Titel_SS06

Figure 3.6: Decision/event tree for prior and posterior decision analysis.

A posterior analysis is in principle of the same form as the prior analysis, however, changes
in the branching probabilities and/or the consequences in the decision/event tree reflect that
the considered problem has been changed as an effect of risk reducing measures, risk
mitigating measures and/or collection of additional information.

A posterior analysis may thus be used to evaluate the effect of activities, which factually have
been performed. For example, for assessment of existing facilities the testing and inspection
of the “as built” facility would be expected to reveal many gross design and construction
errors, leading to a more accurate reliability analysis.

A pre-posterior analysis may be illustrated by the decision/event tree shown in Figure 3.7.

Decisions Random

outcome

Utility /

Consequences

U

Planned

investigations

Results of

investigations

Decisions Random

outcome

Risk reducing

and mitigating

actions

Activity

performance

Figure 3.7: Decision/event tree for pre-posterior decision analysis.

Using pre-posterior analysis optimal decisions in regard to activities that may be performed in
the future, e.g. the planning of risk reducing activities and/or collection of information may be
identified. An important prerequisite for pre-posterior analysis is that decision rules need to be
formulated for specifying the future actions that will be taken on the basis of the results of the
planned activities.

In a pre-posterior analysis the optimal investigation a^4 is identified through:



1

min ' ''[ ((), )] min ' ''((), ) (())

n

aaZZ Z i i

i

E E Caz z E P az zC az



 

(^89)  (3.10)