Page 61 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

2/40 Risk Assessment Process
In the second part of the evaluation, an assessment is made of tion for a section govern the point assignment. The rationale for
the potential consequences of a pipeline failure. Product char- this is discussed in Chapter 1. For instance, if a 5-mile section
acteristics, pipeline operating conditions, and the pipeline sur- of pipeline has 3 ft of cover for all but 200 ft of its length (which
roundings are considered in arriving at a consequence factor. has only 1 ft of cover), the section is still rated as if the entire 5-
The consequence score is called the leak impact factor and mile length has only 1 ft of cover. The evaluator can work
includes acute as well as chronic hazards associated with prod- around this though his choice of section breaks (see Sectioning
uct releases. The leak impact factor is combined with the index of the Pipeline section earlier in this chapter). Using modem
sum (by dividing) to arrive at a final risk score for each section segmentation strategies, there is no reason to have differing risk
of pipeline. The end result is a numerical risk value for each conditions within the same pipeline segment.
pipeline section. All of the information incorporated into this
number is preserved for a detailed analysis, if required. The Relative Unless a correlation to absolute risk values has
higher-level variables of the entire process can be seen in the been established, point values are meaningfid only in a relative
flowchart in Figure 2.4. sense. A point score for one pipeline section only shows how
that section compares with other scored sections. Higher point
Basic assumptions values represent increased safety-decreased probability of
failure-in all index values (Chapters 3 through 6). Absolute
Some general assumptions are built into the relative risk assess- risk values can be correlated to the relative risk values in some
ment model discussed in Chapters 3 through 7. The user, and cases as is discussed in Chapter 14.
especially, the customizer of this system, should be aware of
these and make changes where appropriate. Judgment bused The example point schedules reflect experts’
opinions based on their interpretations of pipeline industry
Independence Hazards are assumed to be additive but inde- experience as well as personal pipelining experience. The
pendent. Each item that influences the risk picture is con- relative importance of each item (this is reflected in the
sidered separately from all other items-it independently weighting of the item) is similarly the experts’ judgments.
influences the risk. The overall risk assessment combines all of If sound, statistical data are available, they are incorporated
the independent factors to get a final number. The final number into these judgments. However, in many cases, useful fie-
reflects the “area of opportunity” for a failure mechanism to be quency-of-occurrence data are not available. Consequently,
active because the number of independent factors is believed to there is an element of subjectivity in this approach.
be directly proportional to the risk.
For example, if event B can only occur if event A has first Public Threats to the general public are of most interest
occurred, then event B is given a lower weighting to reflect the here. Risks specific to pipeline operators and pipeline company
fact that there is a lower probability of both events happening. personnel can be included as an expansion to this system, but
However, the example risk model does not normally stipulate only with great care since a careless addition may interfere with
that event B cannot happen without event A. the objectives ofthe evaluation. In most cases, it is believed that
other possible consequences will be proportional to public
Worst case When multiple conditions exist within the same safety risks, so the focus on public safety will usually fairly
pipeline segment, it is recommended that the worst-case condi- represent most risks.

I Index sum

Figure 2.4 Flowchart of relative risk index system.