Page 333 - Pipeline Pigging Technology

Pipeline Pigging Technology

CALCULATING THE FAILURE PRESSURE OF
CORROSION IN PIPELINES

Structural defects which exceed code tolerances can be assessed using
fitness-for-purpose methods. These methods are well-documented[10], and
have been used extensively in pipeline welding codes[ll]. The ANSI/ASME
B31 Code [5] for pressure piping contains a supplement[12] which allows
pipeline corrosion to be assessed using fitness-for-purpose methods. These
methods are considered acceptable and applicable to pipeline defects.
The failure stress of corrosion in a pipeline can be calculated from [1-3]:

1
O f = 1.15 SMYS (1 - X) {1 - X (M' ) }•' (1)
and
V4 2
M = 1 + {0.4 (2c/(Rt) ) p (2)
where X = d/t or A/A o
= hoop stress at failure
o f
R = pipe radius
A = 2c x t
2c = defect length
t = wall thickness
= defect area
A o
d° = defect depth
SMYS = specified minimum yield strength

This criterion is nearly 20 years old, but a recent review[13] of failure
criteria for defects in pressurized cylinders concluded it was the most
accurate. Various Folios factors, M, are used in the literature but they are all
very similar, with Eqn(2) the most conservative [13].
The accuracy of this criterion can be evaluated by comparing predicted
failure pressures with actual failure pressures of full-scale tests on corroded
pipe [2,14]. The predicted failure pressures are dependent on the use of:

(i) either maximum defect depth (d) or actual defect area (A); and

(ii) actual yield stress (CT) or SMYS in the failure criterion.

The most accurate predictions are obtained using defect area and actual
yield stress [3]. The most inaccurate (and most conservative) predictions are
obtained using SMYS and maximum defect depth. Using the data in Refs 2 and

314