Cracks welds SA213-T23

[fborela]

Hello
Please, I need your help and I thank you in advance.
We Weld several tubes 42,4x4,5 SA213-T23 to header 219,1x30 SA335-P22. We preheat to 200 ºC. We weld with GTAW process and filler metal ER90S-B3. PWHT was made 1,5 hours to 750 ºC.
The joint was the tube straigth beveled and the header machined similar to a socket welding. Gap of 2 mm was respected between the end of the tube and the header.
After PWTH we detect a lot of cracks just up to the Weld in the tube of T23, perhaps in the heat affected zone or upper to it.
We didn´t know the cause. It´s a very serious problem. Hydrogen cracking don´t appear to be the cause (preheat and PWHT was made) in according with ASME I. The cracks are transversal of the axis of the header.
Can you have an idea of the problem??
Thank you
Fernando

 

[metengr]

Nando17;
Based on limited information, your description of the cracks sounds more like hydrogen cracking if they are transverse oriented. I know that you mentioned you followed ASME Section I for preheat and PWHT maintenance. However, this is no guarantee that hydrogen cracking will be avoided. Moisture, cleanliness and electrode maintenance are all factors in working with T23, a creep strength enhanced ferritic steel. I have seen where prior to PWHT, moisture had come in contact with welds in Grade 91 and T23 materials, resulting in stress corrosion cracking after PWHT.

 

[pandithan]

Pl do intermediate SR with less temp and cool it before taking for PWHT. This is because of diference in thickness and will happen in this material if u are not treating properly

 

[stanweld]

What was the time interval between weld completion and PWHT? Were the first tube-to-header welds more subject to cracking? Did you concentrate preheat on the header? Where were T/C placed - on the header? In the furnace?

We have had occasion where similar hydrogen cracking was incurred on T22 to P22 tube-to-header welds when preheat and PWHT was inadequately performed.

 

[stanweld]

Lastly, I recommend that you perfom a metallurgical analysis on one of the cracked tubes as to cause.

 

[fborela]

Hello
Thank you for your answers.
the pwht was made with electrical resistances and blankets of insulation to all the header (About 4 m length). Is almost imposible to say that the problem was HIC, because we preheat to 200 C, we made pwht asap the last weld was cold and the electrodes were maintened hot.The T/C were placed in the welds.
We are thinking that it is a design problem because of The preheat was made with torch of propane before every weld and the temperature very well controlled with infrared pirometer. Too the interpass temperature was controlled.
branches were too short up to the fixed point in the wall of the boiler. So when the length of the header is increased because of the temperature of the pwht the branches try to prevent the displacement creating high strengths.
To clarify. The cracks were parallel to the welds, no transversal,but always in the same position respect to the header (in the side in front 7of the another branch). So this would be the cracks you'd get if you keep fixed the branches and you stretch the header, for example with a pwht.

 

[metengr]

Nando17;
Thanks for the clarification and follow-up. I agree with stanweld you need to have an analysis performed to confirm your suggestion. If the cracks are confirmed as tensile overload, you need to adjust your PWHT to spread the heat over a larger area or wrap the entire header to avoid thermal gradients.

 

[stanweld]

We had major problems on T91 to P91 headers (HRSG) preheated and PWHT's as you describe. Essentially the finned tubes are acting to rapidly cool the tube, especially when preheat is primarily directed on the heavy wall header material. Also the PWHT, similarly perfored, did not have adequate T/C placement and did not account for loss of power to a heating pad. Hardenss values from 270 BHN to 370 BHN were revealed and all headers/tube welds were required to be reheat-treated in the field after installation (severity of the problem was not discovered until mostly intalled). Cost to fix: over $7,000,000.

Our HRSG manufacturer's were able to rectify the problem by preheating the headers with resistance coils (they may augment with torch). PWHT is perfomed in specially designed clam shell furnaces heated electrically. Insulation is also wrapped around the tubes to a distance of about 12 inches to include the initial 2-4 inches of finning.

The previously mentioned tube/header problems were preheated/PWHT'd similar to your description. Preheat by flame was inadequate as well as location and quantity of T/Cs. Similarly PWHT was performed by electric resistance coils. Metallurgical evaluation indicated that the welds actually only acheived about 1100 F with hardness > 300 BHN equivalent. It is noted that the Owner directed Preheat/PWHT. We directed the added repair work without incident.