Page 15 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

xii Introduction
like the best experts and giving it the collective knowledge of the bined has been made more intuitive. In both cases, the variables
entire organization and all the years ofrecord-keeping. to consider are mostly the same as in previous editions.
As with previous editions, the best practice is to assess major
risk variables by evaluating and combining many lesser vari-
Changes from Previous Editions ables, generally available from the operator’s records or public
domain databases. This allows assessments to benefit from
This edition offers some new example assessment schemes for direct use of measurements or at least qualitative evaluations of
evaluating various aspects of pipeline risk. After several years several small variables, rather than a single, larger variable,
of use, some changes are also suggested for the model proposed thereby reducing subjectivity.
in previous editions of this book. Changes reflect the input of For those who have risk assessment systems in place
pipeline operators, pipeline experts, and changes in technology. based on previous editions, the recommendation is simple:
They are thought to improve our ability to measure pipeline retain your current model and all its variables, but build a
risks in the model. Changes to risk algorithms have always been modern foundation beneath those variables (if you haven’t
anticipated, and every risk model should be regularly reviewed already done so). In other words, bolster the current assess-
in light of its ability to incorporate new knowledge and the ments with more complete consideration of all available
latest information. information. Work to replace the high-level assessments of
Today’s computer systems are much more robust than in past ‘good,’ ‘fair,’ and ‘poor,’ with evaluations that combine sev-
years, so short-cuts, very general assumptions, and simplistic eral data-rich subvariables such as pipe-to-soil potential
approximations to avoid costly data integrations are less justifi- readings, house counts, ILI anomaly indications, soil resis-
able. It was more appropriate to advocate a very simple tivities, visual inspection results, and all the many other
approach when practitioners were picking this up only as a measurements taken. In many cases, this allows your ‘as-
‘good thing’ to do, rather than as a mandated and highly scruti- collected’ data and measurements to be used directly in the
nized activity. There is certainly still a place for the simple risk risk model-no extra interpretation steps required. This is
assessment. As with the most robust approach, even the simple straightforward and will be a worthwhile effort, yielding
techniques support decision makmg by crystallizing thinking, gains in efficiency and accuracy.
removing much subjectivity, helping to ensure consistency, and As risks are re-assessed with new techniques and new infor-
generating a host of other benefits. So, the basic risk assess- mation, the results will often be very similar to previous assess-
ment model of the second edition is preserved in this edition, ments. After all, the previous higher-level assessments were no
although it is tempered with many alternative and supporting doubt based on these same subvariables, only informally. If the
evaluation ideas. new processes do yield different results than the previous
The most significant changes for this edition are seen in the assessments, then some valuable knowledge can be gained.
Corrosion Index and Leak Impact Factor (LIF). In the former, This new knowledge is obtained by finding the disconnect-
variables have been extensively re-arranged to better reflect the basis of the differences-and learning why one of the
those variables’ relationships and interactions. In the case of approaches was not ‘thinking’ correctly. In the end, the risk
LIF, the math by which the consequence variables are com- assessment has been improved.