PDA

View Full Version : Stress Corrosion Cracking in NH3 Vessels



US Iceman
20-06-2006, 03:43 PM
I wanted to post some information on a subject that has been under investigation for ammonia systems. Thought some of you may be interested.

http://www.npl.co.uk/lmm/docs/stress.pdf

http://www.npl.co.uk/lmm/docs/stress_corrosion_cracking_basics.pdf

http://www.corrosion-doctors.org/Forms/scc.htm

http://www.osha.gov/SLTC/etools/ammonia_refrigeration/ammonia/index.html
Then click on the link "Keeping Ammonia Pure"

http://www.azom.com/details.asp?ArticleID=102

Have any of you ammonia guys ran into this problem?

NH3LVR
17-07-2006, 10:39 PM
Just ran across this forum and saw your post.
The company I work for is going to be cutting the head off a HP Receiver this week. Three years old and one of the heads is leaking. Hydrogen Embrittlement the most likely cause according to the Mfg.
This is the first time I have ever seen this.
Up until now It had only existed for me as a theoretical possibility.
Unfortunately we will replace the head and that will most likely be the end of it. I do not think that any further investigation will go on.

US Iceman
17-07-2006, 11:51 PM
Hi NH3LVR, ( i like that name:D )

I'm not a metallurgist or pressure vessel design engineer, but if one head is leaking, what happens when the other head starts to leak?

Has anyone checked this closely?

From the limited exposure I have had to this, any weld zone could be a potential problem area. Was the vessel stress-relieved in an oven before it was sent to the job site?

Does the receiver have a good refrigerated purger installed? I think oxygen in the non-condensable gas can also accelerate this cracking.

I guess I would hate to see someone go to the trouble of cutting off a vessel head, re-fitting a new head, re-certifying the vessel, and then find it starts to leak shortly after the repairs are finished.

That is one I would hate to explain to an owner.:(

NH3LVR
18-07-2006, 12:44 AM
I am not a metallurgist either. According to my info the chances of the other head staring to leak are slim. This is the first time I have ever seen this.
The State Inspector is closely involved. We will be going inside the vessel and looking at it when we cut the head off.
I do not know exactly what can be determined from a visual inspection however.
This vessel was not stress relieved. The Mfg now does this routinely.
The system does not have a purger installed. Up until recently they hand purged every day. I believe a Auto Purger will soon be installed.
The original thought was to replace the vessel, however the downtime was not acceptable.
Instead a temporary receiver was installed this weekend.
The owner has shown great patience with the repairs.
I have had little involvement with this job. I took the pictures when the problem first arose. I will get a chance to see the head when it is removed, unless I am working out of town next week.
I will take photos, although I do not think that will be of any practical use.

US Iceman
18-07-2006, 02:43 AM
Have you tried the dye penetrant yet to see if any other cracks are present? Most of the pictures I have seen all show the cracks near the top of the vessel (in the vapor space) and near welds.

If you get the chance to view the inside, is someone going to check it visually, or use a non-destructive method? Odds are you will find a lot of internal cracks, that have not surfaced yet. Literally.

I personally don't think manual purging does that much. It might get some of the non-condensables out, but I think more ammonia is vented than non-condensables. The Auto-Purgers do a very nice job!

If you get some pictures, can you post them here for others to see? That would be helpful.

US Iceman
18-07-2006, 02:49 AM
I forgot to add a comment: Stress relieving on pressure vessels is NOT normally performed unless requested.

I know of one manufacturer who provides this as an optional cost for every vessel quoted. You might mention this to your office designers, etc. to make sure they know this.

Another option is: corrosion allowance on vessels. This might make sense for vessels exposed to wet conditions, or vessels that routinely have their vapor barriers punctured.

Surge drums and intercoolers come to mind.

NH3LVR
18-07-2006, 04:28 AM
We have not tried dye penetrant yet. I understand these cracks start on the inside so I would not think they could be found at this point from the outside.
Someone will go inside the vessel, but the Inspector will ultimately be the one to specifiy the test method.
As to purging I was recently in a plant that had a manual purger installed 30 years ago. ( I worked on the job)
The operator had been using it with a valve shut off for years and had no idea he was accomplishing nothing.
Manual purging is effective if done right. Of course it only removes enough non-condensables to keep the head pressure down. An Auto-Purger is far and away the best bet. I recently forgot to close a valve and blew some N2 into a system through the Discharge check valve. Turned on the Auto purger and forgot about it. Love them, although I do not totaly trust them. They sometimes decide to purge a little too much and fill the water container with NH3.
I instruct customers to run the purger during the day, unless they have a severe problem.
The local Mfgs just started doing the stress relief in the last few years. Not soon enough in this case!
I am attaching a couple of pictures I took of the vessel when the leak started. You will notice two holes very close to each other.

Johnny Rod
18-07-2006, 11:36 AM
Is it possible to use ultrasonic imaging for vessels like this, before you open them up? I would have thought the differences in structure and density caused by cracking would show up. Or perhaps another non-destructive imaging technique? Not my area really as you can tell.

