Reliability Reports

This page provides definitions related to the standard reports provided with Responder's Archive Explorer.

The Reliability Indices DO NOT include planned outages or non-outages by default. An administrator may include planned outages by performing an optional configuration outlined in the Configuration Guide.

Aggregation Interval is a period of time less than or equal to the period covered by a report, within which interruptions are accumulated to compute reliability indices. Each row within a given report corresponds to one aggregation interval. For example, the report known as “Reliability Indices: System-Wide by Week” uses an aggregation interval of one week. If this report is generated for a one year reporting period, then it will have 52 rows, each reporting the SAIFI, SAIDI, etc. as computed for a specific week of the year. The indices given in each row reflect only the interruptions that began within the particular week to which the row corresponds.

For the "To Date" reliability reports there is only one aggregation interval (and hence only one row), and it is implicitly the same as the entire reporting period, namely the entire known outage history of the system to date.

SAIFI (System Average Interruption Frequency Index) is the average number of interruptions per customer during the year. It is calculated by dividing the total annual number of customer interruptions by the total number of customers served during the year:

     SAIFI = Total Number of Customer Interruptions / Total Number of Customers Served

SAIDI (System Average Interruption Duration Index) is the average total downtime (reported here in minutes) per customer during the year. It is calculated by dividing the annual sum of all customer interruption durations over the year by the total number of customers served during the year:

     SAIDI =  Sum of Durations of Customer Interruptions / Total Number of Customers Served

CAIFI (Customer Average Interruption Frequency Index) is the average number of interruptions per customer during the year, for customers that experience at least one interruption during the year. It is calculated by dividing the total number of customer interruptions during the year by the total number of customers affected by interruptions during the year. In determining the total number of customers affected by interruptions during the year, each customer is counted only once regardless of the number of interruptions (greater than or equal to one) that the customer may have experienced during the year.

    CAIFI = Total Number of Customer Interruptions / Total Number of Customers Experiencing Interruptions

CAIDI (Customer Average Interruption Duration Index) is the average total downtime (reported here in minutes) per customer-interruption. It is calculated by dividing the sum of all customer interruption durations by the total number of customer interruptions over some interval.

       CAIDI = Sum of Durations of Customer Interruptions / Total Number of Customer Interruptions

MAIFI (Momentary Average Interruption Frequency Index) is the number of momentary interruptions divided by the total number of customers. (“Momentary interruption” is defined below under “Interruptions Excluded from the Statistics.”) MAIFI is calculated by dividing the total number of momentary customer interruptions within an aggregation interval by the total number of customers.

      MAIFI = Total Number of Momentary Customer Interruptions / Total Number of Customers Served

Interruptions Excluded from the Statistics

Momentary Interruptions

Interruptions of duration shorter than a specified amount of time (an administrative setting whose default is 5 minutes), do not count in the SAIFI, SAIDI, CAIFI and CAIDI statistics, and are instead represented by the MAIFI statistic.

Major Event Days

Interruptions that begin on any Major Event Day do not count in any of the reported statistics. A Major Event Day is defined, as per IEEE standard 1366-2003, as any day whose SAIDI figure (for the day) exceeds a certain threshold value Tmed, which is derived from the previous 5 years' worth of historical outage data (the exact procedure for that derivation is documented in IEEE standard 1366-2003).

Exceptional Values

A reported value of -1 for any statistic indicates that the actual value could not be computed due insufficient data (e.g., the rx_customers table is empty, or a similarly unusual data exception is encountered).