Caustic Cleaning - Concern or Commnon Practice?

[myname_earl]

Hello,

I recently came across our plant's reboiler cleaning procedure as well as our shell-and-tube cleaning procedure. We have various materials from SA-285 to 316SS and have seen historical caustic corrosion in our reboilers... Not BAD, but has pushed us to replace the equipment entirely (60 years of age).

I am wondering what amplitude of corrosion or CSCC can I see from caustic cleaning. The Reboilers are circulated for 8 hours of caustic @180F.. roughly 15% concentrate. The shell-and-tube procedure calls for 16 hour soak in a caustic bath @ ~170F, 15% concentrate.

I am just curious how bad this is for our equipment - SS vs CS, and any recommendations for other cleaning options. I've also heard of scenarios of circulating caustic with residual product in the exchangers and creating hydrogen atoms that then produce hydrogen embrittlement... Is this realistic?

Mechanical Integrity Specialist (Year 1)

 

[0707]

I would say yes. It is realistic.

Regards

luis

 

[EdStainless]

It really depends on how good your rinse is after cleaning.
It needs to be very good.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[myname_earl]

I wouldn't say its a spectacular cleaning job. It's pretty good but to say there isn't any residual caustic would be a lie.

We have almost reached t-min on a heat exchanger that is now ~5 years old with wear at the baffles and some general corrosion internally. Very noticeable at the baffle to shell interface.

First the MI at the time assumed cooling water corrosion, then it was galvanic (SS bundle to CS shell), now it's pushed towards crevice corrosion as the damage mechanism. We did not have this problem with its predecessor which lasted 15 years.

Our water treatment vendor is always changing, but it seems we can rule them out from the problem as they have data from coupons and other sources to prove the water is OK. Then what? Galvanic corrosion makes sense after a long period of time but since CS and SS are fairly close to each other on the galvanic corrosion chart, the 5 years doesn't make sense. We have ruled out MIC based on water tests. We deep clean all of our exchangers roughly once every 5-10 years, otherwise is cleaned with firewater. Maybe we have a caustic corrosion problem - unlikely but not off the table yet. under-deposit attack makes sense since the flow seems to leave deposits at the baffles and could very well develop into worse.

I believe I have my options narrowed down to 1, maybe 2 damage mechanisms. I want to check if this makes sense based on the damage mechanisms from API 571... The caustic isn't necessarily a problem in this case, more of a plant wide concern from someone very new to materials, damage mechanisms, and mechanical integrity in general

Mechanical Integrity Specialist (Year 1)

 

[EdStainless]

Rotating water treatment vendors is bad news.
Each time that treatment is changed it changed you are likely stripping off the existing corrosion films and building new ones.
And this translates into metal lost.
Normally I would think that BFW has low enough conductivity that galvanic corrosion between CS and SS would be minimal.
It is common to see it though in condensate, even if it only has CO2 in it.
Figure out what the deposits are. Get samples and have them analyzed.
SS tubes and CS shell? Are the baffles and tie rods all SS?
Where is the metal loss?


= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[myname_earl]

The baffles and tie rods are also SS. We are seeing metal loss internally on the shell where the baffles sit - that's why their assumption was galvanic corrosion. I am under the impression of a damage mechanism related to flow.

Mechanical Integrity Specialist (Year 1)

 

[EdStainless]

It could be a combination of mechanical wear, galvanic corrosion, and flow effects.
Have you considered using wear pads inside the shell?

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[myname_earl]

I have not, I know we had issues even getting the bundle into the shell, and not sure of the tolerance stack if we can. Our "temporary repair" was to add a layer of belzona/JB weld/industrial epoxy to the internal wall of the shell, then have it milled out back to the original ID. Whether this has worked or not is beyond me, we will find out on the next internal inspection ~2 years from now. Overlay/more permanent repairs were not made based on the demand of the exchanger. What is destroying this exchanger and one other is relatively beyond myself until we open it again.

Mechanical Integrity Specialist (Year 1)