Corrosion of A249 TP316L HT exchager tubes 1

[jsobilo1354]

I am seeing significant corrosion rates (.020"/yr.) on A249 TP316L HT tubes (.750",.060" AW) in rich amine service (CO removal). These shell and tube heat exchangers were replaced 3 years ago(in-kind). The previous exchangers were in service a little over 10 years. I have pulled samples and in the process of completing metallurgical examinations. I have also sampled tubes with eddy-current but was only able to inspect a small portion of the entire length of the tube due to a hard, tight black corrosion product that could not be removed(10,000 psi hydroblasting could not remove this scale completly). The areas that could be tested revealed 50%-60% wall loss from nominal "t".

Has anyone seen these type corrosion rates in this service with this type SS?

 

[EdStainless]

Is this general corrosion or pitting? Concentrated at the inlet? along the weld?

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[jsobilo1354]

Eddy current inspection showed general wall loss in areas where the scale could be hydroblasted to 50%-60% of nominal "t". The actual thru wall ares of corrosion was isolated corrosion, possibly under deposit in nature.

The sampling of the corrosion deposits was just completed on the SEM. The deposits consisted of 47% iron, 34% oxegen, and 17% carbon.

The high oxegen presence in the amine system was recently attributed to an upstream operational problem. Do you know the effects of O2 relative to corrosion rates in these type environments?

 

[EdStainless]

Oxygen will not accelerate corrosion in stainless. My hunch is that it is all iron oxide. Under those deposits there was no available oxygen.
You have pulled tubes? When will you have lab work?
You cannot trust ET to tell you the type of corrosion. You need to see tubes.
There are not a lot of environments that will cause general corrosion of 316. Has this unit ever been acid cleaned? What is the operating pH?

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[jsobilo1354]

Have removed tubes and should have results in 4-5 days. My concern is this grade stainless typically performs very well in this service. With the Moly content inhancing the corrosion resistance and the low carbon content,less likely to sensitize.

The operating PH noted in the last solvent analysis was 10.69. Also, All Anions (Acetate,Formate,Chloride,etc.) and Heat Stable Salts (HSAS,IHSS) were all well within the Upper limit ranges.

My only thought is during original fabrication (after solution annealing), the bundles were not cleaned,leaving scale formed by the HT process, which would not allow the protective "passive" film to form. What are your thoughts?

 

[EdStainless]

Does the unit leak? Do you have general wall loss, or are some places fine and others pitted through the wall?

pH 10.7 should be fine.
I hate to ask, but where were the tubes manufactured? Did you have people witness tubing testing? bundle fab? Was there a base-line ET run on the unit?

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[jsobilo1354]

There are five gas treaters that have three amine shell and tube exchangers in series, associated with each train. Two of the trains have exchangers leaking (both are the first exchanger of the series). Since we could only look at a small portion of the total length of each tube with FEC(3ft. of 20 ft.) due to the inability to remove the corrosion product from the ID of the tubes, and with the triangular desgn of the tubes, the known extent of the corrosion is based upon what little FEC results we have and the two outer tubes that were removed. The two tubes removed did have general wall loss throughout the tube which ranged from 30%-40% wall loss from nominal.

As we did not own this facility at the time of installation I do not know the extent of QA/QC that was performed during fabrication and the equipment file as usual, doesn't have much info.

 

[EdStainless]

Are you sure that the tubes are 316?
You should also do some analysis on the scale to determine its structure, not just composition.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[jsobilo1354]

The tube matrial is being verified at the lab to assure it meets the composition requirements for A249 TPA316L. Additonal testing is in progress, the composition was just part of the preliminary results. Full metallurgical examination/analysis results re still pending.

 

[robsalv]

We operate a DEA gas scrubbing here, though I've never seen anything like what you've described.

Our guiding info warns against acid gas breakout though, which can occur if the DEA is overloaded... which have nothing to do with your situation, but suggests that you may also need to look at the control and quality of the process side in your investigations... covers all your bases.

Come back and let us know what you've found.

Thanks.

Rob

 

[unclesyd]

We operated 2 Girbotol CO2 removal plants for H2 purification for many years with periodic periods of very high corrosion rates until the installation of metal edge filters and carbon towers. Each MEA train consisted of 9 Hxs' with MOC of 316L, both shell and tubes. Installation of the filters immediately helped with the problems of plugging/fouling, but didn't completely eliminate it. The second step was the addition of side stream activated carbon towers, using an SGL carbon. The towers went a long way in mediating the corrosion after several carbon cycles. Initial life of a tower was quite short and required close monitoring even after the initial cleanup period.
I have been unable to find out the specific carbon we used.
I keep thinking it was a Pittsburgh material.

Auto Clean

http://www.cuno.com/industrial/petroleum.shtml

Our biggest problem was still fouling and corrosion of the exchangers and reboilers.
The carbon towers and filters helped but didn’t completely resolve this problem. We were constantly rebuilding these tube bundles in our shop due complete plugging caused by the close tube spacing and the triangular pitch which precluded any type of blast cleaning.
The layout man in Fab shop and myself solved this problem by sneaking, (midnight plant test) in a tube bundle, reboiler, with 25% less tubes and on a square pitch. The square pitch was for cleaning if necessary. I have looked at some notes where it was deemed that preventing the hold up of the rich MEA went a long way in preventing formation additional deposits and corrosion.
When both plants were decommissioned 3 years later all exchangers were of this design. None of the square pitch bundles were ever pulled for cleaning. The onstream time for the carbon towers and filters also increased.

Hydroblasting
There are hydroblasters and there are real hydroblasters. I can't say how many times that I've caught hydroblasters changing the tips on a lance to take some of pressure off the operator when manually blasting tubes. We had one incident where blaster took out over 100 tubes in 1200 tube bundle. It got to the point that we required mechanical lancing for all tubes. Even then the selection of the tip is still very critical to keep from eroding tubes.
The operating parameters of tube blasting have to determined and watched during the each operation on a specific bundle. I developed a specific list of the operating parameters needed to clean a specific bundle especially the more difficult ones.

 

[jsobilo1354]

All metallurgical examinations are complete. The tubes do meet the ASTM criteria for A249. We also calculated the velocity of the Rich Feed at the exchangers and it was 2.8 ft/sec which should not be an issue.

We are looking at other process variables to assure flashing is not occuring. Are records are inconclusive regarding treatment of the tubes after solution annealing during the original fabrication. At this time that is what I am leaning towards but will let everyone know the final results. Thanks for all the input on this matter.

 

[EdStainless]

I don't see that this has anything to do with the inital condition of the tubes.
The fouling and underdeposit corrosion are the issues to deal with.

= = = = = = = = = = = = = = = = = = = =
Rust never sleeps
Neither should your protection

http://www.trent-tube.com/contact/Tech_Assist.cfm

 

[jsobilo1354]

If the tubes will not properly cleaned after solution annealing the iron oxide scale would remain and would not allow the protective film to form leaving the tubes vulurable. As far as the deposits go they are very typical in this service yet other exchangers are not seeing this type of accelerated corrosion with the same A249 material.