US Iceman
18-07-2006, 04:10 PM
They sometimes decide to purge a little too much and fill the water container with NH3.


I suspect this problem is not due to the purger. It is probably due to liquid backing up in the condenser coil (above the purge point). When the purge solenoid opens, liquid is vented into the purger instead of gas.

This sounds a lot more like the condenser coil is stacking up liquid due to poor drainage or lack of sufficient equalization through the equalizing line.

US Iceman
18-07-2006, 04:14 PM
Is it possible to use ultrasonic imaging for vessels like this...


I'm not sure either. I would really like to get someone who does NDT (non-destructive testing) to contribute to this area in the forums.

This is certainly not my area of expertise.

SteveDixey
18-07-2006, 05:50 PM
One area I think known to be a problem with this is areas subject to freeze\thaw action like the entries and exits from \ to tops of low temp vessels that operate both below and above freezing point.

Never heard of an HP receiver having those kind of problems as SCC is linked in most circles to ammonia liquid at low temperatures.

Hydrogen embritlement might be more an issue with inappropriate manufacturing. If there has been some way hydrogen sulfide was formed inside vessel, that might be an issue (NH3 contaminated with other gases?), or corrosive gases have sat in a "dead end".

As far as I am aware, any embrittlement can be contained providing the joints are not subject to appreciable stress, by correctly aligning pipes and ensuring nothing sags.

Had some issues with ultrasonics on "used" plant where corrosion was external but from what I have picked up over the years this cracking is difficult to detect and predict as it tends to be diffuse.

Steve

US Iceman
18-07-2006, 06:53 PM
Hi Steve,



One area I think known to be a problem with this is areas subject to freeze\thaw action like the entries and exits from \ to tops of low temp vessels that operate both below and above freezing point.


I believe this falls under the scope of external corrosion. As you suggest, repeated freeze/thaw cycles are a huge problem.

My concern with this subject is that it could be prevented if the vapor barriers were 100% tight, which is not likely.

Almost all vapor barriers have some very small leak paths, so this will always continue to be a problem. Even if the vapor barrier would 100% water-proof and would not allow water vapor transmission through the material, we would still see problems due to installation of the vapor barriers.

I think the insulation system is extremely critical to provide correctly, however, we know this can be a very expensive proposition.

Here is one rather ingenious solution someone developed to counteract this problem.

http://www.insuldry.com/solutions2004.htm

Some other ideas might include the use of a corrosion allowance in the the design thickness of the vessel. This would build in additional material thickness to increase the useful life of the vessel.

Another idea may be to use stainless steel.

US Iceman
18-07-2006, 07:04 PM
Never heard of an HP receiver having those kind of problems as SCC is linked in most circles to ammonia liquid at low temperatures.


Most of the vessel problems related to SCC I have heard of are in the HP receivers. Some of the thoughts are this is related to the lack of adequate purging and free oxygen, and the lack of stress relieving during manufacturing.

There is a big effort right now here in the US to provide post-weld heat treatment to vessels to relieve the stresses induced by welding and fabrication.



...by correctly aligning pipes and ensuring nothing sags.


This is another important subject too. Additional stresses can be "forced" onto vessel connections that can impact the overall problem. To my knowledge, no manufacturer of refrigeration vessels includes allowances for these "nozzle loads".

I think the safest procedure is to state the nozzles should not have any additional loads supported on them. Which I believe is to the heart of the matter you mentioned.

Adequate support and restraint is very important to provide.

Johnny Rod
19-07-2006, 09:26 AM
The little I know about SCC, from the point of view of the ammonia, is that oxygen, halides and water are the apposite contaminants. Oxygen will tend to accumulate in the high side vapour spaces, and being an oxidant will aid almost any corrosion process. NCG purgers should of course keep this low. Halide ions, which means chloride, bromide, fluoride and iodide, can help accelerate corrosion - chlorides are known to go for ferrous metals, which is one reason seawater is so good at rusting things, but really almost any inorganic chloride will do. These will tend to remain in the liquid phase, and could come from things like flux residues. Finally water, I think you already know it looks like dry ammonia is a factor - doesn't sit well with the IIAR's recommended ammonia spec but I think the usual figure of at least 1000-2000ppm (0.1-0.2%) moisture is generally accepted to rule that one out.

I think greater factors are physical things like manufacturing stresses and temperature cycling really get the process going, it's not a simple process, but that's what I know about the ammonia itself.

SteveDixey
19-07-2006, 07:55 PM
Hi Steve,



I believe this falls under the scope of external corrosion. As you suggest, repeated freeze/thaw cycles are a huge problem.

Another idea may be to use stainless steel.

One company I know feels it can still be an internal problem because one thing that they feel is prudent to do is thoroughly check steel vessels that are being changed over from R22 to ammonia duty as R22 is phased out, as ammonia tends to accelerate the process.

Even stainless is not immune, although chlorides tend to be the main culprit there.

