NDT before and after PWHT low carbon steel

  • NDT before and after PWHT low carbon steel

    Posted by george on 27/11/2021 at 4:49 pm

    PWHT is approx 1 hour @ 650 deg Celsius on P1 (low carbon) steels. Is there a Code requirement (ASME/ANSI/API/EN/SNiP) that such NDT MUST be done after PWHT?
    What do forum members think of a suggestion for NDT before PWHT only. If all unacceptable RT indications are repaired prior to stress relief (1 hour @ 650 deg), is there any likelihood/possibility of stress relief of P1 materials causing metallurgical problems or acceptable/undetected indications to open up which may then be detected by RT and be cause for rejection. Is the latter realistic or is NDT after PWHT on low carbon materials just a convention.
    Thanks in advance for your comments.

    charlie replied 1 year ago 2 Members · 1 Reply
  • 1 Reply
  • Heat Treatment (PWHT)


    28/11/2021 at 2:00 pm
    2 Points

    his was the subject of a recent discussion here and I can make the following comments, which are my personal views – others may add or disagree:
    1. In our case PWHT on carbon steel is generally service related e.g. Carbon steel pipe work in Amine and Caustic service.
    2. The code says that NDT should be performed after welding and as heat treatment is a part of the welding technology, it should be performed after heat treatment.
    3. Many sites will want to perform NDT before heat treatment as this saves the time and money of having to repair welds and again perform heat treatment on the repair. As the owner we are willing for the contractor to do the NDT before heat treatment, but at his own cost.
    4. The NDT carried out before heat treatment should be the same as the NDT carried out after heat treatment. Generally the NDT before heat treatment has been radiography, but as the owner we have substituted this for UT after heat treatment. The theory being, that the only repairable defects likely after heat treatment (if this has not been performed correctly) is cracking and the preferred method for crack detection in this instance would be UT.

Reply to: george
Your information:

Start of Discussion
0 of 0 replies June 2018