Failure Lesson: How Unrealistic Residual Stress Assumptions Can Drive Unnecessary Cut-Outs

During pipeline construction, PAUT and TOFD inspection of a girth weld identified a cluster of embedded indications interpreted as possible lack of fusion. An ECA was performed to determine acceptability. The ECA applied through-wall tensile residual stress equal to material yield strength as the residual stress input — a conservative default assumption in BS 7910. The assessment concluded the indications were unacceptable, triggering a cut-out of the weld and a full reweld.

The reweld was more challenging than the original due to positional access constraints and higher restraint in the as-built configuration. The repair weld subsequently developed a small HAZ crack, requiring additional NDE and another intervention. The overall outcome was increased risk from the repair process relative to what was predicted for the original weld.

The ECA applied through-wall yield-strength residual stress as a blanket conservative assumption without considering whether it was appropriate for the specific girth weld configuration. The pipe had been fabricated with a hyperbaric welding process with controlled heat input, and a post-weld proof test had been performed. Evidence supported that actual residual stresses were likely substantially lower than the assumed value. A more calibrated residual stress assumption, supported by published weld residual stress profiles and the proof test stress reduction credit, may have allowed the original indications to be accepted.

Engineering review of the residual stress assumption against the specific weld configuration. Consideration of the proof test stress reduction credit permitted by BS 7910 where applicable. Sensitivity analysis showing how the ECA result changes with residual stress variation. Discussion of the relative risk of accepting the original indication versus performing a repair weld.

Conservative ECA assumptions are appropriate when uncertainty is genuinely high. However, maximum conservatism applied uniformly without engineering justification is not a rigorous approach — it is a default that may produce decisions with their own risks. Residual stress assumptions should be selected based on weld type, heat input, PWHT history, restraint, and applicable standard guidance. Sensitivity analysis should be performed to show how significant the residual stress assumption is to the ECA outcome. Proof test credit and other engineering evidence should be considered where applicable.

BS 7910 — Guide to Methods for Assessing the Acceptability of Flaws in Metallic Structures, including residual stress guidance and proof test credit. API 579-1/ASME FFS-1 — Fitness-for-Service. DNV-RP-F108 — Assessment of Flaws in Pipeline and Riser Girth Welds.