A crosscountry product pipeline ruptured along the longitudinal ERW seam prior to a planned shut down. The line pipe conformed to API 5LX 46 grade material with 12 inch dia. and 0.25 inch thickness. Detailed failure investigation was carried out in the laboratory on the failed portion of the pipe to etablish the cause of failure. Visual observation, Mechanical property evaluation, Fractographic analysis and Analytical calculation using fracture mechanic based concept were carried out as a part of failure investigation .
Visual investigation revealed a rupture opening of one meter length with localized yielding at the point of maximum opening along the rupture. The external surface in and around the region of failure exhibited localized corrossion in the form of deep pit whereas the internal surface revealed uniform general corrossion. Ultrasonic thickness measurement near the maximum opening area indicated the presence of average residual thickness of 5.4 mm as against the original 6.3 mm. Fractography of the fracture surface at the point of maximum opening of the rupture indicated a lack of fusion opening and running along the inner surface. The dicontinuities in the form of lack of fusion were quantified a 14 mm and 41 mm in length separated at a ditance of 45 mm. The depth of both the dicontinuities from the inner urface of the pipe were meaured a 2.5 mm (max.).Analytical calculation using crack tip opening displacement (CTOD), Mutual interaction (a per PD 6493 code) of the two adjoining dicontinuities and literature reported toughness value for the ERW seam were conidered. The reultant operating pressure at the time of failure was evaluated a 47 Kg/cm2. The pipeline had been hydrotested fifteen years back at a pressure of 75 Kg/cm2. Also, operating logbook prior to failure indicated a lower operating pressure of 35 Kg/cm2. Post failure testing of the line pipe after replacement of the damaged section to enure fitness for purpose revealed further rupture in the line at a hydrotest pressure of 49.5 Kg.cm2. which further supported the analytical calculation.
Though hydrotest of the line pipe is a destructive test for integrity assesment it can ensure the fitness of the line for the intended operation. Periodic hydrotest is a must since defect inferred dormant can grow to critical size with time on account of corrossion. Proper record of failure pressure of the in-service line during hydrotest can help in scheduling the inpection interval of the line through defect sentencing curves.
1.0 FAILURE BACKGROUND
A cross country product pipeline ruptured along the longitudinal ERW seam prior to a planned shut down. The pipeline was in service since sept.1966. The precommissioning hydrotest carried out during contruction stage was at a pressure of 114 Kg/ cm2. During Nov. 1983 the pipe line wa subjected to hydrotest at a pressure of 75 Kg/cm2. The external corrossion noticed in the weld seam and related heat affected area was commensurate with the service exposure.The pipe line was to be taken under planned shut down on 5.12.97 when rupture occurred. The line pipe conformed to API 5LX 46 grade material with 12 inch dia and 0.25 inch thickness. Detailed failure investigation was carried out in the laboratory on the failed portion of the pipe to etablish the cause of failure. Visual observation, Mechanical property evaluation, Fractographic analysis and analytical calculation uing CTOD concept were carried out as a part of failure investigation
2.0 LABORATORY INVESTIGATION:
The ruptured pipe showed a longitudinal opening along the ERW weld seam with a maximum length of about one meter (Fig.1). At the centre of the opening localised yielding of around 2% of nominal circumference was noticed. The external surface in and around the region of failure exhibited localied corrossion attack with deep pitting. A minimum thickness of 5.4 mm a against the nominal thickness of 6.3mm was noticed at the point of yielding. The internal surface of the pipe was seen to have pitting attack along the rupture at the ERW seam (Fig.2).
Fig 1: Photograph of as received pipe sample showing a longitudinal rupture opening along the ERW seam with a maximum length of about one meter.
Fig 2: Internal surface of the pipe sample showing localised corrosion attack near to the weld.|
Mechanical property assesment
Tensile test sample were made from the sound portion of the pipe and the same were tested for the Yield and Ultimate strength and hardness. The test result indicated UT a 48.6 Kg/mm2 , Y a 42.3 Kg/mm2, %el a 25.The bulk hardness of the pipe parent metal and weld were found to be within the normal limit i.e 145VHN and 176VHN respectively.
The fracture surface at the point of maximum opening and yielding was oberved under scanning Electron Microscope after ultrasonic cleaning. The fractograph indicated dicontinuities on the pipe inner surface (Fig.3). The depth of the dicontinuities were measured to be around 2.5mm (Max.). The fracture surface also revealed cleavage facets and river pattern at the other location (Fig.4).
