Professional Documents
Culture Documents
Crack in Corrosion Defect Assessment in Transmission Pipelines
Crack in Corrosion Defect Assessment in Transmission Pipelines
Crack in Corrosion Defect Assessment in Transmission Pipelines
Assessment in Transmission
Pipelines
Cracks may occur coincident with corrosion representing a new hybrid defect in gas and
oil pipelines known as crack in corrosion (CIC) that is not directly addressed in the cur-
1 Introduction defects. They concluded that the behavior of these hybrid defects
fell between that for cracks and corrosion. Furthermore, it was
Pipelines are the safest form of transportation for oil and refined
found that the transition from crack to corrosion behavior
products; however, aging pipeline infrastructure may experience
occurred when the corrosion depth was greater than 75% of the
defects in the form of cracks or corrosion commonly resulting
total defect depth [6].
from coating or cathodic protection degradation, local environ-
The aim of this present study is to review the most commonly
ment, or third party damage during fabrication. Evaluation of
used existing techniques for evaluating cracks and corrosion, and
these defects is important for continued safe operation. Based on
to provide a framework for applying them to CIC defects. For the
the type of defect, there are different codes and standards for
purpose of this study, a series of rupture tests were undertaken to
assessing pipeline defects. For instance, the most popular codes
investigate the failure behavior of a thin walled gas pipe contain-
for cracks assessment in oil and gas pipelines are API 579 [1] and
ing a series of axial flaws such as crack, corrosion, or CIC defects.
BS7910 [2] and common methods for assessing corrosion are
RSTRENG [3] and the CPS method.
In general, significant amounts of corrosion (i.e., greater than
10% of the wall thickness (WT)) and cracks are not found to-
2 Background
gether. This may in part be related to the soil chemistry required 2.1 Corrosion Defect Assessment. Corrosion may occur on
to generate each type of defect [4]. However, new hybrid defects the outside of a pipe due to coating failure, and after some years
including cracking within corrosion have been identified [5]. without protective coating, the pipe will experience external cor-
These include cracking coincident with corrosion, termed CIC rosion. Corrosion can also occur on the internal surface of the
and have to be investigated extensively. pipeline caused by contaminants in the products such as small
A study by Cronin and Plumtree [6] determined the behavior of sand particles or amino acids. There are several methods for
long cracks within long corrosion grooves for these hybrid assessing corrosion occurring on pipelines, with RSTRENG [3]
most commonly used in practice. Developing methods such as
CPS [7] provide improved failure pressure predictions. Most
Contributed by the Pressure Vessel and Piping Division of ASME for publication
in the JOURNAL OF PRESSURE VESSEL TECHNOLOGY. Manuscript received October 9,
existing methods consider that the circumferential or hoop stress
2011; final manuscript received April 17, 2012; published online March 18, 2013. in the pipe dominates the response and estimate the collapse pres-
Assoc. Editor: Saeid Mokhatab. sure based on the area of metal loss.
The Folias factor (M) describes the bulging effect of a shell sur- The stress intensity ratio, Kr , controls the failure of cracks in
face that is thinner in wall thickness than the surrounding shell low toughness materials, whereas Lr controls the failure of cracks
in high toughness materials. It is considered safe, if the current
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
state of the material defined by the assessment point A determined
2c 2 2c 4 for the component containing a crack under an applied load lies
M ¼ 1 þ 0:6275 pffiffiffiffiffi 0:003375 pffiffiffiffiffi (3)
Dt Dt below the FAD curve (Fig. 1). The component may fail by frac-
ture for Kr > 1 or plastic collapse for Lr > LprðmaxÞ . In some cases,
The failure stress of a corroded pipe under internal pressure the component may fail in the transition region between fracture
containing corrosion defect oriented along the axis of the pipe is and plastic collapse.
