230
15 Risk-Informed Decision Making
Maintenance and repairs:
• Some repairs are carried out when damages are identified during the inspection
and after the IGRB discusses them.
15.3.4 Dam 4
In Sect. 11.1.2 the dam situation was qualified as follows for the north section of the
dam (see Tables 11.1 and 11.2):
• Slopes are steeper than designed;
• Failure is incipient toward the Eastern abutment;
• There is a monitoring system, but there is limited information regarding the readings, frequency, etc.
• There is uncontrolled erosion of the downstream (D/S) slopes.
15.4 Success and Failure Criteria
As we explained in Sect. 7.7, a risk assessment should always start with the definition
of the success and failure criteria. Indeed, the failure metric needs to be clearly stated
in order to avoid plunging into a state of confusion from the first contact with the
public (see Sect. 15.11.1). For example, in a dam risk assessment, if the dam “survives” a given hazard but the water source for the local population is contaminated,
it should be considered a failure. Too often we have seen a failure considered only
when a dam is breached, or looked at only from a stability point of view, disregarding
other issues, and forgetting the system’s success metric. This became obvious with
the Oroville hydrodam’s forensic analysis (see Sect. 3.1), which stated:
The Independent Forensic Team reportedly formed the impression that most Department
of Water Resources—DWR—staff and those involved in the Potential Failure Modes Analysis—PFMA—studies considered the use of the emergency spillway in terms of only an
“extreme” flood event. The Independent Forensic Team notes that a “1 in 100” year storm
might be considered an “extreme” event in an operational sense. However, from a dam safety
viewpoint an “extreme” dam safety would be a much larger storm.
Incidentally, the Oroville case was again an example of unclear glossary and definitions leading to misleading evaluations of risks. As the Oroville report states, failure
modes can be excluded from detailed consideration simply due to the definition of
“failure” that is used. In our practice we always start a risk assessment by defining the
system, its success and failure criteria, and multidimensional consequence metrics
(Oboni and Oboni 2018).
For this case study we will use a success criterion as follows: A dam is successful
if it does not breach due to “birth defects”, construction, operations, management
15 Risk-Informed Decision Making
Maintenance and repairs:
• Some repairs are carried out when damages are identified during the inspection
and after the IGRB discusses them.
15.3.4 Dam 4
In Sect. 11.1.2 the dam situation was qualified as follows for the north section of the
dam (see Tables 11.1 and 11.2):
• Slopes are steeper than designed;
• Failure is incipient toward the Eastern abutment;
• There is a monitoring system, but there is limited information regarding the readings, frequency, etc.
• There is uncontrolled erosion of the downstream (D/S) slopes.
15.4 Success and Failure Criteria
As we explained in Sect. 7.7, a risk assessment should always start with the definition
of the success and failure criteria. Indeed, the failure metric needs to be clearly stated
in order to avoid plunging into a state of confusion from the first contact with the
public (see Sect. 15.11.1). For example, in a dam risk assessment, if the dam “survives” a given hazard but the water source for the local population is contaminated,
it should be considered a failure. Too often we have seen a failure considered only
when a dam is breached, or looked at only from a stability point of view, disregarding
other issues, and forgetting the system’s success metric. This became obvious with
the Oroville hydrodam’s forensic analysis (see Sect. 3.1), which stated:
The Independent Forensic Team reportedly formed the impression that most Department
of Water Resources—DWR—staff and those involved in the Potential Failure Modes Analysis—PFMA—studies considered the use of the emergency spillway in terms of only an
“extreme” flood event. The Independent Forensic Team notes that a “1 in 100” year storm
might be considered an “extreme” event in an operational sense. However, from a dam safety
viewpoint an “extreme” dam safety would be a much larger storm.
Incidentally, the Oroville case was again an example of unclear glossary and definitions leading to misleading evaluations of risks. As the Oroville report states, failure
modes can be excluded from detailed consideration simply due to the definition of
“failure” that is used. In our practice we always start a risk assessment by defining the
system, its success and failure criteria, and multidimensional consequence metrics
(Oboni and Oboni 2018).
For this case study we will use a success criterion as follows: A dam is successful
if it does not breach due to “birth defects”, construction, operations, management