A pipeline risk management scheme involves identifying
and assessing the susceptibility and severity of credible
threats to pipeline integrity, determining the likelihood of
various failure scenarios and calculating their likely impact
on people and the environment. The resulting failure
probabilities and consequence values generate risk metrics.
These factors are then used to develop various mitigation
and inspection strategies to maintain pipeline integrity.
There are two groups of people in a pipeline company
who benefit from a risk assessment, namely engineers
and managers, with each having different areas of
responsibility. An effective risk model should generate
information that addresses the responsibilities and
concerns of both these groups.
Risk modelling
A risk model must consider the susceptibility of a pipeline
to various hazards, the probabilities of different failure
scenarios together with the consequences arising from
each of these failure scenarios, i.e. the model needs to
address the following questions:
)
What can go wrong?
)
How likely is it to happen?
)
What are the consequences if it does happen?
Various techniques are available to evaluate risk
factors for different threats and failure scenarios. These
range from descriptive qualitative approaches based on
engineering judgement, which generate relative risk values,
through to detailed mathematical analyses in a fully
quantitative assessment which express risk in absolute
terms. However, the degree of model complexity used to
determine a risk factor does not necessarily reflect the
accuracy of the results.
Semi-quantitative risk models of varying complexity
have been promoted for a number of years, but in the
authors experience the algorithms and weighting factors
commonly found in such models still rely heavily
on engineering judgement, which can lead to
inconsistencies across assessments. They also
often lack sufficient resolution for meaningful
assessments.
In most cases there is insufficient information
available to justify applying a fully quantitative
approach to assessing risk. However, with the
growing availability of detailed pipeline data
such as ILI results, CP survey data and geospatial
information, operators now have an expectation
that pipeline risk models will provide a more
consistent and numerically based assessment of
risk levels. As an example, determination of the
consequences of failure is a key component of risk
models and the availability of GIS-based tools is
improving the definition of hazard zones and the
impact on receptors in these zones.
An effective numerically based pipeline risk
model should incorporate mathematical logic
into its decision making process and generate risk
factors in a consistent manner. Also, in addition
to the standard risk factor (PoF x CoF), individual
and societal risk levels feature in many pipeline
regulations so there is a need for the model to
create the parameters necessary to calculate these
risk values along a pipeline.
It is essential the model incorporates a
transparent decision making process that
provides a clear audit trail. Regulatory authorities
increasingly demand auditability so this effectively
rules out many black-box software solutions where
a single click generates an answer but the process
is opaque.
QPRAM risk model
MACAW Engineering in collaboration with ROSEN
Integrity Solutions developed QPRAM, a numerical
pipeline risk assessment model, as a result of its
Figure 1.
The integrity management process.
Figure2.
Risk model view.
44
World Pipelines
/
JUNE 2015