given by Eq. (4) In general, the FAD approach is applied at three different lev-
2 3 els. Levels 1 and 2 include assumptions regarding the material’s
d response. They are more conservative than level 3, which utilizes
t 6 1
Pf ¼ r6 t 7 7 (4)
the actual material tensile strain behavior. Five methods (A–E)
R 4 d 15 are associated with a level 3 FAD. Hosseini et al. [10] demon-
1 strated that level 3 FAD methods B (K-approach) and D (J-
t M approach), using API 579, were more accurate analytical methods
compared to others assessing longitudinal crack defects. Conse-
The CPS method [7] uses a plastic collapse criterion, where quently, these methods (API 579, level 3 FAD, methods B and D)
failure is initiated when the stresses in the material exceeds a criti- were considered for the current study. The K-approach uses the
cal value (i.e., ultimate tensile strength). The upper (plain pipe) level 2 assessment procedure, except that the FAD is constructed
and lower (long groove) limit solutions provide bounds for the using the actual material properties for level 3. Method D applies
collapse pressure prediction, and the actual failure is determined ductile tearing analysis in which the fracture tearing resistance
based on the metal loss [7]. The main advantage of this method is (J–R curve) is defined as a function of the amount of stable ductile
its ability to make full use of the characteristic corrosion geometry tearing.
and tensile properties of the material. In addition, it is unnecessary The level 3 FAD (Fig. 1) uses the actual true stress–strain curve
to consider defect interaction, since the interaction of adjacent for the material, where the stress intensity ratio is defined as given
corrosion is taken into account [7]. in Eq. (10). The cut-off for localized plastic collapse (Eq. (11))
provides the upper bound for this failure mode
2.2 Crack Assessment. Cracks can occur on the internal or
" #0:5
external surfaces of a pipe and typically grow on the pipe surface Eeref ðLr Þ3 ry
perpendicular to the hoop stress (maximum principal stress) result- Kr ¼ þ (10)
ing in alignment of the crack along the longitudinal direction [9]. Lr ry 2Eeref
There are several methods for assessing for cracks in pipelines.
API level 3 FAD (methods B and D) [1] have been successfully
applied to evaluate the cracks. These methods use linear elastic
fracture mechanics (LEFM) and elastic–plastic fracture mechanics
(EPFM), respectively. LEFM can be used when the plastic zone at
the crack tip is small and the linear elastic stress intensity factor
(K) is given by
pffiffiffiffiffiffi
KI ¼ Yrm pa (5)
On the other hand, EPFM can be used when the plastic zone at
the crack tip is significant. This method is commonly applied
using the J integral (JI), based on the strain energy density (U) in
the vicinity of the crack
1 @U
JI ¼ (6)
t @a Fig. 1 Failure assessment diagram (level 3)
3 Material Characterization
API 5L Grade X52 pipeline steel of external diameter 508 mm
(20 in.), 5.7 mm wall thickness was used as the experimental basis
for this study.
Defect depth
Corrosion Crack
Fig. 4 CIC defect (a) cut slit, (b) fatigue crack, and (c) corrosion defect
Test ID Defect depth (%WT) Pressure (MPa) CPS (MPa) RSTRENG (MPa) CPS (MPa) RSTRENG (MPa)
Crack Collapse API 579 API 579 API 579 API 579
depth (a) pressure level 3 level 3 level 3 level 3
Test ID (%WT) (MPa) Exp. (method D) (method B) (method D) (method B)
Fig. 6 Collapse pressure prediction for cracks using API579 Fig. 7 Collapse pressure prediction for cracks using API579
level 3 FAD (method B) level 3 FAD (method D)
experimental range of 52–66% total defect depth, the CIC results For corrosion defects, the CPS method provided more accurate
were closer to those for corrosion than cracks. Since the crack to collapse pressure predictions compared to RSTRENG. The CPS
corrosion ratio has a major effect on the behavior of CIC defects method uses a more complete description of the material response
[6], further work will be undertaken to expand the results. and surrounding material loss in comparison with RSTRENG.