Fig 3: Fractography of the fracture surface under Scanning Electron Microscope showing a discontinuity on the pipe inner surface. (X15)
Fig 4: Fractography of the fracture surface under Scanning Electron Microscope revealing cleavage facets.(X1500)
An insitu metallography replica was used to tranfer the contour of the fracture surface in order to measure the length of the dicontinuities. The local dicontinuities were measured to be around 14mm at the maximum opening and 41mm at a distance of 45mm from the earlier one. Moreover non quantifiable mall dicontinuities were also seen between these two larger ones.
The hoop stress on the pipe ection / operating pressure required for the propagation of the dicontinuity to ultimate failure was computed using the Crack Tip Opening Displacement (CTOD) concept of Fracture mechanics as follows :
'CTOD' parameter for the crack of length '2a' in a thin walled pipeline i given by 
'a' = Half the length of the original defect; E = Elatic modulus of the pipreline material; s
h = Operating hoop tre; s
ys = Yield strength of the material; Mf = tre intenity magnification factor; R = Radiu of the pipe; B = Thickne of the pipe; P = Operating preure
The defect instability is expected to occur leading to the failure if and only if the CTOD at the crack reaches the critical value of the weld toughness (CTOD)c. Hence
The toughness data of API 5LX 60 material has been considered in view of the tensile test result of the pipe showing a higher tensile value as against the requirement of API 5LX46. The weld toughness (CTOD) for API 5LX 60 is reported a 0.23 mm . As per defect interaction analysis using PD6493 the dicontinuities were found to be of non interacting type. since there were smaller non quantifiable dicontinuities seen in between the larger dicontinuities, the overall length of the dicontinuity has been conidered as the summation of the two larger one along with their distance of separation. The data considered for computation are as follows:
(CTOD)c=0.23 mm; sys=59.74 Ksi;'a' =50 mm;E=29480 Ki.; Mf =1.04 R=167.1mm; B=5.4 mm
Eqn.6 yield a s
h value of 20.9 Ki which in turn reult in operating pressure of 47 Kg/cm2 a estimated from eqn.5. On individual assesment of the defect, the required operating pressure for failure was found to be 155 Kg/cm2. and 132 Kg/cm2. for the smaller and larger defect respectively.
- Considerable localised yielding prior to the rupture has been noticed at the failure location indicating that the operating pressure must have led to hoop stress in excess of the yield trength.
- The cleavage fracture seen is normal in view of the thin walled piping. These type of fracture are often associated with very little plastic deformation. This corrraborated with the 2% of diametral expansion seen on the sample.
- The dicontinuities seen on the fracture surface are parallel to the tube axis and confined to weldment. From the morphology of the dicontinuities, the presence of crack or laminar inclusion are ruledout. The dicontinuities are thus attributed to lack of fusion due to improper edge preparation during the welding of the pipe which have subequently grown due to general internal corrossion.
- From the analytical calculation it was seen that the dicontinuities could lead to failure when the operating pressure of the pipeline reach a value of around 47 Kg/cm2. In the absence of any dicontinuity / defect the expected operating pressure for failure would be much higher than this value.
- The original preservice hydrotest and the subequent hydrotest were done at a pressure of 114 Kg/cm2 and 75 Kg/cm2 repectively which were much less than the required pressure for the individual defect to lead to rupture. The defect have thus survived the preservice hydrotest.
- Originally the defect were separated by a ditance more than required to be considered a interactive type as per PD 6493 code. Thus the defect have not interacted during the preservice hydrotest.
- During the long service life of the pipe, both the lack of fusion area have undergone subcritical growth due to corrossion / erosion on the inner surface which is possible in view of the 6'O clock position of the ERW seam.
- Only a marginal growth (4mm only) can make both the defect to interact as per PD 6493. Thi growth owing to corrossion / erosion doesnot leave any prominent indication on the fracture surface that can be seen visually. When the defect interact with each other, with considerable localised reduction in the wall thickness, the required pressure for rupture was found be around 47Kg/cm2.
From the laboratory analysis and analytical calculation, it is inferred that the failure of the pipe line has occurred due to internal pressurisation leading to increased hoop stress more than the yield strength limit of the material at the location containing lack of fusion.
- 'Elementary Engineering Fracture Mechanic', David Broek, Martinu Nijhoff Publn., 1986, p.p 391 - 394.
- 'A Finite line crack in a pressurised cylindrical shell', Folia E.., Int. Journal of fracture Mechanic, Vol.1, 1965, p.p. 104 - 113.