Page 324 - Pipeline Pigging Technology

Assessment of pipeline defects

and they are inspected with intelligent pigs, corrosion is proving to be a major
problem, causing repair and replacement bills of hundreds of millions of
dollars in European[6] and American[7] pipelines.
Therefore, the combination of on-line inspection data with defect-signifi-
cance calculations is becoming essential as pipelines age and the use of high-
resolution intelligent tools becomes more popular. Such tools present a
pipeline operator with detailed data, ideal for defect-significance calcula-
tions, whereas previous inspection systems could not accurately size or
reliably detect defects. The combination of an accurate inspection tool and
a reliable defect assessment can avoid expensive repairs which, even for
onshore lines, can be in excess of £.100,000 per defect.
This paper presents a methodology for the assessment of corrosion in
pipelines, with particular reference to on-line inspection of heavily-corroded
pipelines. The use of on-line inspection for the condition monitoring of
corroding pipelines is discussed and safety factors for use in the assessment
methods proposed.

ON-LINE INSPECTION DATA

Introduction

A description of the development of intelligent on-line inspection tools (as
exemplified by British Gas) and their capabilities can be found in the
literature [8,9]. This section concentrates on the type of data that can be
obtained from an on-line inspection, and the analysis of bulk data prior to
assessing the significance of the reported corrosion.

Single and repeat runs

On-line inspection tools can give detailed information of a variety of types
of corrosion (Fig. 1) along an entire pipeline length. The data can be processed
to focus attention on sections of the pipeline or individual pits in individual
pipeline spools, Fig.2.
The accuracy of some tools is such that readings from a later on-line
inspection can be superimposed on those from the early inspection, and
corrosion rates obtained for sections of the pipeline, Fig.3(a). Additionally, it
is sometimes possible to compare readings in individual spools to check for

305