Page 360 - Pipeline Risk Management Manual Ideas, Techniques, and Resources

Societal versus individual risk 15/335
to many. Many established criteria are set at or near this alternate modes are less safe than pipelines, society’s risk expo-
value for the consequence of “increased chance of fatality.” sure has actually increased. Similarly, if an individual pipeline
One shidy suggests that North American society is not particu- operator determines that current, regulatory-implied levels of
larly concerned with risks that fall below this level [95]. So, the risk are not acceptable, and they choose to spend resources to
“one chance in a million” can be perhaps be seen as a ‘level of reduce their risk levels, then they may incur some economic
interest’ indicator. This might be a valid basis on which to consequences from competitors who choose to accept higher
establish some risk criteria. See page 341 for more discussion risk levels. The higher-perceived risk operators may or may not
about the “one chance in a million” criterion. incur additional costs due to their tolerance of higher risk levels.
In the en4 risk acceptability is a very personal judgment. It has been expressed in several studies that societal risk
No one wants to accept any risk without some accompanying aversion is inversely related to the number of potential fatali-
benefit. The value of the benefit is very subjective as is the ties: lin and Uno where n = number of fatalities, have each
perception of the level of risk, regardless of how many risk been proposed as “risk aversion functions” [95]. Such func-
measurements are presented. There are many relatively trivial tions can be used to help quantify differences in risk or risk
risks that disturb us greatly, while more threatening risks are of perception, as population densities change.
relatively little concern. For example, every year, thousands are
accidentally electrocuted, yet there are no mass demonstrations
against electricity or demands that distribution voltages be VI. Societal versus individual risk
reduced [57].
Decisions about risk are made in many dimensions. The use A distinction is often made between individual risk and societal
of risk analysis requires interpretation, context, and an under- risk. Individual risk provides an estimate for the risk to an indi-
standing of the analysis itself. When a regulatory body has to vidual at a specific location for a specified period of time. In
determine the “acceptability” of a risk, their determination is many applications, individual risk is equivalent to the risk to
normally based on many things such as the number of people “one or more individuals.” An individual risk for a pipeline,
exposed, societal benefits derived from the activity, precedents with potential consequences expressed in terms of fatalities,
set by other “approved activities, the degree of control over might be expressed like this: “This pipeline presents a risk of
exposure, and many other factors. chance of fatality per year.” This is normally equivalent
An interesting aspect of risk acceptability is that, whether or to saying “This pipeline has a one in amillion chance of causing
not criteria are quantified, a risk tolerance level can be inferred one or more fatalities per year.” The individual risk is insensi-
from regulations or industry actions. The acceptable risk levels tive to the number of individuals present, but the time of
impiied by regulations can be quantified with the assumption exposure for an individual can be considered.
that currently observed accident frequencies are the result of Societal risk is usually taken to mean the relationship
adherence to minimum requirements. This is somewhat between the frequency and number of individuals that could
complicated when regulations are performance based rather suffer a specified harm-for instance, the annual risk of death
than prescriptive, thus requiring “enough” mitigation to offset of a large number of people in one pipeline incident. It does
threats, rather than prescribing exactly which specific consider the number of individuals exposed as well as their
actions and what frequency of action are required. In that case, times of exposure. Because societal risk must aggregate many
common industry practices will normally arise from such per- possible scenarios (such as various fatality count scenarios),
formance-based regulations and those can be used to infer the FN curves such as those shown in Figure 15.1 are often used to
current acceptable risk levels. Similarly, an individual com- display risks. (FN curves are also discussed in Chapter 14.)
pany’s level of “acceptable risk” can be “back calculated from An individual is obviously not exposed to the threat from the
their actions, even when no such quantification is offered by the entire length of a multiple-mile pipeline simultaneously. Her
company. For example, if a company performs actions based on maximum exposure occurs if she is very near (perhaps directly
strict adherence to minimum regulatory requirements and has over) the pipeline 24 hours of every day. She is also exposed
systems that are similar to most other systems, then their risk to pipeline failures some distance along the pipeline to either
levels should be similar to that of all other companies following side. If she moves perpendicularly away from the line, her risk
similar protocols and operating in similar environments. The decreases because she is exposed to less pipeline, based on
performance record of the entire population of pipeline simple geometry. So, under one approach, the risk per unit
systems has therefore been implicitly judged to be acceptable length of a pipeline can be used to estimate individual risk by
by the company. Such estimations will, of course, be very determining the length of pipe that can affect a single point.
uncertain because many assumptions of similarity must be Logically, this length would be determined by using hazard
made and, even then, finding comparative failure rate data in zone calculations.
sufficient quantity to draw meaningful conclusions will be dif- A probabilistic risk assessment (PRA) is traditionally
ficult. Nevertheless, knowledge that actions themselves lead to applied to process industry scenarios where the bounds of
acceptable risk estimates is an interesting concept-intuitive a perceived threat can be clearly defined. Applications to
on some level, but with subtle implications. pipeline become more problematic, especially when compar-
As previously noted, an ironic phenomenon may occur in the isons are to be made with nonlinear facilities. For example, a
quest for risk reduction in pipelining. Because most activities 700-mile pipeline will have a societal risk proportional to 700
are cost driven, money spent in the name of safety may actually times its 1 -mile risk, or 3696 times its 1000-ft risk, and so on, if
increase the overall risks. For example, if safety-enhancing unit-length risks are extrapolated to their full length. Compared
spending is mandated for pipelines, the increased costs may to a chemical plant that can only impact a limited geographical
drive more freight to alternate transportation modes. If these area, the pipeline will appear to present a greater risk.