Steve

US Iceman
19-07-2006, 08:30 PM
Granted, refrigerant change outs add to the complications. In the case of R-22 going to ammonia even more so.

Do you know if the company has had any problems with a powdery residue after the ammonia is introduced into the system?

What type of cleaning procedures are they using?

My comments about using SS were more for oxidation induced corrosion on the external surfaces, but you're right... the use of SS is not a cure-all for material use.

SteveDixey
19-07-2006, 09:45 PM
Do you know if the company has had any problems with a powdery residue after the ammonia is introduced into the system?

Yes, they have had that problem. What they did as a flush I'm not sure, but I think they costed a change of ammonia into the project price, but this was a FrigoPack with about 600kg of NH3 in it and very short pipe runs.

Another less than reputable company used some kind of solvent and then used water to pressure test the coil. This was going from Ammonia to R22!

I heard something about them having a hell of a game getting all the solvent\water mix out the evaporator tubes (a Frigo spiral freezer), and it spent a whole week on vacuum with three pumps pulling, but details were sketchy as the project engineer running the show was obviously keen to keep details quiet.

At the time it was well dodgy to be doing that as R22 should not have been used as it was effectively a new installation. Given the hassle, keeping it on ammonia would have been an easier and cheaper option as four weeks production orders were lost due to delays in getting the plant going!

Steve

US Iceman
19-07-2006, 10:05 PM
I would not want to be that project engineer when the owner asked him to explain the delay.

That cost someone a lot of money.:eek:

Why in the world was someone retrofitting R-22 to an ammonia system? That is the first time I have EVER heard that one.

That must have been a very good "sales job".:rolleyes:

SteveDixey
20-07-2006, 12:17 PM
That must have been a very good "sales job".:rolleyes:

It wasn't a sales job as such. Just a case of find a smaller evaporator for that freezer and fit it as cheap as possible. The fact that it had a bloody great old recip chugging away on it and the rest was "not important".

Glad I got out when I did.

Steve

Andy
21-07-2006, 10:02 PM
I would not want to be that project engineer when the owner asked him to explain the delay.

That cost someone a lot of money.:eek:

Why in the world was someone retrofitting R-22 to an ammonia system? That is the first time I have EVER heard that one.

That must have been a very good "sales job".:rolleyes:

I fitted an ammonia condenser on a large central R22 chilled water system about January time:D

The condenser was galvinized (Baltimore) I left the liquid valve open for a week, purged it with nitrogen and vacumned it out over two days.

No problems to report, system works well, with no build up of deposits in the system. Had my doubts at the time, but I have to say it was fine:)

You can do things like this, but why bother:confused:

Kind Regards Andy:)

US Iceman
21-07-2006, 11:46 PM
I guess my concern was more towards the fact an ammonia system was being changed to R-22 Andy. Usually, it's the other way around R-22 to ammonia.

As long as there is no copper or brass in the system.:o

Andy
22-07-2006, 11:25 AM
As long as there is no copper or brass in the system.:o

:D :D

now that would be funny

Kind Regards Andy:)

Ponca Dave
10-02-2008, 03:41 AM
Just ran across this forum and saw your post.
The company I work for is going to be cutting the head off a HP Receiver this week. Three years old and one of the heads is leaking. Hydrogen Embrittlement the most likely cause according to the Mfg.
This is the first time I have ever seen this.
Up until now It had only existed for me as a theoretical possibility.
Unfortunately we will replace the head and that will most likely be the end of it. I do not think that any further investigation will go on.
The Natl Board has studied this extensively and has concluded that Ammonia should contain at least 0.2 percent water to inhibit SCC.

NH3LVR
10-02-2008, 04:48 AM
It has been a year and a half since I posted those pictures of the vessel. My first post on RE was on this thread, if I recall correctly.
I just wanted to update a bit.
We installed a temporary receiver in the system. When we applied for a permit it was found that the condenser installation did not meet structural codes. This started a very expensive chain of events.
We installed a new receiver and condenser. Another contractor built a new condenser stand and set it in place.
However there has been no time for the plant to shut down, and the final cutovers have not been made. When they are made I hope to get pictures of the inside of the vessel.

US Iceman
10-02-2008, 06:03 AM
It would also be interesting to find out if the new receiver was specified with heat treatment for stress relieving. One of the US vessel manufacturers has been requesting customer to specify this to reduce the residual stresses created by manufacturing. This is in addition to the 0.2% water content.

NH3LVR
10-02-2008, 07:20 AM
Yes, the new receiver was stress relieved. After that experience no one wants to go through that again.
As to the water content, it would be nice to have a simple, but accurate way to measure.
In this case I am sure the plant did have water in the system. Apparently it was not effective in preventing the problem.

hendry
15-02-2008, 07:41 PM
Hola pals, i have read the newsletter by RVS.
as far as i remembered, we have been educated in our Material Science subject that PWHT is a necessity.

why those smart fellows do not comply?
all about $ ....!

now, their incompliances has incurred to us, the user, extra expense/s.

what an unjust event!!!