Considering CIC defects as cracks of equivalent depth, level 3 Moreover, there is no need to consider defect interaction rules,
FAD (method D) in API 579 procedure provided conservative since the interaction of adjacent corrosion is considered without
results between 10% and 30% (18% on average) of the experi- simplification.
mental collapse pressures. Assuming that the CIC defects were of For cracks, the level 3 FAD (method D, API 579) provided the
equivalent depth to the corrosion defects, the CPS method would least conservative predictions (20% average error). Method B
predict that the collapse pressures were between 3% and 22% con- (level 3 FAD, API 579) gave more conservative predictions with
servative (10% on average) of the experimental collapse pres- an average error of 33%; however, method D requires additional
sures. However, this approach is not recommended, since information which may not always be available for defect
numerical analysis [6] has demonstrated that a transition to corro- assessment.
sion dominated failure only occurs for very shallow cracks rela- Rupture testing of five CIC specimens demonstrated the effect
tive to the total depth of the CIC defect. of defect depth on the collapse pressure. Applying API 579 level
3 method D predicted the collapse pressures with an average error
of 18%. The results showed that the failure pressures of CIC were
6 Conclusions lower when their equivalent depths were similar to those for simu-
Experimental rupture tests were conducted to investigate the lated corrosion and cracks.
failure behavior of longitudinally oriented corrosion, crack, and In addition to defect length and depth, the defect profile is an
CIC defects in API 5L Grade X52 pipeline steel external diameter important parameter when determining the failure behavior of a
508 mm (20 inch), 5.7 mm wall thickness. It was found that failure simulated corrosion defect or CIC defect. Future work will
for simulated corrosion, cracks, and CIC defects occurred by plas- include numerical modeling to investigate a wider range of defect
tic collapse and ductile tearing. geometries.
Appendix A: Reference Stress Where b is given by the following equation and the parameters
Aij , are provided in Appendix C of the API 579 code [1].
A.1 API 579-Cylinder Approach. Stress intensity for a cylin- 2u
der under internal pressure containing semi-elliptical surface b¼
p
crack (taken from Appendix D section D.5.10 of API 579). A
sample cylinder is shown in Fig. 11. The G2 , G3 , and G4 influence coefficients can be computed
It should be noted that there is no bending stress in this case; using paragraph C.14.3 or C.14.4 in Appendix C of the API 579
therefore, the reference stress can be written as follows: code [1].
Q is determined using following equation:
rref ¼ MPm
Q ¼ 1 þ 1.464 (acÞ1:65
r0 is the uniform coefficient for polynomial stress distribution
where Pm is the hoop stress (MPa or psi).
References
[1] American Petroleum Institute, API579, 2000, “Recommended Practice for Fit-
ness for Service.”
[2] British Standards Institute, BS 7910, 2000, “Guide on Methods for Assessing
the Acceptability of Flaws in Metallic Structures,” BSI-10.
[3] Escoe, A. K., 2006, Piping and Pipeline Assessment Guide, Elsevier, New
York.
[4] Hosseini, A., 2010, “Assessment of Crack in Corrosion Defects in Natural Gas
Transmission Pipelines,” MASc. Thesis, University of Waterloo, Waterloo,
Ontario, Canada.
[5] CEAP, Canadian Energy Pipeline Association, 2005, “Stress Corrosion Crack-
ing Recommended Practice,” 2nd Edition, 1860, 205 5-Aveniue SW, Calgary,
Alberta, Canada T2P 2V7.
[6] Cronin, D. S., and Plumtree, A., 2008, “Assessment of Crack in Corrosion
Defects in Natural Gas Transmission Pipelines,” 7th International Pipeline
Conference.
[7] Cronin, D. S., and Pick, R. J., 2000, “Prediction of the Failure Pressure for
Complex Corrosion Defects,” Int. J. Pressure Vessels Piping, Vol. 79, pp.
279–287.
Fig. 11 Semi-elliptical Surface crack in longitudinal direction [8] Cosham, A., and Hopkins, P., 2007, “Best Practice for the Assessment of
of a cylinder [1] Defects in Pipelines-Corrosion,” Eng. Failure Anal., Vol. 14, pp. 1245–1265.