Page 378 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Risk communications 151353
Table 15.8 Varying objectives at various stages of risk events evokes a dramatic and frightening mental picture in the
audience.
Stuge Objective Other interesting aspects of risk perception involve the
familiarity of the system and the concept of “signal potential”
Before Reduce the anxiety about potential emergencies that from an event. Society tends to perceive as a greater risk one
the agency considers unlikely.
During Prevent panic in mid-crisis. that is new or unusual or can be viewed as a harbinger of more
After Prevent or reduce outrage about actions (or inaction) to come [33]. For example, a train wreck claiming many lives
taken. might not generate as much social response-being a more
familiar and well-understood system-than a small pipeline
Source- Davis, G.. and Jones, D., “Risk Communication Guide for accident, in which the system is less understood and the conse-
State and Local Agencies,” California Governor’s Office of Emergency quences are perceived to be less controllable and more cata-
Services. October 2001 strophic; hence, the more widespread public outcry from a
pipeline accident (‘‘These lines are everywhere. they are old
and aging, and could be routinely failing!”).
200 ft of the pipeline), the impression of accuracy and preci-
sion is given. In reality, estimates that span several orders of Presenting risk estimates
magnitudc are often accepted as being as precise as practical.
It is also easy to lose sight of what the results of a statistical Risk is a very complex subject not fully understood by techni-
analysis is really telling us. The numbers usually represent our cal managers, much less most public audiences. Differences in
belief as to how a large number of pipelines (or pipeline seg- terminology and practices among the “experts” only increase
ments), or a specific pipeline operating over a very long period the difficulties surrounding risk communications. Therefore. it
of time, will tend to behave. The inferences to a specific system seems prudent to first solidify the communicator’s understand-
or specific time period will be very uncertain. as is always the ing of the risk analysis. Then, considerations must be extended
case when trying to predict the behavior of the individual from to the perceptions, concerns. and potential comprehension
the behavior of the group. limitations ofthe audience.
Challenges to failure estimates should be expected and must In Chapter 1, we noted that a scientific theory must be falsi-
be addressed. It is the position of this text that historical leaks fiable in order to be considered a “theory” rather than an act of
of any pipeline or group of pipelines provide only a very lim- faith, for example. In presenting risk estimates to the public,
ited estimate of future failure potential for a specific pipeline. the presenter probably believes that these estimates are based
They can easily under- or overestimate future leak potentials on scientific methods, are indeed bona fide theories, and can
because conditions have changed and will change from the be proven incorrect or at least demonstrated to be largely
previous operations. Therefore, leak history should not be inaccurate, over time. However, for relatively rare events like
used in isolation for judgments of failure probability. It is used pipeline failures, it might take long periods of time to amass
as evidence of certain conditions that might exist, and this evi- enough contradictory evidence to state that the presented risk
dence, along with all other information that can be obtained, is assessment was incorrect. Therefore, an audience hearing risk
used in a relative risk assessment to present a more realistic predictions regarding a system of concern may appreciate that
view of the risk. from a practical standpoint, they are being asked to take the
The communicator of risks must understand these and other numbers “on faith.” The communicator should be sensitive to
limitations of the assessment being presented. Preparations for this.
both technical and emotional challenges to any information Worst case scenarios are commonly used in a risk analysis,
presented to an audience are warranted. sometimes to demonstrate that such scenarios are almost
unimaginably improbable or that a system is relatively safe
even if the lowest probability event would occur. Of course,
Audience considerations oftentimes the worst case scenario is catastrophic. Such scenar-
Risk perception ios must often be presented if for no other reason than to
demonstrate that the communicator fully understands the
Risk judgments are heavily influenced by how memorable past risk and is not hiding critical information. It is “full disclosure”
events are and the public’s ability to imagine future events. and is needed for credibility. However, such scenarios will
Recent events, especially when covered extensively by media often work against the effective communication of risks
reporting, can seriously distort perception. People tend to over- since such scenarios are often the only fact retained from the
estimate risks from such events (such as homicides or pipeline communication.
explosions), while underestimating less dramatic risks (such as Risk communications always involve the quantification of
diabetes or fatal falls) [33]. Knowledge of recent events or complex issues. This requires that numbers be shown. At first
community-memorable events will be useful in predicting the glance, numbers appear to speak for themselves-they appear
temperament of an audience. to be clean of biases. A false precision is often assigned to an
It has been shown that people try to reduce the anxiety gener- analysis yielding a specific number, especially when such a
ated in the face of uncertainty by discounting that uncertainty number appears to be the result of extensive calculations. In
[33]. Therefore, the tendency might be to hear or retain only the reality, uncertainty and some bias will always be present in risk
consequence potential and not the more difficult-to-appreciate estimates, as is discussed in Chapter 1. High variability is a
improbability that is being communicated. This is especially characteristic of most pipelines and we add to that inherent
true when a worst case scenario, though highly improbable, uncertainty when we measure risk with our imprecise tools.