Frequency Duration Outage Method
NOTE: Outages from this setting will be superseded by data within the Daily Change input table, if that data is modifying the same resource(s).
The Frequency Duration outage method option allows units to fail or return to service at any time step within the simulation, not just at the beginning of a month or a day. The method will employ all or nothing outages for most outages, but will use partial outages at the beginning and end of the outage period. For example, if a unit is determined to be out for 7.5 hours, Aurora will account for this by using a partial outage for 1 of the 8 hours affected, and a full outage for the other 7.
To perform multiple risk iterations with the frequency duration outage method, the Perform General Risk Study option should be checked.
For a more detailed discussion see Forced Outage Modeling Using the Frequency Duration Method: Considerations, Pitfalls, and Solutions.
NOTE: If this option is selected for a Risk run, the existing option of defining binomial resource outages through entries in the Risk Table is not allowed.
To use the frequency duration outage method:
1. Select Frequency Duration for the outage method in the dropdown.
2. Input resource information in the following columns of the Resources Table:
Use this column to specify the Forced Outage Rate (FOR) defined in percent. If a value of FUEL is used in this column, values from the Fuel Table are used.
Mean Repair Time
Use this column to specify the Mean Time to Repair (MTTR) defined in hours. If a value of FUEL is used in this column, values from the Fuel Table are used.
Risk Outage
Use this column to set a flag indicating whether the resource uses the random forced outage with MTTR methodology. A value of (1) will activate the frequency duration methodology. A null (no value) or zero (0) will turn off the frequency duration method. If a value of FUEL is used in this column, values from the Fuel Table are used.
Risk Resource Link
Use this column to link (assign the same outage pattern) the resource specified on the current row to the resource specified in this field.
Frequency and Duration Outage Logic
This logic samples from an exponential distribution for both the time to fail and time to repair variables. The exponential distribution is a single parameter distribution, requiring only a mean to define the sampling distribution. The distribution for repair time is defined through the input value for MTTR. The Mean Time to Failure (MTTF) is calculated using the resource Forced Outage Rate (FOR) and the MTTR. The equation used is MTTF = MTTR*(1/FOR – 1). Input resource FORs can change by month and year and the MTTF is updated monthly during the simulation to reflect this variation.
Initialization
A random variable seed structure is set up that allows the sequence of outages and repair times to be repeated for individual resources from study to study. It is driven by the Risk input seed and will allow repeat-ability from run to run, even with the addition of new resources, as long as any new resources are added at the end of the Resources Table. See Restart Risk Study for more information.
At the initial time step of the study, the current resource FORs are used to determine the initial outage state of the resource. If on outage, a sample is taken from the repair time distribution to determine time to repair for this outage. If the unit is determined to be available, a sample is taken from the failure time distribution to determine time to fail. Because the initial resource states and durations are theoretically associated with events that took place before the study start, these initial samples for repair or failure time are truncated using a simple random sample from a uniform distribution.
Simulation
The simulation process for unit outages is a straightforward sample and wait process. As a unit fails, a sample is taken for the time to repair for this outage, and the unit is unavailable to the dispatch until the repair time has elapsed. As a unit returns to service (repairs are finished), a sample is taken to determine the number of hours of operation until its next failure. The dispatch results are used to determine whether the unit actually operates in any given hour, and only those hours where the unit actually runs are counted against time to failure. For non-contiguous dispatch hours, the breakpoints for changes in operating levels are assumed to be at the hour of dispatch (i.e., the observed operating level from a dispatch will be assumed constant until the next dispatch).
Freq Duration Outages Based on Elapsed Time
This option changes the process used to determine the Mean Time to Fail (MTTF). By default, as described above, only those hours where the unit actually runs are counted against time to failure. In other words, the MTTF is driven by the hours the unit operates. With the Freq Duration Outages Based on Elapsed Time option selected, this logic changes so that the MTTF is driven by elapsed time since the last outage instead of hours of operation.
Additional Reporting
Selecting the Write Frequency Duration Debug Table switch will report an additional FreqDurationStat Output Table with detailed information about the random draws associated with outage events and return-to-service events.
Outage Settings
