Inspecting duplex tube to header welds for root erosion/cracking

[ndeguy]

We have air coolers on our plant (oil & gas) part of which consists of a row of 1 inch dia, 1,6 mm to 3,5 mm wall thickness duplex (22% Cr, UNS31803) tubes TIG welded with matching filler metal to an 8 inch duplex header (wt 11 - 18mm). Weld ferrite content range 35% - 65%.

I am no corrosion engineer but it seems that duplex is susceptible to weld root erosion or cracking in case of wet-upset in an H2S environment. So we wish to check the weld roots for such whilst new equipment fabricated from material more-suited to the potential environment are procured.

A simulation test rig of one header with 4 or 5 tubes welded in place is being made for performance demonstration purposes. Does anybody have experience of such configurations? Is low-frequency eddy current a possibility? Or DC MPI - knowing that the permeability variations may produce non-relevant indications. As the material/configuration set-up is difficult to inspect I expect signal interpretation and evaluation will be a problem. Any ideas?



Nigel Armstrong
Karachaganak Petroleum
Kazakhstan

 

[metengr]

Are the duplex stainless steel Hx tubes roll expanded and seal welded to the header or are the Hx tubes partial penetration welded (eg, strength weld)to the header?

 

[ndeguy]

The latter, though hopefully with full penetration. The tube bundle construction drawing shows tube bevel as square edge with 3mm root gap. The WPS covering this work shows a 40 deg bevel with a 2mm root gap, 5G welded. We are waiting on a response from the manufacturer as to the actual fit-up, i.e. as the dwg, as the WPS or some other! The PQR was on pipe to pipe. At present we have no access for videoborescope to check the penetration achieved. The tube wall thicknesses vary from 1,65mm to 3,05mm.

I made a mistake on the headers dimensions - they are in fact 120mm OD x 16mm wt and 106mm OD x 14mm wt. Material ASME SA182 F51, tube ASME SA 789 UNS S31803.

Thanks

Nigel Armstrong
Karachaganak Petroleum
Kazakhstan

 

[metengr]

ndeguy;

I can't be sure, but I would expect that the tube to header weld joint detail would consist of a partial penetration weld versus a full penetration weld.

The weld detail would probably consist of a shoulder that is machined into the stub tube hole bore at some distance from the OD surface of the header. The sides of the header wall starting above the shoulder would be beveled at some angle to accommodate a partial penetration weld.

If the weld joint configuration mentioned above is confirmed, I would expect that ultrasonic examination using an angle beam approach might be your best option.

 

[EdStainless]

ET will be problematic because of variations in magnetic characteristics. If you are not at full saturation you will have trouble.
i like the sound of UT better, especially considering your geometry.

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

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

 

[ndeguy]

Thanks to you both.

As the tube thickness is only between 1,6mm and 3,5mm the only option I can envisage for UT is from the header OD and bouncing the sound off the header ID up into the weld. It'll have to be with a very small transducer so that it does not rock. And then I shall have to think up a manner of orbiting the transducer around the tube whilst maintaining the correct skip distance. Its a difficult one -that's why I had hopes for eddy current.

When a mock-up is available the various NDT possibilites can be tested. Until then its on paper as I feel it would be unrealistic to expect a definitive root assessment without the NDT tech being able to experiment on the mock-up.

Incidentally, a colleague has had experience of a similar cooler configuration failing in less than a day after going wet!



Nigel Armstrong
Karachaganak Petroleum
Kazakhstan

 

[ndeguy]

metengr

Just to reaffirm - the tube is butted up against holes drilled (?) in the header, with a 3 or 2 mm root gap and either no prep or 40 deg prep respectively on the tube, depending on whether the Italian manufacturer adhered to the construction drawing or the WPS. Still waiting confirmation on that.

I have drawn out the configuration and, even coming from the thicker-walled header, I cannot envisage gaining practical results with ultrasonics. I may be wrong - I certainly won't rule out using ultrasonics until performance demonstration tests prove one way or another.

Ed

I thought a relatively high ferrite content would make eddy current difficult. Again its not ruled out and we will trial it asap.

Thanks gents

Best regards

Nigel Armstrong
Karachaganak Petroleum
Kazakhstan

 

[unclesyd]

Just a little thought about your failure. Have you looked at this joint design from a thermal/mechanical point of view. The reason is that it has been my experience that any type of environmental cracking in a heat exchanger normally takes a little longer to initiate.