2-1/4 Cr 1 Mo Tempering and PWHT Temperatures 2

[spayette]

Forgings produced to ASTM A-182 Gr F22 are normalized at a temperature of 1650F followed by a tempering at 1250F. Welded piping components then receive PWHT between 1300F and 1400F, as per ASME B31.3.

I have been advised that for such components the tempering temperature should be raised beyond the ASTM requirements to approximately 50F above the PWHT temperature. What are the advantages and disadvantages of this approach?

 

[metengr]

Well, the approach is to try to perform post weld heat treatment (PWHT) below the original or specified tempering temperature of the material. The main reason is to avoid any noticeable drop in mechanical properties after the completion of PWHT. Typically, PWHT (if required) of welded components is to provide for adequate stress relief and softening of the weld deposit, and weld heat affected zone without significantly altering the bulk properties of the material.

If you notice, most ASME material specifications for normalizing and temper provide a minimum value for tempering temperature to achieve desired mechanical properties. Increasing the tempering temperature of the original heat treatment by 50 deg F, provides for margin during PWHT to assure that you do not over-temper (or soften) the material away from the weld heat affected zone.

Advantages of the slightly higher tempering temperature - no real change in bulk properties of the original material (away from the weld) if welding and PWHT are required for fabrication.

Disadvantages - none really, other than working closely with your steel supplier on the front end of an order to meet your target tempering temperature requirements.

 

[spayette]

Metengr

Thank you very much for your reply.

The fittings in question are approximately 100 each, 4" Weld-O-Lets that will be welded to a 40" diameter fired heater manifold (insulated from radiant section of heater)and 4" P91 tubes. Design conditions are 275 psig at 1000F. In this case, do you think that there would be value added to send the fittings back to the supplier and require that the normalizing and tempering be re-done?

Thanks again.

 

[metengr]

Spayette;
No. I see no real benefit to be gained by re-heat treatment of F22 weld-o-let fittings, as long as you have CMTR's indicating they have been supplied in the N&T condition. The 2-1/4% Cr-1% Mo is the "work horse" of the Cr-Mo alloy steels in boiler and pressure vessel service. This material is forgiving in terms of PWHT range and original tempering temperature.

On the flip side of things, make sure that you have tight preheat and PWHT parameters when joining the F22 to the P91 material. Are you very familiar and experienced with welding P91 material, and the concerns with preheat, interpass temperature and PWHT?

 

[davefitz]

Spayette:
Make sure the thickness transition from the F22 weldolet to the P91 pipe is not greater than 30 degrees, as per asme I fig PG 42.1 . We just had a failure of such a welded connection between a 12" P91 pipe and a 12" x 22" F22 weldolet, and it seems the thickness transition was an improper 45 degrees ( 2 yr service life ). A better design would use a P91 transition piece instead of causing the thickness transition in the F22 part.

 

[stanweld]

davefitz,
Did you make the dissimilar weld with "B9" filler metal or "B3" filler metal? Did failure occur in weld metal or HAZ of F22 O'let?

 

[davefitz]

stanweld:
I did not make the weld; I arrived in the job a few years after it was built. We are still evaluating the failed section and will advise the nature of teh failure after the sample is analyzed in a lab. It looks like the failure is in the P91 piping immiediately outboard of the HAZ, but lab will confirm .

 

[stanweld]

davefitz,
I have also seen failures in similar joints with improper weld transition geometry. In those cases the root cause of failure (in < 4000 hrs) was extreme thermal gradients about the circumference of the P22 pipe immediately downstream of the desuperheater. Poor joint geometry was also a contributor and failure would have occured due to stress intensity from the poor joint geometry alone in the P91 HAZ.

 

[davefitz]

stanweld:
Yes, you have hit the problem on the head. In our case, in addition to a poor weld geometry, the P91 riser is immediately downstream of the reheater interstage spray. There is inaddequate spray water residence time between the spray water nozzle and the downstream 22" header that is the source for the 12" P91 risers. The thin wall P91 pipe has its temperature cycling radically, whilethe thicker forged saddle has a nearly constant temperature, such that many thermal cycles are imposed on the weld.

 

[spayette]

metengr

Thanks for your very valuable responses. Yes, we have people experienced in welding P91 material. However, if you have any suggestions that you would like to offer, they would be most appreciated.

 

[stanweld]

davefitz,
Very similar high thermal cycling was experienced at the Western site as well. One must truly take into account cyclic operating parameters of combined cycle units and assure proper joint designs and metallurgy transitions to assure freedom from very premature failures.