182
7 Risk Assessment and Management of Chemical Processes
Combining Fault Tree and Event Tree Analyses
As shown in Fig. 7.7, fault trees and event trees can be combined to obtain a full
picture of the causes and consequences of an accident scenario (as depicted earlier
in Fig. 7.2).
In this case, the top event represents the loss event; the causal events of the
fault tree represent process failures, human errors, or external influences; and the
outcomes of the event tree represent the consequences of the loss event, namely,
effects on people, the environment, and property. In this pathway of causes to
consequences, preventive and mitigative safeguards are also considered.
Another related technique that can be used is the bow tie analysis (ISO, 2009).
This approach provides a clear representation of accident scenarios and focuses
especially on preventive and mitigative safeguards present in the system (see the
structure of Fig. 7.2). However, unlike fault and event trees, bow tie analyses cannot
be applied to cases where multiple causes occur simultaneously.
7.7
Risk Evaluation (Step 5)
Once the consequences and the probability of potential accident scenarios have been
estimated, the risk evaluation step then combines these to determine the level of
risk posed and classify it as acceptable or not acceptable. Acceptability can be
determined through a comparison of the risk with protection goals set by internal
corporate guidelines and by legislation, benchmarking with other processes, or
through other predefined criteria. Risks deemed not to be acceptable will require
subsequent safety improvement actions that will be identified and implemented in
the next (and final) step of risk management.
One common way to estimate and communicate the level of risk is through the
use of a risk matrix. These matrices help to visually estimate and present the level of
risk using one axis to show increasing consequence severity and the other to show
increasing probability of occurrence.
Consequences should include the estimated effects on people, the environment,
and property, and its axis in the matrix can be represented through either a numerical
scale or using descriptive terms such as “negligible,” “low,” etc. (see Table 7.8).
Similarly, the probability axis can also be represented by a numerical scale or
through descriptive terms (see Table 7.9).
Figure 7.8 provides a generic example of a risk matrix. In this example, the
range of unacceptable risks is delimited by setting an acceptability line. Risks that
exist above this line are classified as unacceptable and require risk management.
The shading of cells within the matrix helps further visualize the level of risk
and is set depending on how the probability and consequence scales are defined.
The consequence and probability axes themselves can also be designed either
symmetrically or to give more weight to one or the other.
7 Risk Assessment and Management of Chemical Processes
Combining Fault Tree and Event Tree Analyses
As shown in Fig. 7.7, fault trees and event trees can be combined to obtain a full
picture of the causes and consequences of an accident scenario (as depicted earlier
in Fig. 7.2).
In this case, the top event represents the loss event; the causal events of the
fault tree represent process failures, human errors, or external influences; and the
outcomes of the event tree represent the consequences of the loss event, namely,
effects on people, the environment, and property. In this pathway of causes to
consequences, preventive and mitigative safeguards are also considered.
Another related technique that can be used is the bow tie analysis (ISO, 2009).
This approach provides a clear representation of accident scenarios and focuses
especially on preventive and mitigative safeguards present in the system (see the
structure of Fig. 7.2). However, unlike fault and event trees, bow tie analyses cannot
be applied to cases where multiple causes occur simultaneously.
7.7
Risk Evaluation (Step 5)
Once the consequences and the probability of potential accident scenarios have been
estimated, the risk evaluation step then combines these to determine the level of
risk posed and classify it as acceptable or not acceptable. Acceptability can be
determined through a comparison of the risk with protection goals set by internal
corporate guidelines and by legislation, benchmarking with other processes, or
through other predefined criteria. Risks deemed not to be acceptable will require
subsequent safety improvement actions that will be identified and implemented in
the next (and final) step of risk management.
One common way to estimate and communicate the level of risk is through the
use of a risk matrix. These matrices help to visually estimate and present the level of
risk using one axis to show increasing consequence severity and the other to show
increasing probability of occurrence.
Consequences should include the estimated effects on people, the environment,
and property, and its axis in the matrix can be represented through either a numerical
scale or using descriptive terms such as “negligible,” “low,” etc. (see Table 7.8).
Similarly, the probability axis can also be represented by a numerical scale or
through descriptive terms (see Table 7.9).
Figure 7.8 provides a generic example of a risk matrix. In this example, the
range of unacceptable risks is delimited by setting an acceptability line. Risks that
exist above this line are classified as unacceptable and require risk management.
The shading of cells within the matrix helps further visualize the level of risk
and is set depending on how the probability and consequence scales are defined.
The consequence and probability axes themselves can also be designed either
symmetrically or to give more weight to one or the other.