Crack at tube to tube sheet weldment

[leedrong]

Dear All

We have very urgent and critical problems on operation starting stage of boiler feed water heater. The heater was erected after hydrostatic pressure test at fabricator's shop without any problems.

* Shell --- SA516-60 ,ID1130 * 44t
* Tubes ---SA268-TP430 , OD25.4*2.108t*3250L
* Tube sheet --- SA182F304L , 110T
* Weld Rod --- Inconel 82

But on comissioning stage we found few tube end welded with tube sheet were leaking.
On rehydrostatic pressure test after repair them , There was 10 moer another new tubes leaking.
We have no idea why the tubes leak are spreading because the tube end weldment was confirmed by radiographic test every each tube joint with tubesheet.

Design of shell and tube sheet and tubes re-confimed by calculation that all components have sufficient thickness and stregth.

One thing is we did not post weld heat treatment of the weldment , tube to tubesheet joint.

Firstly we want know the cause of leak , crack by hydrotest.

Anybody have idea and good suggestion?

 

[unclesyd]

You have a major problem if your tubes are truly TP430 SS.
This is first for me.

430 SS has to be annealed after welding as it forms a very brittle phase on welding.

Please verify that the tubes are 430 SS.

What was the weld metal?

Were the tubes rolled and welded or just welded?

 

[metengr]

What was your hydrotest pressure, and is this a high pressure or low pressure feedwater heater?

Normally, for feedwater heaters the feedwater tubes are either roll expanded and seal welded into the tubesheet or explosively bonded and seal welded to the tubesheet. If this is just a seal weld - PWHT is not required.

This sounds like a very serious fabrication problem. I would contact the OEM immediately because this is a warranty issue.

 

[SnTMan]

I agree with syd, if tubes are 430, you got problems.

I'm curious, how were you able to RT tube-ts joint? Is it a buttweld detail? Is it possible your RT didn't really tell you anything?

Mike

 

[leedrong]

Thanks ~!

Tube to Tubesheet joint is strenth weld with RT except expand roll .It was a licenser's standard. PWHT was not recomende by licenser. The weld joint figure is tubesheet side J-groove 3mm depth 2.2mm width ,tube end is 1mm below than tubesheet surface, the total weldment thickness 4mm.

RT of tube end weldment is not difficult when we use special source and fixture.
Shell side test pressure is 106 bar.g and design is 80 bar.g
Would deflection of tube sheet by test pressure be a cause of crack ?
Required thickness is 106mm by calculation but used is 110mm.
Anybody know how to calculate exact deflection amount of fixed tube sheet on the test pressure ? and what value could be cause of crack of tube weldment?
Is it be a counter measure to change joint figure tubesheet side J -groove and fillet weld with make tube extrude 3mm from tube sheet surface and plus heavy expand roll ?
There would be many possibilities but we have no idea of exact cause.
Sudden steam supply from tube side could be cause of tube welment crack ?

 

[metengr]

leedrong;
One last but important question, where exactly are the leaks - along the tube side of the partial penetration weld or thru the weld itself?? This is very important information to confirm. The location of the leak along the heat affected zone of the Type 430 ss tube would imply crack initiation and propagation in the region where martensite could form from welding. The formation of martensite in the base metal HAZ coupled with residual stresses would render this region very susceptible to crack initiation and propagation.

If the crack is in the centerline of the weld, this could imply wrong filler material or porosity in the weld joint from poor cleanliness during fabrication.

 

[GenB]

I believe the weld gave-in because the weld along does not have sufficient strength to hold the shell pressure; expanded tubes give better result, then the weld becomes just a seal-weld.
genb

 

[rmw]

I have two questions.

First, were the tubes seal welded and then rolled or rolled and then seal welded. (that is one question)

Second, how far behind the seal weld did the rolling process start?

If you answer these questions, depending upon the answers I may have further comments.

And, I guess I just thought of a third question that is simply a matter of curiosity. Is the entire TS 304 or is it just a 304 overlay?

rmw

 

[leedrong]

Thanks ~!

Mr.metengr
The crack exist in circle along with welment mostly . But some of them are cross crack .
Do you think the selection of welding rod Inconel 82 was bad choice than ER309 ?
We will change tube material to ss304L from ss430 When we fail stop the crack.

Mr.rmw
We did strength weld only without expand rolling .
Tube sheet material was solid SA182F304L .

 

[GenB]

I believe leedrong
wants to hear that the inconel is the responsible one for his mishaps,
he did not rolled the tubes, therfore it is very difficult to assess the weld failure.
I still believe the weld alone could not hold the pressure.

leedrong again said that the tubes were not rolled.
further he called strength weld to the welded-only tubes
in the case w/o rolling the welds are not strength welds or seal welds.

genb

 

[unclesyd]

As asked before is the cracking in 430 tube proper or the filler metal?

What was the tubesheet hole clearance?

How many tubes in the heater?

Did you test the tube welding procedure prior to assembly?


If the cracking is in the 430 tube this is one problem. If the weld metal is cracking then there is another problem.
If the tube is cracking that tube is lost, either replace it or plug it.

If the weld metal is cracking then you have essentially the same option as above as any welding on 430 can produce the brittle phase in the 430 even if you start over again with a different filler.

Either condition brings in the question operational reliability.

The removal and replacement of the affected tubes
sounds like the best approach.

I believe rolling the tubes now without serrations in the tube sheet will present just as many or more problems than exist now.

In our nitric acid plant
We used Inco 82 (A) to join 430 SS to 304 and 310 SS but with a post weld anneal. I don't recall any problems.

 

[metengr]

leedrong;
Your question "Do you think the selection of welding rod Inconel 82 was bad choice than ER309 ?" No, I don't.

From your recent observation, it appears that the cracks are caused from welding the Type 430 stainless steel with no preheat and no anneal (PWHT), if these were strength welds. The Type 430 stainless, as-welded, exhibits poor ductility in the weld HAZ. This would explain the circumferential and even transverse cracks (from residual stresses).

You could attempt to salvage this heater by roll expanding the existing tubes in the tubesheet and remove the cracked strength welds. Seal weld the rolled tubes using a preheat and Inco 82. If this does not work, about the only advice I could give is to replace the Type 430 tubes with Type 439 (if you need ferritic stainless steel in this application).

 

[vesselguy]

TP430 is a P7 material and per UHA-32, the HAZ required PWHT. Martensitic S.S. is brittle unless heat treated. It is possible you guys fractured the tubes. How did your fabricator get away with no PWHT on P7 material? Did they prove to you by micro harness testing on the HAZ to prove that the hardness is less than say 350 Vickers with the WPS to get your exception to PWHT?

I think you need to 1. Hire a metalugist to review on the WPS. and 2. Get the fabricator to come to your site and examine what is wrong. 3. At the same time, hire a NDE specialist to do some MT or PT of the tube and tube to tube sheet joints to find out where the cracks are.

 

[leedrong]

Our final decision is 1) change the type of tube material from TP430 to 304L 2). change the welding of tube to tube sheet from strength weld only to heavy expanding plus seal welding .
Many thanks for your good advice.