Interaction of Adjacent Corroded Regions

Introduction

• Interaction between two, or more, separate corroded regions is defined when failure occurs at a lower pressure than would cause failure for any of the individual corroded regions.
• Adjacent corroded regions have been divided into three general types of groups to facilitate the analysis to determine failure interaction.

Type I Defects with Interactions

Type I defects consist of flaws that are separated circumferentially but overlap when projected into a single plane. These flaws should be treated as a single defect so long as a single separation does not exceed 6 wt. Interaction becomes significant for closely spaced, overlapping short flaws.
[_Drew RSTRENG Powerpoint slide 125]

Limitations to All Corrosion Analysis

(v5)[_Drew RSTRENG Powerpoint slide 126]

• “These procedures should not be used to evaluate the remaining strength of corroded girth or longitudinal welds or related heat-affected zones, defects caused by mechanical damage, such as gouges and grooves, and defects introduced during pipe or plate manufacture, such as seams, laps, rolled ends, scabs, or slivers.”
• Other limitations include pipe vibration, fittings, wrinkle/ripple bends, etc.
  • Crack-like defects
  • Combined corrosion and crack-like defects
  • Combined corrosion and mechanical damage
  • Metal loss defects due to mechanical damage (e.g. gouges)
  • Metal loss in indentations and buckles, or metal loss that is coincident with other damage
  • Metal loss in fittings
  • Pipelines that operate at temperatures outside their original design envelope or operating at temperatures in the creep range

Interactions – Type I & II

[_Drew RSTRENG Powerpoint slide 127]

• Type II defects consist of multiple flaws on the same axial line but are separated by a full wall thickness pipe. Use RSTRENG to analyze the individual flaws and the overall combination. The lowest calculated failure pressure should be used. Tests suggest that flaws must be closer together, axially, than one-half a flaw length to interact.
• Type III defects consist of shorter, deeper defects within longer, shallower defects. RSTRENG provides adequate predictions based on the worst-case projected corrosion area. For very long corroded areas, RSTRENG analysis can be limited to one dia length, or about 20 inches, whichever is greater, so long as the deepest pitting is included.

RSTRENG ECDA_Interactions Type I Type II.png

Interactions – Long, Narrow Defects

[_Drew RSTRENG Powerpoint slide 128]

• Tests indicate that the analysis of long, narrow, near-uniform defects can be limited to a length of two pipe diameters. Probably one pipe diameter is sufficient, so long as the deepest point is in the center of the region.
  RSTRENG ECDA_Interactions LND.png

Interactions – Pits

[_Drew RSTRENG Powerpoint slide 129]

Criteria for interaction is whether any array exhibits failure pressure less than that of a single pit. The deepest, longest pit in a circumferential array is the one to analyze.
For diagonally arrayed pits, the projected length always gives a conservative result. For longitudinal arrays of pits, if touching or separated by < 1 wt, analyze the entire defect. Single pits separated by more than 1 wt do not interact significantly.
RSTRENG ECDA_Interactions Pits.png

Interactions – Spiral Oriented Machined Groves

[_Drew RSTRENG Powerpoint slide 130]

• These defects failed at very high pressures, 80% to 100% of the burst pressure of a defect-free pipe specimen. All ruptured along the spiral grooves, but at much greater pressure levels than for axial flaws of similar length and greater than would have been predicted on the basis of projected length.
• It has been suggested that a “spiral angle factor” be validated to use to multiply the remaining strength based on projected length to generate more accurate, less conservative estimates. This empirically generated spiral angle factor, SAF, might be as follows:

30 degrees1.2

• Currently, it is recommended that the projected axial length be used for the analysis, but the results will likely be very conservative.
  RSTRENG ECDA_Interactions SPOMG.png

Helical Patterns > 45 degrees

[_Drew RSTRENG Powerpoint slide 134]

• For helical patterns that lie more than 45 deg, it is sufficient to consider the most severe longitudinal section through the corroded area.
  RSTRENG ECDA_HOPC.png

Interactions – Patches of Metal Loss

[v3][_Drew RSTRENG Powerpoint slide 137]

• Patches or pits and patches separated longitudinally should be assumed to interact if the separation is less than 1.5 inches. Apply RSTRENG over the entire length.
• The failures were all axial ruptures and the research indicates that the circumferential extent of the defects plays little or no role in the failure behavior.
  RSTRENG ECDA_Interactions PML.png

Interactions – Compound Defects

[slide 139]

Type III interaction is examined here where shorter, deeper defects within longer, shallower defects are examined. Some rupture and/or leaks did occur when pressure reversal was experienced, but not at predicted failure pressures below 100% of SMYS.
RSTRENG ECDA_Interactions CD.png

Interactions Rules (3t and 6t)

[slide 140]
RSTRENG ECDA_IR3t6t.png

If the circumferential separation distance is less than six times the wall thickness, the composite area (A1 + A2 − A3) and the overall length, L, should be used.

Corrosion Interaction Distances

[slide 141]
RSTRENG ECDA_CID.png

Flaw Spacing

[slide 142]
Flaws spaced further apart than three times the wall thickness should continue to be evaluated separately.
RSTRENG ECDA_FS.png

Effects of Large Axial Stresses on Corroded Pipes

[slide 143]

• The Southwest Research Institute performed five experiments using 48″OD x .480″ wt, X65 line pipe. The intent was to examine the effects of longitudinal stress and internal pressure on corroded pipe. The corrosion was simulated by machining 25% to 50% of the wall thickness over rectangular areas of various sizes. Two identical metal loss areas were machined on opposite sides of each pipe specimen.
• Therefore, if a pipe specimen were subjected to a bending moment, one machined side would be in compression and the opposite machined side would be in tension.
• The tests were performed in basically two ways. First by pressuring to a high level, then applying bending moment, and then pressuring to failure. Second by pressuring to a high level and then applying bending moment to failure.
• The first type of failure would result in axial fracture on the compression side. The second type of failure would result in circumferential fracture on the tension side.
• For the axial rupture cases, the RSTRENG-predicted failure pressures were reasonably close to the actual rupture pressures, and RSTRENG would have predicted safe operating pressures. For the circumferential rupture cases, i.e., the bending moment applied until rupture, the actual ruptures occurred at pressures significantly lower than predicted by RSTRENG. The axial stress and circumferential extent of the metal loss controlled the results.

Related Links