am talking about would be only in the Chimney tray area that is where we have liquid water accumulation in presence of H2S. This is about 3m heigh. I wouldn't be worried about HIC in other locations in the column. Aside that we extend the inspection to other areas of the column and no more indicaions were found.

EdStainless, thanks for your reply. I am not saying that there are coating impervious to hydrogen. My process thinking on this is that if I isolate the base metal from the service then the corrosion reaction that produces atomic hydrogen will not take place and therefore there won't be any hydrogen to diffuse. Is this incorrect?

Kind regards,

 

[TGS4]

Your material - Belzona 1321 - does not indicate that it is impervious to hydrogen diffusion. I'm not sure how it would help.

I will be a third on the recommendation for cutting out a sample section and replacing with a flush patch. I completely agree with rationale provided by metengr and EdStainless.

A properly-designed bakeout and PWHT ought not require additional support. The column can be left in-place and heated with a fully-circumferential band. A full cicumferential band will avoid bending. We assist PWHT companies all the time to perform this type of work - it is rather common.

 

[NBrink]

It's not that it's impervious to hydrogen diffusion, it's that the hydrogen is created by the corrosion reaction on the steel surface. Isolate the steel surface and you stop the hydrogen production (and hence diffusion). It's not an uncommon approach at all. See it on fat amine flash drums all the time...

That said, there's really no way to know when that liner fails and you start charging again.

 

[r6155]

I insist. Consult with acoustic emisión company. No removing a section, no welding. Acoustic emission can be made in one day, and a 100% volumetric examination obtained. According to results you could go into service again.

Regards
r6155

 

[EdStainless]

There is no reason to trust UT or AE unless you have a reference piece of material that matches the vessel, and matches the types and sizes of defects in the vessel.
Theoretical and extrapolated calibrations are just that, they can easily be off by 50% or more. Unless you have ground truth samples you are estimating.

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

 

[dhengr]

Cespi82:
Laminations and slag inclusions were not all that uncommon in that thick a stl. pl. back in the 70’s. These usually showed up in about the middle third of the pl. width. You could pay a premium to get the pl. UT inspected at the mill, but that premium often turned into a longer lead time to get that more perfect piece of pl., because some fair share of the pls. turned up showing some anomalies. Then the mills would always (almost always) insist that they were only seeing minor slag inclusions, which they claimed should be looked on with less disfavor. I have several smaller prepared samples, than are suggested above, which have turned into paper weights over the years. They always brought the steel guys up short when they started arguing that there was nothing wrong with their pls., and the difference btwn. laminations and slag inclusions, etc. The reason these anomalies are usually not too serious a problem is that the primary stresses in the structures run parallel to the lamination, no appreciable loss of stl. area, no significant stresses perpendicular to the lamination, etc. So they did not usually cause a serious stress/strain or growth problem on relatively static structures. On the one hand, I agree with TGS4, ‘how do you know how long they’ve been there?’ But, if I were a bettin-man, I’d bet they are as old as the vessels, and you say there is no indication that they have caused any problems. I don’t know the ASME codes the way you guys do, but a couple things come to mind. Code size limits on something like this are not absolute or perfectly defined in theory, they are set by a code committee which has no incentive/reason to stick their necks out. There may also be stress conditions, etc. which can make them grow and become a problem, so that should be evaluated too. But, after 35 years, they’re all of a sudden a problem? That’ll teach you to inspect at that location again. Is there provision in the code which would allow you to map these anomalies and then monitor that they don’t grow over time? It seems to me that this ends up being an engineering judgement and experience call, which is then passed on to management for a final decision , weighing costs for various fix and monitoring approaches vs. the future risks and liabilities.

 

[TGS4]

r6155 - insist all you want - but what would you expect to discover from AE that the OP hasn't already found with PAUT? And, like EdStainless says, unless and until you know the truth, you don't have a calibrated NDE technique.

dhengr - I would likely agree that there is a high probability of those laminations being from original plate flaws. Nevertheless, if the environment is hydrogen-charging, then there is a likelihood of hydrogen accumulation in the mid-wall flaw. This ought to demonstrate the necessity of baseline scans, and periodic inspection - to catch these things growing.

 

[r6155]

@ TGS4. The principal objectives of AE examination are to detect, locate, and assess emission sources caused by surface and internal discontinuities in the vessel wall, welds, and fabricated parts and components.

Acoustic emission (AE) is a meticulous nondestructive examination (NDE) method that exposes deficient areas in pressure vessel integrity. It is the only NDE method capable of assessing volumetric integrity during a vessel pressure test period. When AE is used as a primary examination method during hydrostatic testing, it supports all other NDE methods.
Please, see more in ASME V and google

Regards
r6155

 

[TGS4]

r6155...

T-1220.2 All relevant indications caused by AE sources shall be evaluated by other methods of nondestructive examination.

Again, what would you expect to discover from AE that the OP hasn't already found with PAUT?

AE cannot characterize lamination flaws (or differentiate them from hydrogen blisters) insofar as their extent nor their thru-wall depth. It cannot ascertain whether the edges are smooth or rough. All it can do is generally locate a flaw. Hence T-1220.2.

 

[cespi82]

Thank you all for your comments/suggestions.

NBrink, thanks... It seemed to me that I wasn't explaining my idea properly and that I was being misunderstood about the non metallic coating and hydrogen permeation.

EdStainless / TGS4 / metengr, I got your point regarding cutting a piece of plate and the accuracy and calibration of our UT machines.

Not sure if management will buy the idea of cutting the plate for further examination, but at the end, as dhengr well mentioned: "this ends up being an engineering judgement and experience call, which is then passed on to management for a final decision , weighing costs for various fix and monitoring approaches vs. the future risks and liabilities"

I will incorporate a good summary of these recommendations/observations in my report mentioning all the scenarios and considerations. I will provide management with arguments and decision making criteria, and they will decide.

In case they go for the any option implying performing any welding in the vessel, either installing a cladding or the flush patch, could any of you provide me with some guidelines about how to design/outline the bake-out procedure and the local application?

Kind regards!

 

[r6155]

@ TGS4. Acoustic emission (AE) tell me if the pressure vessel can continue in use or need repair, irrespective of imperfections detected. AE define the integrity.

Regards
r6155

 

[TGS4]

cespi82 - hire a reputable heat treatment contractor. They will design it for you.

r6155 - no. No it doesn't. NDE (of any type) can do no such thing. Only a proper engineering assessment of flaws, once they have been properly characterized can provide guidance on that - see API-579/ASME FFS-1.