PSI - Issue 80

G. Mubarak et al. / Procedia Structural Integrity 80 (2026) 157–168 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

159 3

2.3 Mechanical Modelling To assess mechanical integrity, corrosion was assumed to cause uniform wall thickness reduction based on the lowest measured value, rather than modeling individual defects. This approach is informed by CSA Z662-15 [25], which allows interacting defects to be treated as a single continuous defect, and supported by standards like ASME B31G [26] and DNV-RP-F101 [27]. The triaxial yield (Mises equivalent stress) criterion from API 5C3 was used to estimate failure conditions, focusing on internal pressure and axial load from the tubing’s weight. Bending, shear, and transient stresses were excluded due to lack of data. Fa ilure was defined as the point where equivalent stress equals the material’s flow strength, approximated as the average of yield and tensile strengths. Key equations and assumptions are based on tube geometry and material properties summarized in Table 1.

Table 1: Summary of relevant parameters for mechanical modelling

( ( 1 1 ) )

( ( 1 1 ) )

σ f (MPa)

Material

Linear density (kg/m)

Drilled depth (m)

ID (mm)

Measured minimum thickness (mm)

(MPa)

(MPa)

API 5CT J55

379

517

448

6.99

2,885

50.7

0.72

3. Results and discussion 3.1 Experimental testing

Visual inspection in Figure 2 of both “long” and “short” tubing samples showed consistent OD measurements away from the fracture regions, with OD reductions of about 1 – 1.75 mm at the fractures, likely due to combined metal loss and plastic deformation. The collar of the short sample displayed a 5.5 mm OD increase, consistent with a Type U finish. While the lon g sample's WT remained above the 4.23 mm API 5CT minimum, the short sample’s WT fell below this limit due to corrosion. Figure 3 shows severe corrosion and scale, including multi-colored, flaky deposits on the internal surfaces, lighter corrosion products near perforations (suggesting chloride presence), and mesa-type attack morphologies indicative of CO₂ corrosion. Flow -assisted corrosion was evident with the longitudinal corrosion patterns. Fracture face analysis as shown in Figure 4 revealed through-wall pits, deep metal loss, and evidence of ductile overload from plastic deformation features like jagged fracture planes and shear lips. Chemical spot tests confirmed the presence of CO₂ corrosion (positive for iron carbonates), chloride presence on internal surfaces, and the absence of H₂S -related corrosion (iron sulfide negative), as supported by XRD results. Along with magnetite, lepidocrocite, and chloride-containing compounds, XRD-EDS analysis in Figure 5 confirmed siderite as the dominant corrosion product, consistent with CO₂ -induced degradation. Composition analysis summarized in Table 2 verified the tubing met API 5CT J-55 specifications, ruling out chemical deficiencies as a causal factor in the failure. Finally, hardness measurements in Table 3 confirmed the steel was of expected mechanical strength, with no significant difference between locations near or far from the fracture. This indicates that the material ’s mechanical properties did not contribute to the failure.